Sox3 dosage regulation is important for roof plate specification during central nervous system development
A thesis submitted to the University of Adelaide for the degree of Doctor of Philosophy
By
Kristie Pan Yu Lee, B.Sc. (Hons)
Department of Biochemistry
School of Molecular and Biomedical Sciences
The University of Adelaide
Adelaide, Australia
June 2011 Errata
The following corrections have been made in the final version of the thesis: --
Table of Contents
“5.4.1 Bmp and Wnt RP expression was diminished transgenic embryos” to “5.4.2 Bmp and Wnt RP expression was diminished in transgenic embryos”
“5.4.2” to “5.4.3”
“5.4.3” to “5.4.4”
“5.4.4” to “5.4.5”
“5.4.5” to “5.4.6”
“APPENDIX IV: AN OUTLINE OF THE MICROARRAY EXPERIEMNT” to “APPENDIX IV: AN OUTLINE OF THE MICROARRAY EXPERIMENT”
Summary
Third paragraph, line 8: “have similar role” to “have a similar role”
Abbreviations
Title: “Abbreviations Abbreviations” to “Abbreviations
Anatomical terms: “MZ mantel zone” to “MZ mantle zone” Chapter 1: Introduction
Page 3, line 6: “subfamily” to “subfamilies”
Page 4, line 14: “to minor groove of DNA helix” to “to the minor groove of the DNA helix”
Page 5, line 6: “activate” to “activates”
Page 5, line 11: “These indicate” to “These data indicate”
Page 5, line 29: “target genes” to “target gene”
Page 6, line 1: “target genes” to “target gene”
Page 6, line 1-2: “function specialisation” to “the functional specialisation”
Page 7, line 7: “study on the abnormal development” to “study the abnormal development”
Page 7, line 8: “the discussion” to “the following discussion”
Page 7, line 25: “the formation of DV is” to “the formation of DV axis is”
Page 8, line 16: “with respects to” to “with respect to”
Page 8, line 21: :”(blastocoels)” to “(blastocoel)”
Page 8, line 24: “process which some epithelial cells” to “process by which some epithelial cells”
Page 10, line 13: “mesenchyme” to “mesoderm”
Page 10, line 26: “a node explants from a late-streak embryo has” to “node explants from late-streak embryos have”
Page 10, line 27: “the expression of neural marker” to “the expression of the neural marker”
Page 11, line 14: “The following few section” to “The following section”
Page 13, line 10: “binds to Ephrin” to “binds to Eph’
Page 14, line 3: “NC cells migration” to “NC cell migration” Page 15, line 21: “as well as constitutive” to “as well as constitutively active”
Page 15, line 25: “dysplasic” to “dysplastic”
Page 16, line 4: “part of” to “part of the”
Page 16, line 7: “with onset of” to “with the onset of”
Page 16, line 18: “domain of dorsal” to “domain of the dorsal”
Page 17, line 2: “may involve in” to “may be involved in”
Page 18, line 1: “CNS needs to be” to “the CNS needs to be”
Page 18, line 5: “referred as” to “referred to as”
Page 18, line 6: “controversial that whether” to “controversial whether”
Page 18, line 9: “is not incorporated in” to “is not incorporated into”
Page 18, line 23: “with diminished AP spanning as progressing dorsally” to “with diminishing AP spanning towards the dorsal domain”
Page 18, line 30: “In cooperation of” to “In cooperation with”
Page 18, line 31: “within diencephalon” to “within the diencephalon”
Page 19, line 3: “Division along AP” to “Division along the AP”
Page 19, line 4: “composing” to “comprised”
Page 20, line 20: “present at” to “present in”
Page 21, line 1” “inner cell mast” to “inner cell mass”
Page 22, line 8: “a role of” to “a role for”
Page 24, line 7: “portion” to “proportion”
Page 24, line 29: “pharyngeal arches” to “pharyngeal arch”
Page 25, line 5: “(Minoux and Rijli)” to “(Minoux and Rijli, 2010) Page 25, line 20: “Together, these” to “Together, these studies”
Page 26, line 1: “describes” to “describe”
Page 26, line 2: “the discussion Sox3” to “the discussion of Sox3”
Page 26, line 21: “there lack any” to “there is lack of any”
Page 26, line 30: “transcription regulation” to “the transcriptional regulation”
Page 27, line 2: “this study” to “the study in Zhang et al. (2004)”
Page 27, line 4: “secrets” to “secretes”
Page 27, line 16: “throughout CNS” to “throughout the CNS”
Page 27, line 23-24: “when both ectopically expressed, xSox3 counteracted POU class V protein oct-91 (xOct91) neuralising” to “when both xSox3 and POU class V protein oct-91 (xOct91) are ectopically expressed, xSox3 counteracted xOct91 neuralising”
Page 27, line 26: “function” to “functions”
Page 29, line 13: “A requirement of” to “A requirement for”
Page 29, line 17: “resulting in the loss” to “resulted in the loss”
Page 30, line 22: “xenopus” to “Xenopus”
Page 31, line 20: “a effect” to “an effect”
Page 31, line 24: “heavily implicated the” to “heavily implicated in the”
Page 32, line 29: “also promote gliogenesis” to “also promoting gliogenesis”
Page 33, line 15: “capable to induce” to “capable of inducing”
Page 33, line 19: “identity” to “identify”
Page 34, line 2: “150kp” to “150kb”
Page 34, line 12: “capable to driving” to “capable of driving” Page 34, line 28: “further strengthen by” to “further strengthened by”
Page 35, line 17: “NC cells migration defect towards the” to “NC cell migration defect towards”
Page 36, line 1: “activate” to “activates”
Page 37, line 10: “neural during” to “neural identity during”
Page 40, line 6: “functional compromisation” to “functional compromise”
Page 40, title 1: “1.5.3 The subcommisurral organ” to “1.5.3 The subcommissural organ”
Page 40, line 10: “differentiated from” to “differentiate from”
Page 40, line 12: “fenestrated wall” to “fenestrated walls”
Page 40, line 17: “aspect of SCO” to “aspect of the SCO”
Page 41, line 10: “SCO is” to “the SCO is”
Page 42, line 8: “SCO primordium” to “the SCO primordium”
Page 43, line 5: “poses” to “pose”
Page 43, line 9: “neuroepithelium” to “neuroepithelial”
Page 44, line 19: “possess a thrombospondin type I repeats and a low density lipoprotein receptor” to “posses thrombospondin type I repeats and low density lipoprotein receptor”
Page 44, line 20: “extracellular matrix” to “extracellular matrix proteins”
Page 44, line 28: “function SCO” to “functional SCO”
Page 45, line 7: “islet of cuboidal cells was no present although remanent of SCO ependymal cells” to “islets of cuboidal cells was no longer present although a remanent of SCO ependymal cells”
Page 45, line 24: “author” to “authors”
Page 45, line 28: “as evident” to “as evidenced” Page 46, line 4: “damages” to “damage”
Page 46, line 17-18: “individuals contracting mumps encephalitis led to damages to the CNS ventricular ependymal lining and hence causing non-communicating hydrocephalus” to “contraction of mumps encephalitis can cause damage to the ventricular ependymal lining, which may lead to the development of non-communicating hydrocephalus within affected individuals”
Page 46, line 22: “evidence from” to “evidence from a”
Page 46, line 26: “characteristic to” to “characteristic of”
Page 46, line 27: “the presence of dome-shaped cranium” to “the presence of a dome-shaped cranium”
Page 46, line 29: “SCO defect” to “SCO defects”
Page 47, line 12: “suffered” to “suffering”
Page 47, line 25: “which is the followed” to “which is followed”
Page 48, line 8: “the SCO development” to “SCO development”
Page 48, line 10: “Recent study” to “A recent study”
Page 48, line 15: “In addition of” to “In addition to”
Page 50, line 13: “junction” to “junctions”
Page 51, line 21: “epithelial” to “epithelia”
Page 51, line 30: “the Spag6” to “Spag6”
Page 52, line 15: “precedes the ependymal” to “precedes ependymal”
Page 52, line 16: “by the defective ependymal” to “by defective ependymal”
Page 52, line 20: “associated to” to “associated with”
Page 53, line 6: “signalling event” to “signalling events” Page 53, line 18: “regulation the” to “regulation of the”
Page 53, line 27: “indicating the Msx1 is only indispensable for the development of RP” to “indicating that Msx1 is only indispensable for development of the RP”
Page 53, line 29: “null mutant” to “null mutants”
Page 54, line 30: “known function to” to “known function for”
Page 56, line 16: “6-8 of” to “6-8 weeks of”
Chapter 2: Materials and methods
Page 62, line 13: “50mg/kg” to “50 mg/kg”
Page 62, line 13: “10mg/ml” to “10 mg/ml”
Chapter 3: Characterisation of CH in Sox3 transgenic mouse model
Page 74, line 9: ‘the possibility a defective” to “the possibility of a defective”
Page 75, line 24: “no evident” to “no evidence”
Page 77, line 18: “accounted for this” to “accounted for in this”
Page 77, line 22: “Sr and Nr Sox3 transgene” to “Sr and Nr Sox3 transgenes”
Page 78, line 18: “Haemorrphrage” to “Haemorrhage”
Page 79, line 20: “recapitulated this” to “recapitulated Sox3 endogenous”
Page 80, line 2: “SCO primordium” to “The SCO primordium”
Page 83, line 16: “assoiction” to “association” Page 84, title 1: “in other area of dorsal” to “in other areas of the dorsal”
Page 85, line 27: “inhibitor or” to “inhibitor of”
Page 85, line 30: “study the” to “study of the”
Page 86, line 8: “Sox3 transgene has” to “the Sox3 transgene has”
Page 86, line 10: “integration induced” to “integration-induced”
Page 86, line 20: “inhibits conformational change” to “inhibits the conformational change”
Page 86, line 22: “The lack of poly-A tail” to “The lack of a poly-A tail”
Page 87, line 20: “brought by” to “brought about by”
Page 88, line 14: “telencephalic transgenic expression” to “telencephalic transgene expression”
Page 88, line 26-29: “both the Sr/+ and Nr/+ embryos lack telencephalic Sox3 transgene expression at 10.5 to 12.5 dpc (N. Rogers, unpublished data) arguing against the case of a positional effect despite this explanation is still reserved.” to “both Sr/+ and Nr/+ embryos lack telencephalic Sox3 transgene expression from 10.5 to 12.5 dpc (N. Rogers, unpublished data), which argues against positional effect as an explanation for this observation.”
Page 89, line 17-18: “there cannot rule out the possibility of a positional effect whereby integration site residing transcription enhancing element might act on the transgene and upregulate its expression.” to “it is possible that local transcription enhancing element at the integration site might act on the transgene and upregulate its expression.”
Page 90, line 5: “with identical” to “with an identical”
Page 90, line 28: “spatial domain, and the” to “spatial domain and the”
Page 91, line 17: “these indicate” to “these observations indicate”
Page 93, line 5: “To further this” to “To further investigate this”
Page 93, line 6: “proliferation study” to “a proliferation study” Page 93, line 6: “demonstrated that adult rat SCO” to “demonstrated that the adult rat SCO”
Page 93, line 7: “the SCO is one of all” to “the SCO is the only”
Page 93, line 7: “express” to “expresses”
Page 93, line 11: “When transplanted in vivo” to “When transplanted in mouse in vivo”
Chapter 4: Analysis of SOC development in Sox3 transgenic mice
Page 95, line 13: “As shown in prevous chapter” to “As shown in the previous chapter”
Page 95, line 18: “ependymals” to “ependymal”
Page 95, line 25: “ventriculear” to “ventricular”
Page 95, line 30: “basal localisation” to “basal nuclear localisation”
Page 96, line 14: “invaginated SCO” to “the invaginated SCO”
Page 96, title 2: “SCO dyspalsia” to “SCO dysplasia”
Page 93, title 1: “functioally” to functionally”
Page 98, line 13: “dysplasic” to “dysplastic”
Page 98, line 16: “by which the SCO” to “by which time the SCO”
Page 98, line 20: “indicate” to “indicates”
Page 98, line 28: “exercebated” to “exacerbated”
Page 99, line 8: “Habenuclar” to “Habenular”
Page 100, line 17: “using on alternative technique” to “using alternative techniques”
Page 102, line 8: “previous study” to “a previous study” Page 102, line 18: “proliferative” to “the proliferative”
Page 102, line 19: “The VZ” to “the VZ”
Page 103, line 25: “but also in the” to “but is also found in the”
Page 104, line 4: “intense and was patchy” to “intense and patchy”
Page 104, line 6: “at its neighbouring section” to “in a neighbouring section”
Page 104, line 15: “dysplasic” to “dysplastic”
Page 106, line 12-13: “was present at the LV ChP was detected from WT” to “was present in the LV ChP of WT”
Page 106, line 16: “EGFP indicative” to “EGFP, indicative”
Page 107, line 22: “apparent in those of the diencephalon.” to “apparent in the VZ of diencephalon”
Page 108, line 18: “causative role of defective ChP” to “causative role for defective ChP”
Page 109, line 10: “transformed” to “transform”
Page 109, line 24: “accompanied the” to “accompanied by the”
Page 109, line 25: “neuroepithelial cell” to “neuroepithelial cells”
Page 109, title 1: “sepcific” to “specific”
Page 110, line 8: “projected” to “projecting”
Page 111, line 13: “mediodorsally” to “mediodorsal”
Page 111, line 16: “embryos are in contrast” to “embryos is in contrast”
Page 112, line 1: “dysplasic” to “dysplastic”
Page 112, line 29: “transgene” to “transgenes”
Page 113, line 5: “levels Sox3” to “levels of Sox3” Page 113, line 7: “response” to “respond”
Page 113, line 9: “In addition to the diencephalon” to “In addition to the diencephalic RP,”
Page 113, line 11: “Fasciculation of corpus callosum” to “Fasciculation of the corpus callosum”
Page 113, line 13: “corpus callosum” to “a corpus callosum”
Page 113, line 23: “in the RP development” to “in RP development”
Page 114, line 31: “at 12.5 dpc” to “in the 12.5 dpc”
Page 115, line 18: “fails RP development” to “fails to develop RP”
Page 115, line 23: “a Sox3 dosage” to “Sox3 dosage”
Page 115, line 30: “to be an indirect” to “to be indirect”
Chapter 5: Molecular function of Sox3 in SCO development
Page 118, line 4: “atlas” to “analysis”
Page 120, line 4: “statistically” to “statistical”
Page 120, line 16: “(assumed most gene expressions did not change between samples)” to “(assuming most gene expressions did not change between samples)”
Page 120, line 22: “but the other” to “but absent from the other”
Page 121, line 15: “analysed” to “analysis”
Page 121, line 22: “the direction for future investigation” to “the direction of future investigation”
Page 125, line 12: “might due” to “might be due” Page 126, line 20-21: “In contrast, Sr/+; Nr/+ SCO Dab2 expression had a 1.25-fold decrease from the two microarray samples. This did not correlate with the microarray, which indicated a 3-fold decrease in expression.“ to “On the other hand, the microarray experiment detected a 3-fold decrease of Dab2 expression in Sr/+; Nr/+ SCO . However, qRT-PCR analysis using identical SCO samples as the microarray indicated that Dab2 expression has only downregulated for 1.25-fold.”
Page 129, line 10: “accuratly” to “accurately”
Page 129, title 1: “5.4.1” to “5.4.2”
Page 129, line 27: “SMAD complex” to “the SMAD complex”
Page 131, line 10: “in vitro study” to “an in vitro study”
Page 133, title 1: “5.4.2” to “5.4.3”
Page 134, line 1: “expressed at” to “expressed in”
Page 134, line 3: “these prompt” to “these observations prompt”
Page 134, line 16: “encode a protein that is” to “encode proteins that are”
Page 134, title 1: “5.4.3” to “5.4.4”
Page 135, title 1: “5.4.4” to “5.4.5”
Page 135, line 20: “to implicate in” to “to be implicated in”
Page 136, line 7: “similar to” to “similar function to”
Page 136, line 11: “Perkinje” to “Purkinje”
Page 136, line 12: “GABAnergic” to “GABAergic”
Page 138, title 1: “5.4.5” to “5.4.6”
Page 139, line 3: “rhombencephalon an anteriorly” to “rhombencephalon anteriorly”
Page 139, line 12: “by an prior defect” to “by a prior defect” Chapter 6: Discussion
Page 141, line 9: “mild defect was” to “mild defects were”
Page 142, line 1: “within the dorsal neural developing RP” to “within the developing dorsal neural RP”
Page 142, line 5: “unlikely due to” to “unlikely to be due to”
Page 142, line 12: “These suggest” to “These observations suggest”
Pate 142, line 13: “functional redundant with Sox2 such that the function of Sox3 is substituted by Sox2 in Sox3 null embryos” to “functionally redundant with Sox2”
Page 143, line 20: “these indicate” to “these data indicate”
Page 143, line 23: “SOX3 expression difference were found” to “SOX3 expression were found”
Page 143, line 25: “intensity that those” to “intensity than those”
Page 144, line 1: “progneitor” to “progenitor”
Page 144, line 4: “stem cells” to “stem cell”
Page 144, line 25: “upregulate” to “upregulates”
Page 145, line 7: “in consistent” to “consistent”
Page 145, line 12: “effect” to “effects”
Page 145, line 14: “is due its dynamic function in different” to “is due to its dynamic function in different”
Page 145, line 15: “context” to “contexts”
Page 145, line 18: “Sox3 direct targets” to “direct Sox3 targets”
Page 146, line 4: “Since the RP is an organising centre” to “The RP is an organising centre” Page 146, line 16: “recently study” to “a recent study”
Page 146, line 18: “the transcription activity of Hes1, which Hes1 then acts” to “the transcription activity of Hes1. In return, Hes1 then acts”
Page 146, line 30: “repressing” to “repression”
Page 148, line 25: “leaving the dorsal NT with cells making up the definitive RP” to “and only cells that give rise to the definitive RP are left at the dorsal NT”
Page 149, line 15: “understanding this defect” to “understanding of this defect”
Page 149, line 16: “Cranial NC cells” to “Cranial NC cell”
Page 149, line 17: “begin migrate” to “begin to migrate”
Pag e149, line 23: “were observed” to “was observed”
Page 151, line 1: “due the functional redundancy” to “due to the functional redundancy”
Page 152, line 10: “lethally” to “lethality”
Page 152, line 24: “immuneprecipitation” to “immunoprecipitation”
Figures
Figure 1.18, legend, line 3: “trasngenic” to “transgenic”
Figure 1.6, legend, line 1: “within an developing” to “within a developing”
Figure 3.5 A, the note at the bottom of the figure, line 1: “frequecny” to “frequency”
Figure 3.14, the figure at the top right corner is labelled as (K)
Figure 4.1, legend, line 7: “ependymals” to “ependymal”
Figure 4.15, legend, line 1: “at a varing low level” to “at varying low levels”
Figure 4.16, bottom of the figure: “12.5 dpc” to “15.5 dpc” CHAPTER 1: INTRODUCTION ...... 1
1.1 The central nervous system ...... 1
1.2 SOX3 – a member of the SOX transcription factor family ...... 3
1.2.1 Sox3 belongs to SOXB1 subgroup ...... 3
1.2.2 SOX transcription factor interact with partnering proteins to achieve specificity in transcription regulation ...... 4
1.3 An overview of CNS development ...... 7
1.3.1 Axes formation and patterning of the developing CNS domains ...... 7
1.3.2 Neural induction ...... 8
The development of early embryos ...... 8
Anterior organisers protect the default neural fate of the epiblast ...... 9
AVE and its role in early neural development ...... 9
GO and its derivative (the node) are critical for neural development ...... 10
Summary ...... 11
1.3.2 Neurulation and dorsal midline maintenance ...... 11
NT closure ...... 11
Dorsal NT development and NC migration ...... 13
Summary ...... 14
1.3.3 RP formation and its role in dorsal CNS development ...... 14
The molecular regulatory network of RP development ...... 15
Bmp and Wnt signalling are required for both NC cells and definitive RP development ...... 16
Other genes that lead to RP dysgenesis within the diencephalon ...... 17
Summary ...... 17
1.3.4 Patterning of forebrain – diencephalon ...... 17
1.4 SOX3 in the developing CNS ...... 20 1.4.1 SOXB1 expression during neural development ...... 20
SOXB1 expression during anterior neural development in mouse ...... 20
SOX3 expression during anterior neural development in other vertebrates ...... 21
Summary ...... 22
1.4.2 SOX3 and human disease – X-linked hypopituitarism ...... 22
1.4.3 Sox3 null mouse model ...... 24
1.4.4 SOX3 has an important role in early and neural development ...... 25
SOX3 in early embryonic development ...... 26
SOX3 in neural plate determination ...... 27
SOX3 and neurogenesis in CNS ...... 28
Other SoxB1 members have also been implicated in early embryonic and neural development in CNS ...... 29
Genetic interactions of SOX3 in neural development ...... 31
Summary ...... 32
1.4.6 SOX3 in sex determination ...... 33
1.4.7 Regulation of SOX3 expression ...... 33
1.4.8 Functional redundancy between SoxB1 subgroup members ...... 35
1.5 Cerebrospinal fluid, hydrocephalus and the subcommissural organ ..... 38
1.5.1 CSF production and circulation ...... 38
1.5.2 CNS barriers and CSF homeostasis ...... 39
1.5.3 The Subcommisurral organ ...... 40
1.5.4 Origin and development of mouse SCO ...... 41
1.5.5 SCO and its role in RF production ...... 43
1.5.6 Human SCO ...... 44
1.5.7 Hydrocephalus – Congenital hydrocephalus ...... 45
Relationship between SCO and CH ...... 46
1.5.8 CH in human – L1CAM ...... 47 1.5.9 Genetic and molecular etiology of CH in mouse models ...... 48
Cell adhesion defects ...... 49
Cilia structure/function defects ...... 50
Signal transduction defects ...... 52
Developmental defects ...... 53
Proposed model for the developmental regulation of prosomere 1 RF and SCO ...... 55
Summary ...... 55
1.6 Thesis aims and approaches ...... 56
1.6.1 Aim: To understand the molecular mechanism of Sox3 overdosage that underlies the physiological pathogenesis of CH...... 56
1.6.2 Approach 1: Identification and characterisation of the morphological and cellular defect within the CNS that is responsible for CH related phenotype ...... 57
1.6.3 Approach 2: Understanding the molecular role of Sox3 in SCO development through identification of its downstream effectors ...... 57
CHAPTER 2: MATERIALS AND METHODS ...... 58
2.1 General Solution(s) ...... 58
2.2 Animal husbandry ...... 58
2.3 Mouse embryos generation and tissue collection ...... 58
2.4 Genotyping of animals ...... 59
2.4.1 Sox3 transgene genotyping ...... 59
2.4.2 Sr transgene specific genotyping ...... 59
2.4.3 Nr trangene specific genotyping ...... 60
2.5 Genome walking ...... 60
2.6 H & E stains ...... 61
2.7 Nissl stains ...... 61
2.8 qPCR for Sox3 transgene copy number analysis ...... 61 2.6 Immunofluorescence ...... 62
2.7 BrdU analysis ...... 62
2.8 TUNEL assay ...... 63
2.9 In Situ Hybridisation ...... 63
2.9.1 Template plasmid production ...... 63
2.9.1 Riboprobe synthesis ...... 64
2.9.2 Section in situ hybridisation ...... 64
2.9.3 Whole mount in situ hybridisation ...... 64
2.10 SCO microdissection and RNA extraction ...... 64
2.11 Microarray data processing ...... 65
2.12 qRT-PCR analysis for microarray validation ...... 65
2.13 Image analysis ...... 66
2.14 Primers table ...... 67
2.15 Plasmids table ...... 73
CHAPTER 3: CHARACTERISATION OF CH IN SOX3 TRANSGENIC MOUSE MODEL...... 74
3.1 Introduction ...... 74
3.2 Results ...... 75
3.2.1 Sox3 transgene integration site in Nr mouse model ...... 75
3.2.2 The Sox3 transgene is not expressed in the developing dorsal telencephalon of Nr/+ embryos ...... 76
3.2.3 CH is dosage dependent on Sox3 ...... 77
3.2.4 CNS morphology of Nr/+ and Sr/+ mice suggested non-communicating hydrocephalus ...... 78
3.2.5 Endogenous and transgenic Sox3 expression in the SCO ...... 78 3.2.6 Lower level of SOX3 expression in prosomere 1 RP ...... 81
3.2.7 Dysmorphic SCO in post-weaning Nr/+ and Sr/+ mice ...... 82
3.2.8 Endogenous and transgenic Sox3 expressions in other area of dorsal diencephalon ...... 84
Endogenous and transgenic Sox3 expression in 12.5 dpc dorsal diencephalon ...... 84
Endogenous and transgenic Sox3 expression at 15.5 dpc dorsal diencephalon ...... 85
3.3 Discussion ...... 86
3.3.1 Sox3 transgene integration site in Nr mouse line ...... 86
3.3.2 Transgene incorporated Sox3 regulatory sequences were insufficient for complete recapitulation of endogenous telencephalic expression...... 87
3.3.3 Transgenic Sox3 expression in the diencephalon ...... 89
3.3.4 Sox3 has a role in SCO development ...... 90
3.3.5 Sox3 overexpression leads to defective SCO development and CH ...... 91
3.3.6 Proper Sox3 dosage regulation is important during SCO development ...... 91
3.3.7 The role of endogenous SOX3 in mature SCO ...... 92
CHAPTER4: ANALYSIS OF SCO DEVELOPMENT IN SOX3 TRANSGENIC MICE ...... 94
4.1 Introduction ...... 94
4.2 Results ...... 94
4.2.1 LV expansion and SCO aplasia in 18.5 dpc Sr/+; Nr/+ embryos ...... 95
4.2.2 Normal morphology of the SCO at 12.5 dpc Sr/+; Nr/+ embryos ...... 96
4.2.3 SCO dyspalsia from 13.5 dpc in Sr/+; Nr/+ embryos ...... 96
4.2.4 Morphologically and functioally defective SCO from 15.5 dpc Sr/+; Nr embryos ...... 97
4.2.5 Developmental defects in the diencephalon of Sr/+; Nr/+ embryos are restricted to the RP ...... 99
4.2.6 Absence of apoptosis in SCO from WT and Sr/+; Nr/+ embryos ...... 99 4.2.7 Overproliferation of the SCO primordium of 12.5 dpc Sr/+; Nr/+ embryos .... 101
4.2.8 Loss of SCO progenitors from 12.5 dpc in Sr/+; Nr/+ embryos ...... 103
4.2.9 Downregulation of diencephalic midline markers in 12.5 dpc Sr/+; Nr/+ embryos ...... 104
4.2.10 Characterisation of ChP in Sox3 transgenic mice ...... 105
4.3 Discussion ...... 107
4.3.1 Defective SCO development is the primary cause of CH ...... 107
4.3.3 SCO dysplasia was a consequence of loss of SCO cellular identity ...... 108
4.3.2 Morphological defect was not restricted to SCO but was RP sepcific ...... 109
Loss of PC ...... 110
HbC failed to cross midline ...... 111
Absence of PG development ...... 111
General RP defect ...... 112
4.3.4 Sox3 regulates Msx1 and Wnt1 for RP development ...... 113
CHAPTER 5: MOLECULAR FUNCTION OF SOX3 IN SCO DEVELOPMENT ...... 117
5.1 Introduction ...... 117
5.2 The design of microarray experiment ...... 117
5.3 Results ...... 119
5.3.1 Microarray data processing and quality control ...... 119
5.3.2 Microarray statistical analysis ...... 120
5.3.3 Candidate genes selection ...... 121
5.3.4 Validation of transcriptional profiles in microarray through qRT-PCR...... 124
Category 1: genes with high fold-change ...... 125
Category 2: Genes implicated in Wnt and Bmp6 signalling pathways ...... 125
Category 3: Genes with SCO specific expression ...... 126 5.3.5 The expression domains of pretectal markers are expanded ...... 127
5.3.6 Low penetrance of exencephaly observed in Sr/+ and Sr/+; Nr/+ embryos ..... 128
5.4 Discussion ...... 128
5.4.1 Some candidate genes displayed variable fold-changes between the microarray and qRT-PCR experiments ...... 128
5.4.1 Bmp and Wnt RP expression was diminished transgenic embryos...... 129
Bmp ligands ...... 131
Wnt ligands ...... 131
5.4.2 Fmo1, Adamts3, Hmgcs2 – a possible function in the synthesis of SCO secretions ...... 133
5.4.3 Fibcd1– a possible endocytic activity in SCO development ...... 134
5.4.4 Disrupted dorsal diencephalic patterning due to defective RP development ..... 135
5.4.5 Failure of cranial NT fusion might be associated with disrupted dorsal NT tube patterning ...... 138
CHAPTER 6: DISCUSSION ...... 141
6.1 The knowledge of Sox3 function in mammlian CNS development was limited ...... 141
6.2 Diencephalic RP development was sensitive to overdosage but not underdosage of Sox3 ...... 142
6.3 Sox3 dosage regulation is important for proper RP specification ...... 142
6.4 Sox3 activity was repressed to upregulate Bmp and Wnt expression in RP development ...... 144
6.5 Hes1 – a direct target of Sox3 in RP development? ...... 145
6.6 The role of Sox3 in RP neural differentiation ...... 147
6.7 Sox3 in earlier dorsal NT development ...... 148
6.7.1 Dorsal NT induction ...... 148 6.7.2 NT closure ...... 149
6.7.3 Cranial NC development ...... 149
6.8 Similarities and differences in overdosage of Sox3 between mice and human ...... 150
6.9 Normal SCO development and CSF homeostasis in human ...... 151
6.10 Future directions ...... 152
6.11 Concluding remarks ...... 153
APPENDIX I: THE SEQUENCES OF BTBD11 LAST EXON ...... 155
APPENDIX II: THE SEQUENCES OF THE SOX3 TRANSGENE INTEGRATION SITE IN NR MOUSE MODEL ...... 156
APPENDIX III: SOX3 ANTIBODY SPECIFICITY ...... 167
APPENDIX IV: AN OUTLINE OF THE MICROARRAY EXPERIEMNT ...... 168
APPENDIX V: THE PAIRWISE MA PLOTS GENERATED FROM PARTEK® ANALYSIS ...... 169
APPENDIX VI: MICROARRAY CEL FILES AND GENE LISTS ...... 170
APPENDIX VII: MICROARRAY – A TABLE OF GENES WITH A FOLD-CHANGE IN MODULUS OF 2 OR ABOVE ...... 171
APPENDIX VIII: PLASMID MAPS ...... 176
REFERENCES ...... I
Summary
The central nervous system (CNS) is one of the most complicated organs in the body. Its development requires spatially and temporally dynamic but yet coordinated proliferation and differentiation of neural progenitors in order to allow the formation of the neural tube (NT) from a sheet-like neural plate, as well as the specification of different CNS domains from a homogeneous group of neural precursors. A number of genes have been identified to be important for the robust genetic and molecular regulation of the CNS development. One of these genes is SOX3, which encodes a SOX family transcription factor within a single exon. In humans, SOX3 duplications and mutations have been implicated in X-linked hypopituitarism (XH) with variable mental retardation, suggesting that CNS development is sensitive to SOX3 dosage. To recapitulate the condition in XH patients with SOX3 duplications, a transgenic Sox3 overexpression mouse model was generated in the Thomas laboratory. Two Sox3 transgenic lines were established. Intriguingly, they presented not with XH associated phenotypes, but with hallmarks of hydrocephalus.
This thesis focuses on the characterisation of the hydrocephalus phenotype and the elucidation of the molecular pathogenesis that underlines the defect. Comprehensive morphological and expression analyses of single hemizygous and double hemizygous (mice with two transgene alleles, one from each Sox3 transgenic line) Sox3 transgenic mice demonstrated that the hydrocephalus has a congenital origin (congenital hydrocephalus, CH) and is associated with defective development of the subcommissural organ (SCO), a diencephalic roof plate (RP) derivative that has previously been implicated in CH. Moreover, both the CH and the SCO dysmorphology of Sox3 transgenic mice are dependent on Sox3 dosage. In addition to the SCO, defective development was also identified in other diencephalic RP derivatives in Sox3 double hemizygous embryos. In fact, SCO maldevelopment appears to be the consequence of a general RP defect, in which the identity of the RP cells (or SCO precursors) is lost and is coupled with an elevation of cellular proliferation. Further molecular analysis of the SCO development supported that Sox3 overexpression leads to the loss of RP cellular identity as the expression of genes implicated in both Bmp and Wnt signalling pathways, which are important for normal RP development, were significantly downregulated in Sox3 transgenic embryos. In addition to its role as the precursor of RP derivatives, the RP is also essential to maintain dorsal identity for proper DV
patterning within the NT. Moreover, dorsal NT patterning was disrupted within the developing Sox3 transgenic embryos. Together these suggest that the Sox3 overdosage leads to the loss of RP identity and the failure of RP dependent processes.
In summary, this thesis indicates that, at least in the mouse, Sox3 dosage regulation is important for proper RP development. In fact, a low level of Sox3 is permissive for RP specification. It is not known what cellular identity is taken up by the RP cells in the Sox3 double hemizygous embryos. However, given endogenous Sox3 is highly expressed in the lateral tissue that flanks the RP, it is logical to propose that the RP cells of Sox3 double hemizygous embryos may take on a cellular identity similar to their laterally neighbouring tissue. Interestingly, a CNS defect that may be due to RP dysfunction has been identified in some XH patients with SOX3 duplications, suggesting that SOX3 may have similar role in humans. CH is a heterogeneous disorder in which identification of its genetic basis in humans is difficult. This thesis proposes SOX3 as a candidate gene for CH and other RP associated defects in humans, which may assist the development of relevant prognostic techniques for clinical application.
Statement
This work contains no material which has been accepted for the award of any other degree or diploma in any university or other tertiary institution and, to the best of my knowledge and belief, contains no material previously published or written by another person, except where due reference has been made in the text.
I give consent to this copy of my thesis, when deposited in the University Library, being made available for loan and photocopying, subject to the provisions of the Copyright Act 1968.
I also give permission for the digital version of my thesis to be made available on the web, via the University’s digital research repository, the Library catalogue and also through web search engines, unless permission has been granted by the University to restrict access for a period of time.
Signed: ………………………….………(Kristie Pan Yu Lee)
Date: ………………………………….…
Acknowledgments
My PhD project would not be possible without my supervisor, Paul Thomas, for offering me a position in his laboratory, as well as his support for the last four years. Here, I would like to thank him for the numerous hours that he spent in reading my manuscript and thesis, in our marathon meeting where he gave me valuable advice, and in undertaking double-blinded tests of my many in situ hybridisation experiments. Also, I would like to thank him for his moral support and faith during times in my PhD when experiments did not work as well as we hoped for. In addition, I would like to thank my co-supervisor Kirk Jensen, whose scientific advice has been greatly beneficial to my PhD project.
I am deeply grateful to those who provided professional help and opinion that made me to become a better scientist. To James Hughes, who had frequent lengthy discussion with me about the SCO development; and to Mike Morris, who gave me great assistance and valuable opinion in those last experiments of my PhD. To Sandie, who performed the BrdU injection and made my mouse work so much easier for looking after the colonies; to Dale, who provided his unlimited pool of knowledge in my microarray statistical analysis, to Pike- See, Rob Moyer from Obstetrics and Gynaecology and Gail Hermanis from Anatomical Sciences who offered great help in the my histology work; and to Rosalie Kenyon from the Adelaide Microarray Centre for performing my Microarray experiment. Moreover, I would like to express my appreciation to those whose data I presented in this thesis, Chloe Shard and again Mike Morris and James Hughes.
Doing a PhD means that you probably spend most of your conscious time in the lab, and you probably see your colleagues at work more often than your rabbit at home. Thus, I would like to thank everyone who shared their time (and laughter and tears if there were any) with me in the past four years, especially James, Sandie, Bryan, Dale, Nick, Anne, Sophie, Pete and Adrienne (and Shwetha and Edwina who are greatly missed!), all current and past member of the Thomas lab, as well as my fellow PhD candidates. I would like to extend my appreciation to everyone on the third floor whose help I received throughout the years.
Last but not least, I would like to thank my friends and family. To Angela, Rosalie, Wee-ching, Tracy and Zoe, thank you for all the support that you have given me, prior and during my PhD (and hopefully years from now on).Your friendship is and will be mostly
appreciated. To my mum and dad, thank you very much for your unconditional support, both emotionally and financially, since I was born and of course during my PhD. I would never make to where I am without you. Thanks for being there whenever I needed and for making sure that I was properly fed all the time. To my brother Eric, thanks for your patience to listen to my work and to put up with an occasional grumpy sister. Finally, to Nick, thank you very much for your patience, encouragements and understanding in the past years. Thanks for being there and giving me a reason to smile for even at the end of a day when none of my experiment worked.
Abbreviations Abbreviations
Standard terms
% percentage
⁰C degree Celsius
cDNA complementary deoxyribonucleic acid
DNA deoxyribonucleic acid
g gravitational force
kb kilobase
kDa kilodalton
kg kilogram
RNA ribonucleic acid
M molar
mg milligram
ml millilitre
μl microlitre
mm millimetre
μm micrometer (micron)
mM millimolar
μM micromolar
ms millisecond
ng nanogram
N normality
w/v weight per volume
v/v volume per volume
Non-standard terms
a.a. amino acid
AH acquired hydrocephalus
BAC bacterial artificial chromosome
bp base pair
CH congenital hydrocephalus
dpc days post coitum
ES embryonic stem
HCNE(s) highly conserved non-coding element(s)
IRES internal ribosome entry site
NC neural crest
ORF open reading frame
P postnatal
RF Reissner’s fibre
RIN RNA integrity number
SCBE SOX core binding sequence
UTR untranslated region
WT wildtype
XH X-linked hypopituitarism
Mouse lines
Nr Sox3 non-sex reversal line
Sr Sox3 sex reversal line
Materials and methods
BrdU bromodeoxyuridine
DIG dioxygenin
EtOH ethanol
gDNA genomic DNA
H & E haematoxylin and eosin staining
IgG immunoglobulin G
MeOH methanol
MQ milliQ
NaAc sodium acetate
OCT optimal cutting temperature
O/N overnight
PBS phosphate buffered saline
PCR polymerase chain reaction
PFA paraformaldehyde
qPCR quantitative polymerase chain reaction
qRT-PCR quantitative real-time polymerase chain reaction
RT room temperature
TUNEL terminal deoxynucleotidyl transferase dUTP nick end labelling
U unit
Statistical and bioinformatics terms
ANOVA analysis of variance
BLAST Basic Local Alignment Search Tool
FDR false discovery rate
Limma linear models for microarray analysis
MGI Mouse Genome Database
NCBIm37 mus musculus genome assembly NCBIm37
Anatomical terms
3V third ventricle
4V fourth ventricle
AP anterior-posterior
Aq Sylvian aqueduct
AVE anterior visceral endoderm
BBB blood-brain barrier
BCSFB blood-CSF barrier
CCx cerebral cortex
Cer cerebellum
ChP choroid plexus
CNS central nervous system
CSF cerebrospinal fluid
CSFBB CSF-brain barrier
CVO circumventricular organs
DLHP dorsolateral hinge point
DTh dorsal thalamus
DV dorsoventral
EpTh epithalamus
FP floor plate
GO gastrula organiser
HbC habenular commissure
Hpc hippocampus
LGE lateral ganglionic eminence
LV lateral ventricles
MGE medial ganglionic eminence
MHP medial hinge point
MR mesocoelic recess
MZ mantel zone
NT neural tube
PC posterior commissure
PG pineal gland
PR pineal recess
PS primitive streak
Pt pretectum
RP roof plate
SCO subcommissural organ
TrG trigeminal ganglia
VTh ventral thalamus
VZ ventricular zone
ZLI zona limitans intrathalamica
Cell lines
293T human embryonic kidney 293 cells
HeLa human cervical cancer cells
NTERA2 human embryo carcinoma cells
Protein families and protein domains
APC adenomatous polyposis coli
Bmp bone morphogenic protein
Fgf fibroblast growth factor
Gli GLI-kruppel family member
GSK3 glycogen synthase kinase
HMG high motility group
JAK Janus family of tyrosine kinases
MAP mitogen associated protein mTOR mammalian target of rapamycin
NICD notch-intracellular domain
Shh sonic hedgehog
SNARE N-ethylmaleimide sensitive fusion attachment protein receptors
SOX sry-related HMG box
STAT signal transducers and activators of transcription
TA transactivation
Tgfβ transforming growth factor beta
Wnt wingless-related MMTV integration site
Gene Nomenclature
Human
Gene: italic, all capital, e.g. SOX3
Protein: non-italic, all capital, prefixed with ‘h’ indicative of human origin, e.g. hSOX3
Mouse
Gene: italic, first letter capital, e.g. Sox3
Protein: non-italic, all capital, e.g. SOX3
Other model organisms
Gene: italic, first letter capital, prefixed with a relevant letter indicative of animal origin, i.e. ‘c’ for chicken; ‘z’ for zebrafish; ‘x’ for Xenopus, e.g. cSox3, zSox3, xSox3
Protein: non-italic, first letter capital, prefixed with a relevant letter indicative of animal origin, i.e. ‘c’ for chicken; ‘z’ for zebrafish; ‘x’ for Xenopus, e.g. cSox3, zSox3, xSox3
Note: The nomenclature of the highest organism is used if describing a gene or protein from more than one model organisms. 1
Chapter 1: Introduction
1.1 The central nervous system
The central nervous system (CNS) is composed of numerous different types of neurons and glia cells and is regarded as the most complex organ within the mammalian biological system. Within the CNS, neurons with similar function(s) often group together to form a domain that specialises in particular task(s). Thus, the CNS is composed of numerous domains, which each of them is specific to a subset of function(s). In return, these domains co-ordinate and interact with each other, mostly through synapses, to allow the input, relay, processing and output of information required for regulation of physiological homeostasis and high cognition. In addition to this complex collection of domains, CNS is also very plastic in terms of neuronal network, since synapses are established, abolished or reinforced constantly in response to memory and everyday experiences. As a result of its high level of integral and dynamic function, an orchestrated and co-ordinated development hence, is important for the proper function of the CNS.
CNS development initiates from neural induction, a process to specify the neuroectoderm from the general mass of ectodermal cells. A neural plate is then formed from the neuroectoderm through a series of proliferation and cellular re-organisation. Subsequently, neurulation starts in order to form the tubular neural tube (NT) structure by “rolling up” the sheet-like neural plate. This is immediately followed by neurogenesis and gliogenesis to generate the bulk of neurons and glia cells that populate the CNS. Although, the majority of neurons and glia cells were born after neurulation, neural proliferation, specification and differentiation take place from the moment when neuroectoderm is induced. It is the progressive neural differentiation accompanying each phase of the neural development that allows the step-wise specification of the CNS, from the earliest anterior-posterior (AP) and dorsoventral (DV) patterning to CNS domain partitioning, such as the formation of the multilayered cerebral cortex (CCx).
In order to achieve this complex process, CNS development employs the function of a number of genes, with their expressions robustly regulated to orchestrate the intricate and 2 dynamic molecular and genetic interactions. One of these genes is the Sry-box containing gene 3 (SOX3), which encodes a transcription factor highly expressed within the developing CNS (Collignon et al., 1996; Solomon et al., 2004; Wood and Episkopou, 1999). 3
1.2 SOX3 – a member of the SOX transcription factor family
SOX3 is a member of the Sry-related HMG box (SOX) transcription factor family. All members of the SOX family encode transcription factors that demonstrate high sequence homology across their DNA binding domain, high motility group (HMG) box, which was firstly identified in the protein of the male sex determining gene, sex determining region of Chr Y (SRY). In fact, all SOX members share at least 50% HMG sequence homology to its founder SRY. SOX family members are further divided into subfamily based first on the homology of their HMG box sequences followed by the homology of their flanking sequences. Currently there are 30 SOX genes across all phyla, with 20 of them found in mammals. SOX members have been demonstrated to be involved in a wide range of biological functions including sex determination, neural induction, neural crest (NC) formation, neural differentiation, cardiogenesis, haematopoiesis, skeletogenesis, osteogenesis and sensory placode formation (Figure 1.1) (De Martino et al., 1999; Dy et al., 2008; Hoser et al., 2008; Kiefer, 2007; Lefebvre et al., 2007; Schepers et al., 2002).
1.2.1 Sox3 belongs to SOXB1 subgroup
SOX3 belongs to the SOXB subfamily, which is further subdivided into SOXB1 and SOXB2. In addition to over 90% amino acid sequence identity shared by SOXB members across the HMG box, their flanking sequences are well conserved. SOXB1 members include transcriptional activators sry-box containing gene 1 (SOX1), sry-box containing gene 2 (SOX2) and SOX3, while SOXB2 members include transcriptional repressors sry-box containing gene 14 (SOX14) and sry-box containing gene box 21 (SOX21). SOXB is the only SOX subfamily that has both transcriptional activators and repressors (Dy et al., 2008; Hoser et al., 2008; Kiefer, 2007; Lefebvre et al., 2007). Moreover, SOXB members have been strongly implicated in neural development (discussed in 1.4.1 and 1.4.4), with B1 and B2 members demonstrated to have opposing effect in this process (Sandberg et al., 2005).
SOX3 is a single exon gene residing on the X chromosome (Bylund et al., 2003; Gubbay et al., 1990; Holmberg et al., 2008; Schlosser et al., 2009; Stevanovic et al., 1993; 4
Thomsen et al., 2009). Human SOX3 encodes a 446 amino acid (a.a) transcription factor characterised by the presence of a HMG box spanning the region of 137-215 a.a. at its N- termini. HMG box has been demonstrated to be important for DNA binding, protein-protein interaction and nuclear transport. Human SOX3 has a transactivation (TA) domain and four polyalanine tracts across its C-terminal. In addition, it has a nuclear localisation sequence flanked by the HMG box and the TA domain. In comparison to human SOX3, mouse SOX3 lacks the first 75 a.a. at the N-termini. However, mouse SOX3 shares 93% a.a. identity with the human SOX3 for the rest of the protein, including the HMG box, the TA domain and the polyalanine residues (Figure 1.2) (Kamachi et al., 2000; Kiefer, 2007; Miyagi et al., 2009; Schepers et al., 2002).
1.2.2 SOX transcription factor interact with partnering proteins to achieve specificity in transcription regulation
Several studies have indicated that the HMG box in SOX family transcription factor specifically binds to the SOX core binding sequence (SCBE), 5’-A/T A/T CAA A/T G-3’ (Denny et al., 1992; Kanai et al., 1996; Mertin et al., 1999). The HMG box has an L-shaped surface, which upon binding to minor groove of DNA helix, leads to bending of the DNA molecule and confers conformational change for transcriptional regulation (Connor et al., 1994; Lefebvre et al., 2007; Scaffidi and Bianchi, 2001; Wegner, 2009). However, it was later shown that the sole binding of SOX to SCBE is not enough to exert the regulation on target gene transcription, since the HMG box has low affinity to the SCBE and fails to induce DNA bending upon binding (Lefebvre et al., 2007; Wegner, 2009). Moreover, one would imagine this short and loosely-defined SCBE, which can be bound by all SOX HMG box, is insufficient to confer the SOX transcription factor specific target genes regulation across a wide range of biological processes. Indeed, the binding specificity of the SOX HMG box is also influenced by differences in SCBE flanking sequences (Bergstrom et al., 2000; Denny et al., 1992; Kanai et al., 1996; Mertin et al., 1999).
In addition to recognition sequences, it appears that SOX proteins achieve their specific target gene regulations through the availability of binding partners, which is cell- context dependent. It has been proposed that SOX proteins physically interact with other transcription factors to co-operatively and/or synergistically regulate target genes. For SOX Group Subgroup Biological function gene
SOXA - Sry • Testis determination, regulate brain fucntion
SOXB SOXB1 SOX1 • Lens development, neural differentiation • Embryoinc stem cell maintenance, neural induction, lens development, SOX2 sensory neuron differentiation • Neural differentiation, lens induction, gonadogenesis, pituitary SOX3 development SOXB2 SOX14 • Interneuron specification, neural differentiation SOX21 • CNS patterning, neural differentiation
SOXC - SOX4 • Heart, lymphocyte, thymocyte development, neurogenesis SOX11 • Organogensis, neurogenesis SOX12 • Cardiogenesis; neurogenesis
SOXD - SOX5 • Skeletogenesis, cardiogenesis, gliogenesis, neural crest development
L-SOX5 • Chordrogenesis
SOX6 • Cardiac function, skeletogenesis, gliogenesis, erythropoiesis
SOX13 • Lymphopoiesis
• Gliogenesis, osteogenesis, neural crest induction and maintenance, SOXE - SOX8 testis development
• Sex determination, chondrogenesis, neural crest survival, gliogenesis, SOX9 inner ear formation, cardiogenesis, progenitor maintenance of hair follicle and pancreas, epithelium specification in gut
SOX10 • Neural crest development, inner ear formation
SOXF - SOX7 • Cardiogenesis SOX17 • Endoderm formation, angiogenesis SOX18 • Cardiogenesis, angiogenesis, hair follicle development
SOXG - SOX15 • Skeletal muscle regeneration
SOXH - SOX30 • Neural development?
Figure 1.1: SOX members participate in a wide range of mammalian biological funcitons. Note that SOXB is the only group that is subdivided into SOXB1 and SOXB2. L-SOX5 is an isoform of SOX5. Figure adapted from De Martino et al. (1999); Dy et al. (2008); Lefebvre et al. (2007); and Schepers et al. (2002). 137 215 HMG Human SOX3 (446 a.a.)
67 145 HMG Mouse SOX3 (375 a.a.)
Figure 1.2: Human and mouse SOX3 share 93% a.a. identity. A schematic diagram of human and mouse SOX3. The HMG box spans 137-215 a.a. in the human SOX3, while the HMG box spans 67-145 a.a. in mouse SOX3. Green box: HMG box. Orange box: nuclear localisation sequence. Black vertical bars: polyalanine tracts. Blue horizontal bar spans the TA domain. Diagram not drawn to scale. Figure adapted from Kamachi et al. (2000); and Lefebvre et al. (2007). 5
example, from a yeast two-hybrid screen, transcription factors encoded by both SOXE members, Sry box-containing gene 8 (SOX8) and Sry box-containing gene 10 (SOX10), were found to be able to physically interact with a number of DNA binding proteins through the C- terminal of the SOX HMG domain (Wissmuller et al., 2006). Furthermore, during testis development, SRY binds to the transcription factor encoded by splicing factor 1 (Sf1), and then transcriptionally activate sry-box containing gene 9 (Sox9), an important downstream effector for testis development. In return, SOX9 physically interacts with SF1 to upregulate and maintain its own transcription activity. Moreover, an in vitro experiment has suggested that the hSOX9/hSF1 interacting complex is capable of initiating the transcription of anti- Mullerian hormone (AMH), a further downstream target gene important for testis development in humans (De Santa Barbara et al., 1998). Together, these indicate that SOX proteins partner with other transcription factor in the regulation of testis development during mammalian sex determination.
Besides sex determination, the interaction of SOX2 with its binding partners for stem cell maintenance is well characterised. In vitro studies in both human and mouse cell lines found that hSOX2 heterodimerises with its co-binding partner, POU domain, class 5, transcription factor 1 (POU5F1, formerly known as Oct3/4) proteins, in order to transcriptionally upregulate fibroblast growth factor 4 (FGF4) and nanog homeobox (Nanog) (Rodda et al., 2005; Yuan et al., 1995). Crystallography studies revealed SOX2 interacts with the POU domain of POU5F1 through its HMG box (Remenyi et al., 2003). Similar physical interaction was also observed in chicken, where cSOX2 and POU domain, class3, transcription factor 2 (cPouf2, formerly known as cBrn2) proteins heterodimerise and then bind to Nes258 enhancer to activate the transcription of nestin (cNes), a gene encoding neurofilament in neural progenitors (Tanaka et al., 2004). Furthermore, cSOX2 interacts with the paired domain of paired box gene 6 (cPax6) proteins through the HMG box. The cSOX2/cPAX6 complex then binds to the DC5 enhancer element of argininosuccinate lyase (cAsl, also known as δ-crystallin) and induces chicken lens placode formation (Kamachi et al., 2001).
Together, the above evidence suggests that despite all SOX proteins are capable of binding to the common SCBE sequences, they trigger specific target genes regulation through a combination of binding partner(s) interaction and selective recognition of SCBE flanking 6 sequences. This specific, cell-context dependent target genes regulation allows function specialisation of different SOX proteins within a wide range of biological systems. 7
1.3 An overview of CNS development
In order to understand the function of SOX3 in CNS development, it is necessary to have a thorough appreciation of the cellular and molecular interactions of the CNS development from neural induction during early embryonic development, to neurulation and subsequently neurogenesis and gliogenesis. Moreover, it is important to note that these events are tightly coupled to cellular proliferation, differentiation and specification. The following paragraphs describe these key events in CNS development. As the main aim of this thesis is to study on the abnormal development of the diencephalon and the dorsal midline in Sox3 transgenic mice, the discussion will have particular focus on these aspects of CNS development. The mechanisms described below are based on mouse model unless otherwise stated.
1.3.1 Axes formation and patterning of the developing CNS domains
Prior to the detailed discussion on CNS development, it is important to define the establishment of CNS axes and the terminology for the developing CNS domains. One should bear in mind that CNS axes formation is not an independent biological process, but is rather dependent on the proper patterning, specification and differentiation of cells within the whole CNS. In fact, the establishment of the AP axis is coupled to neural induction, as it takes place within the anterior embryo. After neurulation, patterning along the AP axis allows the CNS to be segmented arbitrarily into four parts along the NT, they are: the prosencephalon (the most anterior CNS that gives rise to the forebrain), the mesencephalon (immediately posterior to the prosencephalon and gives rise to the midbrain), the rhombencephalon (immediately posterior to the mesencephalon and gives rise to the hindbrain) and the spinal cord (immediately posterior to the hindbrain) (Figure 1.3). Moreover, the prosencephalon is further divided into two domains, the rostrally residing telencephalon and the caudally positioned diencephalon (Figure 1.3) (Beddington and Robertson, 1999; Levine and Brivanlou, 2007; Zaki et al., 2003).
Similarly, the formation of DV is associated with neurulation initiation (discussed below). Although the DV axis of the developing CNS is not apparent until the formation of the NT, but it is defined before neurulation. Prior to neurulation initiation, at the neural plate 8
stage, the notochord, a thin long structure that resides ventrally to the ventral midline of the future NT, emanates signals to induce neurulation (Figure 1.3) (Bassuk and Kibar, 2009; Vieira et al., 2010; Zaki et al., 2003), as well as to establish the ventral identity and hence, the DV axis of the future NT. In addition, the dorsal midline, a structure opposing the ventral midline, is also formed at the site of NT fusion. Both the dorsal and the ventral midline play an essential role in further DV patterning and cellular organisation during later CNS development (Beddington and Robertson, 1999; Chizhikov and Millen, 2005).
1.3.2 Neural induction
Neural induction is a process that defines the neural fate of the anterior ectoderm. Thus, proper neural induction relies on the robust embryonic patterning orchestrated by the complex genetic and molecular interaction of signals emanating from the organiser(s), developmental domain(s) that secrete morphogens to instruct fate and behaviour of the surrounding tissues. Briefly, there are three organisers that are essential for anterior patterning and hence, neural development during early embryogenesis. They are the anterior visceral endoderm (AVE), the gastrula organiser (GO) and its derivative, the node (Beddington and Robertson, 1999; Levine and Brivanlou, 2007; Robb and Tam, 2004). The following section describes the process of early embryonic development with respects to anterior patterning and neural induction by these three anterior organisers.
The development of early embryos
During early embryogenesis, the zygote undergoes rapid cell divisions to ultimately generate a multicellular blastocyst, an early embryonic structure that contains an inner cell mass (gives rise to the epiblast and the extraembryonic endoderm), a trophectoderm (gives rise to the extraembryonic ectoderm) and a blastocyst cavity (blastocoels). The epiblast is a group of pluripotent cells that later generates all cells of the foetus as well as the extraembryonic mesoderm. The formation of blastocyst is followed shortly by gastrulation, a process which some epithelial cells from the epiblast transform into mesenchyme and migrate anteriorly in order to produce the three germ layers, i.e. ectoderm, mesoderm and endoderm (Beddington and Robertson, 1999; Levine and Brivanlou, 2007). After gastrulation, the three germ layers further differentiate into various specific progenitor cells, e.g. induction of A B
Figure 1.3: A schematic diagram demonstrating the AP patterning of the developing mouse CNS. A: Ventral view of the neural plate at 8.5dpc. The arrows indicate the direction of NT closure. B: Lateral view of the mouse NT at 10.5 dpc. Di: diencephalon. PCP: prechordal plate. Tel: telencephalon. Colour codes are as follows: diencephalon (light purple); mesencephalon (green); notochord (dark purple); rhombencephalon (yellow); prechordal plate (red); telencephalon (blue). Di: diencephalon. PCP: prechordal plate. Tel: telencephalon. Top to bottom: dorsal to ventral. Left to right: anterior to posterior (B). Figure adapted from Zaki et al. (2003). 9
neuroectoderm from ectoderm, in which these cells are later specialised during organogenesis, e.g. generation of neurons during CNS development. In addition to the formation of the three germ layers, gastrulation is also important for neural development as it is associated with the establishment of anterior organisers that are essential for anterior patterning and hence, neural induction (Beddington and Robertson, 1999; Levine and Brivanlou, 2007).
Anterior organisers protect the default neural fate of the epiblast
Recent evidence has suggested that the epiblast has a default neural fate. A spontaneous differentiation assay demonstrated that embryonic stem (ES) cells that were isolated and cultured as single cell with minimal growth factors and the absence of extrinsic neuralising factors adopt a neural fate (Levine and Brivanlou, 2007; Smukler et al., 2006). To further this proposition, Sox2 and orthodenticle homolog 2 (Drosophila) (Otx2), two neural fate markers, are initially expressed in all cells within the epiblast but are later restricted to the neuroectoderm (Ang et al., 1996; Avilion et al., 2003; Levine and Brivanlou, 2007; Wood and Episkopou, 1999). Hence, it is believed that posteriorising signals such as wingless- related MMTV integration site (Wnt), bone morphogenic protein (Bmp) and Nodal are required to inhibit the default neural fate and allow proper development of the AP axis and other non-neural tissue (Camus et al., 2006; Levine and Brivanlou, 2007). By the same token, neural induction is then a process of “preserving” the neural fate inherited from the epiblast rather than “instructing” the uptake of neural fate. Thus, anterior organisers are important to secrete neuralising factors that antagonise the posteriorising signals and protect the default neural fate within the anterior embryos for proper neural induction (Levine and Brivanlou, 2007).
AVE and its role in early neural development
The AVE is a small domain of extracellular endoderm that is adjacent to the future neural plate and defines the AP axis (Figure 1.4). AVE can be distinguished from the visceral endoderm by the expression of visceral endoderm-1 (VE-1), cerberus-1 homolog (Xenopus) (Cer1), LIM homeobox protein 1 (Lhx1),Goosecoid homeobox (Gsc), Hesx1 and haematopoietically expressed homeobox (Hhex) at 6.5 days post coitum (dpc) prior to gastrulation (Belo et al., 1997; Rosenquist and Martin, 1995; Thomas and Beddington, 1996; Thomas et al., 1998). It is still unknown what cue(s) determine AVE formation although 10
early Nodal expression seems to be required, since there lacks AVE specification in Nodal null mouse embryos (Camus et al., 2006). Removal of the AVE or manipulation of its gene expressions resulted in compromisation of forebrain development (Ang et al., 1996; Dattani et al., 1998; Matsuo et al., 1995; Shawlot and Behringer, 1995; Thomas and Beddington, 1996). However, transplantation of AVE from an early-streak embryo to a late-streak embryo could not establish an ectopic body axis nor induce the expression of early neural markers within the primitive ectoderm, despite the expression of posterior markers were repressed (Kimura et al., 2000). In addition, early neural marker expression is expanded posteriorly in Nodal null mice regardless of the lack of AVE specification (Camus et al., 2006). Taken together, the above evidence suggests that the AVE is not responsible for neural induction although it is required to maintain the neural fate of the anterior CNS (Belo et al., 1997; Kimura et al., 2000; Levine and Brivanlou, 2007; Robb and Tam, 2004).
GO and its derivative (the node) are critical for neural development
The primitive streak (PS) is the site of migrating mesenchyme during gastrulation. The GO is a group of cells at the anterior proximity of the early/mid-streak within an early embryo (Figure 1.4) and is the predecessor of the node (Robb and Tam, 2004). As a result, the position of GO is defined by the PS, which forms on the opposite side of the AVE of an early-streak embryo. As gastrulation progresses, the GO migrates anteriorly with the distal end of the PS and differentiates into the node at late-streak stage (Figure 1.4) (Beddington and Robertson, 1999; Levine and Brivanlou, 2007; Robb and Tam, 2004). Interestingly, unlike the AVE, an early study has found that the GO is sufficient to induce an ectopic AP axis when transplanted onto mid-streak embryo, despite the lack of anterior feature of this secondary AP axis (Beddington, 1994). Moreover, the GO is critical for neural induction as demonstrated by the lack of neural tissue in mouse mutants that fail to form the GO (Huelsken et al., 2000; Levine and Brivanlou, 2007; Liu et al., 1999). In contrast to the GO, neither neural induction nor axis formation is affected in mutants lacking a node (Davidson et al., 1999; Klingensmith et al., 1999), although a node explants from a late-streak embryo has successfully induced the expression of neural marker engrailed-2 (En2). Collectively, these indicate that the GO but not its descendent, the node, is indispensable for neural induction, despite the node has a function in promoting anterior neural fate after induction of the neural fate by the GO. dpc 6.0 6.5 7.0 7.5 Pre-streak Early streak Mid-streak Late streak
Figure 1.4: Gastrulation and neural induction within an early mouse embryo. The AVE (yellow) is firstly defined at 6.0 dpc in a pre-streak embryo. Subsequently, the formation of the PS (orange) is initiated and defines the posterior side of the embryo at (6.0-6.5 dpc). The GO (red) then is induced and co-migrates with the distal end of the PS towards the anterior side of the embryo (6.5-7.0 dpc). At the mid-streak stage (7.0 dpc), neural ectodermal fate (blue) is established at the anterior end of the embryo, while the AVE also migrates further proximally. Finally at the late streak stage (7.5 dpc), the GO-derived node is formed at the most distal end of the PS. Mesoderm (turquoise) that is derived from the node now resides anteriorly distal to the streak. Top to bottom: proxi- mal to distal. Left to right: anterior to posterior. Figure from Levine and Brivanlou (2007). 11
Summary
In summary, the anterior organisers, the AVE, the GO and the node, are important for neural fate promotion and maintenance during early neural development. However, only the GO is critical and sufficient for neural induction.
1.3.2 Neurulation and dorsal midline maintenance
Subsequent to neural induction but prior to neurulation, the neuroectoderm undergoes proliferation and cellular reorganisation to form an expanded and thickened ectodermal layer, termed the neural plate. Neurulation is an important event during embryogenesis to allow the ultimate formation of a tubular NT by “rolling up” the sheet-like neural plate. In mouse embryos, neurulation initiates around 7.75 dpc by sonic hedgehog (Shh) signalling emanating from the notochord at the posterior CNS and the prechordal plate (a specialised mesoendoderm that derives from anterior notochord) at the anterior CNS (Bassuk and Kibar, 2009; Vieira et al., 2010). There are three critical processes encompassed by neurulation that are required for further CNS development, which are a) NT closure for the formation of an enclosed CNS; b) NC migration for craniofacial and other non-CNS development, and c) dorsal NT induction and patterning. The following few section discusses these three major processes in detail, with particular focus on NT closure and dorsal NT induction and patterning.
NT closure
In order to prepare for NT closure after neurulation induction, the neural plate undergoes further cellular proliferation to ultimately transform the neural plate into a long thickened sheet-like structure that is narrow mediolaterally but has a long AP axis, such that a tubular NT will form upon fusion of the lateral edges of the neural plate. NT closure can be divided into three main stages – furrowing, folding and NT fusion. As its name suggests, furrowing involves the formation of a furrow and hence, the bending of the neural plate along the AP axis at different positions. Indeed, study of NT closure within the spinal cord found that there are three hinges that are mechanically important for furrowing, i.e. the medial hinge point (MHP) residing at the ventral midline of the future NT, and the two dorsolateral hinge 12
(DLHP) points near the dorsal aspect close to the lateral limits of the neural plate (Figure 1.5). At these hinge points, the columnar neuroepithelial cells are transformed into wedged-shaped neural folds by cellular convergence and extensions through regulation by non-canonical Wnt/planar cell polarity signalling pathway. (Bassuk and Kibar, 2009; Colas and Schoenwolf, 2001; Copp and Greene, 2010; Copp et al., 2003; Vieira et al., 2010). Subsequent to furrowing, folding takes place to further bend the MHP and DLHP for NT closure. Folding at the MHP elevates the lateral sides of the neural plate in order to establish the DV spanning of the NT lumen. In fact, the NT appears as a “V”-shape when viewed upon the AP axis after MHP elevation (Figure 1.5). Concurrently, folding at the DLHPs brings the lateral edges of the neural plate in juxtaposition at the dorsal midline (Figure 1.5) in preparation for NT fusion. As a result, the cooperative folding at MHP and DLHP together allows the neural plate to “roll up” into a hollow tubular structure for NT fusion.
Live imaging techniques have allowed the identification of three NT fusion initiation sites along the entire dorsal midline of the future NT. NT fusion for the caudal CNS initiates at the posterior hindbrain (closure 1), from where it progresses anteriorly into the mid- hindbrain boundary and posteriorly towards the caudal end of the spinal cord (Figure 1.6). In contrast to the caudal CNS, the rostral CNS has two separate sites for NT fusion initiation, one at the fore-midbrain boundary (closure 2) and one at the anterior forebrain (closure 3) (Figure 1.6). NT fusion initiates from closure 2 progresses bi-directionally, with the posterior progression meeting up with the NT fusion initiated from closure 1 at the neuropore, a site where the NT fusion event completes, at the mid-hindbrain boundary (hindbrain neuropore), and the anterior progression joining with the NT fusion initiated from closure 3 at the neuropore within the forebrain (anterior neuropore) (Figure 1.6) (Copp and Greene, 2010).
The molecular event(s) that regulate the NT fusion are not well understood. However, recent discoveries found that the cellular event of NT fusion initiation within the midbrain and hindbrain involves the extension of filopodial-like protrusions from the non- neuroepithelial cells that surrounds the dorsal tip of the neural folds. These filopodial-like protrusions extends towards their opposing neural fold, establish a “bridge-like” connection and bring the lateral edges to proximity to allow fusion of the neuroepithelium (Pyrgaki et al., 2010). Simultaneously, their neighbouring epidermal ectoderm fuses together and pinches off from neuroepithelium to form a single layer, leaving the neural plate to form a continuous tubular structure (Figure 1.7). Figure 1.5: Coronal view of a closing NT. Neurulation initiates at ~8.5 dpc in mouse. Neural plate furrowing and folding take place along the AP axis for NT clsoure and fusion. MHP (red) is immediately dorsal to the notochord (n) while DLHP (blue) are at the lateral edges of the neural folds (boxed). Furrowing at both MHP and DLHP allows the formation of bending points along the closing NT. Folding leads to further bending at the MHP for the elevation of the lateral neural plate (red arrows), forming a “V”-like shape. This is cooperated by the folding at the DLHP, which allows further bending of the neural folds (boxed) towards the midline (black arrows) in preparation for NT fusion. ee: epidermal ectoderm. Top to bottom: dorsal to ventral. Figure from Colas and Schoenwolf (2001)
NOTE: This figure is included in the print copy of the thesis held in the University of Adelaide Library.
Figure 1.6: Sites of NT fusion initiation within an developing mouse embryo. Closure 1 resides at the caudal end of the hindbrain. Closure 2 resides at the bound- ary of midbrain and forebrain. Closure 3 positions at the anterior forebrain. The arrows indicate the direction of NT fusion upon initiation from closure sites. The anterior, posterior and hindbrain neurpore are NT fusion termination sites. Lateral view. Top to bottom: dorsal to ventral. Left to right: anterior to posterior. Figure from Copp and Greene (2009).
NOTE: This figure is included in the print copy of the thesis held in the University of Adelaide Library.
Figure 1.7: NT fusion and midline remodelling during mouse NT closure. A: NT fusion is initiated by filopodial-like protrusions from non- neuroepithelial cells (orange) extending to towards the opposite neural fold. Note that neural folds (red) are still continuous with their neighbrouing epidermal epithelia (green). B: Neuroepithelial adhesion, forming a fused and enclosed NT. Note that the the neuroepitheliua are no longer continuous with the epidermal epithelia. C: Cells at the two lateral-most edges of the neural folds now are adjacent to each other at the dorsal most aspect of the NT. Remodelling, specification and differentiation of these cells establishes the dorsal midline and later RF identity within the dorsal NT. The epidermal epithelia remodels to form a continuous layer of tissue residing dorsally to the fused NT. Coronal view. Top to bottom: dorsal to ventral. Figure from Copp et al. (2003). 13
Several studies have attempted to identify the molecular mechanism(s) that regulate the separation of epidermal and neural ectoderm after NT fusion. A model in Xenopus has been proposed such that this process is dependent on the differential expression of adhesion molecules, i.e. neuroepithelium expresses neural cell adhesion molecule 1 (xNcam1) while epidermal ectoderm expresses cadherin 1 (xCdh1), which allows separation of the two distinct cell lineages. However, loss of function of xNcam1 resulted in an undulated but yet closed NT. Thus, it is still controversial to whether these molecules do play a role in selective NT fusion and epidermal separation (Copp et al., 2003; Cremer et al., 1994; Detrick et al., 1990; Fujimori et al., 1990; Levine et al., 1994; Radice et al., 1997). In addition, NT fusion appears to require ephrin A5 (Efna5), which encodes a ligand that binds to Ephrin transmembrane receptors. Loss of function of Efna5 in mouse embryos demonstrated a failure of NT fusion, and co-expression Efna5 and its cognate receptor eph receptor A7 (Eph7a) protein promoted in vitro cellular aggregation. This suggests that Efna5 is important for adhesion during NT fusion (Holmberg et al., 2000). Nevertheless, the exact mechanism of how Efna5 and Eph7a function in NT fusion and whether they have a role in the separation of neuroepithelial and epidermal neuroectoderm after NT fusion is still yet to be elucidated.
Dorsal NT development and NC migration
During neurulation, interaction between the most lateral neuroectoderm and its adjacent epidermal ectoderm at the dorsal neural folds leads to the specification of a more specialised neuroectoderm with multipotent progenitors. Upon NT fusion, these lateral-most cells occupy the mediodorsal position along the NT and undergo cellular remodelling, which triggers regulated apoptosis in some cells, in order to form a specialised group of cells at the dorsal midline of the NT (Figure 1.7) (Chizhikov and Millen, 2004b; Copp, 2005; Copp et al., 2003). Subsequently, cell cycle progression is downregulated in these cells in order to specify them into an early roof plate (RP), or sometimes referred as the dorsal midline. Ample studies have demonstrated that the withdrawal from cellular proliferation is a feature of an early RP and that a failure in the inhibition of dorsal NT cell cycle progression is associated with the lack of RP development (Baek et al., 2006; Chizhikov and Millen, 2004b; Kahane and Kalcheim, 1998; Sah et al., 1995).
The early RP cells initially consist of precursors that can differentiate into two lineages, the definitive RP cells and the migrating NC cells (Chizhikov and Millen, 2004b; 14
Copp, 2005; Copp et al., 2003). During neurulation, NC cells delaminate and migrate out of the dorsal NT, giving rise to a diverse range of cell fates. As a result, a definitive RP that only consists of RP precursors is left at the dorsal NT upon completion of NC cells migration (Chizhikov and Millen, 2004c; Copp et al., 2003; Krispin et al., 2010a). Interestingly, a recent lineage study has found that cellular destiny at the dorsal NT has been arranged in a spatial manner shortly after NT closure in chick. Prior to NC delamination and migration, early RP cells that are destined to become NC reside at the dorsomedial aspect of the NT, while those that give rise to the definitive RP locate to a more ventrolateral position within the early RP. Thus, it is thought that upon the dorsolateral migration of NC precursor cells, the remaining ventrolaterally residing early RP cells then gradually relocate dorsomedially to make up the definitive RP (Krispin et al., 2010a; Krispin et al., 2010b).
It is beyond the scope of this introduction to discuss NC migration in detail. However, it is worth mentioning that unlike the spinal cord where NC cells migration only initiates after complete NT fusion, cranial NC cells delaminate and migrate before complete NT fusion. Thus, due to the close spatial and temporal relationship between cranial NC migration and NT fusion, it is thought that emigration of NC cells out of the cranial dorsal NT is a requirement for successful cranial NT fusion (Chizhikov and Millen, 2004b; Copp, 2005; Copp et al., 2003).
Summary
In summary, the formation of an enclosed NT requires a successful neurulation initiation, which leads to the juxtapositioning of the dorsal neural folds through furrowing and folding, and is then followed by NT fusion. Failure of NT closure will ultimately affect the other neurulation events that are subsequent to NT closure, i.e. dorsal NT development, in particular RP formation, and NC migration.
1.3.3 RP formation and its role in dorsal CNS development
The RP is an organising centre that secretes dorsalising morphogens, Bmp and Wnt, to antagonise the ventralising signal, Shh, secreted by the floor plate (FP) (Chizhikov and Millen, 2004c; Chizhikov and Millen, 2005). Thus, RP secreted signals are important to 15
specify the dorsal neuronal identities and hence, maintain the proper DV patterning. In addition, RP cells are also responsible for giving rise to the dorsal midline CNS domains during development. In fact, these functions of RP are readily demonstrated in mouse mutants with RP ablation, which these mice display developmental failure in RP derivatives, as well as a lack of dorsal neuronal specifications and a partial or complete NT ventralisation (Bach et al., 2003; Chizhikov and Millen, 2004c; Chizhikov and Millen, 2005; Lee et al., 1998; Lee et al., 2000b). The genetic and molecular interactions that are implicated in the development of RP and its derivatives, as well as the role of RP in dorsal NT specification will be discussed below.
The molecular regulatory network of RP development
In early chick embryos, dorsal NT identity is induced by signals encoded by bone morphogenic protein 4 (Bmp4) and bone morphogenic protein 7 (Bmp7), emanating from the epidermal ectoderm before NT fusion (Liem et al., 1995). In return, Bmp signalling is initiated within the dorsal NT to instruct early RP development (Hébert et al., 2002) as well as dorsal NT specification. The importance of RP-derived diffusible Bmp factors in dorsal NT patterning and neuronal specification was readily demonstrated by in vitro and in ovo experiments in chick NT (Liem et al., 1997; Liem et al., 1995). Not surprisingly, in vivo studies have also shown that Bmp signalling is indispensable for dorsal CNS patterning in both mouse and chick (Hébert et al., 2002; Liu et al., 2004).
Homeobox, msh-like 1 (Msx1), a homeobox transcription factor, is a downstream factor of Bmp signalling in RP development and dorsal NT patterning. Msx1 and Msx2 were induced by Bmp signalling through ectopic expression of Bmp4, Bmp7 as well as constitutive Bmp receptors in chick and mouse dorsal NT (Liem et al., 1997; Liem et al., 1995; Timmer et al., 2002). Moreover, ectopic cMsx1 expression was sufficient to induce downstream dorsal NT marker expressions within the chick NT (Liu et al., 2004), while ablation of Msx1 in mouse embryos resulted in disruption of dorsal diencephalon patterning as well as dysplasic development of RP derivatives (Bach et al., 2003).
LIM homeobox transcription factor 1 alpha (Lmx1a) was later found to be upstream of Msx1 and Msx2 in RP and dorsal NT development, as ectopic Lmx1a induces expression of a number of dorsal NT markers, including Msx1. Mutations in the Lmx1a coding region in the dreher mouse leads to a loss of RP as well as disruption of dorsal NT patterning in the spinal 16
cord (Chizhikov and Millen, 2004b; Millonig et al., 2000). Expression analysis showed that Lmx1a is expressed in the neural fold prior to complete NT closure at 8.5 dpc. Its expression is maintained throughout the dorsal NT during 9.5 dpc and 10.5 dpc. Subsequently, Lmx1a expressing domains within cranial RP then become gradually restricted to part of telencephalon, mesencephalon, rhombencephalon and anterior diencephalon (Failli et al., 2002). Although not well studied, the dynamic expression pattern of Lmx1a coincides temporally with onset of neurogenesis, i.e. from 10.5 dpc onwards, suggesting that Lmx1a may have an instructive role in RP directed dorsal CNS patterning of the developing brain.
wingless-related MMTV integration site 1 (Wnt1) expression can be observed in chick neural fold before NT closure and in both NC and RP progenitors at the dorsal NT (Hollyday et al., 1995; Megason and McMahon, 2002). It is unlikely that Wnt1 has an inductive role in dorsal NT identity as an early RP was present in Wnt1 mouse mutant (Shimamura et al., 1994). However, Wnt1 is definitely required to maintain RP and dorsal NT identity, since the development of diencephalic RP derivatives was perturbed in Wnt1 conditional mutant. Loss of Wnt1 in mice allowed the expression of Cdh1 in the medial region of the RP, an area normally devoid of Cdh1 (Chizhikov and Millen, 2004c; Shimamura et al., 1994). Further evidence suggests that Wnt1 may be regulated by Msx1 in RP development as Msx1 null mice lack Wnt1 expression within a domain of dorsal diencephalon (Bach et al., 2003).
Bmp and Wnt signalling are required for both NC cells and definitive RP development
Interestingly, the expression of the above mentioned early RP genes persists and becomes more restricted to differentiated RP cells after NC cell migration is completed. This is in contrast to Sox9, Forkhead box D3 (Foxd3) and Snail homolog 2 (Drosophila) (Snail2), which are initially expressed in the dorsal NT and early RP but are later switched-off in differentiated RP cells as NC cells migration completes (Krispin et al., 2010b). This suggests that Bmp4, Bmp7, Msx1 and Wnt1 are not only required for the development of the dorsal NT and early RP but also the definitive RP. The exact molecular regulation for the development of definitive RP from the dorsal NT or early RP has not been elucidated. However, growth differentiation factor 7 (Gdf7) is a Bmp family member and is so far the only gene identified with expression restricted to the definitive RP at the spinal cord from 9.5 dpc and at the diencephalon from 12.5 dpc (Louvi and Wassef, 2000). Loss of Gdf7 function leads to a failure in differentiation of the three dorsal most neuronal types in mouse spinal cord (Lee et 17 al., 1998), a phenotype similar to the loss of RP (Lee et al., 2000a). It is thus logical to speculate that Gdf7 may involve in definitive RP specification from early RP or dorsal NT precursors.
Other genes that lead to RP dysgenesis within the diencephalon
Due to the clear DV neuronal specification within the spinal cord, functional studies of most of the genes mentioned above were performed within the caudal CNS. Only Msx1 and Wnt1 have been specifically implicated in the diencephalic RP development and specification of dorsal neurons (Bach et al., 2003; Louvi and Wassef, 2000). In addition to Msx1 and Wnt1, Regulatory factor X, 4 (Rfx4) is also important for the proper development of diencephalic RP derivative. Rfx4 belongs to the helix turn helix family of transcription factor. Rfx4 transcript 3 (Rfx4_v3) is the CNS splice variant and it is expressed at the closing cranial neural folds from 8.5 dpc with its expression later restricted to the RP. Rfx4_v3 transcript can be detected at the diencephalic RP at 16.5 dpc and Rfx4_v3 null mutant mice lack dorsal but not ventral midline derivatives in the forebrain (Blackshear et al., 2003). Moreover, Notch gene homolog 2 (Notch2), a member of the Notch transmembrane receptor family, has been shown to be involved in the inhibition of CNS differentiation and regulation of cell cycle progression. Notch2 null mice are embryonic lethal. Notch2 chimeric mice although survived postnatally, Notch2 null cells failed to contribute to diencephalic RP formation in chimeric embryos. Thus, this strongly suggests Notch2 as an essential component for RP development (Kadokawa and Marunouchi, 2002).
Summary
In summary, dorsal NT is induced by Bmp signals released from the neighbouring epidermal ectoderm at the dorsal neural folds before complete NT fusion. Upon dorsal NT induction, Bmp4, Bmp7, Lmx1a, Msx1 and Wnt1 are expressed within the dorsal neural folds and persist into definitive RP. Loss of function of these genes leads to the lack of dorsal NT specification and development of RP derivatives, demonstrating their importance in dorsal NT, as well as early and definitive RP development.
1.3.4 Patterning of forebrain – diencephalon 18
After NT closure, CNS needs to be patterned correctly for cellular specification and functional specialisation. Unlike the clear AP segmentation in the rhombencephalon and spinal cord, or the DV multilayered organisation of the telencephalon, functionally similar cells within the diencephalon are collected in clusters referred as nuclei. Intriguingly, cells from the same diencephalic nucleus do not necessarily share the same lineage. As a result, it has been controversial that whether the diencephalon can be regarded as a compartmentalised structure. Until recently, a more widely accepted model that divides the diencephalon based on molecular marker expression and spatial position has been used. However, it is necessary to bear in mind that lineage identity is not incorporated in the construction of the model.
Based on the model, the diencephalon is divided dorsoventrally as follows (from dorsal to ventral): RP, alar plate, basal plate and the FP (Figure 1.8) (Lim and Golden, 2007; Puelles and Rubenstein, 2003). Additionally, the diencephalon has been divided into neuromere units, termed prosomeres, along the AP axis. There are three prosomeres along the AP length of diencephalon, all spanning the entire DV axis. Prosomere 1 is the caudal most domain and comprises mostly of pretectum (Pt), prosomere 2 is immediately rostral to prosomere 1 and it is mostly made up of epithalamus (EpTh) and dorsal thalamus (DTh), while prosomere 3 is mainly composed of ventral thalamus (VTh) (Figure 1.8) (Lim and Golden, 2007; Puelles and Rubenstein, 2003). Diencephalic AP patterning is dependent on an organising centre, zona limitan intrathalamica (ZLI). The function of ZLI has been extensively studied in chick embryo due to the ease of manipulation and access to developing embryos. When observed laterally, ZLI is an inverted wedge-like domain situated at the ventral boundary between prosomere 2 and prosomere 3. It extends from the FP towards the basal plate and alar plate with diminished AP spanning as progressing dorsally (Figure 1.8) (Lim and Golden, 2007; Puelles and Rubenstein, 2003). The ZLI is defined by the interaction between the prechordal neural plate, i.e. anterior neural plate that lies dorsal to the prechordal plate, and epichordal neural plate, i.e. the caudal neural plate that extends posteriorly from midbrain and lies dorsal to notochord, as demonstrated in chick embryos (Vieira et al., 2005). ZLI formation is dependent on Shh signalling and in return, ZLI emits Shh that is both required and sufficient to induce fates of VTh and DTh nuclei (Lim and Golden, 2007; Zeltser, 2005). In cooperation of GLI-kruppel family member (Gli) zinc finger proteins, Shh from ZLI defines nuclei identities within diencephalon (Hashimoto-Torii et al., 2003; Lim and Golden, 2007). P2
P1 P3
Figure 1.8: The patterning of the diencephalon. ZLI (blue and as indicated) is positioned between P2 and P3 with an expanded AP domain ventrally from the FP and basal plate (light orange). Its AP span is progressively restricted from the alar plate (yellow) towards the RF. Bmps and Wnts signallings are important for RP development and dorsal NT identity, while Shh and Bmps signallings have have opposing effect and maintain ventral identity. DT: dorsal thalamus. Hy: hypothalamus. Mes: mesencephalon. P1: prosomere 1. P2: prosomere 2. P3: prosomere 3. PT: pre-thalamus. PrT: pre-tectum. T: thalamus. Tel: telencephalon. VT: ventral thalamus. Left to right: anterior to posterior. Top to bottom: dorsal to ventral. Figure from Lim and Golden (2007). 19
To summarise, unlike the caudal CNS, the diencephalon does not display a segmented structure. However, the diencephalon is divided into four domains along the DV axis, namely the RP, the alar plate, the basal plate and the FP. Division along AP axis has also been established, suggesting a model composing of three prosomeres, namely prosomere 3, prosomere 2 and prosomere 1 in an order from the anterior to posterior. 20
1.4 SOX3 in the developing CNS
1.4.1 SOXB1 expression during neural development
Since SOXB1 genes are highly expressed in the developing CNS, the following section will discuss SOXB1 developmental expression with particular focus on SOX3 expression during neural development.
SOXB1 expression during anterior neural development in mouse
SOX3 is highly expressed along the neuraxis, especially in the presumptive brain and sensory placodes during mouse development (Figure 1.9). Sox3 expression is initiated at 6.0 dpc throughout the epiblast of the mouse embryo before gastrulation (Wood and Episkopou, 1999). During gastrulation (6.5-7.5 dpc), Sox3 expression becomes restricted to the presumptive neuroectoderm, which resides at the anterior of the embryo, and is absent at the posterior embryo where non-neural ectoderm is present (Figure 1.10). Intriguingly, Sox3 expression is re-expressed at the posterior embryo as demonstrated by its transcript detection in both the neuroectoderm and primitive ectoderm at 7.5-8.0 dpc (Figure 1.10). During neurulation (8.0-9.5 dpc), Sox3 expression is detected along the entire NT, with a markedly reduced intensity found at the dorsal neural fold (Collignon et al., 1996; Wood and Episkopou, 1999). Between 9.5-12.5 dpc, Sox3 expression resolves to specific domains within the CNS. As neurogenesis progresses from 12.5 dpc, Sox3 expression becomes restricted to the ventricular zone (VZ), the ventricle lining where neural progenitors reside during neurogenesis. Sox3 transcripts can be detected at the VZ of the dorsal telencephalon, the ventral diencephalon and the roof of midbrain (Figure 1.9) (Collignon et al., 1996; Solomon et al., 2004; Wood and Episkopou, 1999). In agreement with the above Sox3 expression analysis, both Sox3 transcript and SOX3 protein are present at the developing neuroepithelium from 11.5 dpc of dorsal anterior CNS (Figure 1.11 C) (N. Rogers, unpublished data), and this VZ restricted SOX3 expression is maintained from 12.5 to 15.5 dpc as neurogenesis progresses (N. Rogers, unpublished data; K. Lee, unpublished data).
Sox2 is the earliest SoxB1 family member to be expressed during development. Sox2 transcript can be detected in the morula at 2.5 dpc. Its expression is progressively restricted to 21 the inner cell mast at 3.5 dpc and it was found to be present within the neuroectoderm at 7.5 dpc during gastrulation (Figure 1.10) (Avilion et al., 2003; Wood and Episkopou, 1999). Unlike Sox3, which is re-expressed at the posterior embryo at 7.5-8.0dpc, Sox2 expression is limited to the neuroectoderm at this stage (Figure 1.10) (Wood and Episkopou, 1999). During neurulation, its expression within the CNS almost overlaps completely with that of Sox3, except that it is also highly expressed within the dorsal neural folds (Wood and Episkopou, 1999). In addition, Sox2 is present in the VZ from both the dorsal and ventral neuroepithelium within the anterior CNS during neurogenesis at 12.5 dpc (Figure 1.11 B) (N. Rogers, unpublished data).
In contrast, Sox1 is the last SoxB1 family member to be expressed in the developing CNS. SOX1 is initially expressed at the anterior neuroectoderm from 8.0 dpc (Figure 1.10) (Wood and Episkopou, 1999). Similar to Sox2, its expression is highly detected along the neuraxis, including the dorsal neural folds, during neurulation (Wood and Episkopou, 1999). In contrast to Sox3, Sox1 expression at the VZ is low at the dorsal but high at the ventral neuroepithelium (Figure 1.11 A) (Pevny et al., 1998; Wood and Episkopou, 1999)
The progressive restriction of SoxB1 to neural progenitor cells suggests that SoxB1 proteins may be required for maintenance of neural progenitors. In fact, it has been demonstrated SOX2 and SOX3 are expressed in the subventricular zone and dentate gyrus, the two identified neural stem cell niches within an adult brain (N. Rogers, unpublished data). This observation is further supported by a study using a SOX1 and SOX3 cross-reactive antibodies demonstrating the presence of SOXB1 members in neural stem cell niches (Wang et al., 2006).
SOX3 expression during anterior neural development in other vertebrates
Limited data is available for SOX3 expression in human embryonic or adult tissues. Northern blot analysis has shown that SOX3 is expressed at a very high level in the foetal brain and at a lower level in the foetal spinal cord, adrenals, liver, thymus and spleen and pancreas. A low level of SOX3 expression was also detected in human adult testis, spleen, heart and liver (Stevanovic et al., 1993).
Interestingly, SOX3 orthologs have been identified in chicken (Gallus gallus), zebrafish (Danio rerio), medaka fish and Xenopus. Both Xenopus Sox3 (xSox3) and zebrafish Embryo age
(dpc) Sox1 Sox2 Sox3
• Epiblast3 • Epiblast3 6.5 extraembryonic extraembryonic (pre streak) • • tissues3 tissues3
7 • Anterior ectoderm3 • Anterior ectoderm3 (early streak) • Anterior half of embryos2 • Recedes form streak 7.5 3 Columnar ectodermal • Anterior ectoderm3 ectoderm (late streak) • cells (future neural • Posterior epiblast3 plate cells)2
8 • Neural plate ectoderm3 • Anterior ectoderm3 • Posterior ectoderm3
• Restricted 1 • Neuroectoderm 8.5 neuroectoderm 2 3 • Restricted expression (neurulation • Neuroepithelium expression 1 3 at neuroectoderm starts) • otic placode 3 3 • posterior foregut • nasal placode
• Peripheral nervous 9 system (PNS)3
9.5 • Neuroectoderm1 • Neural tube2 • Olfactory placode1
• Expression downregulates in • CNS (varying levels 10.5 motor neurons, 1 within rhombomeres) interneurons and floorplate2
• Foetal brain excludes optic cup1 1 • Urogenital ridge • Urogenital ridge (at 11.5 • Urogenital ridge 1 (within germ cells) somatic sertoli cell precursor)1 • Low level at limb bud1
• VZ of presumptive • VZ in telencephalon • VZ of telencephalon cerebral cortex and 12.5 9 and diencephalon2,9 and diencephalon9 hippocampus • Ventral diencephalon9
• presumptive cerebral • Around lateral • VZ of the cortex and 13.5 ventricles in presumptive cerebral 4 hippocampus telencephalon2 cortex8 • ventral diencephalon4
• Cerebellum5 • Ependymal layers in 1 • Ependymal layers in 9 Adult • Testis 9 CNS (P2) CNS • Testis (Spermatogonia)6,7,
Figure 1.9: Expression patterns of SOXB1 transcription factors. Shaded: protein expression; unshaded: gene expression. 1Collingon et al. (1996). 2Pevny et al. (1998). 3Wood & Episkopou (1999). 4Solomon et al. (2004). 5denotes data from Weiss et al. (2003). 6Rizzoti et al.(2004). 7Raverot et al. (2005). 8Bani-Yaghoub et al. (2006). 9personal communication, P. Thomas & N. Rogers. dpc 6.0 7.0 - 7.5 7.5 8.0 Stage Pre streak Mid- to late streak Neural plate Head-fold Sox1 Sox2 Sox3
Figure 1.10: Sox1, Sox2 and Sox3 dispaly similar, but dynamic expressions during early mouse development. In situ hybridisation of Sox1, Sox2 and Sox3 at various stages pre-, during and post-gastrulation. Sox1 expression is only evident from the head-fold stage. Sox2 and Sox3 are expressed throughout the epiblast at the pre-streak stage. Their expressions are restricted anteriorly at the mid-streak to the late streak stage, with Sox2 expression remains at the anterior embryo at post-late streak stage, while that of Sox3 is re-expressed at the posterior end of the embryo. Top to bottom: dorsal to ventral. Left to right: anterior to posterior. Scale bar: 200μm. Figure from Wood and Episkopou (1999).
NOTE: This figure is included in the print copy of the thesis held in the University of Adelaide Library.
Figure 1.11: Expression of Sox1, Sox2, Sox3 partially overlap in the developing telencephalon at 12.5 dpc. In situ hybridisation demon- strated that SoxB1 are expressed at the VZ of various domains within the telencephalon. A-B: Sox1 and Sox2 have higher expression at the MGE and LGE of the ventral telencephalon than the developing CCx of the dorsal telencephalon (blue brackets). C: Stronger Sox3 expression is present at the dorsal telencephalon (blue brackets) in comparison to the MGE and LGE of the ventral telencephalon. Coronal sections. Top to bottom: dorsal to ventral. Scale bar: 1mm. Figure from N. Rogers (unpublished data). 22
Sox3 (zSox3) encode a protein having 68% amino acid identity with hSOX3. In comparison, the Sox3 proteins encoded by chick (cSox3) and Medaka fish (mfSox3) share 66% and 84% amino acid identity with hSOX3 respectively. Similar to Sox3, cSox3, xSox3, zSox3 and mfSox3 are expressed predominantly in the developing CNS. For cSox3 particularly, its expression is regionalised to the VZ, where the proliferating neural precursors are within the CNS as neurogenesis starts (Bylund et al., 2003; Koster et al., 2000; Neves et al., 2007; Okuda et al., 2006; Penzel et al., 1997; Rex et al., 1997b; Uwanogho et al., 1995).
Summary
Together, the above evidence strongly indicates a role of SOXB1 during neural development, since they are highly and widely expressed in the developing CNS. In particular to SOX3, its expression at the developing CNS is rather dynamic and appears to be associated with neural progenitors.
1.4.2 SOX3 and human disease – X-linked hypopituitarism
X-linked hypopituitarism (XH) is a congenital disorder characterized by deficiencies in the production of pituitary hormones. XH-patients display short stature and retarded growth due to growth hormone deficiency. The SOX3 containing region, Xq26-27 was duplicated in a number of XH-affected families (Figure 1.12) (Hol et al., 2000; Lagerstrom- Fermer et al., 1997; Solomon et al., 2002; Stankiewicz et al., 2005; Zucchi et al., 1999). Moreover, SOX3 was the only gene among the three genes identified within the submicroscopic duplication of a XH-affected family to be expressed in both the human foetal brain and the mouse infundibulum, a primordium that gives rise to the posterior pituitary and its axonal connections with the hypothalamus. In fact, two of the XH patients in this family displayed anterior pituitary hormone deficiency associated with anterior pituitary hypoplasia, an ectopic posterior pituitary, and an absence of infundibulum (Figure 1.12) (Woods et al., 2005). In addition to duplication, SOX3 mutations that introduce either a 7-polyalaine expansion (SOX322Ala) (Burkitt Wright et al., 2009; Woods et al., 2005) or 11-polyalaine expansion (SOX326Ala) (Laumonnier et al., 2002) at the most N-terminal polyalanine tract of hSOX3 have been identified in two independent XH families with variable defects along the hypothalamic-pituitary axis (Figure 1.12) (Woods et al., 2005). Molecular studies on the 23
localisation of SOX322Ala and SOX326Ala performed in various cell lines found that both SOX322Ala and SOX326Ala form cytoplasmic and perinuclear aggregates and fail to translocate into the nucleus for transcription activity (Albrecht et al., 2004; Wong et al., 2007), demonstrating that the association between the loss of SOX3 function and XH in humans. Hypothalamic innervations are essential to induce the development and function of the anterior pituitary. Intriguingly, Sox3 is only expressed at the developing hypothalamus and infundibulum, but not the Rathke’s pouch, the primordium for the anterior pituitary. Thus, the anterior pituitary defects in XH patients are likely to be secondary to infundibulum aplasia. Together, the above evidence further supports SOX3 as the causative gene in XH, and that a normal dosage of SOX3 is required for the proper development and function of the hypothalamic-pituitary axis.
XH patients associated with Sox3 mutations and duplications sometimes display other symptom(s) in addition to deficiencies of pituitary hormones (syndromic XH). In fact, SOX3- related syndromic XH patients frequently display varying levels of mental retardation (Figure 1.12). Interestingly, XH patients carrying SOX322Ala had normal intellectual performance (Woods et al., 2005) which is in contrast to those carrying SOX326Ala (Laumonnier et al., 2002). This suggests that different CNS regions, e.g. the forebrain and the posterior pituitary, have different sensitivities to SOX3 dosage or activity (Woods et al., 2005). Alternatively, mental retardation may be a result of a more severe loss of normal SOX3 function as the frequency of aggregate formation of SOX326Ala (88%) was significantly higher than that of SOX3Ala22 (42%) (Wong et al., 2007).
Nonetheless, the presence of mental retardation or other CNS related symptoms in SOX3 associated XH patients prompts the speculation that SOX3 may be required for the normal development and/or function of CNS regions other than the posterior pituitary (Hamel et al., 1996; Hol et al., 2000; Lagerstrom-Fermer et al., 1997; Laumonnier et al., 2002; Raynaud et al., 1998; Solomon et al., 2002; Stankiewicz et al., 2005; Woods et al., 2005; Zucchi et al., 1999). Within the developing mouse CNS, both Sox3 and SOX3 are highly expressed at the dorsal telencephalon, which gives rise to the CCx (Solomon et al., 2004; K. Lee, unpublished data; N. Rogers, unpublished data). CNS defect has also been identified within the CCx of XH patient(s). Indeed, dysgenesis of the corpus callosum, an axonal fibre from the CCx that connects and allows the communication between the left and right Symptom(s) Source Region mapped
• Mental retardation • Hamel et al. (1996) • Linkage to Xq24-27.3 • Isolated hypopituitarism
• Panhypopituitarism • Hol et al.(2000) • Duplication: Xq26-27
• Panhypopituitarism • Lagerström-Fermér • Duplication: Xq25-26 • Mild to moderate mental et al. (1997) retardation
• Hypopituitarism • Solomon et al. • Duplication: Xq26.1- (2002) Xq27.3
• Mental retardation and growth • Laumonnier et al. • 11-polyalanine hormone deficiency (2002) expansion
• Mental retardation with isolated • Raynaud et al. • Linkage to Xq22- growth hormone deficiency (1998) Xq27.2
• Mental retardation and growth • Laumonnier et al. • Inversion breakpoints: hormone deficiency (2002) Xp21.3 & Xq27.1
• Hypopituitarism • Stankiewicz et al. • Duplication: Xq26.2- (2005) 27.1
• Variable hypopituitarism, • Woods et al. (2005) • Duplication: Xq27.1 • Corpus callosum dysgenesis, (685.6kb) • Anterior pituitary hypoplasia, • 9 polyalanine • Ectopic posterior pituitary expansion • Absence of infundibulum
Figure 1.12: Summary of the symptom(s) observed in XH patients that have SOX3 mutations or chromosomal abnormalities that are linked to SOX3. Studies shaded in yellow are from one affected family, while studies shaded in grey are from a seperate affected familiy. Unshaded studies are from unrelated affected families. Studies from Laumonnier et al. (2002) and Woods et al. (2005) have independently identified SOX3 mutations with polyalanine expansions. 24
hemispheres of the forebrain, was observed in one SOX3-linked XH patient (Woods et al., 2005).
1.4.3 Sox3 null mouse model
In order to investigate the role of SOX3 in murine CNS development as well as to understand the molecular pathogenesis of SOX3 associated XH patients, Rizzoti et al. (2004) generated Sox3 null mice (Sox3ΔGFP) by replacing the endogenous Sox3 open reading frame (ORF) with that of the green fluorescence protein (GFP) through homologous recombination. A portion of Sox3ΔGFP adult mice displayed dwarfism as a result of pituitary dysmorphogenesis, which is secondary to the defective hypothalamic development and function in these mice. Adult Sox3ΔGFP mice had lower concentrations of pituitary growth hormones. Further study of Sox3ΔGFP embryos found defective morphogenesis of the hypothalamus, infundibulum and Rathke’s pouch. Misexpression of Rathke’s pouch patterning genes, Bmp4, fibroblast growth factor 8 (Fgf8) and homeobox gene expressed in ES cells 1 (Hesx1) were evident during pituitary development. The Sox3ΔGFP hypothalamic and pituitary phenotypes resembled those from the XH patients, further supporting the link between SOX3 and this disorder. Moreover, Sox3ΔGFP postnatal mice exhibited defects in the dorsal hippocampal commissure, a collection of axonal fibres that connects the two hippocampi, as well as corpus callosum agenesis (Rizzoti et al., 2004). Forebrain midline glia cells are shown to be important for the expression of surface molecules that exert axonal repulsion or attraction in order to direct proper corpus callosum and hippocampal commissure migration across the midline (Lindwall et al., 2007). Thus, the agenesis of corpus callosum and hippocampal commissure in midline crossing may indicate a potential defect in development or function of these glia cells.
In addition to its extensive expression in the developing CNS, Sox3 is also expressed in the developing pharyngeal arches (Wood and Episkopou, 1999), indicating a role for craniofacial development. In fact, in the original study of Sox3ΔGFP mice and in another independently generated Sox3 null mouse line, an abnormal positioning or absence of pinna, as well as an overgrowth of teeth were detected, suggesting a craniofacial defect (Rizzoti et al., 2004; Weiss et al., 2003). Further study by Rizzoti and Lovell-Badge (2007) described these phenotypes and identified the role of Sox3 during pharyngeal arches development. 25
Indeed, the migration of NC cells from rhombomere 3 of the hindbrain to the pharyngeal arch 2, which later gives rise to middle ear and hyoid bone, was disrupted (Rizzoti and Lovell- Badge, 2007). The author suggested that this might be a result of pharyngeal arch 2 identity compromisation (Rizzoti and Lovell-Badge, 2007), since pharyngeal arches are required to signal for proper NC cell migration (Minoux and Rijli). Moreover, genetic screenings has identified fibroblast growth factor (Fgf) signalling as a possible mechanism interacting with Sox3 in pharyngeal arch 2 development, as an enhanced phenotype was observed at the pharyngeal arch 2 of Sox3 and fibroblast growth factor receptor 1 (Fgfr1) compound heterozygotes (Rizzoti and Lovell-Badge, 2007).
Furthermore, Sox3 has also been implicated in the process of spermatogenesis (Raverot et al., 2005). Compromised fertility characterised by the appearance of empty tubules in male gonads, and smaller ovaries with a higher incidence of atretic follicles in female gonads were detected from two independently generated Sox3 null mouse models (Laronda and Jameson, 2011; Raverot et al., 2005; Rizzoti et al., 2004). Molecular analysis of hemizygous Sox3 null male from one of the models found that testis development and primary germ cell counts during early postnatal stages were normal. In preparation for puberty, immature spermatogonia are required to differentiate into mature spermatogonia, which later give rise to spermatozoa during spermatogenesis. However, characterisation of these Sox3 hemizygous null mice at the prepubertal stage found that they lacked mature spermatogonia, resulting in a loss of germ cells at later postnatal stages. Together, these suggest that Sox3 is dispensable for primary germ cell formation, but is required for the differentiation of mature spermatogonia from immature germ cells within the male gonads (Laronda and Jameson, 2011; Raverot et al., 2005).
1.4.4 SOX3 has an important role in early and neural development
SOXB1 members were previously shown to be important in early embryonic and neural development. Indeed, with the exception of mouse, Sox3 is the earliest SoxB1 member to be expressed during the embryonic development of chick, Xenopus and zebrafish (Okuda et al., 2006; Penzel et al., 1997; Rex et al., 1997a), and has important function(s) for their early development. Moreover, Sox3 has been implicated in the neural development in all vertebrates studied, including mouse (Bylund et al., 2003; Dee et al., 2008; Lee et al., 2006; 26
Rizzoti et al., 2004; Rogers et al., 2009). The following sections describes the function of SOXB1 in neural development with a particular focus first on the discussion Sox3 function in the early embryonic development of chick, Xenopus and zebrafish models, followed by the discussion on the role of Sox3 in neural development in all animal models mentioned above.
SOX3 in early embryonic development
The role of SOX3 in early embryo development has been mostly studied in Xenopus and zebrafish models. In Xenopus, maternal xSox3 transcript is regionalised to the zygotic animal pole (Penzel et al., 1997), suggesting that it may play a role in early embryonic events. Axis formation is one of the early events in embryonic development and is essential for numerous position-dependent downstream processes including patterning, specification and differentiation. Thus, xSox3 may be involved in axis formation by maintenance of ventral identity as ectopic sox3 expression within the dorsal region at the early zygotic stage led to embryo ventralisation. Interestingly, in vitro study in 293T and HeLa cells suggested that xSox3 may physically interact with the canonical Wnt signalling effector catenin (cadherin- associated protein)-beta 1 (xCtnnb11) through its C-terminal TA domain in order repress Wnt signalling for ventral identity maintenance (Zhang et al., 2003; Zorn et al., 1999). Later and more detailed analysis indicated that although xSox3 is capable of physical interaction with xCtnnb1, its ventralising effect, when expressed ectopically, is likely through the direct binding and transcriptional repression of nodal homolog 5 (xNodal5), a gene that encodes the dorsalising signal that belongs to the nodal related, transforming growth factor beta (Tgfβ) family (Zhang et al., 2003). Together, these data suggest that although ectopic xSox3 is sufficient to induce ventralisation within an early embryo, there lack any loss of xSox3 function experiment to demonstrate whether xSox3 is necessary for ventral identity induction or maintenance.
Zhang et al. (2004) attempted to address the above question by injecting antibody that ablates xSox3 function in early Xenopus embryos. In agreement with previous studies, injected embryos had elevated nodal5 expression and dorsalisation. They also showed aberrant gastrulation and failure of three germ layer formation (Zhang et al., 2004). However, the anti-xSox3 antibody used although had higher affinity to xSox3, it cross-reacted with xSox2. It is evident that SOX2 is critical in maintenance of pluripotency in epiblast through transcription regulation of nanog homeobox (Nanog), and Sox2 null mice fail to progress into 27
gastrulation due to the loss of epiblast maintenance (Kuroda et al., 2005; Rodda et al., 2005). Thus, this study should be interpreted with caution as the observed phenotype could be due at least in part to a partial loss of xSox2 function.
In addition, Spemann’s organiser, a centre that secrets morphogens for patterning and axis formation, plays a critical role in axis determination. Loss of zSox3 function by injection of zSoxB1 nuclear localisation or DNA binding dominant negative mutants resulted in ectopic organiser formation and axis duplication in zebrafish. Coimmunoprecipitation experiment using a fusion of zSox3 HMG box with Drosophila repressor domain, EnR, demonstrated that zSox3 acts as a direct repressor for organiser markers to inhibit its formation (Shih et al., 2010).
SOX3 in neural plate determination
The role of Sox3 in neuroectoderm formation is rather dynamic. Within the zebrafish, Sox3 enhances neural fate uptake in the primitive ectoderm, since ectopic zSox3 expression at zebrafish morula stage led to neuroectoderm expansion and CNS duplication. In contrast, zebrafish with morpholino knock-down of zSox3 at 1-2 cell stage still displayed CNS development, although largely dysmorphic as affected embryos showed midbrain and hindbrain with significantly reduced lumen and elevated apoptosis throughout CNS (Dee et al., 2008). Thus, zSox3 although is capable of promoting neural fate uptake, but is dispensable for neural induction.
In an overexpression study, Rogers et al. (2009) demonstrated xSox3 is required for noggin (xNog) induced neural fate uptake in Xenopus ectodermal explants, as it promotes the expression of xSox2 and geminin (xGem), a neuralising molecule, in early neural development. In addition, xSox3 increases neural plate cell proliferation and inhibited pro- epidermal Bmp signalling (Rogers et al., 2009). Interestingly, when both ectopically expressed, xSox3 counteracted POU class V protein oct-91 (xOct91) neuralising activity during gastrula and neurula stage at non-neural ectodermal tissue (Archer et al., 2010). This demonstrated that xSox3 may have multiple and dynamic function in tissue specification, depending on spatial circumstances and cellular context. At last, in support of the function of Sox3 in promoting neural fate from uncommitted pluripotent cells, forced SoxB1 expression in mouse ES cells led to an enhanced uptake of neural fate upon induced differentiation (Zhao et al., 2004). 28
SOX3 and neurogenesis in CNS
The ancient role of SOX3 in neural differentiation has been conserved in Drosophila. The Drosophila SoxB1 homolog SoxNeuro (SoxN) is required for the proper specification of neuroblast (Overton et al., 2002). Genetic screens in Drosophila have found that SoxN represses Wnt signalling through transcriptional inhibition of its signalling effector during neural development (Chao et al., 2007). Moreover, the function of SoxB1 during Drosophila neuroblast development is dependent on the posttranscriptional regulation of SoxN activity. SoxN was found to be sumoylated at lysine 439. SUMOylation-deficient SoxN displayed a significant upregulation in its transcription activity in vitro. Besides, the overexpression of SUMOylation-deficient SoxN in Drosophila has led to a defective neural development, demonstrating the importance of this posttranslational modification in the regulation of SoxN activity (Savare et al., 2005).
The first functional study of SOX3 in neurogenesis came from a series of in ovo experiments within the chick spinal cord and found that overexpression of cSoxB1 members, particularly cSox3, repressed the expression of neuronal markers and inhibits neural differentiation. Furthermore, a fusion protein that combines the cSox3 HMG box and Drosophila transcription activator VP16 mimicked the effect of cSox3 overexpression in chick spinal cord, suggesting that cSox3 acts as a transcriptional activator in this context. Interestingly, investigation of cell-cycle mechanism demonstrated that cSox3 is unlikely to maintain neural progenitor pool through regulation of cell-cycle progression (Bylund et al., 2003). Moreover, NES is a neural progenitor and early neuronal marker widely used in studies of CNS development. It encodes a neurofilament protein that functions in neural progenitors and immature neurons. Sequence analysis and experiments in various chick neural stem cell lines unravelled possible regulation where cSox3 binds to the enhancer element and activates cNes expression when dimerised to cPou3f2, suggesting that cSox3 may regulate expression of cNes in its role in neural progenitor maintenance (Tanaka et al., 2004).
In contrast to Bylund et al. (2003), it was later found that overexpression of Sox3 in neural explants grown from telencephalic neural progenitors extracted at 14.0 dpc mouse embryos did not inhibit neuronal fate (Kan et al., 2004). However, this result may not be a true representation of the situation in vivo. First, the presented experiments lacked both positive and negative controls, which rendered it difficult to interpret. Moreover, 29
telencephalic neurogenesis starts from 12.5 dpc. Although SOX3 is still highly expressed at the VZ of the developing telencephalon at 14.5 dpc (N. Rogers, unpublished data), neural progenitors from 14.0 dpc telencephalon might have been triggered towards a more specific lineage and become less responsive to Sox3 in comparison to earlier neural progenitors. At last, it is possible that Sox3 function is cell-context and animal model dependent.
Evidence indicates that Xenopus Sox3 shares similar function as cSox3 in inhibition of neural differentiation. Overexpression of xSox3 in Xenopus repressed the expression of neuronal markers, indicating a block in neurogenesis (Rogers et al., 2009). Intriguingly, Sox3 although inhibits neurogenesis, it appears to be required for the initiation of this process, since the transient loss of zSox3 function through morpholino knockdown during neurogenesis has significantly repressed the expression of differentiating neuronal markers (Dee et al., 2008).
A requirement of zSox3 in neurogenesis is further evident in zebrafish hypothalamus. Lymphoid enhancer factor 1 (zLef1), a downstream effector gene of canonical Wnt signalling, is regulated by diffusible morphogen wingless-related integration site 8b (zWnt8b) for posterior hypothalamic neurogenesis. Knock-down of zLef1, and hence zWnt8b induced Wnt signalling, resulting in the loss of proneural and neuronal gene expressions within the developing hypothalamus. Rescue study found that zSox3 overexpression was capable of restoring normal hypothalamic development in zLef1 knock-down embryos. Indeed, zLef1 was found to bind to zSox3 upstream regulatory sequence in vivo (Lee et al., 2006). These data together suggest that zSox3 is activated and regulated by Wnt signalling during neurogenesis for zebrafish hypothalamic development (Lee et al., 2006).
Other SoxB1 members have also been implicated in early embryonic and neural development in CNS
Sox2 is the first SoxB1 member to be expressed during early embryonic development in mouse. Indeed, Sox2 is expressed throughout the inner cell mass of a blastocyst (Avilion et al., 2003). The function of Sox2 has been demonstrated to be critical for early mouse embryonic development, since Sox2 null mice are embryonic lethal before implantation due to the loss of identity of the inner cell mass (Avilion et al., 2003). In vitro studies of ES cells have readily demonstrated that SOX2 heterodimerise with POU5F1 to transcriptionally 30
activate Nanog in order to maintain stem cell identity and pluripotency (Kuroda et al., 2005; Rodda et al., 2005).
In Xenopus, ectopic xSox2 enabled Fgf induced neuroectoderm formation in epiblast explants (Mizuseki et al., 1998; Rogers et al., 2009). Archer et al. (2010) later found that the requirement for Fgf in neural induction in the system can be replaced by the co-expression of xPou5f1, which encodes the Xenopus homolog of POU5F1, a well known SOX2 partnering protein in mouse and human ES cell maintenance (Archer et al., 2010; Chew et al., 2005; Rodda et al., 2005). On the other hand, the loss of xSox2 led to the failure of primitive ectoderm to uptake a neural fate but did not enhance the uptake of the alternative epidermal fate, suggesting that xSox2 is required to maintain the ability of the primitive ectoderm to respond to neural induction signals (Kishi et al., 2000). As a result, it is yet to be determined whether the enhanced uptake of neural fate during Sox2 overexpression was directly promoted by the increased Sox2 expression, or a secondary effect of Sox2-enhanced pluripotency within the primitive ectoderm such that it is more responsive to neural inductive signals.
During CNS development, neurogenesis initiates prior to gliogenesis. Thus, during neurogenesis, neural progenitors have to be properly regulated to allow proper neuronal differentiation as well as to retain the progenitor pool for gliogenesis. During telencephalic development, Sox2 interacts with Notch signalling to promote gliogenesis, presumably through maintenance of the proliferative state of neural progenitors. Intriguingly, Sox2 inhibits neurogenesis but does not impede glia differentiation (Bani-Yaghoub et al., 2006). In agreement with this, constitutive expression of xSox2 in xenopus retained neural progenitors within the proliferative cell cycle and inhibited neurogenesis (Archer et al., 2010). Furthermore, SOX2 expression is present within the neurogenic region in adult mice. Sox2 mouse model that harbours a regulatory sequence deletion of the neural cell specific enhancer within the Sox2 regulatory sequence although is viable, but displays impaired adult neurogenesis and progressive neurodegeneration due to a significant decrease in proliferating neural progenitor cells (Ferri et al., 2004). Together, these data indicate that Sox2 maintains neural progenitor identity by retaining them within the active cell cycle.
SOX1 also has instructive activity in early neural differentiation. The first evidence suggesting the inductive role of SOX1 in neural differentiation came from a study in mouse embryonic carcinoma cells. Ectopic expression of Sox1 in these cells replaced the 31 requirement of retinoic acid in directing these pluripotent cells into neural fate (Pevny et al., 1998). Not unexpectedly, ectopic xSox1 expression in Xenopus during 1 to 2- cell stage caused neural plate expansion (Archer et al., 2010). Furthermore, overexpressing xSox1 in Xenopus ventral ectoderm at 32-cell stage was sufficient to induce expression of neural markers (Nitta et al., 2006). Surprising for such a potent effect in neural differentiation, Sox1 null mice are viable but display epilepsy and severe defects in ventral telencephalon (Malas et al., 2003).
Forced Sox1 expression does not only enhance neural fate uptake, but also neuronal differentiation (Archer et al., 2010; Nitta et al., 2006; Zhao et al., 2004). This agrees with later experiments where Sox1 was found to promote neurogenesis in developing mouse telencephalon (Kan et al., 2004; Kan et al., 2007). In contrast, evidence from isopropyl-β-d- thiogalactase- (IPTG-) induced transient expression of Sox1 in mouse embryonic carcinoma cells showed enhanced neural fate but not enhanced neurogenesis (Pevny et al., 1998). However, in this experiment, the transient Sox1 expression might only be sufficient to promote uptake of neural fate from pluripotent carcinoma cells but not enough to provide instructive signals for further differentiation from neural progenitors into neurons.
Genetic interactions of SOX3 in neural development
SOXB2 family member SOX21 is a transcription repressor that promotes neurogenesis. Chick in ovo study has demonstrated that cSox21 is capable of counteracting cSox3 function in spinal cord when overexpressed. Transient knock-down of cSox21 inhibited neurogenesis and increased proliferation of neural progenitors, a effect in contrast to that of cSox3 (Sandberg et al., 2005). Given the sequence homology of the HMG box of SOXB1 members, it seems possible that SOX3 and SOX21 have overlapping transcriptional targets and may regulate each other through competitive binding of their common target genes.
Notch signalling has been heavily implicated the inhibition of neurogenesis and maintenance of neural progenitor pools in various systems. Strong evidence indicated that SOXB1 members also possess similar functions. Thus, an in ovo study in chick spinal cord was carried out to determine the interplay and possible genetic interaction between these two mechanisms. cSox3 was previously shown to act as a transcriptional activator to maintain neural progenitor pool when overexpressed in spinal cord (Bylund et al., 2003; Holmberg et al., 2008). Similarly, Notch-intracellular domain (NICD), a truncated Notch receptor that 32
confers constitutive activation of Notch signalling, repressed the expression of proneural genes neurogenin 2 (cNeurog2) and transcription factor 3 (cTcf3), and inhibited neurogenesis when overexpressed within the same system. However, NICD overexpression was unable to maintain neural progenitor pool in the absence of cSox3 transcription activation activity in this context. In contrast, the overexpression of cSox3 was able to inhibit neural differentiation in the absence of Notch signalling. This study demonstrates the dependence of Notch signalling on cSox3 in neurogenesis inhibition within the chick spinal cord.
Interestingly, further proneural gene expression studies of the above experiment displayed selective downregulation of cNeurog2 and cTcf3 only by NICD but not cSOX3 overexpression. In contrast, overexpression of cNeurog2 and cTcf3 was able to release NICD- but not SOX3- induced neurogenesis inhibition. Collectively, these results indicate that a common regulatory system for neurogenesis inhibition shared between Notch signalling and Sox3 within the chick NT is unlikely (Holmberg et al., 2008). However, it should be noted that SOXB1 regulation is highly dependent on cellular context and thus the possibility of Notch and Sox3 interaction in other systems cannot be excluded.
Summary
Together from all of the above evidence, within the Xenopus and zebrafish, Sox3 is required for early embryonic patterning and axes formation by maintaining ventral identity of the embryos through its function as a transcription repressor. It is important to note that this function of Sox3 is unlikely to be conserved in mouse. This is evident by the fact that Sox2, rather than Sox3 is the first SoxB1 member to be expressed within the early mouse embryo, and that Sox3 null mice are viable and do not appear to have early embryonic patterning defect. On the other hand, Sox2 is a critical factor for early embryonic development as it is required to maintain the pluripotency of the inner cell mass in both mice and humans.
Moreover, SoxB1 family members are indispensable for normal CNS development. Sox1 and Sox3 are capable of instructing neural fate from uncommitted ectoderm in the absence of other growth factors. In addition, xSox2 also appears to enhance neural fate when overexpressed in Xenopus, despite it is not known whether this is a consequence secondary to its role in pluripotency maintenance. In addition, Sox2 and Sox3 both inhibit neurogenesis, with Sox2 also promote gliogenesis in mouse, while Sox1 promotes neurogenesis. 33
Intriguingly, despite its function in neurogenesis inhibition, Sox3 also seems to be required for neurogenesis initiation in Xenopus.
1.4.6 SOX3 in sex determination
In mouse, the dynamic expression of Sry between 10.5 dpc to 12.5 dpc, the developmental timeframe for sex determination, allows the different differentiation of the bipotential gonad. Interestingly, endogenous Sox3 is expressed at the urogenital ridge of both male and female mouse embryos during 11.5 dpc, despite this expression is no longer present at 12.5 dpc to 13.5 dpc (Collignon et al., 1996). SOX3 is the closest relative of SRY in the SOX family (Stevanovic et al., 1993) and is thought to be the evolutionary precursor of SRY (Graves, 1998; Graves, 2001). This has prompted speculation that SOX3 may play a role in sex determination (Graves, 1998; Graves, 2001). However, two independent Sox3 null mutant mice generated by Rizzoti et al. (2003) and Weiss et al. (2003) showed no evidence of sex reversal, indicating that Sox3 is dispensable for sex determination. Interestingly, a recent study where Sox3 was ectopically expressed in the developing gonad during 12.5 dpc to 13.5 dpc leads to XX male sex reversal. The author concluded that although SOX3 has evolved to specialise in CNS development, it is still capable to induce male gonadal development by acting as an analogue of SRY when ectopically expressed at the right place and time (Sutton et al., 2011).
1.4.7 Regulation of SOX3 expression
SOX3 has a specific and dynamic expression in CNS as demonstrated in mouse and other animal models. Thus, it is important to identity the cis-regulatory elements that control SOX3 expression as loss or gain of function in these elements can be detrimental to any SOX3 dependent processes. Using a lacZ transgene reporter approach Brunelli et al. (2003) demonstrated that Sox3 CNS expression regulatory sequences mostly reside 3kb upstream and downstream of the mouse gene. Some of these regulatory sequences were capable of driving similar expression patterns when injected into Xenopus embryos (Brunelli et al., 2003). 34
Moreover, a number of bioinformatics studies have identified a few long range highly conserved SOX3 regulatory sequences that reside up to 450kb upstream and 150kp downstream from the locus. In fact, a recent study has discovered eight highly conserved non-coding elements (HCNE) between human and zebrafish, which are located up to 350kb upstream and 150kb downstream of the SOX3 locus. These SOX3 regulatory elements are all functionally significant as they were able to drive reporter expression when injected into zebrafish (Navratilova et al., 2009). Similar searches for highly conserved regulatory sequences were also performed between human, mouse, rat and pufferfish and have identified two highly conserved sequences flanking the SOX3 locus. One of the conserved sequences resides 350kb upstream of the locus, while the other one is also upstream but at 170kb from the locus (Ahituv et al., 2007; Pennacchio et al., 2006). In vivo analysis of these regulatory sequences in mouse found all of these highly conserved sequences are capable to driving Sox3 expression within the developing CNS (Ahituv et al., 2007; Pennacchio et al., 2006). Together, these data indicate that the dynamic regulation of SOX3 expression during CNS development can be regulated by long range regulatory sequences residing up to hundreds of kilobases away from the locus.
Retinoic acid is able to induce neural differentiation in in vitro cell line. Interestingly, in NTERA2 cell line, SOX3 was upregulated but SOX2 is downregulated during retinoic acid induced neural differentiation (Stevanovic, 2003). Since then, a number of studies have been carried out to identify the regulatory components responsible for driving this SOX3 expression. The first study found that SOX3 upregulation is partially dependent on the binding of retinoic X receptor alpha (RXRα) to the 400bp sequence immediate upstream of the SOX3 coding region and a nearby novel regulatory motif (Mojsin et al., 2006; Nikcevic et al., 2008). In addition, sequence analysis has identified three CCAAT Nuclear factor Y (NF-Y) specific binding sequence lying between 200bp to 400bp upstream to the SOX3 coding sequence. Binding of NF-Y to these CCAAT sequences upregulates SOX3 expression as demonstrated by an in vitro co-transfection assay (Krstic et al., 2007). The importance of NF- Y in SOX3 expression regulation is further strengthen by the conservation of NF-Y binding sequence in a range of mammalian SOX3 loci (Kovacevic-Grujicic et al., 2007). However, studies of all of the above SOX3 regulatory sequences and components were performed in vitro. Thus, the in vivo significance of these elements is yet to be elucidated. 35
1.4.8 Functional redundancy between SoxB1 subgroup members
Given the broad CNS expression of SOX3, one may expect its absence would result in embryonic lethality. However, as described earlier, Sox3 null mice are viable (Rizzoti et al., 2004; Weiss et al., 2003) and exhibit relatively mild and specific defects in corpus callosum, dorsal hippocampal commissure and hypothalamus. While these defects indicate that SOX3 is specifically required for different aspects of CNS development, it is also likely that the comparatively mild phenotypes of Sox3 null mice are due to functional redundancy between the SoxB1 family members.
SOXB1 members share >50% and >90% a.a. identity overall and within the HMG box respectively. Moreover, their expression patterns have considerably overlapped in the developing CNS. Evidence for their partial functional redundancy in CNS development is readily available from studies using gain- and loss- of function analysis. Craniofacial development requires the proper migration of cranial NC cells from the dorsal NT to the pharyngeal arches. Deficiency study clearly demonstrated the redundant function of Sox2 and Sox3 in the development of pharyngeal arch. No craniofacial or pharyngeal arch 2 developmental defect was evident in Sox2 null or Sox3 heterozygous mice. However, Sox2 and Sox3 compound heterozygotes or Sox2 heterozygote in Sox3 hemizygous background had a lower viability and an increased severity of NC cells migration defect towards the pharyngeal arch 2 (Rizzoti and Lovell-Badge, 2007).
Moreover, similar to Sox3 null mice, Sox1 null mice also display relatively mild phenotype despite its broad expression pattern in the developing CNS (Pevny et al., 1998). The expression of SoxB1 members overlap extensively and thus, suggesting that some Sox1 functions may be compensated by Sox2 or Sox3 during CNS development. In addition, all chick cSoxB1 members were found to be able to inhibit neural progenitor differentiation in the developing chick spinal cord when overexpressed (Bylund et al., 2003). Besides, cSox2 is responsible for the maintenance of the neural progenitor pool within the chick spinal cord while loss of its function resulted in precocious cell-cycle exit and neuronal differentiation of the neural precursors. Interestingly, forced co-expression of cSox1 in this cell context was able to rescue this loss of cSox2 function phenotypes during chick neural development (Graham et al., 2003). Furthermore, in vitro analysis has demonstrated that SOXB1 members can bind to common interacting protein(s) within the neural cell context. In fact, cSox2 and cSox3 are able to physically interact with cPou3f2, which the heterodimer then binds to the 36
regulatory sequence and activate the transcription of cNes in chick neural stem cells (Tanaka et al., 2004)
In addition to neural development, SOXB1 members have been implicated in lens development. In fact, all cSoxB1 members are expressed during lens development (Kamachi et al., 1998). On the other hand, in mouse, although the expression of Sox1 and Sox2 are present at the early optic vesicle, with Sox2 expression later switched off upon the formation of the lens vesicle, Sox3 is not expressed at any stage of the lens development (Collignon et al., 1996; Nishiguchi et al., 1998). The requirement of SOX2 in lens development was further demonstrated in human as patients with SOX2 mutation display eye defect (Kelberman et al., 2008).
The overlapping expression of SOXB1 members during lens development suggests they may be functionally redundant in this process. Actually, SOXB1 members have overlapping molecular interactions as numerous in vitro studies showing that they share a list of common binding sequences and proteins during lens development (Kamachi et al., 1998). Pax6 is critical for lens formation and the loss of its function in mouse mutant resulted in small eye phenotype (Hill et al., 1991). Interestingly, Sox2 and Sox3 could interact synergistically to enhance PAX6 expression in human lens epithelial cells in vitro (Aota et al., 2003). Additionally, cSox2 was the first cSoxB1 member identified that recognises and binds to DC5 enhancer element of the crystallin gene, cAsl, in chicken lens development. Later, cSox1 and cSox3 were also shown to recapitulate this interaction in vitro (Kamachi et al., 1995; Kamachi et al., 1998). Similar to cAsl, the expression of crystallin, gamma A (Cryga) is upregulated during mouse lens induction. Again, Sox1 is able to physically interact with the promoter region and upregulate the expression of Cryga during lens induction (Nishiguchi et al., 1998). Intriguingly, cSox2 also interacted and upregulated the promoter region of another mouse crystalline gene, crystalline gamma F (Crygf), when it was transfected into chick epithelial lens cells (Kamachi et al., 1995). The cross-activation between SoxB1 proteins in activation of a number of crystallin genes in lens development suggests that the function of SoxB1 in transcriptional activation of crystallin genes is a long conserved mechanism from a common SOXB1 precursor, and was retained along evolution during later genes duplication and divergence.
In an attempt to further unravel the complicated redundant roles of SOXB1, Archer et al. (2009) studied the overlapping function of xSoxB1 members in early Xenopus 37 development. They overexpressed xSoxB1 members with xOct91, the Xenopus homolog of mammalian POU5F1 that dimerises with SOX2 in the maintenance of stem cell pluripotency. In Xenopus, the ectopic expression of xSoxB1 alone or partnered with xOct91 in early Xenopus embryos all resulted in neural plate expansion at the expense of epidermal ectoderm, despite differences in downstream expression profiles amongst them. In contrast to the identical consequence in the promotion of neural fate uptake during early Xenopus embryonic development, xSoxB1 members had different effect on neuronal differentiation when overexpressed. Indeed, xSox2 and xSox3 inhibited neurogenesis, while xSox1 promoted neurogenesis (Archer et al., 2010). Thus, the author concluded that xSoxB1 members elicit identical outcome in the uptake of neural during early Xenopus development, yet, they trigger different molecular profiles upon overexpression, which then confers their different actions at later neuronal differentiation and hence, allows specialisation of xSoxB1 function (Archer et al., 2010). 38
1.5 Cerebrospinal fluid, hydrocephalus and the subcommissural organ
This thesis focuses on the characterisation of the Sox3 transgenic mouse model (further discussion in 1.6) that displays hydrocephalus, a condition of excessive cerebrospinal fluid (CSF) retention. Hence the following section describes the mode of CSF circulation and the physiological, cellular and molecular mechanisms that regulate CSF homeostasis using the mouse as a model unless otherwise stated.
CSF fills the inner ventricular space within the brain and the central lumen within the spinal cord. In addition, CSF is also found in the subarachnoid space, a cavity between the brain and the arachnoid (outer meninges). The CSF-filled subarachnoid space acts as a cushion to protect the brain from external mechanical shock. Except the immature CNS, where the presence of spiral junctions between ependymal cells lining the ventricles creates a brain-CSF barrier, CSF is in direct contact with the CNS extracellular environment in fully developed CNS. Thus, CSF is important for regulating osmolarity, pH and solute concentration for CNS homeostasis. In addition, the CSF provides means of transporting signals, in the form of hormones, neuropeptides and glycoproteins, that are important for orchestrating development as well as maintenance of proper physiological function (Brown et al., 2004; Lowery and Sive, 2009; Mashayekhi et al., 2002; Milhorat, 1975; Miyan et al., 2003; Miyan et al., 2006; Pérez-Fígares et al., 2001; Rodríguez et al., 1999).
1.5.1 CSF production and circulation
The CSF is produced and secreted by the choroid plexuses (ChP), which are present in each ventricle within the brain. ChP are responsible for 70%-80% of CSF production. It has been documented that about 20-30% of CSF is also produced by the nervous parenchyma depending on different animal models (Brown et al., 2004; Lowery and Sive, 2009; Milhorat, 1975; Pérez-Fígares et al., 2001). Upon secretion from ChP, the bulk of CSF circulates in a rostrocaudal direction from the most anterior lateral ventricles (LV) through the foramen of Munro, an opening posterior to the LV, into the third ventricle (3V). Upon leaving the 3V, CSF flows into the narrow Sylvian aqueduct (Aq) into the fourth ventricle (4V). It exits the 39
4V at the posterior opening and enters the subarachnoid space (Figure 1.13). The majority of CSF is drained back into circulation within the subarachnoid space through arachnoid villi, small pockets formed by the arachnoid that is in close proximity to the veins. A small portion of the CSF also flows into the lumen of the spinal cord (Brown et al., 2004; Meiniel, 2007; Pérez-Fígares et al., 2001).
It is believed the rostrocaudal flow of CSF requires a combination of the following three aspects to maintain homeostasis. 1) The hydrostatic pressure difference between the anterior CNS, site of CSF production, and the posterior CNS, site of CSF drainage; 2) the unidirectional beating of cilia on the ventricular lining ependymal cells to enforce the laminar flow within and between the ventricles. The beating of cilia is particularly important for CSF flow through narrow passages, notably the Aq; 3) the arterial pulsation that contributes to the CSF flow from highly vascularised subarachnoid space into the spinal cord (Lowery and Sive, 2009; Milhorat, 1975; Pérez-Fígares et al., 2001).
1.5.2 CNS barriers and CSF homeostasis
The CNS is the most fragile organ within the body. It is thus reasonable to minimise the effect of perturbation from external environment in order to maintain homeostasis. Unlike other organs where materials from the interstitial fluid are freely exchanged with the circulation, both CNS interstitial fluid and CSF are restricted from material exchange from the circulation, which is constantly under external stress. There are three sites of permeability restriction in the CNS: the blood-brain barrier (BBB), the blood-CSF barrier (BCSFB) and the CSF-brain barrier (CSFBB). The BBB is characterised by tight junctions between endothelial cells lining blood vessels proximal to the CNS to restrict material exchange between the circulation and the CNS. BBB is present throughout the CNS with the exception of the circumventricular organs (CVO). The CVO is a collection of highly vascularised brain organs derived from ventricular wall protrusion of the 3V and 4V and are made up of germinal astrocytes. The neighbouring blood vessels of these CVO have fenestrated wall and are devoid of BBB. Thus, this allows material exchange between the CVO with the circulation (Balslev et al., 1997; Duvernoy and Risold, 2007; Engelhardt and Sorokin, 2009). The BCSFB is formed by apical tight junctions between specialised ependymal cells. The BCSFB is only found in CVO and it restricts material exchange between the CVO and the
NOTE: This figure is included in the print copy of the thesis held in the University of Adelaide Library.
Figure 1.13: Rostrocaudal flow of CSF within the CNS. Schematic diagram shows CSF flow from the LV into the 3V, then Aq , 4V and then the subarachnoid space and the spinal cord canal. Left to right: anterior to posterior. Top to bottom: dorsal to ventral. Figure from Picketts (2006). 40
CSF (Duvernoy and Risold, 2007; Engelhardt and Sorokin, 2009; Rodriguez et al., 1998). Moreover, CSFBB, characterised by the permeable apical spiral junction between neighbouring ependymal cells that line the ventricles, is present only in the immature brain. The CSFBB regulates exchange between CSF and interstitial fluid in the immature brain during CNS development (Engelhardt and Sorokin, 2009; Halfter and Schurer, 1998; Saunders et al., 2000). Functional compromisation of either BBB, BCSFB and CSFBB is detrimental to CSF homeostasis and leads to CNS pathogenesis (Engelhardt and Sorokin, 2009).
1.5.3 The Subcommisurral organ
The subcommissural organ (SCO) is composed of pseudostratified ependymal cells that differentiated from the neuroepithelium lining the RP of the prosomere 1 within the diencephalon (Meiniel, 2007; Pérez-Fígares et al., 2001; Rodriguez et al., 1998). It is a special member of the CVO as its neighbouring blood vessels do not possess fenestrated wall. Thus, the SCO is the only CNS organ that possesses both functional BBB and BCSFB and is restricted in material exchange with the circulation as well as the CSF (Duvernoy and Risold, 2007; Rodriguez et al., 1998).
The mammalian SCO can be morphologically divided into two layers, the ependymal layer lining the apical surface and the hypendymal layer at the dorsal aspect of SCO. SCO cells are highly polarised. In fact SCO nuclei undergo basal migration during development. In the mature SCO, nuclei are found basally, while their elongated cell bodies extend towards the apical membrane with clear zonation. In summary of the numerous ultrastructural studies of SCO cells, Rodriguez et al. (1998) divided the SCO cell body into four zones: the perinuclear region, the intermediate region, the subapical region and the apical region (Figure 1.14). The perinuclear region is clustered by rough endoplasmic reticulum with dilated cisternae. The intermediate region is occupied by central Golgi apparatus surrounded by rough endoplasmic reticulum. The subapical region is filled by microtubules, mitochondria, smooth endoplasmic reticulum and a few secretory granules. The apical region has similar subcellular make up as the subapical region except it is has an abundance of secretory granules, indicating ventricular secretory function (Rodriguez et al., 1998). In addition, large protrusions and cilia from SCO cells into the ventricular cavity can be found at the apical 41
region (Figure 1.14). (Fernandez-Llebrez et al., 1996; Galarza, 2002; Grondona et al., 1994a; Grondona et al., 1994b; Hoyo-Becerra et al., 2006; López-Avalos et al., 1995; Lösecke et al., 1986; Meiniel, 2007; Nualart and Rodriguez, 1996; Pérez et al., 1995; Rodríguez et al., 1986; Rodríguez et al., 1984a; Rodriguez et al., 1998; Sterba et al., 1982).
On the other hand, a single process, termed the basal process, is sent from the SCO cell towards the basal surface in contact with the subarachnoid. Numerous secretory granules and rough endoplasmic reticulum have been observed under electron microscopy at the very basal ending of the basal process, suggesting SCO cells are capable of subarachnoidal and vascular secretion (Fernández-Llebrez et al., 1987; Lösecke et al., 1986; Pérez-Fígares et al., 2001; Rodríguez et al., 1984b; Rodriguez et al., 1998; Sterba et al., 1982). Moreover, SCO is innervated as the basal process forms synapses with axonal ending from the innervating neurons (Figure 1.14). However, due to the scope of this thesis, the different types of SCO innervations and their function in regulating SCO activity will not be discussed.
As suggested by ultrastructural studies, the SCO has a secretory function. Although both hypendymal and ependymal layer possess similar secretory function (Pérez et al., 1995; Rodríguez et al., 1984a; Rodríguez et al., 1984b; Rodriguez et al., 1998), there are compositional differences in the basal/leptomeningeal secretion and the apical/ventricular within SCO cells. Immunocytochemical and lectin binding experiments revealed that basal secretory granules share similar positive lectin binding properties and immunoreactivities to those of apical secretory granules in some vertebrates (Fernández-Llebrez et al., 1987; López- Avalos et al., 1995; Lösecke et al., 1986; Peruzzo et al., 1990; Rodríguez et al., 1984a; Sterba et al., 1982). It is still not known the exact composition of the basal secretion. However, it is clear that SCO basal secretion is soluble in CSF as no precipitated material has been observed. Based on the similar lectin binding properties and immunoreactivities, it is possible that the basal secretion contains glycosylated-protein(s) similar to those in apical secretory granules as described below in section 1.5.5.
1.5.4 Origin and development of the murine SCO
The SCO primordium arises from the diencephalic RP neuroepithelium and its development is relatively early during neural development in comparison to other CNS innervations
Nucleus RER Perinuclear
Golgi Intermediate RER
SER Microtubules Subapical ISG Mitochondria
Apical MSG Soluble RF Ventricular protrusion
Figure 1.14: Ultrastructure of a mature SCO ependymal cell. The SCO cell body is divided into four zonations from basal to apical as follows: the perinuclear, intermediate, subapical and apical region. Soluble RF is secreted from the mature secretory granules into the ventricle at the apical zone. SCO cell has ventricular protrusions at its apical pole. SCO cell can be innervated by other neurons at the basal process. BP: basal process. ISG: imma- ture secretory granule. MSG: mature secretory granule. RER: rough endoplasmic reticulum. SER: smooth endoplasmic recticulum. Top to bottom: basal to apical. Figure from Rodriguez et al. (1998). 42
domains. In fact, SCO cell proliferation is at its peak from 10.0 dpc to 13.0 dpc (Rakic and Sidman, 1968). Interestingly, at 11.0 dpc, the SCO neuroepithelium is still indistinguishable from its neighbouring RP neuroepithelium. However, bundles of axonal tracts, termed the posterior commissure (PC) have formed immediately dorsal to the thinner ependymal of the SCO primordium, and allows the SCO primordium to be easily identified at this early stage of development (Figure 1.15). As development progresses, the SCO invaginates ventrally. This is also accompanied by the ventral expansion of the PC, due to the increase in its collection of axonal bundles. At 12.0 dpc, SCO primordium appears as an invaginated dorsal neuroepithelium feature within the prosomere 1 of the diencephalon. At this early stage, nuclei of SCO progenitors reside more or less equally along the basal apical axis, similar to its neighbouring neuroepithelial cells (Figure 1.15). Also, the SCO neuroepithelium display no difference to its neighbouring RP neuroepithelium.
From 13.0 dpc, the SCO subsides from proliferation (Rakic and Sidman, 1968) and initiates differentiation. The exact time of SCO differentiation initiation has not been established. However, it is clear that SCO differentiation initiates before 14.5 dpc, since secretory organelles, ultrastructure characteristic to SCO cells but not the neighbouring neuroepithelial cells, and SCO-spondin, a specific secretory material by the SCO, are already present in some SCO cells (Baas et al., 2006; Rakic and Sidman, 1968). Morphologically, the SCO invagination becomes more prominent by 13.0 dpc. In addition, an early pineal recess (PR), a recess immediately anterior to the SCO, and an early mesocoelic recess (MR), a recess posterior to the PR and intercepts AP domain of the SCO, are evident. Between 14.0- 15.0dpc, the SCO morphology becomes well-defined with clear PR and MR (Figure 1.15). At this stage, the SCO cells start becoming more elongated and densely packed. Simultaneously, some of the SCO nuclei have migrated basally in order to achieve the pseudostratified ependymal morphology of a mature SCO (Figure 1.15). These are in contrast to its neighbouring RP cells, where the differentiated cells become flattened and sparsely packed with no nuclei localisation (Figure 1.15). As SCO differentiation progresses at 16.0-16.5 dpc, there are more SCO cells start expressing SCO-spondin and other secretory proteins. At the same time, the SCO morphology becomes more defined, and it finally obtains its adult shape by 17.5 dpc (Baas et al., 2006; Rakic and Sidman, 1968). By 18.5 dpc, all SCO cells are capable of expressing secretory material. Moreover, the basal localisation of SCO nuclei is evident across most of the SCO when analysed histologically (K. Lee, unpublished data). 43
An external contribution of cellular lineage is frequently observed within the developing CNS. One of the examples is the tangentially migrated GABAnergic interneurons, which arise from the medial ganglionic eminence (MGE) of the ventral telencephalon and migrate to the dorsal telencephalon to make up the future CCx (Zaki et al., 2003). Thus, one may poses a question of whether there is any external lineage of SCO cells. Indeed, Rakic and Sidman (1968) has used tritiated thymidine to study an extensive range of developmental timepoints and found no evidence of inbound or outbound progenitor migration during SCO development. They proposed SCO cells are likely to arise locally and like many other neuroepithelium, SCO cells undergo radial migration for proliferation (Rakic and Sidman, 1968).
Together, these indicate that SCO cells proliferate and differentiate locally within the neuroepithelium of the diencephalic prosomere 1. In brief, the SCO primordium subsides from peak proliferation and initiates differentiation from 13.0 dpc onwards. As SCO cells differentiate, the neuroepithelial cells elongate and basally relocate their nuclei in order to form the pseudostratified ependymal layer of the mature SCO. SCO secretory material is expressed in some differentiated SCO cells from 14.5 dpc. The SCO becomes more differentiated and by 18.5 dpc, all SCO cells are differentiated and expressing SCO secretions.
1.5.5 SCO and its role in RF production
Reissner, an anatomist at the University of Dorpat, first identified a long fibrous structure with approximately 1.5 μm in diameter that extends from the SCO along the Aq into the 4V, and then into the spinal cord central canal of the river lamprey in 1860. This observation was confirmed in 1866 by Kutschin and this structure was termed the Reissner’s fibre (RF) (Galarza, 2002; Meiniel, 2007; Meiniel and Meiniel, 2007; Rodriguez et al., 1998). Immunohistochemical studies found polyclonal antibodies that were raised against bovine spinal cord canal RF cross-reacted with the SCO in most vertebrates (Grondona et al., 1994a; Grondona et al., 1994b; Hoyo-Becerra et al., 2006; Nualart and Rodriguez, 1996; Pérez et al., 1995; Rodríguez et al., 1984a; Sterba et al., 1982; Sterba et al., 1981). This implicated the SCO as the source of RF secretion. A number of lectin binding and immunocytochemical studies have identified that the RF material secreted by SCO is mostly glycosylated protein Stage (dpc)
10.0
Diencephalon Mesencephalon
11.0
12.0 * *
14.0
PR MR SCO
15.0
PR MR
16.0 PR
MR
Figure 1.15: The ontogeny of mouse SCO development. SCO primordia (dotted area in 10.0-12.0 dpc; dotted area with short underlying vertical lines in 14.0-16.0 dpc) is evident at 10.0 dpc as the neural epithelium immediately ventral to and intermingled with the PC (white tear-drop shapes). Ventral evaginations (asterisk) are apparent by 12.0 dpc. SCO nuclei undergo basal migration during development such that nuclei basal localisation (dotted area) with apically positioned cell bodies (short ventral lines under the dotted area) are apparent in some SCO cells from 14.0 dpc. Both PR and MR are evident from 13.5 to 14.0dpc, and become more distinct by 15.0 dpc. As SCO differ- entiation further progresses at 16.0 dpc, the SCO ependymal layer becomes thicker, while the PR and MR become more well defined. Sagittal view. Top to bottom: dorsal to ventral. Left to right: anterior to posterior. Figure from Rakic and Sidman (1968). 44
(glycoproteins) with a high level of cysteine and sialic acid residues, which gives these glycoproteins a negative charge (Meiniel and Meiniel, 1985; Rodriguez et al., 1998).
Comparisons between bovine and chick extracts from the SCO, the spinal cord RF and the CSF extracts by immunoblot and immunocytochemical analyses suggested that the thread-like RF that extends along the Aq into the 4V and the spinal cord canal, termed RF- proper, is made up of mainly two processed macromolecules, i.e. one with a higher molecular weight of 450kDa and the other with a lower molecular weight of 190kDa. The two macromolecules are believed to arise from their respective precursors of 540kDa and 320kDa such that upon proteolytic digestion they polymerise through their cysteine residues to form the RF-proper. (Cifuentes et al., 1996; Creveaux et al., 1998; del Brío et al., 2000; Gobron et al., 1999; Nualart and Hein, 2001; Nualart et al., 1991; Nualart and Rodriguez, 1996; Pérez et al., 1995). A handful of polypeptides with various sizes have also been identified from the CSF and the SCO extracts from different developmental stages and animal species. The function of these various soluble peptides and their possible role in the formation of the two macromolecules that make up the RF-proper is still yet to be elucidated (Didier et al., 1995; Grondona et al., 1994a; López-Avalos et al., 1995; Rodriguez et al., 2001; Vio et al., 2008).
SCO-spondin (SSPO) is the only gene identified that encodes one of the many RF peptides. It was firstly isolated from the screening of bovine SCO cDNA as a 2.6kb transcript. SCO-spondin possess a thrombospondin type I repeats and a low density lipoprotein receptor type A repeats, which both are characteristic domains for extracellular matrix (Gobron et al., 1996). Sequence analysis showed that SSPO is highly conserved amongst vertebrates including mouse and human. Moreover, SCO-spondin expression is also evident in a number of vertebrates (Creveaux et al., 1998; Gobron et al., 1999), including the mouse SCO (Baas et al., 2006). In agreement with SCO-spondin being a structural component of RF, its expression in mouse persists from embryonic into postnatal stage and then adulthood (Baas et al., 2006; Gobron et al., 1999; Goncalves-Mendes et al., 2003).
1.5.6 Human SCO
Unlike the murine, bovine and a number of other vertebral SCO, human SCO regresses from the late prenatal stage such that a function SCO is no longer present in adult 45 humans. Investigation of a series of human foetal brains, however, has discovered a fully developed and functional SCO with visible apical secretory granules during 3-5 months of gestation. However, the SCO in a 9-month old human foetus was shortened in its height in both the ependymal and hypendymal layer. By 1-year of age, cuboidal SCO cells were sparse and intermingled with islets of cells with irregular morphology. Between 2-9 years of age, SCO cells gradually receded to only surrounding the boundaries of the SCO. Finally, in adult humans, islet of cuboidal cells was no present although remanent of SCO ependymal cells were evident at the very apical side of the dorsal 3V. Interestingly, this SCO regression was also reported in anthropoid apes (chimpanzee) and bats (Castañeyra-Perdomo et al., 2004; Castaneyra-Perdomo et al., 1985; Galarza, 2002; Rodriguez et al., 2001; Rodríguez and Caprile, 2001).
Interestingly, SCO secretion from human does not appear to be conserved with other animal models, since an antisera raised against the bovine RF-proper crossreacts with the SCO of a number of vertebrates except human, bats and anthropoid apes (Rodríguez et al., 1990; Rodríguez et al., 1984a; Rodriguez et al., 2001). The insoluble thread-like RF-proper was also not observed in human, anthropoid apes and bats. Interestingly, antibodies raised against the SCO extract from two 3 to 4-month human foetuses has identified a 500kDa glycoprotein from the human foetal CSF (Rodriguez et al., 2001). Further study found that the secretion recognised by these antibodies is abundant within the rough endoplasmic reticulum and secretory granules of human foetal SCO cells. Moreover, similar to the glycoproteins that make up the RF-proper in other chordates, this 500kDa glycoprotein is rich in sialic acid residues. In addition, it shares similar lectin-binding profile(s) with the RF material in other chordates, indicating they have similar carbohydrate side chain(s) (Rodriguez et al., 2001). The author thus speculated that the lack of cross-reactivity between the anti-bovine RF sera and the human glycoprotein is due to the difference in the polypeptide backbone possibly caused by evolutionary divergence (Rodriguez et al., 2001). Thus, human SCO, although regresses from late embryogenesis, appears functional during early embryonic development, as evident by its secretory function of high molecular glycoprotein(s). However, the exact role of SCO and its secretion during early human embryogenesis is yet to be elucidated.
1.5.7 Hydrocephalus – Congenital hydrocephalus 46
Hydrocephalus is the abnormal accumulation of CSF within the CNS. Individuals affected by hydrocephalus often display mild to severe dilated brain ventricles due to increased hydrostatic pressure generated by the excessive build up of CSF. There is evidence that excess CSF retention results in cell damages and altered cell behaviour by increased physical stretch and modified CSF biochemical contents. This is especially severe in periventricular axons and ependymal cells due to their proximity to the ventricles (Del Bigio, 2010; Del Bigio and Zhang, 1998; Fukushima et al., 2003; Huh et al., 2009; Mao et al., 2006; Meiniel, 2007; Pérez-Fígares et al., 2001).
Hydrocephalus can be a mechanistic consequence of one or a combination of either an overproduction of CSF, a blockade of CSF flow within the ventricular system by physical obstruction and/or an insufficient drainage of CSF in the subarachnoid space. Hydrocephalus associated with a blockade of CSF ventricular flow, which often happens within the narrow Aq of the brain, is termed non-communicating hydrocephalus. On the other hand, hydrocephalus that has no CSF flow obstruction is classified as communicating hydrocephalus. Hydrocephalus can have either congenital or acquired origin. Acquired hydrocephalus (AH) arises after CNS development. Some AH does not have a genetic basis. For example, individuals contracting mumps encephalitis led to damages to the CNS ventricular ependymal lining and hence, causing non-communicating hydrocephalus. (Galarza, 2002; Huh et al., 2009; Overholser et al., 1954; Pérez-Fígares et al., 2001; Zhang et al., 2006). Alternatively, congenital hydrocephalus (CH), as its name suggests, appears in the congenital period. CH often has a genetic basis and is a consequence of improper development of CNS domain(s) that are responsible for CSF regulation. Yet, evidence from rat model indicates that a lack of folic acid and vitamin B12 in maternal diet also causes CH. Interestingly, if CH develops before the closure of calvarial bone sutures, the excessive accumulation of CSF always leads to an enlarged cranium and prevents the fusion of calvarial bones during postnatal development, resulting in a dome-shaped head that is characteristic to CH. Thus, the presence of dome-shaped cranium often suggests a prenatal or early postnatal onset of hydrocephalus and is then termed overt CH.
Relationship between SCO and CH
The frequent coexistence of SCO defect and Aq stenosis (non-communicating hydrocephalus) in CH mice, as well as the close proximity of the SCO and RF structures to 47 the Aq have prompted Overholser et al. (1954) to propose that SCO secretion, in particular RF-proper, is important for the maintenance of Aq patency. Moreover, studies in most mouse or rat CH models displayed restricted or obstructed Aq with SCO dysplasia (Blackshear et al., 2003; Bruni, 1998; Dietrich et al., 2009; Jones and Bucknall, 1988; Jones et al., 1987; Overholser et al., 1954; Pérez-Fígares et al., 2001; Rodriguez et al., 2007; Takeuchi et al., 1987; Takeuchi et al., 1988; Takeuchi and Takeuchi, 1986). In fact, a definitive link of the loss of normal SCO function as the primary cause of CH was established in the study when Vio et al. (2000) demonstrated new born pups suffering from non-communicating CH as a result of maternal transfer of anti-RF antibodies during gestation in rats (Vio et al., 2000). This link was further strengthened by the observation of a severely dysmorphic SCO with reduced anti-RF immunoreactivity in H-Tx hydrocephalic rats preceding the onset of CH and complete Aq obstruction (Somera and Jones, 2004). Moreover, two human foetuses suffered from CH displayed diminished SCO while one of them was morphologically defective in comparison to two non-hydrocephalic human foetuses (Castaneyra-Perdomo et al., 1994). Collectively, these data suggest that the SCO is required for the regulation of proper CSF flow by maintaining the patency of the Aq.
1.5.8 CH in human – L1CAM
CH has an incidence of 1-3 per every 1000 live births and has a mortality of about 50% (Huh et al., 2009; Pérez-Fígares et al., 2001). However, only one gene, L1 cell adhesion molecule (L1CAM), has been linked to CH in human. L1CAM encodes a transmembrane adhesion glycoprotein (L1 protein) that belongs to the immunoglobulin superfamily. Various mutations that lead to the loss of L1CAM function have been implicated in X-linked hydrocephalus within a wide spectrum of clinical manifestations. L1 is composed of 6 immunoglobulin-like domains and 5 fibronectin III-like domains in its extracellular portion. It anchors on the cell surface by spanning the cytoplasmic membrane with one stretch of transmembrane domain, which is the followed by a C-terminal intracellular tail. Cell surface L1 proteins undergo either homodimerisation or heterodimerisation to initiate intracellular signalling, which ultimately promotes cell-cell adhesion, neurite outgrowth, axonal growth and fasciculation, as well as neuronal migration (Fransen et al., 1995; Fransen et al., 1996; Jackson et al., 2009; Jouet et al., 1993; MacFarlane et al., 1997; Okamoto et al., 2004; Okamoto et al., 1997; Weller and Gartner, 2001; Yamasaki et al., 1997). 48
Two independent L1cam null mouse models for human L1CAM related disease have been established. Both of them phenocopied the human condition and display CH with incomplete penetrance, along with axonal growth disruption and a failure of axonal midline crossing at various parts of the CNS (Cohen et al., 1998; Dahme et al., 1997; Rolf et al., 2001). Ultrastructural study of these hydrocephalic mice indicated normal ventricular ependymal lining. This is in contrast to the CH mouse models caused by mutation in genes that are involved in maintaining cell-cell adhesion and junctional integrity (discussed in 1.5.9). Moreover, the SCO development in these L1cam null mice did not appear to be affected. As a result, the molecular mechanism that underlies the CH in L1cam null mouse models is still unknown. Recent study to identify modifier genes for hydrocephalic phenotype in L1cam null mouse models has identified a strongly linked gene, L1cam hydrocephalus modifier 1 (L1hydro1), residing on chromosome 5 (Tapanes-Castillo et al., 2010). Further investigation of the molecular function of L1hydro1 may help to elucidate the molecular basis underlying the pathogenesis of L1cam-related hydrocephalus.
1.5.9 Genetic and molecular etiology of CH in mouse models
In addition of L1CAM, there are a handful of mouse models demonstrating CH as a result of mutations in genes not yet identified in human. The rapid breeding, the ease of genetic manipulation, the good availability of molecular tools and the similarity in physiological pathogenesis of CH to human allow thorough investigation of the CH in mouse models. Identification of novel genes involved in murine CH may help to shed light in the search of genetic basis in human CH as well as to expand our very limited pool of knowledge in both the genetic and molecular pathogenesis of CH.
Laminar CSF flow requires the uniform beating of cilia on the apical surface of the ventricular lining ependyma, as well as a clear passage through the ventricular system (Fliegauf et al., 2007; Pérez-Fígares et al., 2001). Moreover, adhesion molecules are essential to maintain tight junction integrity of the BCSFB in order to protect the CNS from external perturbation and to maintain CSF homeostasis (Engelhardt and Sorokin, 2009; Huh et al., 2009). Thus, it is not surprising that the primary cause of CH in a number of mouse models results from the failure of ciliogenesis, absence of functional adhesion molecule(s) and/or disrupted SCO development/function. In fact, Huh et al. (2009) has categorised these murine 49
CH mutations into four groups based on the cellular function of these mutated genes, i.e. cell adhesion defects, cilia structure/function defects, signal transduction defects and developmental defects (Huh et al., 2009). The following sections describe these categorised CH defects in various mouse models. Discussion of CH-causing developmental defect will focus on the importance of proper SCO development.
Cell adhesion defects
One of the phenotypes observed in human CH is denudation of ependyma (Dominguez-Pinos et al., 2005). Besides L1cam, there are a number of CH mouse models that display an ependymal defect and are associated with mutation in genes implicated in cell adhesion. Myosin, heavy polypeptide 10, nonmuscle (Myh10) (formerly known as Nmhc-IIb), a gene encoding the heavy chain for a motor protein complex, nonmuscle myosin, is responsible for cell migration, neurite outgrowth and cell adhesion. CH was readily detected in mutant mice with Myh10 ablation (B-/B-) or with reduced expression of the motor-impaired MYH10 (BCN/BCN). Although the development of the SCO appeared normal, denudated ependyma with underlying neural cells protruding into the ventricles were evident as early as 12.5 dpc in these mice. In addition, both mouse models lacked spinal cord lumen, causing restricted CSF drainage and was believed to be a major cause of CH. Detailed immunofluorescence study revealed that MYH10 forms a transient mesh-like structure at the apical boarder of the ventricular lining ependyma within the spinal cord from 8.5 dpc. This apical border expression of MYH10 overlaps with those of other adhesion molecules, N- cadherin and β-cadherin. Thus, it was proposed that MYH10 is required for the proper adhesion and hence, the integrity of the developing ependyma during spinal cord lumen formation. To further support the loss of MYH10 adhesion function rather than motor function as the cause of CH, elevated expression of a motor-impaired MYH10 has successfully rescued the CH phenotype but not the defective neuronal migration phenotype in the mutant mice (Ma et al., 2007; Tullio et al., 2001).
The CH mouse model hydrocephalus with hop gait (hyh) shares similar morphological pathogenesis to that of B-/B- and BCN/BCN, except that Aq stenosis was also found at late postnatal stage when severe hydrocephalus developed in hyh (Perez-Figares et al., 1998; Wagner et al., 2003). The gene responsible for CH phenotype in hyh mice, N- ethylmaleimide sensitive fusion protein attachment protein alpha (Napa), encodes the alpha 50
subunit that binds to the signal transducers and activators of transcription (SNARE) protein complex, which is important for membrane fusion during intracellular vesicular trafficking. The hyh mouse harbours a hypomorphic missense mutation in Napa which causes an abnormal localisation of the adhesion molecules, CDH1 and CTNNB1. Moreover, ependymal denudation is apparent in hyh mice before the onset of CH. Together, these suggest that the CH in hyh mice is a consequence of the loss of adherence junctions and thus, integrity of the ependymal cells. (Batiz et al., 2006a; Batiz et al., 2006b; Chae et al., 2004; Hong et al., 2004)
Another mouse model that is related to adherence junction stability is the Disc, large homolog 5 (Drosophila) (Dlg5) null mouse model (Dlg5-/-). Dlg5-/- mice display kidney and brain defects due to a lack of epithelial tube maintenance. Moreover, they display severe CH at postnatal day (P) 0 with deterioration of ependyma, especially along the Aq, as well as Aq obliteration. Dlg5 encodes a membrane associated guanylate kinase that is important for the localisation and stabilisation of cadherin-catenin junction through SNARE protein complex interactions (Nechiporuk et al., 2007).
Cilia structure/function defects
Cilia are thin rod-like protrusions from a cell. The core structure of a cilium is supported by microtubules that are connected with motor proteins dynein and kinesin for movement. Cilia can be found on lung and kidney epithelia, as well as on the apical side of the CNS ventricular lining ependyma. Ependymal cilia have a core made up with microtubules arranged as a 9+2 axoneme, i.e. 2 microtubules forming the central pair with 9 outer doublet microtubules surrounding the central pair. These microtubules are held in place by connective protein link nexin and various radial spokes, and are attached by motor protein complex, dynein and kinesin, in order to confer motility (Figure 1.16). The beating of a cilium is initiated by a forward exertion for an effective stroke followed by backward retraction for a recovery stroke. It is this synchronised beating of cilia that enables the repulsion and thus movement of the CSF (Fliegauf et al., 2007; Lechtreck et al., 2008).
Several mouse models with an absence of ependyma cilia formation have convincingly demonstrated the critical function of cilia in CSF regulation. In order to initiate cilium development, a basal body, termed centriole, needs to migrate and then dock on the cellular membrane. The docking of the centriole leads to a series of downstream events that recruit activators for ciliogenesis. Forkhead box J1 (Foxj1) (formally known as Hfh-4) 51
encodes a winged-helix/forkhead transcription factor highly expressed in epithelial cells, including the CNS ependyma during late embryogenesis and adult (Blatt et al., 1999). Foxj1 null (-/-) mice were born with communicating CH. Ependymal cells appeared normal, although there was an absence of cilia formation suggesting the requirement of Foxj1 in ciliogenesis. Foxj1 null (-/-) mice appeared to have a deficiency in centriole migration or docking onto the membrane as centrioles were scattered throughout the cytoplasm rather than docked underneath the cell membrane (Brody et al., 2000; Chen et al., 1998).
In addition to formation of cilia, defective motility function is another major cause of cilia related-CH. The spontaneous hy3/hy3 CH mouse model has single nucleotide deletion within the coding region of hydrocephalus inducing (Hydin), leading to a premature stop codon. Hydin encodes a protein responsible for the formation of a couple of central spokes within the axoneme. Although ependymal cilia formation appeared normal in hy3/hy3, the absence of Hydin-encoding central spokes caused inefficient beating of the cilia (Davy and Robinson, 2003; Lechtreck et al., 2008; Lechtreck and Witman, 2007). Interestingly, deletion of a HYDIN-containing region on chromosome 16 was associated with hydrocephalus in a human patient (Callen et al., 1990). Besides, Dynein, axonemal, heavy chain 5 (Dnahc5) (formally known as Mdnah5) encodes one of the heavy chains for the motor protein complex, dynein. DNAHC5 is expressed in the ependyma of the CNS ventricles and Aq. Dnahc5 deficient mice displayed non-communicating CH caused by CSF flow obstruction with a morphologically normal SCO (Ibanez-Tallon et al., 2002; Ibanez-Tallon et al., 2004). Similarly, cilia from respiratory epithelial of DNA polymerase λ (Pol λ) null mice (Pol λ-/-) lacked the inner dynein arms between the nine surrounding doublet microtubules and display CH. Yet, the exact mode of action of Pol λ in ciliogenesis is still not known (Kobayashi et al., 2002).
Moreover, Sperm associated antigen 6 (Spag6) deficient mice suffer from CH. Spag6 is the murine orthologue of PF16 central pair associated protein (PF16) in Chlamydomonas reinhardtii (C. reinhardtii). PF16 is responsible for the maintenance of axonemal central pair integrity in C. reinhardtii. Interestingly, sperm analysis suggested Spag6 may possess similar central pair maintenance property in mice, since a normal axoneme with a central pair was found in the immature testicular sperms but not in the mature sperms from the Spag6 deficient mice. Thus, the development of CH in Spag6 deficient mice might be a result of the loss of cilia function due the lack of axonemal central pair maintenance (Sapiro et al., 2002).
NOTE: This figure is included in the print copy of the thesis held in the University of Adelaide Library.
Figure 1.16: Schematic diagram indicating the ultrastructure of the core of a cilium. The ependymal cilium core is organised as a 9+2 axoneme, with a central pair of mictortubulues residing in the centre surrounding by 9 outer doublets. Inner and outer dynein arms, nexin links and radial spoke are present to hold the axoneme in place. Figure from Dawe et al. (2007). 52
Furthermore, mice deficient for primary ciliary dyskinesia protein 1 (Pcdp1) displayed CH. It is still not clear the exact function of Pcdp1 encoding protein. However, Pcdp1I is highly expressed in respiratory epithelial and ependymal cilia, as well as sperm flagella, suggesting its potential involvement in axoneme assembly, maintenance or function. In fact, Pcdp1 shares 22% homology with Hydin and was thus, speculated to have similar function in structural support for the central pair (Lee et al., 2008).
Lastly, CH was also found in mice with deficiency in Intraflagella transport 88 homolog (Chlamydomonas) (Ift88). Ift88 encodes a component of a complex that is required for the intracellular transport between the base and the tip of the cilia. Loss of Ift88 causes retarded cilia beating in ependymal cells. Moreover, Ift88 is required for the proper function of ChP cilia because disruption of polarity and elevated intracellular activity were identified in ChP cilia of the Ift88 deficient mice. Further ion analysis displayed possible upregulated ChP activity due to an increased chloride concentration. The authors concluded that the hyperactivity of ChP was likely to be the initiating factor of CH , since CH could be detected as early as P1, which precedes the ependymal cilia development at P7 (Banizs et al., 2005). Yet, the possibility of an exacerbation by the defective ependymal cilia after the initial CH onset could not be excluded. Together, the hy3/hy3, Dnahc5 deficient, Pol λ null, Spag6 deficient, Pcdp1 deficient and Ift88 deficient mouse models indicate that the lack of axonemal assembly, or the loss of axonemal connective or motor proteins lead to disrupted cilia motility and are associated to CH development.
Signal transduction defects
Mutations in signalling molecules in mice have also been shown to cause CH. Suppressor of cytokine signalling 7 (Socs7) is a molecule that suppresses cytokine signalling and is regulated by the Janus family of tyrosine kinase (JAK) and the signal transducers and activators of transcription (STAT) pathways. Its exact function in cytokine signalling is still unknown. It has, however, been implicated in insulin signalling and cellular proliferation (Banks et al., 2005; Krebs et al., 2004). Socs7 null mice develop CH that can be detected from 3-15 weeks from birth. Although a horse-shoed shape SCO with pseudostratified cells was present in hydrocephalic Socs7 null mice, it was morphologically disorganised. Intriguingly, no other histological abnormality was detected in other organs that have been implicated in the regulation of CFS homeostasis (Krebs et al., 2004). 53
Adenylate cyclase activating polypeptide 1 receptor 1 (Adcyap1r1) is a glycoprotein- coupled receptor that is expressed in the RP and FP of the developing CNS. In fact, Adcyap1r1 expression is present in 11.5 dpc and 13.5 dpc SCO according to online brain gene expression atlas (Allen atlas, http://www.brain-map.org/) (Lein et al., 2007). Adcyap1r1 encodes the receptor that preferentially binds to neuropeptide Adenylate cyclase activating polypeptide 1 (Adcyap1). Activation of Adcyap1r1 leads to a series of signalling event including protein kinase A and protein kinase C pathways. Adcyap1r1 has been implicated in a wide range of physiological functions such as the regulation of neurotransmitters and hypothalamic hormones, as well as vasodilatation (Lang et al., 2006). Overexpression of human ADCYAP1R1 in developing murine CNS led to CH. Histological study revealed a diminished SCO in 15.5 dpc affected mouse embryos, possibly through the increased apoptosis and decreased proliferation. Shortened and disorganised cilia were observed at lateral ventricular ependyamal cells and thus, concluded by author to be the main contributing factor to CSF flow disruption (Lang et al., 2006).
Both of these signalling molecules have displayed SCO dysmorphology. However, it is still currently unknown how these signalling molecules lead to the dysmorphic SCO and its associated CH. An important question is that whether these molecules have a role in the development and/or regulation the physiological function of the SCO, e.g. RF secretion. Further studies are required to answer this question and to unravel how these molecules interact with each other during these processes.
Developmental defects
SCO is one of many dorsal midline structures that differentiate from the RP ependyma. As a result, proper RP specification and maintenance is crucial for SCO development. In fact, loss of Msx1 led to disorganised SCO and its closely positioned PC at 15.5 dpc. The lack of Msx1 resulted in the loss of bone morphogenic protein 6 (Bmp6) and Wnt1 RP markers and the misexpression of dorsal markers, Pax6, Paired box gene 7 (Pax7) and Lhx1, within the dorsal diencephalon. Intriguingly, these defects were restricted within the prosomere 1, indicating the Msx1 is only indispensable for the development of RP that gives rise to the SCO. Homeobox msh-like 2 (Msx2) is a closely related homeobox transcription family member to Msx1. Msx1 and Msx2 double null mutant however, had complete absence of Wnt1 RP expression, suggesting the limited RP disruption in Msx1 null 54
was likely to be a consequence of functional redundancy. Not surprisingly, conditional Wnt1 swaying mutant (Wnt1sw/sw) also suffered from CH and lacked a proper SCO development from 12.5 dpc (Louvi and Wassef, 2000).
Rfx4_v3 +/- heterozygotes displayed CH and SCO agenesis at P0.5. Interestingly, similar to the Msx1 null mutant, SCO was the only RP derived organ that has morphological defect identified despite endogenous Rfx4_v3 is expressed along the entire RP from 8.5 dpc to 12.5 dpc (Blackshear et al., 2003). Thus, it was suggested that the prosomere 1 RP is more prone to developmental perturbation in comparison to the rest of the RP. Regulatory factor X, 3 (Rfx3), a closely related member of Rfx4 is also associated with CH in a deficiency mouse model. Affected individuals displayed severely disorganised SCO with a lack of SCO- spondin expression at 14.5 dpc. In addition, ChP of these mice appeared underdeveloped (Baas et al., 2006).
Pax6 is expressed in the RP of prosomere 1 and prosomere 2 as early as 12.5 dpc (Lein et al., 2007). Pax6 spontaneous mutant small eye homozygous Sey/Sey displayed CH at 12.5 dpc, with the loss of SCO and the pineal gland (PG), a RP derivative from prosomere 2. Again, diencephalic ventral features appeared histologically normal (Estivill-Torrus et al., 2001). In vitro study of Sey/Sey cortical neurons demonstrated Pax6 inhibits cell cycle progression (Estivill-Torrus et al., 2002). Interestingly, the proper regulation of cell-cycle progression is essential during RP development as loss of Lmx1a, a gene expressed during RP induction, resulted in the failure of RP cells to withdraw from proliferation and thus, a lack of RP derivatives development. (Chizhikov and Millen, 2004b). Hence, the SCO defect of the Sey/Sey mice may be a consequence of a perturbed cell cycle regulation in the absence of Pax6 function.
Similary, endogenous engrailed 1 (En1) is a marker for mes-metencephalic region. Ectopic En1 expression in the RP completely ablated both SCO and PG development at 13.5 dpc, possibly due to disregulated cell cycle as observed by an increased RP cell death. Not surprisingly, mice with an ectopic En1 expression within the RP developed CH (Louvi and Wassef, 2000). Conditional ablation of huntingtin (Htt) in RP cells using a Wnt1-Cre deletion mouse strain also resulted in CH with SCO dysgenesis, as well as a possible ChP and midline ependymal maldevelopment. At the moment, there is no clear known function to Htt (Dietrich et al., 2009). As a result, unravelling the role of Htt in SCO development may help to further understand its function. 55
Collectively, the above mouse models further strengthen the link between the abnormal SCO development and CH. A number of these CH mice also displayed other RP defect, e.g. PG, suggesting a general RP defect, and demonstrating that the formation of a normal SCO is dependent on a robust RP development.
Proposed model for the developmental regulation of prosomere 1 RF and SCO
Based on the Msx1, Rfx4_v3, Rfx3 and Wnt1 loss-of-function mouse models and the importance of Wnt and Bmp signalling in the development of RP in other CNS domain, Huh et al. (2009) has proposed a model for the RP development in prosomere 1. Similar to the RP of other CNS domains, Bmp and Wnt signalling are expressed at the dorsal NT upon induction from neighbouring epidermal ectoderm (Figure 1.17). Bmp and Wnt signalling are retained within the prosomere 1 developing RP and activate the expression of downstream transcription factors, Msx1, Rfx4_v3 and Rfx3, in order to inhibit the dorsal lateral markers, e.g. Lhx1 and Pax6, from expressing at the RF (Figure 1.17). In return, Msx1 and perhaps Rfx4_v3 and Rfx3 are required to up-regulate or maintain Bmp and Wnt signalling pathways (Figure 1.17). As a result, a robust expression of these RP genes is required for RF development which then allows the proper derivation of the SCO.
Summary
Briefly, CH is frequently observed in mouse models with mutation in genes implicated in the formation of adherence junctions between ventricular lining ependymal cells, or in the cilia development and/or function. Moreover, CH is detected in mouse mutant with defective prosomere 1 and hence, SCO development. Furthermore, SCO dysmorphology is apparent in CH affected mice that have a loss of function or an overexpression of genes encoding signal transduction molecules. On the other hand, perturbation of ChP polarity and activity has also led to CSF overproduction as well as defective ependymal development in a CH mouse model. In summary, the above evidence suggests that CH is a heterogeneous disorder and can be caused by the functional perturbation of a number of genes within multiple causative CNS domains. Thus, a robust regulation of CSF flow and production is essential to the maintenance of CSF homeostasis.
NOTE: This figure is included in the print copy of the thesis held in the University of Adelaide Library.
Figure 1.17: Molecular mechanism that regulates prosomere 1 development. During prosomere 1 development, Wnt and Bmp signalling are activated and induce the expression of Msx1, Rfx3 and Rfx4_v3. In return, Wnt and Bmp signallings are mutually regulatd by their downstream effectors, Msx1 and Rfx4_v3. The RP also secrets morphogens and other factor to inhibit the expression of dorsolateral markers, such as Pax6 and Lim1, from expressing at the RP. Figure from Huh et al. (2009). 56
1.6 Thesis aims and approaches
SOX3 duplications have been linked to XH patients (Solomon et al., 2002; Woods et al., 2005), suggesting that CNS development is sensitive to SOX3 dosage. In order to elucidate the genetic and molecular mechanism of SOX3 function during CNS development, a Sox3 overexpressing mouse model has been established to recapitulate the condition of these XH patients. Sox3 overexpressing lines were generated by previous members of the Thomas laboratory through pronuclear injection of bacteria artificial chromosome (BAC) clone derived Sox3 transgene construct (Figure 1.18 A). The transgene contains the Sox3 ORF, which is contained on a single exon, and 26kb of upstream and 8kb of downstream flanking sequences. Of the five Sox3 transgenic founders generated with this construct, two displayed a dome-shaped cranium and expanded LV (Figure 1.18 B and C) (P. Thomas, unpublished data).
In order to further investigate this phenotype, an internal ribosome entry site (IRES)- enhanced green fluorescence protein (EGFP) reporter cassette was inserted into the 3’ untranslated region (UTR) of Sox3 transgene (Sutton et al., 2011). Similarly, six out of twenty Sox3 transgenic founders displayed severe dysmorphic dome-shaped cranium with expanded LV and thinning of CCx (Figure 1.18 B and C). These founders died within 6-8 of birth (P. Thomas, unpublished data). Only two transgenic founders, without dome-shaped cranium, were able to be propagated into independent transgenic lines, namely sex reversing (Sr) and non-sex reversing (Nr). These transgenic lines were maintained as single hemizygotes, i.e. Sr/+ and Nr/+. Dome-shaped cranium, expanded LV and cerebral cortical thinning were evident in some Nr/+ and Sr/+ offspring. The frequent observation of the above phenotypes from Sox3 transgenic founders and in a proportion of Sr/+ and Nr/+ mice indicated a strong link between these phenotypes and Sox3 transgene. Thus, this thesis focuses on the characterisation of these phenotypes, which are later defined to be secondary to CH, in relationship to the increased dosage of Sox3 in Nr/+ and Sr/+ mice. The overall aim of this thesis is thus: --
1.6.1 Aim: To understand the molecular mechanism of Sox3 overdosage that underlies the physiological pathogenesis of CH. 57
In order to achieve the above aim, two approaches were employed which are outlined below.
1.6.2 Approach 1: Identification and characterisation of the morphological and cellular defect within the CNS that is responsible for CH related phenotype
Characterisation of CH related phenotypes requires identification of the causative domain(s) within the CNS. Previous literature has demonstrated a strong link between defective SCO development and CH (Baas et al., 2006; Bach et al., 2003; Blackshear et al., 2003; Dietrich et al., 2009; Lang et al., 2006; Louvi and Wassef, 2000). Moreover, endogenous SOX3 expression was observed in the adult SCO by another lab member (N. Rogers, personal communication). Thus, molecular and morphological studies were performed in CNS domain(s) that were previously implicated in CH, with a particular emphasis on the SCO. In order to strengthen the causative role of defective SCO in the CH of Sox3 transgenic mice, transgenic Sox3 expression analysis and embryonic morphological study were performed. Moreover, the implication of Sox3 overdosage in the proliferation, apoptosis and differentiation of the developing SCO has also been interrogated in Sox3 transgenic mice for a thorough understanding of the normal function of Sox3 and its role in the cellular mechanism during SCO development.
1.6.3 Approach 2: Understanding the molecular role of Sox3 in SCO development through identification of its downstream effectors
Sox3 target genes in mammalian CNS development have not been identified. The Sox3 transgenic mouse model thus provides an excellent system to identify endogenous Sox3 downstream effectors within a physiologically relevant in vivo context. Sox3, specifically the dosage of Sox3, is important for normal SCO development as later demonstrated by analysis outlined in section 1.4.2 above. Microarray analysis was performed with microdissected SCO tissue from mouse embryos for an attempt to discover Sox3 downstream target(s). Verification of microarray results was performed using both quantitative and qualitative assays to further explore molecular pathway(s) that were regulated by Sox3 during SCO development. A EGFP
= Sox3 genomic flanking sequences IRES
5’ Sox3 3’ BAC
26kb 8kb 36kb
NOTE: This figure is included in the print copy of the thesis held in the University of Adelaide Library.
Figure 1.18: Sox3 transgenic mouse model displayed LV expansion. A: Schematic diagram of the Sox3 transgenic construct. Note that it is not drawn to scale. B-C: LV dilatation associated with cerebral cortical thinning were apparent in some Sox3 trasngenic founders displaying CH (C). H&E stains. Coronal sections. Top to bottom: dorsal to ventral. Scale bar: 2mm. Figure (B) and (C) from P. Thomas (unpublished data). 58
Chapter 2: Materials and methods
2.1 General Solution(s)
PBS: 136mM NaCl, 2.6mM KCl, 1.5mM KH2PO4, 8mM NaHPO4, pH7.4
2.2 Animal husbandry
All animals were approved by ethics committees, the University of Adelaide (Adelaide, Australia). All mice were housed and kept in the Laboratory Animal Services, the University of Adelaide (Adelaide, Australia) in 12 hour day:night cycle with sufficient water and food under specific pathogen free environment. Animals were sacrificed by cervical dislocation (2 weeks or older) or decapitation (neonates and embryos) for rapid death and to minimise distress to animals.
Generation and genotyping of Sr transgenic line has been described (Sutton et al., 2011). Nr transgenic line was generated with the same protocol as Sr. Both Sr and Nr mouse lines were maintained in C57BL/6 x CBA mixed background.
2.3 Mouse embryos generation and tissue collection
Mouse embryos were generated by either of the following crosses: a) Male Sr/+ x Female WT (+/+) for Sr/+ and WT (+/+) embryos; b) Male Nr/+ x Female WT (+/+) Male for Nr/+ and WT (+/+) embryos; and c) Male Sr/+; +/+ x Female +/+; Nr/+ for Sr/+; Nr/+, WT (+/+), Sr/+ and Nr/+ embryos. Upon collection, embryos were fixed in 4% PFA/PBS, washed 3 times in PBS for 10 minutes. These fixed tissues were then either further processed as frozen tissues for sectioning or as whole mount embryonic brain tissues. 59
For frozen sections, collected embryos were cryoprotected in 30% sucrose, embedded in OCT and stored at -80⁰C. Frozen tissue sections were prepared by Leica CM1900 cryostat. For collection of whole mount 12.5 dpc embryonic brain tissues, each embryo was dissected transversely with a straight incision immediately above the eye. This was then followed by the removal of telencephalic vesicles and surface ectoderm, resulting in the rest of the tissue containing only the diencephalon, mesencephalon and part of anterior rhombencephalon. The dissected tissue was then sequentially washed in 25%, 50% and 75% v/v MeOH:PBS (0.1% v/v Triton) each for 5 minutes. Then the tissue was further washed in 100% MeOH for 5 minutes twice, and kept in 100% MeOH at -20⁰C.
2.4 Genotyping of animals
gDNA was isolated from tail tissue as per High Pure PCR Template Purification kit (Roche, cat. #: 11796828001).
2.4.1 Sox3 transgene genotyping
Sox3 transgene genotyping was as described in Sutton et al. (2011) with the following primers: EGFP Forward and EGFP Reverse; Gapdh short Forward and Gapdh Reverse; and Sry Forward and Sry Reverse.
2.4.2 Sr transgene specific genotyping
Specific genotyping of Sr was undertaken using extracted gDNA with the following primers: Sr Forward and Sr Reverse; and Gapdh long Forward and Gapdh long Reverse. Sr genotyping PCR was performed as: 1 cycle of 95⁰C 2 minutes; 35 cycles of the followings: 95⁰C 30 seconds, 60⁰C 30 seconds and 72⁰C 3 minutes; 1 cycle of 72⁰C 7 minutes; 1 cycle of 25⁰C 5 minutes. 60
2.4.3 Nr trangene specific genotyping
Specific genotyping of Nr was performed with extracted gDNA with the following primers: Nr/WT Forward, Nr Reverse and WT Reverse; and Sry Forward and Sry Reverse. Nr genotyping PCR was performed as: 1 cycle of 95⁰C 30 seconds; 35 cycles of the followings: 95⁰C 30 seconds, 60⁰C 1 minute and 72⁰C 40 seconds; 1 cycle of 72⁰C 5 minutes; 1 cycle of 22⁰C 10 minutes.
2.5 Genome walking
Protocols as described in Genome WalkerTM kits (Clontech Technologies, cat. #: 638904). Briefly, Nr/+ gDNA was fragmented with four restriction enzymes independently and then randomly cloned into bacterial vectors to create four gDNA libraries, which served as templates for later PCR reactions. For mapping of Nr 5’ integration site, four independent PCR were first carried out with the primer pair: AP1 and Sr Forward. The primer sequence of AP1 is complementary to the vector sequences, while the primer sequence of Sr Forward is complementary to the sequence near the 5’ end of the Sox3 transgene, such that it allows the amplification of the junctional sequence across the 5’ end of Sox3 transgene and the bacterial vector. The PCR products were separated by gel electrophoresis (1% w/v agarose). The strongest band for each PCR was excised and purified using QIAquick Gel Extraction Kit (Qiagen, cat. #: 28704) as per manual. Four independent PCR were carried out using the purified PCR products as templates with the nested primer pair AP2 & GSP2 in order to further amplify the junctional sequence. Again, the PCR products were separated by gel electrophoresis (1% w/v agarose), and the strongest band for each PCR was excised and gel purified by QIAquick Gel Extraction Kit (Qiagen, cat. #: 28704) as described in the manual. Each PCR product was then subcloned into pGEM®-T Easy vector system I (Promega, Cat. #: A1360) as per manual. Subsequently, each clone was sequenced with T7 or Sp6 primers independently. Obtained sequences were aligned in CLC Sequence Viewer and analysed using online sequence alignment software BLAST (Altschul et al., 1997). For mapping of Nr 3’ integration site, the above protocol was employed with the Sr Forward primer replaced with 3’ Sox3 transgene v1, and GSP2 primer replaced with 3’ Sox3 transgene v2. 61
2.6 H & E stains
For H & E stains, tissue sections were air-dried for 2 hours at RT, stained in 0.1% w/v haematoxylin for 20 seconds, rinsed in water for 5 minutes, immersed in 0.5% w/v eosin for 15 seconds, rinsed in water for 5 seconds, dehydrated in 50% v/v EtOH followed by 70% v/v EtOH for 10 seconds each, equilibrated in 95% v/v EtOH for 30 seconds, 100% EtOH for 1 minute and histolene for 15 seconds. Tissue sections were then mounted in xylene based DePex mounting medium Gurr® (Merck, cat. #: 36125).
2.7 Nissl stains
Tissue sections were air-dried for 2 hours at RT, stained at 0.1% w/v cresyl violet for 5 minutes, washed in water for 3 minutes. They were then dehydrated in 95% v/v EtOH for 10 seconds, 100% for 10 seconds, equilibrated in histolene for 30 seconds and mounted in xylene based DePex mounting medium Gurr® (Merck, cat. #: 36125).
2.8 qPCR for Sox3 transgene copy number analysis
gDNA was isolated from tail tips as described in section 2.4. qPCR was performed on the extracted gDNA on a ABI 7500 StepOne Plus platform using Fast SYBR Green Master Mix (Applied Biosystems, cat. #4385610) as described in manual. In a single qPCR run, one sample of each genotype was assayed with primers against Sox1, Neurog3 and Sox3 using the following primers: Sox3 Forward and Sox3 Reverse (Sox3); Sox1 Forward and Sox1 Reverse (Sox1); and Neurog3 Forward and Neurog3 Reverse (Neurog3). Three independent sets of gDNA were subjected to qPCR analysis. Neurog3 served as the reference gene. The relative quantity of Sox1 to Neurog3 for WT XY was set as 1 and all other genotypes were normalised to this. No genotypes deviated significantly from WT XY when compared pair- wise using a student’s t-test, illustrating that the technique is capable of accurately measuring gene dosage. The relative quantity for Sox3 versus Neurog3 was then normalised to one for WT XY and the relative quantity of Sox3 for other genotypes was taken as a measure of gene 62
dosage. Estimation of gene dosage did not alter when Sox1 was used as the reference gene instead of Neurog3.
2.6 Immunofluorescence
Immunofluorescence was performed on tissue sections by protocols described previously (Sutton et al., 2011). Antibodies dilutions are: goat anti-SOX3 (R&D systems cat. #: AF2569, 1:100), rabbit anti-EGFP (Abcam cat. #: ab290, 1:2000), rabbit anti-GAP43 (Chemicon® International, cat. #: AB5220), rabbit anti-RF (kindly donated by A. Meiniel, 1:1000), rabbit anti-Ki67 (Novocastra, cat. #: NCL-Ki67p), rabbit anti-activated Caspase-3 (BD pharmingen cat. #: 559565, 1:1000), donkey anti-goat IgG (Jackson ImmunoResearch Laboratories Inc., 1:400), donkey anti-rabbit IgG Cy3 (Jackson ImmunoResearch Laboratories Inc. cat. #: 711-165-152, 1:400), donkey anti-sheep IgG Cy3 (Jackson ImmunoResearch Laboratories Inc. cat. #: 713-165-147), donkey anti-goat Cy5 (Jackson ImmunoResearch Laboratories Inc. cat. #: 705-175-147).
2.7 BrdU analysis
Pregnant mice were administered 50mg/kg BrdU (Sigma cat. #: 59-14-3) at 10mg/ml in PBS intraperitoneally 2 hours before sacrifice. Embryos and tissue were processed as described above. Frozen sections were air-dried for 2 hours then washed with PBS (0.1% v/v Triton) for 10 minutes, denatured in 2N HCl at 37⁰C 30 minutes, neutralised in 0.1M borate buffer at RT, washed 3 times in PBS for 10 minutes and then proceed to the blocking step of immunofluorescence protocols as described in Sutton et al. (2011). Antibodies used are: sheep anti-BrdU (Biodesign International cat. #: M20107S, 1:1000) and donkey anti-rabbit IgG Cy3 (Jackson ImmunoResearch Laboratories Inc. cat. #: 711-165-152, 1:400). Protocols for BrdU images capturing and signal quantification are described here. First, positive controls were used to standardise the microscope and camera before photo capture for each experiment. AxioVision Rel 4.7 was used to obtain fluorescence images at 200ms exposure with gain factor of 1. For each image of the SCO, an image of MGE, a region without transgenic Sox3 expression, from the same section was captured at 200ms as a control for 63 variance in labelling and washing efficiency. Cell counting was performed using ImageJ software (Abramoff et al., 2004) without image manipulation and knowledge of the genotype. The quantity of SCO or MGE cells with positive signal in each section was scored as a ratio, i.e. number of positive SCO cells/number of total SCO cells or number of positive MGE cells/number of total MGE cells. The ratio of positive SCO cells was then normalised to the ratio of positive MGE cells. Three sections separated at an interval of 20μm were counted per SCO. All counting was repeated three times and an average was taken. Two-tailed unpaired student T-test was performed for statistical analysis.
2.8 TUNEL assay
Embryos and tissue were collected as described above. TUNEL assay was performed on frozen sections using In Situ Cell Death Detection Kit, TMR red (Roche Applied Science, cat. #: 12 156 792 910) with protocols as described by manual. Tissue sections were then mounted in Prolong® Gold Antifade Reagent with DAPI (Invitrogen, cat. #: P-36931).
2.9 In Situ Hybridisation
2.9.1 Template plasmid production
Template plasmid production was performed for the following genes: Sox14, Gata3 and Lhx5. For each gene, a fragment of about 400-700bp was amplified with the following primers: Sox14 long Forward and Sox14 long Reverse (Sox14 ); Gata3 long Forward and Gata3 long Reverse (Gata3 ); and Lhx5 long Forward and Lhx5 long Reverse (Lhx5). PCR products were separated by gel electrophoresis (1% w/v agarose), then excised and gel purified by QIAquick Gel Extraction Kit (Qiagen, cat. #: 28704) as per manual. Subsequently, purified DNA fragments were independently cloned into pGEM®-T Easy vector system I (Promega, Cat. #: A1360) as described in manufacturer’s manual. 64
2.9.1 Riboprobe synthesis
Anti-sense riboprobes were synthesised as described below. First, plasmid templates were prepared by linearising non-expressing plasmids with appropriate restriction enzymes (see 2.14 for plasmid table). In order to purify linearised plasmids, each digested sample was mixed with 1 volume of phenol/chloroform and spun at 18,000g for 2 minutes. Top aqueous layer was removed and precipitated with 1/10 volume of 3M NaAc (pH 5.2) and 2 volumes of 100% EtOH at -20⁰C O/N. Sample was then spun at 18,000g for 30 minutes and washed with 70% v/v EtOH before resuspended in MQ water. Riboprobes were synthesized in-vitro using purified linearised plasmid templates with DIG RNA Labelling Mix (Roche Applied Science, Cat #: 11 277 073 910) according to manufacturer’s recommendations with template at a final concentration of 60ng/μl.
2.9.2 Section in situ hybridisation
Frozen sections were air dried for 2 hours prior to in situ hybridisation. Protocols were as described in Wilson et al. (2005).
2.9.3 Whole mount in situ hybridisation
Whole mount tissue was rehydrated in MeOH series in the following order. First in 75%, then 50% and 25% v/v MeOH:PBS (0.1% v/v Triton) for 5 minutes each. They were then processed as described in Thomas et al. (1998).
2.10 SCO microdissection and RNA extraction
SCO microdissection was performed using Nikon SMZ1000 GFP dissecting microscope. For microarray SCO collection, embryos were stage-matched to 53-55 somites. For qRT-PCR SCO collection, embryos were stage-matched to 52-56 somites. The steps of SCO microdissected was as described in briefly below and in details in Figure 5.1. Mouse 65
embryos were placed laterally and an incision horizontal to AP axis was made above the optic placode. Excised tissue was then placed with ventral side into the page and dorsal side out of the page. A V-shaped cut was made to remove the telencephalon followed by removal of surface ectoderm. A “cup-like” tissue with intact diencephalon and mesencephalon and most dorsal rhombencephalon was left. Small incisions were made at the dorsal midline from both the posterior and anterior end of the cup. This allowed the tissue to lay flat on the surface of the microscope stage like a “book”. PC could be visualised as a collection of a more translucent tracts around the centre of the “book”. SCO was excised by cutting around the anterior, posterior and lateral limits of PC. SCO tissues from four animals of each genotype were pooled for each sample. SCO tissue was temporarily stored in RNAproctect Cell Reagent (Qiagen, cat. #: 76526). They were homogenised by brief sonication and QIAshredder (Qiagen, cat. #: 79654) treatment. RNA was isolated with RNeasy Plus Mini Kit (Qiagen, cat. #: 74134) as described per manual. The integrity and concentration of the RNA was assessed using the Agilent RNA 6000 Nano Kit with Agilent 2100 Bioanalyser, a chip-based capillary electrophoresis method, by the Adelaide Microarray Centre. Only RNA with an RIN number of ≥8 was used for further analysis.
2.11 Microarray data processing
Microarray data was processed either using Affy package from Bioconductor Project (Ihaka and Gentleman, 1996) with the following algorithms: RMA (background subtraction), PMOnly (probe specificity normalisation), and Quantile (data normalisation); or Partek® software with RMA import normalisation algorithm. The normalised data was then subjected to further statistical analysis described in text.
2.12 qRT-PCR analysis for microarray validation
In order to synthesis cDNA, reverse transcription reaction was carried out with Omniscript Reverse Transcription kit (Qiagen, cat. #: 205110) as per manual, with 80ng of
RNA, 5.6μM oligo(dT)12-18 (Invitrogen, cat. #: 18418-012), 1.25μM dN6 (Geneworks, cat. #: RP-6), 10U Proctector RNase inhibitor (Roche Applied Science, cat. #: 03 335 399 001). The 66
qRT-PCR was performed with Fast SYBR Green Master Mix (Applied Biosystems, cat. #4385610) as described on manual, on ABI 7500 StepOne Plus platform. All primers amplify across intron boundaries. Primers used are as follows: Adamts3 Forward and Adamts 3 Reverse (Adamts3); Bmp5 Forward and Bmp5 Reverse (Bmp5); Bmp6 Forward and Bmp6 Reverse (Bmp6); Dab2 Forward and Dab2 Reverse (Dab2); EGFP short Forward and EGFP short Reverse (EGFP); Fibcd1 Forward and Fibcd1 Reverse (Fibcd1); Fmo1 Forward and Fmo1 Reverse (Fmo1); Gapdh short Forward and Gapdh short Reverse (Gapdh); Hmgcs2 Forward and Hmgcs2 Reverse (Hmgcs2); Lhx5 short Forward and Lhx5 short Reverse (Lhx5); Msx1 Forward and Msx1 Reverse (Msx1); Rspo3 Forward and Rspo3 Reverse (Rspo3); Sdha Forward and Sdha Reverse (Sdha); Sox14 short Forward and Sox14 short Reverse (Sox14); Sox3 Forward and Sox3 Reverse (Sox3); Wnt1 Forward and Wnt1 Reverse (Wnt1); and Wnt2b Forward and Wnt2b Reverse (Wnt2b). The relative quantity of each gene was derived from the difference in intensity (ΔCT) in comparison to the house keeping gene, Sdha. Gapdh also served as another endogenous control. Subsequently, all WT measurements were set to 1 and all Sr/+; Nr/+ measurements were normalised to WT for data presentation. Controls for no reverse transcriptase and no template controls were performed for product specificity. Melt curve analysis was undertaken to ensure amplification efficiency.
2.13 Image analysis
All whole mount images were captured by Nikon SMZ1000 GFP dissecting microscope with AnalySIS software. Zeiss Axioplan 2 Imaging upright microscope with Axiovision software was used to obtain immunofluorescence images, while Zeiss Axiophot upright microscope with AnalySIS software was used to obtain colour images of tissue sections. All captured images were processed by Adobe Photoshop CS. 67
2.14 Primers table
Name Primer sequence (5'->3') Use Source
Flanking near 3' of Sox3 3' Sox3 AGTGACACACTGTCGCTTG transgene, used in Genome Primer 3 transgene v1 walking
Flanking near 3' of Sox3 3' Sox3 CAAGGACCCACGTCAAGT transgene, used in Genome Primer 3 transgene v2 walking
Adamts3 GTTTCCATGGCAAAGAGCAT qRT-PCR analysis Primer 3 Forward
Adamts3 TCATTCGTACCAGGACCACA qRT-PCR analysis Primer 3 Reverse
Adaptor primer from Genome Genome AP1 GTAATACGACTCACTATAGGGC WalkerTM kit WalkerTM kit
Nested adaptor primer from Genome AP2 ACTATAGGGCACGCGTGGT Genome WalkerTM kit WalkerTM kit
Bmp5 CTCCATGTGGAATCTGGGTC qRT-PCR analysis Primer 3 Forward
Bmp5 Reverse CATGGTCATGAGCTTTGTCA qRT-PCR analysis Primer 3
Bmp6 GTTCTCCCCACATCAACGAC qRT-PCR analysis Primer 3 Forward
Bmp6 Reverse CCAGCCAACCTTCTTCTGAG qRT-PCR analysis Primer 3
Dab2 Forward TCTTCTGATCCCCACAGGAG qRT-PCR analysis Primer 3 68
Name Primer sequence (5'->3') Use Source
Dab2 Reverse CCAAAGCCAGTCTTTTGCTC qRT-PCR analysis Primer 3
EGFP Forward ATGGTGAGCAAGGGCGAGGAGCTGTT Genotyping of Sox3 transgene P. Thomas
EGFP Reverse CTGGGTGCTCAGGTAGTGGTTGTC Genotyping of Sox3 transgene P. Thomas
EGFP short ACGACGGCAACTACAAGACC qRT-PCR analysis Primer 3 Forward
EGFP short GTCCTCCTTGAAGTCGATGC qRT-PCR analysis Primer 3 Reverse
Fibcd1 CAGCTGGCTTCCAGGTCTAC qRT-PCR analysis Primer 3 Forward
Fibcd1 CCAACCTCGGAAAAAGTTCA qRT-PCR analysis Primer 3 Reverse
Fmo1 Forward CTCAACCCAGTTCAACCTTCA qRT-PCR analysis Primer 3
Fmo1 Reverse CCACTGTCCAGAGACAGCAA qRT-PCR analysis Primer 3
Gapdh long Internal positive control for Sr TGGCTGTGGTAACCATTCATAAGGTAG K. Lee Forward transgene genotyping
Gapdh long Internal positive control for Sr GGTCGGTGTGAACGGGTGAG K. Lee Reverse transgene genotyping
Gapdh short ATGCCAGTGAGCTTCCCGTTCAGC qRT-PCR analysis Unknown Reverse 69
Name Primer sequence (5'->3') Use Source
Internal positive control for Sox3 Gapdh short ACCCAGAAGACTGTGGATGG transgene genotyping and qRT- P. Thomas Forward PCR analysis
Gapdh Internal positive control for Sox3 CTTGCTCAGTGTCCTTGCTG P. Thomas Reverse transgene genotyping
Gata3 long CCGAAACCGGAAGATGTCTA Template plasmid synthesis Primer 3 Forward
Gata3 long GCCCATTTGGACATCAGACT Template plasmid synthesis Primer 3 Reverse
Gata3 short CCCACCACGGGAGCCAGGTA qRT-PCR analysis Primer 3 Forward
Gata3 short CGGTGTGGTGGCTGCTCAGG qRT-PCR analysis Primer 3 Reverse
Flanking near 5' of Sox3 GSP2 CTTCTTTGTGCGGACCCTTATTCCAAG transgene, used in Genome P. Thomas walking
Hmgcs2 GCGGCACAGCCTCCCTCTTC qRT-PCR analysis Primer 3 Forward
Hmgcs2 GGGGCGGGCGTTACCACTC qRT-PCR analysis Primer 3 Reverse
Lhx5 long GTATCTCTCCGAGCGACCTG Template plasmid synthesis Primer 3 Forward
Lhx5 long CCCCAACATCTCAGACTCGT Template plasmid synthesis Primer 3 Reverse
Lhx5 short GTGCGCGAAGAAATCGTAAT qRT-PCR analysis Primer 3 Forward 70
Name Primer sequence (5'->3') Use Source
Lhx5 short ACGAGTCTGAGATGTTGGGG qRT-PCR analysis Primer 3 Reverse
Msx1 Forward GCCCCGAGAAACTAGATCG qRT-PCR analysis Primer 3
Msx1 Reverse TTGGTCTTGTGCTTGCGTAG qRT-PCR analysis Primer 3
Neurog3 CCTCTCAGACGGTGGAGTTATATT qPCR analysis Primer 3 Reverse
Neurog3 ATGTTGTACATAGCCTGCAAGTC qPCR analysis Primer 3 Reverse
Genotyping of Nr specific Nr Reverse GAGTGTTGGAGGGGGTTGAG K. Lee transgene
Genotyping of Nr specific Nr/WT CTGGGTTAGAGAGCAGCATCC transgene/WT chormosome 10 K. Lee Forward sequences
Rspo3 AGTGTGTCTCTCTTCGTGTCCA qRT-PCR analysis Primer 3 Forward
Rspo3 Reverse GCAGAAATTTTTGTTGAAACAGG qRT-PCR analysis Primer 3
Fischer et al. Sdha Forward AACACTGGAGGAAGCACACC qRT-PCR analysis (2003)
Fischer et al. Sdha Reverse GCACAGTCAGCCTCATTCAA qRT-PCR analysis (2003)
Sox1 Forward CCCCAGAGACACAACAACCT qPCR analysis Primer 3 71
Name Primer sequence (5'->3') Use Source
Sox1 Reverse AGTCACCCACTTCTGCTTCG qPCR analysis Primer 3
Sox14 long ACATCGATGAAGCCAAAAGG Template plasmid synthesis Primer 3 Forward
Sox14 long TGCCTGGGAAGAGGATGTAG Template plasmid synthesis Primer 3 Reverse
Sox14 short TATGCCAGTCTTGGTCATGC qRT-PCR analysis Primer 3 Forward
Sox14 short TACGTGGTGCCCTGTAACTG qRT-PCR analysis Primer 3 Reverse
Sox3 Forward GAACGCATCAGGTGAGAGAAG qPCR or qRT-PCR analysis Primer 3
Sox3 Reverse GTCGGAGTGGTGCTCAGG qPCR of qRT-PCR analysis Primer 3
Sp6 ATTTAGGTGACACTATAG Sequencing primer Unknown
Genotyping of Sr specific Sr Forward ACAGCCTTGTGAGTAGGTATGCTCTTG P. Thomas transgene
Genotyping of Sr specific Sr Reverse ACAGCCTTGTGAGTAGGTATGCTCTTG P. Thomas transgene
Sry Forward CACTGGCCTTTTCTCCTACC Genotyping of Sry P. Thomas
Sry Reverse CATGGCATGCTGTATTGACC Genotyping of Sry P. Thomas 72
Name Primer sequence (5'->3') Use Source
T7 GTAATACGACTCACTATAGGGC Sequencing primer Unknown
Wnt1 Forward TACTGGCACTGACCGCTCT qRT-PCR analysis Primer 3
Wnt1 Reverse GAATCCGTCAACAGGTTCGT qRT-PCR analysis Primer 3
Wnt2b CGGGACCACACTGTCTTTG qRT-PCR analysis Primer 3 Forward
Wnt2b ACCACTCCTGCTGACGAGAT qRT-PCR analysis Primer 3 Reverse
Genotyping of WT chromosome WT Reverse GTCCTACTCCCTCAACACCTGTC K. Lee 10 sequences 73
2.15 Plasmids table
Restriction enzyme RNA polymerase used for antisense used for anti- Name of plasmid Vector Gene of interest Source riboprobe sense riboprobe production production
pGEMTEasy-Sox14 pGEMTEasy Sox14 NcoI Sp6 K. Lee
pGEMTEasy-Lhx5 pGEMTEasy Lhx5 NcoI Sp6 K. Lee
pGEMTEasy-Gata3 pGEMTEasy Gata3 AleI Sp6 K. Lee
pYX-Asc-Fibcd1 pYX-Asc Fibcd1 SpeI T3 IMAGE clone ID: 6848779
Kindly gifted from M. Wassef pBluescript-KS+-Wnt1 pBluescript-KS+ Wnt1 HindIII T7 (Bally-Cuif et al., 1995) Kindly gifted from Y. Lallemand pGEMT-Msx1 pGEMT Msx1 BamHI Sp6 & C. Ramos 74
Chapter 3: Characterisation of CH in Sox3 transgenic mouse model
3.1 Introduction
Dorsally extended cranium, which led to disruption of craniofacial morphology and hence, pinna displacement (Figure 3.1 A-B) was frequently observed in Sr/+ and Nr/+ mice. CNS histology from these mice further indicated expanded LV and thinning of CCx (Figure 3.1 C-D), which are classic CH hallmarks (Baas et al., 2006; Bach et al., 2003; Blackshear et al., 2003; Huh et al., 2009; Krebs et al., 2004; Lang et al., 2006; Picketts, 2006). Together they strongly suggested CH (or overt CH in a case with dome-shaped cranium) in these affected animals. However, Sox3 is highly expressed at the developing telencephalon during neurogenesis (Figure 1.11) (Wood and Episkopou, 1999; N. Rogers, unpublished data). Thus, the possibility a defective cerebral cortical development in Sox3 transgenic mice could not yet be excluded, such that the thinner CCx observed might partly be a consequence of CH and partly be a result of the defective cerebral cortical development. Alternatively, the dome- shaped cranium and expanded LV could be solely secondary effects of CH.
In order to test the above possibilities, characterisations of Sr and Nr transgenic mice were carried out. Prior to any expression analysis, the Sox3 transgene integration site(s) were mapped such that PCR amplification across the integration site(s) could be used to genotype animals used in all further expression and molecular studies. One would expect that Sox3 is present within a developmental domain if its development were to be affected in the transgenic Sox3 mouse model. As a result, Sox3 transgene expression was also analysed in the developing telencephalon in order to investigate whether the cerebral cortical defect is a consequence of local transgene expression or a secondary effect of CH. Similarly, Sox3 transgenic expression mapping was also carried out in CH related CNS domain(s), with a particular focus on the SCO. This was then followed by morphological characterisation of the SCO defect and its relationship with CH pathogenesis in Sox3 transgenic mice. Since CH phenotypes were transgene specific in both Sr/+ and Nr/+ mice, Nr/+ mice was used 75
primarily during preliminary stage of expression analysis due to the breeding difficulty in Sr mouse model (E. Sutton and P. Thomas, personal communication).
3.2 Results
3.2.1 Sox3 transgene integration site in Nr mouse model
In order to perform study on the Nr/+ and Sr/+ mouse models, their transgene integration sites were firstly mapped such that Sr and Nr specific genotyping could be performed by PCR amplification across the integration site. Moreover, transgene integration into coding and regulatory sequences can lead to phenotype that is not specific to the integrating transgene but rather is a result of haploinsufficiency or positional effect. It was thus a priority to map Sox3 integration site(s) in order to confirm that any phenotype studied was indeed specific to Sox3 transgene expression. In fact, the transgene integration site has already been mapped to chromosome 19 within the 5’UTR of Aldh1a1 in Sr line (Sutton et al., 2011). It was later found that the proximal integration of Sox3 transgene near Aldh1a1 might cause ectopic expression in the developing gonads that led to the sex reversal phenotype reported (Sutton et al., 2011). The mapping of transgene integration site in Nr mouse model is described below.
In order to map the Sox3 transgene integration site in Nr mice, genome walking, a PCR based technique, was used. Briefly, restriction enzyme-fragmented Nr/+ genomic gDNA was cloned into bacterial vectors in order to create a Nr/+ gDNA library as the PCR template. The mappings of the integration site at the 5’ and 3’ end of the Sox3 transgene were carried out independently. For the mapping of each end, PCR amplifications were performed across the transgene and bacterial vector with a transgene specific primer (complementary to either the 5’ or 3’ end of the transgene) and a vector specific primer. PCR products were then sequenced and aligned with known genomic assembly using Basic Local Alignment Search Tool (BLAST) (Altschul et al., 1997).
Interestingly, there was no evident of amplification across the boundary between two transgenes, suggesting that there is only one copy of the transgene within the integration site Nr/+ Nr/+ overt CH A B P42
P26 C E Cer CCx Hpc LV
WT 3V 4V Cer 4V ChP
D F CCx Hpc LV Cer Cer
4V Nr/+ overt CH ChP
3V 4V
Figure 3.1 Overt CH in single hemizygous Sox3 transgenic mice. A-B: Nr/+ individuals affected by overt CH (B) displayed dorsally raised cranium (black arrows) in comparison to same age Nr/+ mice without overt CH (A). Apparent ventral pinna displacement (white arrowheads) was observed in affected mice (B). C-D: Nr/+ with overt CH (D) demonstrated dilated LV and 3V, thinning of CCx (double-headed arrows), posterior displacement of Hpc and CNS haemorrphages (black arrowheads), but not 4V expansion (boxed). E-F: (E) and (F) are boxed regions of (C) and (D) respectively. H & E stains (C-F). Sagittal sections (C-F). Top to bottom: dorsal to ventral. Left to right: anterior to poste- rior. Scale bar: 2mm (C-D), 500µm (E-F). 76
(further discussed in 3.2.3). In fact, the Sox3 transgene had integrated into chromosome 10qC1. The 5’ end of the transgene is immediately upstream of 85,122,881bp on NCBIm37 NCBI mouse genomic assembly. Subsequent mapping of the 3’ transgene integration site (performed by Chloe Shard) identified an integration-induced deletion of 13bp in chromosome 10 between 85,122,868-85122880bp (NCBIm37). The 3’ end of the transgene was immediately downstream of 85,122,867bp (NCBIm37) (Figure 3.2) (see Appendix I and II for sequences). The integration site fell within the last 200bp of the 3’ UTR within the common last exon of a number of BTB (POZ) domain containing 11 (Btbd11) transcripts (Hubbard et al., 2006) (Fig 3.2 and 3.3). Collectively, these results indicated that Sox3 transgene integration did not result in a significant deletion of endogenous chromosomal sequence at the integration site. In addition, the Sox3 transgene integration site in the Nr mouse model did not disrupt the ORF of Btbd11.
3.2.2 The Sox3 transgene is not expressed in the developing dorsal telencephalon of Nr/+ embryos
Endogenous Sox3 is highly expressed in the telencephalon at 12.5 dpc (Figure 1.11) (N. Rogers, unpublished data). In order to investigate whether there was a primary defect in cerebral cortical development of the Sox3 transgenic mice, Sox3 transgene expression during telencephalic development was firstly analysed in the Nr/+ mice. Previous study demonstrated that cSox3 has a role in neurogenesis through inhibition of neuronal differentiation when overexpressed in the chick NT (Bylund et al., 2003). Since telencephalic neurogenesis initiates around 12.5 dpc in mouse (Zaki et al., 2003), Sox3 transgene expression analysis was conducted at this timepoint. Agreeing with its predicted role as a transcription factor, endogenous SOX3 protein was localised to the nuclei within the developing telencephalon. SOX3 endogenous expression was highest at the VZ of the dorsal telencephalon, the dorsal hippocampus (Hpc) and the lateral ganglionic eminence (LGE) (Figure 3.4 A and G). Endogenous SOX3 was also present in the ChP of the LV (discussed in section 4.2.10), the lamina terminalis, the MGE and the ventral Hpc, albeit with lower intensity (Figure 3.4 A, G and J). In addition, endogenous SOX3 was also observed at a low level in the diencephalic domains of EpTh and VTh (discussed in 3.2.8) (Figure 3.4 G). 77
To detect the expression of transgenic Sox3, an antibody specific to EGFP, a Sox3 transgene reporter, was used. Surprisingly, EGFP expression was not detected within the dorsal telencephalon, MGE or LGE. A very low level of stripy EGFP signal was present in the dorsal and ventral Hpc (Figure 3.4 B, E and H). The strongest EGFP signal was however, detected in the laminal terminalis, ChP of the LV (discussed in 4.2.10), EpTh and VTh (discussed in 3.2.8). In fact, EGFP expression was strongest at regions with low endogenous SOX3 but failed to recapitulate the expression pattern at regions where the endogenous SOX3 expression was high. Furthermore, there was no obvious ectopic expression of the transgenic Sox3 within the telencephalon. As a result, the absence of EGFP and hence, Sox3 transgene expression in the developing dorsal telencephalon that gives rise to the future CCx during neurogenesis suggests that the cerebral cortical thinning in Nr/+ mice was likely to be a secondary consequence of CH rather than a primary defect of dorsal telencephalic development.
3.2.3 CH is dosage dependent on Sox3
Statistical analysis was performed to determine the penetrance of the CH phenotype in Nr/+ and Sr/+ hemizygous, as well as Sr/+; Nr/+ double hemizygous mice. Both Nr/+ and Sr/+ hemizygotes displayed overt CH with incomplete penetrance at 3 weeks of age. In fact, only 18.69% of the Nr/+ and 30.64% of the Sr/+ hemizygotes displayed overt CH (Figure 3.5 A). It is important to note that cases of non-overt CH were not accounted for this statistical analysis. Thus the above figures may underestimate the actual penetrance of CH, which includes both overt and non-overt forms, in Sr/+ or Nr/+ hemizygotes. To determine whether CH is dosage dependent on Sox3, crossing of two single hemizygotes, i.e. Sr/+ and Nr/+, was carried out in order to generate mice that have both the Sr and Nr Sox3 transgene. Interestingly, Sr/+; Nr/+ double hemizygous mice demonstrated a marked increase in overt CH penetrance to 98.36% (Figure 3.5 A). Statistical analysis of Mendelian ratio at 3-week weaning age suggested the presence of pre-weaning lethality in Sr/+ and Sr/+; Nr/+ individuals, since their survival rate was 44.76% and 17.31% respectively. This observed lethality was likely due to the earlier onset and hence, increased overt CH penetrance or exencephaly (discussed later in 5.3.6) in Sr/+ and Sr/+; Nr/+ mice. Sox3 transgene
Chromosome 10 B5.3 Chromosome 10 C2 Btbd11
Figure 3.2: Schematic diagram showing Sox3 transgene integration site on chromosome 10 in the Nr mouse model. The Sox3 transgene has integrated towards the very 3’ end of Btbd11 UTR. Btbd11 neighbouring sequences are in grey. Diagram is not drawn to scale. A
B
Figure 3.3: The integration site of the Sox3 transgene in Nr mouse model. The transgene resides within the last 200bp of the 3’UTR from the last exon of Btbd11. A: Ensembl snapshot indicat- ing various splice variants of Btbd11, which most of them share the same last exon. The region containing transgenic Sox3 integration site is boxed in red. B: Zoomed in boxed region in (A) demon- strating the last Btbd11 exon (green, red and yellow colour boxes). The Sox3 transgene integration site is indicated by the black line. Figure from Ensembl Genome Browser (www.ensembl.org) and Hubbard et al. (2007). SOX3 EGFP DAPI A B C Tel Tel Tel Tel DHpc DHpc Tel DHpc DHpc Tel DHpc DHpc
LV LV LV LV LV LV
WT LGE LGE LGE LGE LGE LGE MGE MGE MGE MGE MGE MGE
D E F Tel Tel Tel Tel Tel DHpc DHpc Tel DHpc DHpc DHpc DHpc 12.5 dpc LV LV LV LV LV LV Nr/+
LGE LGE LGE LGE LGE LGE
MGE MGE MGE MGE MGE MGE
WT Nr/+ WT
G H DHpc I J DHpc EpTh DHpc EpTh EpTh DHpc VHpc VHpc VHpc VTh VTh VTh
VHpc ChP ChP ChP LaT LaT LaT SOX3 EGFP SOX3 SOX3
Figure 3.4: Endogenous SOX3, not EGFP, indicative of transgenic Sox3 expression in the devel- oping telencephalon at 12.5 dpc. A, G and J: Endogenous SOX3 was highly expressed in the devel- oping telencephalon (white brackets), dorsal Hpc (yellow bracket) and MGE. Endogenous SOX3 was also present in the ventral Hpc (yellow dotted bracket), lamina terminalis and MGE at a lower level. Moreover, it was also detected at the EpTh, VTh and LV ChP. (G) is the boxed region of (A), while (J) is the boxed region of (G). B, E and H: EGFP and hence transgenic Sox3 was not expressed in the developing telencephalon, but in the ventral Hpc, ChP, EpTh, VTh and lamina terminalis of Nr/+ individuals. A lower expression of EGFP and thus, transgenic Sox3 was present in the dorsal Hpc (E and H). (B) and (E) are EGFP signals of (A) and (D) respectively. (H) is the boxed region of (E). D, I: Stronger SOX3 expression was found at areas correlating with EGFP expression. C and F: DAPI stainings of (A) and (D) respectively. Non-specific signals by hematoma (nucleated red blood cells) autofluorescence (arrowheads) and speckles (arrows) were observed with SOX3 signals (A and D). They were also apparent with EGFP signals (B and E) but with lower intensity. Immunofluorescence. DHpc: dorsal Hpc. VHpc: ventral Hpc. LaT: lamina terminalis. Tel: telencephalon. Coronal sections. Top to bottom: dorsal to ventral. Scale bar: 400µm (A-F), 200µm (G-I), 100µm (J). A Overt Expected n Frequency (%) hydrocephalus Frequency (%) penetrance (%)
Sr/+ 963 50 44.76* 30.64
Nr/+ 900 50 47.56 18.69
Sr/+; Nr/+ 364 25 17.31^ 98.36
* and ^ The observed frequecny was significantly lower than expected based on Mendelian ratios (*p=<0.0305; ^p=0.001; Chi Square test).
B
Figure 3.5: CH was dosage dependent on Sox3. A: There was an increased penetrance of CH in Sr/+; Nr/+ double hemizygotes in comparison to Sr/+ or Nr/+ single hemizygotes. Both Sr/+ and Sr/+; Nr/+ lines did not follow Mendelian inheritance for transgene transmission. Analysis performed in mice more than 3 weeks of age. B: qPCR discovered that there were two copies of Sox3 transgenes integrated within chromosome 19 in the Sr mouse line, while there was only one copy of the Sox3 transgene integrated at chromosome 10 for the Nr line. Quantifications were performed to demonstrate the differ- ence in Sox3 copy number between XX and XY individuals within the same transgenic line. Sox3 copy number from Sox3 null (KO) mouse model served as a negative control. Sox1 quantifiation was carried along as an endogenous autosomal control. 78
Since the analysis of overt CH penetrance and its association with single or double hemizygotes demonstrated a strong link to Sox3 dosage, the copy number of transgene was then determined for both Sr and Nr mouse models (performed by James Hughes). Verification of Sox3 transgene copy number by qPCR found that Nr/+ mice had one copy of transgene integrated (n=3), consistent with the observation from the genome walking experiment, while Sr/+ mice had two copies of transgene integrated in tandem (n=3) (Figure 3.5 B). The higher Sox3 transgene copy number correlates with the higher overt CH penetrance in Sr/+ in comparison to Nr/+. Moreover, double hemizygotes Sr/+; Nr/+ achieved almost complete penetrance of overt CH. This is in contrast to the 18-30% penetrance in single hemizygotes, displaying a positive correlation between the number of transgene alleles and overt CH penetrance.
3.2.4 CNS morphology of Nr/+ and Sr/+ mice suggested non-communicating hydrocephalus
In order to further characterise the CH in Sox3 transgenic mice, histological analysis of post-weaning brains from overt CH Nr/+ mice was performed. Nr/+ mice with overt CH showed gross dysmorphic brain features (n=3) in comparison to the WT (Figure 3.1 C-D). As similar to its founders, affected mice displayed thinning of the CCx (Figure 3.1 C-D). The LV expansion has displaced the Hpc, which normally resides ventrally to the CCx. In fact, the Hpc was shifted posteriorly and the AP distance between the Hpc and the cerebellum (Cer) was markedly reduced. The 3V was grossly dilated. Haemorrphrage was also widespread within the CNS, possibly due to increased intracranial pressure (Figure 3.1 C-D). Interestingly, the 4V was not expanded in Nr/+ with overt CH (Figure 3.1 C-F). Similar observations of 4V morphology were also observed in Sr/+ mice (P.-S. Cheah, personal communication). Severely dilatated LV and 3V but not 4V strongly suggested a case of Aq restriction or obstruction, representing non-communicating CH in these mice.
3.2.5 Endogenous and transgenic Sox3 expression in the SCO
Defective SCO development and/or function has previously been associated with non- communicating hydrocephalus in a number of mouse models (Blackshear et al., 2003; Bruni, 79
1998; Dietrich et al., 2009; Jones and Bucknall, 1988; Jones et al., 1987; Overholser et al., 1954; Pérez-Fígares et al., 2001; Rodriguez et al., 2007; Takeuchi et al., 1987; Takeuchi et al., 1988; Takeuchi and Takeuchi, 1986). Moreover, endogenous SOX3 expression was detected in WT adult SCO (N. Rogers, personal communication), suggesting SOX3 has a function within the mature SCO. Together, these data suggest that SCO function may be affected in transgenic Sox3 CH mice. In order to further investigate this possibility and the relationship between increased Sox3 dosage, CH and SCO development, both endogenous and transgenic Sox3 expression were investigated in the SCO during development through analysis of SOX3 and EGFP protein expressions respectively.
The SCO primordium can be distinguished from the neighbouring neuroepithelium from 11.5 dpc through identification of the PC, which is immediately dorsal to the SCO primordium along the dorsal midline. The Growth associated protein 43 (Gap43) gene encodes a membrane glycoprotein that is highly expressed in the growth cone of developing axons (Benowitz and Routtenberg, 1987). Thus, it is an excellent marker for visualisation of the PC which in turn serves as a tool to identify the SCO primordium at 11.5 dpc. SOX3 and EGFP expressions were investigated in neighbouring sections where PC axonal bundles were identified (Figure 3.6 A and B). Endogenous SOX3 was ubiquitously expressed throughout the prosomere 1 RP neuroepithelium, including the SCO, as well as the RP neuroepithelium anterior and posterior to SCO (Figure 3.6 C and G). The presence of EGFP within the SCO neuroepithelium of Nr/+ embryos indicated that the Sox3 transgene recapitulated this expression pattern (n=1) (Figure 3.6 E and F). SOX3 expression was stronger in the prosomere 1 RP of Nr/+ embryos in comparison to WT, agreeing with EGFP reporter expression that Sox3 transgene was present in the SCO (Figure 3.6 C, D and G). Moreover, similar to the endogenous spatial expression, transgene-derived SOX3 was also detected along the RP neuroepithelium that is anterior to the SCO (Figure 3.6 F). In contrast to the high ubiquitous EGFP expression and the robust recapitulation of endogenous spatial expression found in the anterior SCO and RP, EGFP expression became stripy towards the SCO posterior limit (Figure 3.6 D, F and H-J). In fact, no expression of the transgene-derived SOX3 was detected in the RP neuroepithelium posterior to the SCO (Figure 3.6 D, F and H-J). Interestingly, this stripy pattern was recapitulated by the pattern showing stronger SOX3 intensity (Figure 3.6 D, F and H-J). WT Nr/+ A PC B PC SCO GAP43
SCO
Aq Aq
PC C PC D
SCO SOX3 SCO
Aq Aq
E F PC PC 11.5 dpc SCO SCO EGFP
Aq Aq
Nr/+ WT H I J G PC SOX3
SCO
SOX3 EGFP SOX3+EGFP
Figure 3.6: Endogenous SOX3 and EGFP, indicative of transgenic Sox3, were present within the SCO primoridum at 11.5 dpc. A-B: Strong GAP43 signals at the PC along the RP. C and G: Endog- enous SOX3 was detected in the prosomere 1 RP including the SCO primordium (C, G). (G) is the boxed region of (C). (C) is the neigbrouing section of (A). D-F: EGFP indicating transgenic Sox3 expression was found througout the prosomere 1 RP and the RP aneterior to the SCO (F), with corre- sponding higher level of SOX3 (D) in Nr/+ embryos. (E) and (F) are EGFP signals of (C) and (D) respectively. H-J: Stripy EGFP expression was observed towards the posterior limit of Nr/+SCO (I). The stripy EGFP expression correlated with stronger SOX3 intensity (H). (H) and (I) are boxed regions of (D) and (F) respectively. (J) is a merge of (H) and (I). Immunofluorescence. Arrowheads: posterior limit of SCO. Sagittal sections. Top to bottom: dorsal to ventral. Left to right: anterior to posterior. Scale bar: 200µm (A-F), 50 µm (G-J). 80
By 12.5 dpc, PC has ventrally expanded and can be visualised without the aid of GAP43 staining. SCO primordium can also be easily identified as a ventrally invaginated neuroepithelium residing immediately beneath the PC. Ubiquitous endogenous SOX3 was again detected in 12.5 dpc SCO primordium (Figure 3.7 A and 3.8 A). EGFP representing Sox3 transgene was detected from both Nr/+ (n=3) (Figure 3.7 B and D) and Sr/+ (n=1) (Figure 3.8 B and D) SCO primordia. Similarly, SOX3 expression was stronger in these SCO primordia in comparison to that of WT (Figure 3.7 C and Figure 3.8 C). Moreover, stripy EGFP expression similar to that of 11.5 dpc, Nr/+ embryo described above was observed at the posterior limit of both Nr/+ and Sr/+ SCO primordia. EGFP has again demonstrated to be a robust transgene reporter, since these stripy EGFP expressions corresponded to higher SOX3 intensity at these transgenic SCO (Figure 3.7 C, D and Q-S and 3.8 C-G).
By 15.5 dpc, SCO has adopted a more mature morphology. Interestingly, endogenous SOX3 was still detected in all SCO cells, albeit with varying intensities across the SCO (Figure 3.7 E). This high-low intensity pattern had prompted speculation that there may be a correlation between the level of SOX3 expression and a given sub-SCO spatial domain or cellular phenotype. However, further analysis found that these patterns of high-low SOX3 intensity varied significantly between embryos and hence, failed to establish an association with a particular sub-SCO spatial domain. Moreover, a correlation between the differentiation status and the difference in endogenous SOX3 intensity was also testified through colocalisation study of SOX3 and RF, a differentiated SCO marker. Unfortunately, due to the differences in cellular compartmentalisation, i.e. nuclear SOX3 versus cytoplasmic RF, and the intermingling of SCO processes and elongated cell bodies, no clear co-localisation could be identified. Nevertheless, the persistent endogenous SOX3 expression in the SCO at 15.5 dpc was in contrast to its surrounding regions, where SOX3 began to downregulate and became more exclusive to cells near the ventricular area. Similar to 12.5 dpc, EGFP and hence, Sox3 transgene expression was detected within the developing SCO at 15.5 dpc (n=3) (Figure 3.7 G-H).
From 18.5 dpc, all SCO cells have terminally differentiated (Rakic and Sidman, 1968). Interestingly, endogenous SOX3 was still expressed in the mature SCO cells at this stage (Figure 3.7 M), and was maintained into adulthood (Figure 3.7 I and 3.8 L). Again, EGFP with correspondingly higher SOX3 intensity demonstrated that Sox3 transgene recapitulated the endogenous expression both temporally and spatially in Nr/+ prosomere 1 SCO at 18.5 81
dpc embryos (n=3) and adult (n=3) (Figure 3.7 J-L and N-P). Similarly, further study at the prosomere 1 RP of Sr/+ embryos at 15.5 dpc (n=1) and adult mice (n=1) found identical Sox3 transgene expressions (Figure 3.8 I-K and M-O).
In order to further validate the increased dosage of Sox3 in the SCO of transgenic embryos, qRT-PCR was performed to quantify the Sox3 dosage in Sr/+; Nr/+ prosomere 1 RP. Levels of Sox3 transcripts from microdissected 12.5 dpc SCO were compared (SCO microdissection discussed in 5.1). Sr/+; Nr/+ SCO demonstrated a significantly higher level of Sox3 and EGFP transcripts, confirming an increased Sox3 dosage. In fact, Sox3 transcript from Sr/+; Nr/+ embryos was about 2.5 fold higher than that of the WT (n=5) (Figure 3.9), indicating that the Sox3 transgene indeed led to an increased level of Sox3 transcript expression in the SCO.
In summary, endogenous SOX3 was expressed in the SCO primordium, as early as 11.5 dpc and this expression pattern persisted into adulthood. The Sox3 transgene recapitulated this expression both spatially and temporally in both the Nr/+ and Sr/+ lines, further indicating that the SCO transgenic expression is likely to be independent of integration site. Furthermore, the expression of EGFP was spatially associated with higher SOX3 intensity. This was particularly apparent at the posterior limit of the SCO where stripy EGFP expression was observed. Interestingly, the stripy expression pattern appeared in both Nr/+ and Sr/+ transgenic embryos, suggesting that this expression pattern is likely to be under the control of the incorporated transgene regulatory sequences and is not due to a positional effect. Moreover, the robust spatial association between higher SOX3 intensity and EGFP expression in Nr/+ and Sr/+ prosomere 1 RP indicates that EGFP serves as a robust reporter for Sox3 transgene in this context.
3.2.6 Lower level of SOX3 expression in prosomere 1 RP
Expression analysis described above demonstrated that endogenous SOX3 was lower in WT prosomere 1 RP than that of its lateral flanking domains. To further characterise this difference in SOX3 expression, coronal sections across the prosomere 1 RP were studied during different stages of the SCO development. Interestingly a difference in SOX3 expression was observed in 12.5 dpc, 15.5 dpc and 18.5 dpc SCO when compared to its WT Nr/+ WT Nr/+ A C PC PC SOX3 M O PC PC
Aq SOX3 SCO Aq Aq B D
PC EGFP Aq PC
12.5 dpc N PC P
SCO 18.5 dpc PC Aq Aq
EGFP E G Aq
PC SOX3 PC SCO Aq SCO Aq Aq F H
PC EGFP Nr/+ PC 15.5 dpc SCO Q SCO Aq PC Aq SOX3 I K 3V 3V SCO
S Aq
OX3 PC PC R SCO SCO PC
EGFP
Aq Aq
12.5 dpc SCO J L 3V Aq Adult 3V S
SOX3
E
GFP PC PC PC
SCO +
SCO EGFP SCO Aq Aq Aq
Figure 3.7: Both endogenous SOX3 and EGFP, indicative of transgenic Sox3, were highly expressed in the developing and mature SCO. A: Endogenous SOX3 was expressed ubiquitously in 12.5 dpc SCO. E: Endogenous SOX3 was present with varying intensity at 15.5 dpc SCO. M and I: SOX3 was still present in 18.5 dpc (M) and adult (I) SCO. B, D, F, H, J, L, N and P: EGFP and hence, transgenic Sox3 expression has reca- pitulated these expression spatially and temporally (D, H, L and P). (B), (D), (F), (H), (J), (L), (N) and (P) are EGFP signals of (A), (C), (E), (G), (I), (K), (M) and (O) respectively. C, G, K and O: Stronger SOX3 expression was observed in Nr/+ SCO, correlating with transgenic Sox3 expression. Q, R, S: Stripy EGFP and hence Sox3 transgene expression was present towards the posterior limit of SCO primordium at 12.5 dpc (R). This stripy EGFP expression (R) was correlated with the pattern of SOX3 intensity variation (Q). (Q) and (R) are boxed regions of (C) and (D) respectively. (S) is a merge of (Q) and (R). Immunofluorescence. Arrowheads: MR. Yellow arrows: points of SCO ventral invagina- tion. Sagittal sections. Left to right: anterior to posterior (A-L, Q-S). Top to bottom: dorsal to ventral. Scale bar: 200µm (A-L), 100µm (M-S). Aq Aq Aq WT Sr/+ SOX3 EGFP SOX3+EGFP A C E F G 12.5 dpc PC PC PC SOX3 PC PC
SCO Sr/+ Aq SCO SCO SCO SCO Aq
B D 12.5 dpc PC PC EGFP SCO Figure 3.8: Transgenic Sox3 was expressed in 12.5 SCO Aq Aq dpc, 18.5 dpc and adult SCO from Sr/+ mice. A-D: EGFP, indicative of transgenic Sox3, was expressed in Sr/+ 12.5 dpc SCO (D). Higher SOX3 intensity was also detected in the Sr/+ SCO (B) in comparison to the SOX3 EGFP WT (A). (B) and (D) are EGFP signals of (A) and (C) respectively. Similar to , stripy EGFP (F) L M E-G: Nr/+ SOX3 EGFP with corresponding SOX3 intensity variation (E) was PC H PC I PC SCO apprent at the SCO posterior limit of embryos at SCO SCO SCO Sr/+ Aq 12.5 dpc. (E) and (F) are boxed regions from (C) and Aq Aq WT Aq (D) respectively. (F) is the EGFP signal of (E). (G) is a merge of (E) and (F). H-O: EGFP and transgenic Sox3 WT were also present in the SCO of 15.5 dpc (J and K) and N O adult (N and O) Sr/+ mice. Higher SOX3 intensity was 15.5 dpc adult found at Sr/+ (J) SCO at 15.5 dpc in comparison to the WT (H). Note that the adult SCO was dysmorphic J PC K PC Sr/+ SCO SCO SCO PC SCO PC with very few SCO cells due to CH (N). (I), (K), (M) and (O) are EGFP signals from respecitve sections of Aq Aq Aq Aq (H), (J), (L) and (N). Immunofluorescence. Yellow Sr/+
Sr/+ arrows: points of SCO invagination. Sagittal sections (A-G). Coronal sections (H-O). Top to bottom: dorsal to ventral. Left to right: anterior to posterior (A-G). Scale bar: 200µm (A-D and H-O), 100µm (E-G). p=0.0499
p=0.0142
Figure 3.9: Sr/+; Nr/+ SCO had a 2.5-fold increase in the level of Sox3 transcript when compared to the WT. Both Sox3 and EGFP were significantly elevated in the prosomere 1 RP or the SCO of Sr/+; Nr/+ embryos. Gapdh served as a control. 82 flanking regions (Figure 3.10 A, E and I). At 12.5 dpc, SOX3 was widely expressed within the VZ of the SCO flanking regions. Although a ubiquitous SOX3 expression was detected with the SCO, SOX3 expression was evidently more intense in the lateral tissue than that of the SCO (Figure 3.10 A). This mediolateral difference was maintained into 15.5 dpc when endogenous SOX3 expression in the lateral regions was restricted to the 3V lining ependymal cells. A markedly lower SOX3 expression was again observed in the SCO cells in comparison to mediolateral ependymal cells that were immediately flanking the SCO (Figure 3.10 E). Similar SOX3 intensity difference was also detected in 18.5 dpc (Figure 3.10 I). In fact, an obvious high-low SOX3 expression boundary at both lateral limits of the SCO was observed in the above 12.5 dpc, 15.5 dpc and 18.5 dpc embryos studied (Figure 3.10 A, E and I).
Interestingly, SCO from Sr/+; Nr/+ at 12.5 dpc no longer displayed a lower SOX3 intensity (n=3). Instead, its level of SOX3 expression was similar to that expressed by its laterally flanking tissue (Figure 3.10 C). Likewise, at 15.5 (n=3) dpc and 18.5 dpc (n=3), the SOX3 intensity difference was no longer detected and the SCO SOX3 expression was overall comparable in intensity to those of 3V lining ependymal cells (Figure 3.10 G and K). The high-low SOX3 boundaries on either side of SCO lateral limits were noticeably absent in 12.5 dpc, 15.5 dpc and 18.5 dpc Sox3 transgenic embryos (Figure 3.10 C, G and K). These data indicate that overexpression of SOX3 in the SCO of Sox3 transgenic embryos has resulted in level of expression that is comparable to that of flanking neuroepithelial cells in WT embryos. Thus, Sox3 transgene expression appears to disrupt the SOX3 high-low intensity difference between SCO cells and its laterally flanking tissue.
3.2.7 Dysmorphic SCO in post-weaning Nr/+ and Sr/+ mice
Since both endogenous and transgenic Sox3 are robustly expressed in the developing and mature SCO, and that both Nr/+ and Sr/+ mice display CH with incomplete penetrance, SCO morphology was then studied in Nr/+ and Sr/+ post-weaning mice. In coronal sections of WT mice, the SCO displayed the typical horseshoe shape due to the thick lining of the dorsal 3V plate by the elongated ependymal cells. SCO nuclei were localised basally to the ventricles, with their elongated cell bodies positioned apically next to the 3V (Figure 3.11 A and E). In contrast, no thickened psuedostratified ependymal layer was found in either 83
Nr/+(n=3) or Sr/+ (n=1) mice with overt CH, indicating SCO aplasia (Figure 3.11 C and D). In order to exclude that SCO defects in these animals were simply secondary to the increased intracranial pressure associated with CH, four Nr/+ animals without overt CH were also subjected to histological analysis. Interestingly, one of the brains displayed mild LV dilatation upon sectioning, indicative of late CH onset. Only a minimal SCO remnant was evident in this mouse, demonstrating severe SCO dysplasia (Figure 3.11 F and H). There was no detectable expansion of the LV and 3V in the other three brains, consistent with the absence of CH. One of these non-CH mouse displayed SCO dysplasia with a markedly thinner ependymal layer. Moreover, the typical horseshoe-shaped morphology of SCO was no longer observed (Figure 3.11 B).
In summary, Nr/+ mice displayed widely varied CH and SCO dysmorphology phenotypes (summarised as a table in Figure 3.12). These variations can be generally grouped into four catagories: 1) non-CH Nr/+ mice with a morphologically normal SCO; 2) non-CH Nr/+ mice with a morphologically thinner SCO; 3) non-overt CH Nr/+ mice with a morphologically thinner SCO; and 4) overt CH Nr/+ mice with SCO aplasia. To summarise, the assoication of SCO aplasia or dysplasia with both overt or non-overt CH suggests a causative role of defective SCO in the development of CH in Sox3 transgenic mice. Together, the presence of SCO morphological defect in some Sox3 transgenic mice with mild or no LV or 3V expansion indicates that the SCO defect is unlikely to be a mere consequence of the increased intracranial pressure induced by CH.
To further assess the SCO function in Sox3 transgenic mice, the expression of RF in the SCO was analysed using immunofluorescence in some of the mice described above: the non-overt CH Nr/+ mice; and two of the non-CH mice with morphologically normal SCO (Figure 3.12). Furthermore, RF expression was also investigated in the SCO of a Sr/+ non- CH mouse for comparison (Figure 3.12). Moreover, SOX3 expression was studied alongside. Similar to the WT SCO, SOX3 was present in cells within the SCO from both Sr/+ or Nr/+ mice with either a dysplasic or normal SCO (Figure 3.12 and 3.13 A, C, E and G). On the other hand, RF expression was exclusively but ubiquitously expressed in all mature SCO cells from WT post-weaning mice (Figure 12 and Figure 3.13 B). Not surprisingly, RF expression was detected and overlapped with that of SOX3 in the morphologically normal SCO from both of the non-CH Nr/+ mice (Figure 3.12 and 3.13 D) and the non-CH Sr/+ mice (Figure 3.12 and 3.13 H), suggesting that these SCO were functionally normal. Sr/+; Nr/+ WT A C
Nr/+ WT E G SCO SCO Aq Aq
SOX3 SCO SCO
SOX3 Aq Aq F H D B
DAPI SCO SCO Aq Aq SCO SCO
DAPI Aq Aq ventral. Scalebar:100µm. and (K)respectively. Immunofluorescence. Coronalsections. Top tobottom: dorsalto observed. B,D,F, H,JandL: DAPIstainingcorresponding to(A),(C),(E),(G),(I) In addition,theboundary ofSOX3intensitydifference (arrowheads) wasnolonger genic mice,SOX3expression wascomparableintheSCOtoitslateralflanking tissue. difference onbothlateralsidesoftheSCO(arrowheads). at 12.5dpc,15.5dpcand18.8dpc. This createsaclearboundaryofSOX3intensity was observedintheSCOorprosomere1RP incontrasttoitslateralflankingtissue comparison toitsflankingtissue. A, EandI: Figure 3.10:EndogenousSOX3expression wasdownregulated intheSCO
Nr/+ WT K I SCO SCO SOX3 Aq Aq A lowerlevelofendogenousSOX3 L J C, GandK:InSox3trans-
DAPI SCO SCO Aq Aq WT Nr/+ non-CH Nr/+ overt CH Sr/+ overt CH A B C D PC PC PC # PC G SCO #
Aq Aq Aq SCO Aq P26 WT SCO
Aq P34
H WT Nr/+ non-overt CH E F 3V 3V
Nr/+ non-overt CH SCO Aq P34 SCO PC Aq SCO
Aq Figure 3.11: SCO dysmorphology was present in post-weaning Sox3 transgenic mice that were without CH, with non-overt CH/overt CH. A, E and G: WT SCO displayed a thickened ependyaml layer with nuclei localising basally (A, E and G). WT SCO displayed a horseshoe shaped when observed at coronal section (A). (G) is the boxed region of (E). B: Thinner SCO with less elongated ependymal cells were found in Nr/+ mice without CH (B) in comparison to WT (A). C, D, F and H: SCO tissue was minimal in Nr/+ mice with non-overt CH (F and H). The SCO had a very thin layer of SCO ependymal cells that showed nuclei basal localisation. SCO was absent in both Nr/+ and Sr/+ individuals with overt CH (C and D). (H) is the boxed region in (F). H & E stains. Arrowhead: PR. Doublet arrow: MR. Hash: miss- ing SCO. Coronal sections (A-D). Sagittal sections (E-H). Top to bottom: dorsal to ventral. Left to right: anterior to posterior (E-H). Scale bar: 400µm (A-D), 100 µm (E-F), 50µm (G-H). Number of Genotype CH phenotype SCO phenotype RF expression mice studied 3 Nr/+ overt aplasic N/A 1 Sr/+ overt aplasic N/A 1 Nr/+ non-overt dysplasic present 1 Nr/+ non-CH dysplasic N/A 2 Nr/+ non-CH normal present 1 Sr/+ non-CH N/A present Figure 3.12: Summary of the varying CH phenotypes and SCO defects in Nr/+ and Sr/+ single hemizygous mice. Note that N/A means the analy- sis was not performed. 100µm. Sagittal sections. Top tobottom: dorsaltoventral. Left toright:anteriorposterior. Scalebar: (B), (D),(F)and(H)are RFstainingsof(A),(C),(E),and(G)respectively. Immunofluorescence. non-overt CH(E). A lowRFexpressionwasalsoobserved inthesecells(yellowarrow)(EandF). SOX3 positiveSCOremnant survivingaround thedorsalcurvatureofSCO in detected intheseSOX3positive SCOcells (B, D,FandH).Notetherewas asmallisolatedgroupof bouring ependymalcellswerevoidofSOX3 expression (arrows).B,D,F andRFexpressionwas H: in SCOremnantofnon-overtCH Nr/+ or Sr/+hemizygousmice Figure 3.13:SCOremnant wascapableofRF secretion innon-CHor non-overtCHmicefrom Sr/+ non-CH Nr/+ non-overt CH Nr/+ non-CH WT E C A G 3V 3V 3V SCO Aq SCO 3V SCO Aq Aq SOX3 Aq PC PC SCO . A,C,EandG:Similartothe WT SCO(A),SOX3wasexpressed PC Nr/+ (E)andnon-CH(C)Sr/+(G)mice,whileneigh- PC D B F H 3V 3V 3V Aq SCO SCO 3V SCO Aq RF Aq PC Aq SCO PC Nr/+ mousewith PC
PC >5 weeks >5 84
Intriguingly, RF was also found in the SOX3 positive SCO remnant of the non-overt CH mouse (Figure 3.12 and 3.13 F), suggesting that those surviving SCO cells resembled WT behaviour. Together, these indicate that the SCO ependymal cells in non-CH or non-overt CH Nr/+ and Sr/+ post-weaning mice retain at least some normal SCO secretory function.
3.2.8 Endogenous and transgenic Sox3 expressions in other area of dorsal diencephalon
In addition to the SCO, other dorsal diencephalic domains with both endogenous and transgenic Sox3 expression were noted. Due to their proximity and thus a potential influence to SCO development, both the endogenous and transgenic Sox3 expression were analysed through the study of SOX3 and EGFP expression in these dorsal diencephalic domains prior to further characterisation of the SCO defect.
Endogenous and transgenic Sox3 expression in 12.5 dpc dorsal diencephalon
Previous analysis described in 3.2.2 showed that endogenous SOX3 was present in a range of dorsal diencephalic domains other than the SCO, i.e. the VTh and EpTh. In fact, robust expression of SOX3 was detected within the VZ of the 3V, the Aq and the 4V of 12.5 dpc WT embryos (Figure 3.14 A-B and E-J). Within the diencephalon, endogenous SOX3 was observed in EpTh, DTh, VTh and Pt only in medial and parasagittal sections (Figure 3.14 G and I), i.e. sections flanking the medial section but not in more lateral sections (Figure 3.14 E), indicating SOX3 was present in the VZ but not the mantle zone (MZ) within these domains. Interestingly, endogenous SOX3 was expressed in the RP within the diencephalon and mesencephalon (Figure 3.14 I). This observation agrees with previous studies, demonstrating that SOX3 is expressed by neural progenitors along the VZ (Solomon et al., 2004; N. Rogers, unpublished data). Moreover, a low level of endogenous SOX3 was detected in the 12.5 dpc PG primordium (Figure 3.14 C), with some cells displaying stronger SOX3 intensity than others, resulting in a salt and pepper-like expression pattern. In addition, endogenous SOX3 was also observed within the developing hypothalamus at the ventral diencephalon, agreeing with previous literatures (Rizzoti et al., 2004; Solomon et al., 2004). 85
Interestingly, EGFP indicating the presence of Sox3 transgene was present in the VZ of VTh, DTH, EpTH and Pt (Figure 3.15 A, C and E) from 12.5 dpc Nr/+ (n=5) and Sr/+ (n=1) embryos. Similarly, EGFP transgene was found along most of the diencephalic RP, from the most anterior prosomere 3 RP to prosomere 2 then prosomere 1 RP in Nr/+ and Sr/+ embryos (Figure 3.15 A, C and E). The expression of EGFP, however, terminated abruptly at the SCO posterior limit, and was no longer detected in the mesencephalon RP. This was in contrast to endogenous SOX3, which was found throughout the RP of the diencephalon and mesencephalon (Figure 3.14 I). In addition, the expression of Sox3 transgene at the 12.5 dpc PG (Figure 3.15 K) led to a stronger SOX3 level within the PG primordium in general, resulting in the loss of salt and pepper-like endogenous SOX3 expression pattern (Figure 3.15 G and J). Overall, the Sox3 transgene has recapitulated spatially most of the endogenous SOX3 expression within the 12.5 dpc dorsal diencephalon of both Nr/+ and Sr/+ embryos.
Endogenous and transgenic Sox3 expression at 15.5 dpc dorsal diencephalon
By 15.5 dpc, endogenous SOX3 was downregulated in most dorsal diencephalic domains. Endogenous SOX3 expression was absent from most areas of the DTh, VTh and Pt (Figure 3.16 A). Isolated cells with nuclear SOX3 signal were observed in the DTh and VTh (Figure 3.16 A, C and G-I). In contrast, robust endogenous SOX3 expression was still present within ependymal cells lining the 3V and the Aq (Figure 3.16 A) and the diencephalic RP. EGFP was detected in the ventricle-lining ependymal cells and RP, but was absent from the Pt and DTh of Nr/+ embryos (n=3) (Figure 3.16 B and E). Again, the total SOX3 expression was higher in regions with EGFP expression, agreeing with the EGFP reported Sox3 transgene expression (Figure 3.16 D).
Together, the above evidence indicates that at 15.5 dpc, most endogenous SOX3 is restricted to the RP and the ventricular lining ependymal cells of the dorsal diencephalon. Transgenic Sox3 has recapitulated the former endogenous expression in Nr/+ mice. On the other hand, only sporadic endogenous SOX3 is present within the DTh, VTh and Pt. Given the role of SOX3 as an inhibitor or neural progenitor differentiation (Bylund et al., 2003; Holmberg et al., 2008), this suggests that most neural progenitors within these domains have differentiated by this stage. It was not clear whether or not the Sox3 transgene has recapitulated the sporadic endogenous expression in the DTh, VTh and Pt and thus, study the SOX3 DAPI WT A B C PG K Mes Mes E, F 12.5 dpc SOX3 12.5 dpc Aq Aq Aq G, H
WT I,J D Rb Rb PG Aq 4V DAPI Aq
WT
E DTh Pt G I Tel EpTh PC VTh EpTh PCPC EpTh Pt DTh SCO LV 3V DTh PT Di DTh Mes SOX3 Sp MGE VTh 12.5 dpc Di LT Di
F H J DTh Pt Tel EpTh PC PCPC DTh VTh EpTh EpTh LV PtPT DTh 3V SCO Di DTh
DAPI LT Sp Mes VTh MGE Di Di
Figure 3.14: Endogenous SOX3 was expressed at the VZ along the NT at 12.5 dpc including the dorsal diencephalon. A-B: Endogenous SOX3 was detected at the VZ of the mesencephalon and rhombencephalon (A). (B) is the DAPI signal of (A). E, G and I: Endogenous SOX3 was not present at the MZ (E), but was found at the VZ (G and I) of the dorsal diencephalon including the DTh, VTh, EpTh and Pt. Note that sections from (E) to (G) to (I) is in a lateral to medial direction. F, H and J: DAPI of (E), (G) and (J) respectively. K: A transverse section from Schambra (2008) indicating the plane of sectioning for (E) to (J). C-D: Endogenous SOX3 was expressed in the PG neuroepithelium, with some cells displaying stronger expression (arrowheads). Note there were non-specific signals from heamatoma (arrows) in (C). (D) is the DAPI signal of (C). Immunofluorescence. Sp: septal neuroepithelium. Tel: telencephalon. Di: Diencephalon. LT: limana terminalis. Mes: mesencephalon. Rb: Rhomencephalon. Sagittal sections (A-B and E-J). Coronal sections (C-D). Top to bottom: dorsal to ventral. Left to right: anterior to posterior (A-B and E-J). Scale bar: 400µm (A-B, E-J), 100µm (C-D). WT Nr/+ Sr/+ C A EpTh E EpTh EpTh SCO SCO SCO Pt DTh Pt Pt DTh DTh Mes Mes EGFP Mes Di Di VTh Di VTh 12.5 dpc VTh
B D F EpTh EpTh EpTh SCO SCO SCO Pt Pt DTh DTh Pt DTh Mes
DAPI Mes
Di Mes Di VTh Di VTh VTh
WT Nr/+ Pt G J PG PG Figure 3.15: Transgenic Sox3 recapitulated endogenous spatial expression at the VZ of the dorsal diencephalon and PG during 12.5 dpc. A, C and E: EGFP and thus
SOX3 transgenic Sox3 expression was present in the VZ of VTh, DTh, EpTh and Pt of Nr/+
Aq Aq and Sr/+ mice. B, D and F: DAPI stainings of (A), (C) and (E) respectively. G-K: The PG from 12.5 dpc Nr/+ displayed EGFP (K), indicative of transgenic Sox3 expression. H K In the Nr/+ PG, SOX3 intensity was general higher (J) and the WT salt and pepper-like PG PG expression pattern (demonstrated by cells with stronger isolated SOX3 expression indicated by arrowheads in (G)) was no longer observed. (H) is the EGFP signal from EGFP (G) and (K) is the EGFP signal from (J). DAPI signals of (G) and (J) respec- Aq I and L: Aq tively. Note the non-specific SOX3 signals from heamatoma (arrows) (G-H). Immuno- I L fluorescence. Di: diencephalon. Mes: mesencephalon. Sagittal sections (A-F). Coronal PG sections (G-L). Top to bottom: dorsal to ventral. Left to right: anterior to posterior PG Aq (A-F). Scale bar: 400µm (A-F), 100µm (G-L). DAPI
Aq Nr/+ WT A D CCx CCx VTh VTh SOX3 PG SCO Aq SCO Aq Pt DTh Pt DTh CCx CCx B E CCx CCx VTh VTh EGFP SCO PG Aq SCO Aq Pt DTh Pt DTh CCx CCx F C CCx CCx VTh VTh DAPI SCO SCO PG Aq Aq Pt DTh Pt DTh
CCx CCx 15.5 dpc 15.5
(A-F), 50µm(G-I). sections (A-I). Top tobottom: dorsaltoventral.Scale bar:400µm was observed withNr/+SOX3staining (whitearrow)(D).Coronal Immunofluorescence. Note thatnon-specificsignalfromheamatoma respectively. CandF:DAPIstainingsof(A) (D)respectivley. Nr/+ embryos(D).(B)and(E) areEGFP signalsof(A)and(D) expression wasagreedby agenerallyhigherleveloftotalSOX3in mingle withthePCand the dorsalmidlinePt(E). Trasngenic Sox3 cytoplasmically localisedEGFP attheSCOprocesses,whichinter- embryos, theEGFP signalaroundthePt(yellowarrow)wasfrom the dorsalmidline,butnotinPt,DThand VTh (E).InNr/+ domains ofendogenousSOX3,i.e.the Aq liningependyamlscellsand thus, transgenic staining of(G).(I)isamerge of(G)and(H).B,DE:EGFP and arrowhead) (G).(G)istheboxedregionof(A).(H) istheDAPI cells werestillobservedsporadicallyinthePt,DTh and VTh (yellow negative forendogenousSOX3expression,although SOX3positive and dorsalmidlinecells.MostareasofthePt,DTh VTh were SOX3 wassignificantlyrestrictedtotheventricularliningependymal cephalon. A, C,G,HandI:Within thediencephalon,endogenous restricted totheependymalcellswithin15.5dpcdorsaldien- Figure 3.16:EndogenousandtransgenicSox3expressions were WT G SOX3 Sox3 expressionwaspresentwithinthesamespatial H DAPI I SOX3+DAPI
15.5 dpc 15.5 86
transgene expression within these domains was no longer pursued as it is beyond the scope of this thesis.
3.3 Discussion
3.3.1 Sox3 transgene integration site in Nr mouse line
Previous studies have demonstrated that random integration of transgenes introduced by pronuclear injection can trigger deletion of chromosomal sequences flanking the integration site and lead to transgene independent phenotypes (Gabriel et al., 1999; Zhou et al., 2009). A relevant example includes a deletion of exon 17 to 21 of the Dnahc5 gene that was induced by integration of an unrelated transgene and resulted in CH (Ibanez-Tallon et al., 2002; Ibanez-Tallon et al., 2004). In Nr/+ mice, Sox3 transgene has integrated into the last 120bp of the 3’UTR of Btbd11 without inducing a significant chromosomal deletion. Thus, this allows the exclusion of integration induced deletion as a cause of any identified phenotype(s). Moreover, since Btbd11 ORF has not been perturbed, functional protein should still be synthesised.
Nonetheless, Sox3 integration has disrupted the locus of Btbd11. Transcription activation requires the accessibility and hence, a close proximity of the transcription initiation site to the RNA polymerase. Thus, mammalian transcription regulation employs strategy that modifies the accessibility of the transcription initiation site to RNA polymerase through conformational change of the regulatory sequences upon binding of trans-inducing factor(s) (Millevoi et al., 2001; Naar et al., 2002; Taatjes et al., 2002). As a result, the insertion of the 26kb Sox3 transgene may disrupt the spatial alignment between Btbd11 and its flanking regulatory sequences, which then inhibits conformational change that is required for the normal Btbd11 expression regulation. Moreover, the integration site is likely to ablate the polyadenylation signal(s) of Btbd11. The lack of poly-A tail may render the Btbd11 transcripts to be more prone to degradation and hence, reduce its abundance. Collectively, these suggest that the abundance of Btbd11 transcripts may be affected by Sox3 integration. 87
Btbd11 encodes a protein with predicted ankryin repeats and BOZ/BTB domains and a histone-fold domain that is often found in transcription factor (Hunter et al., 2009). According to online brain gene expression atlas of CNS development (Allen atlas, http://www.brain-map.org/), Btbd11 expression is not present at 11.5 dpc within the CNS. By 13.5 dpc, it is strongly expressed in the developing EpTh and Cer and is also present in the RP and the VTh at a very low level. From 15.5 dpc, its expression domains expand to other areas of the CNS, especially in the posterior developing brain, albeit at a very low level. By P4, it is almost ubiquitously expressed within the CNS. By P14, its ubiquitous expression becomes more intense and this pattern is further maintained into adulthood (Lein et al., 2007). The exact function of BTBD11 is not known. Interestingly, a search in the online Mouse Genome Informatics (MGI) database (http://www.informatics.jax.org/) found that there are 101 gene trap alleles for Btbd11, indicating that this locus is prone to foreign DNA integration. Moreover, Btbd11 expression is upregulated in retinoic acid-induced human neuroblastoma cells neural differentiation (Merrill et al., 2004). This suggests Btbd11 may have a pro-neural differentiation role in CNS development. As a result, there exists a possibility that the perturbation of Btbd11 expression may lead to an exacerbation of the established Sox3 transgenic CH phenotypes, or cause a non-CH related CNS phenotype(s) that are yet to be identified. However, the former proposition is very unlikely as Nr/+ in fact has a much lower CH penetrance than that of Sr/+. Thus, the CH is likely to be a consequence solely brought by the Sox3 transgene, which this is also consistent with the overt CH observed in both transgenic Sox3 and Sox3-EGFP founders.
3.3.2 Transgene incorporated Sox3 regulatory sequences were insufficient for complete recapitulation of endogenous telencephalic expression
The transgene construct has incorporated endogenous regulatory sequences that are 26kb upstream and 8kb downstream of the Sox3 ORF, which resides within a single exon. However, they were not sufficient to recapitulate the telencephalic expression of Sox3 at 12.5 dpc. Previous studies found cis-regulatory elements residing in regions up to hundreds or even thousands of kilobases away from gene loci (Chandler et al., 2009; Chandler et al., 2007; DiLeone et al., 2000; Mortlock et al., 2003). In fact, identification of HCNEs that are shared between human, mouse, zebrafish and Xenopus found that regulatory sequences capable of inducing SOX3 CNS expression are located up to 350kb upstream and 150kb downstream 88 from the SOX3 locus (Ahituv et al., 2007; Navratilova et al., 2009; Pennacchio et al., 2006). Interestingly, a 300bp HCNE residing at 350kb upstream of Sox3 ORF alone is able to drive a telencephalic lacZ expression at 11.5 dpc in mouse (Ahituv et al., 2007; Pennacchio et al., 2006). Intriguingly, deletion of this HCNE does not ablate the endogenous Sox3 expression within the telencephalon (Ahituv et al., 2007), demonstrating that this HCNE is sufficient but not essential for 11.5 dpc telencephalic expression. Furthermore, incorporation of as little as 3.5kb up- and downstream endogenous Sox3 regulatory element is enough to allow lacZ reporter expression within the developing mouse telencephalon from 8.5 dpc to 10.5 dpc (Brunelli et al., 2003). Together, these data illustrate the requirement for the complicated and yet intricate cooperation of different cis-regulatory elements for the dynamic spatial and temporal expression of endogenous Sox3 during CNS development.
If as little as 3.5kb of flanking regulatory sequences are sufficient for mouse telencephalic expression at 11.5 dpc (Brunelli et al., 2003), why was there a lack of telencephalic transgenic expression in Sox3 transgenic mice, despite the considerably larger and completely overlapping transgene that was incorporated? With the above knowledge in mind, there are three explanations proposed. First, the incorporated endogenous regulatory sequence might have omitted an essential element that is responsible for telencephalic Sox3 expression particularly at 12.5 dpc. However, telencephalic Sox3 transgene expression is absent in both 10.5 dpc and 11.5 dpc Sox3 transgenic embryos (N. Rogers, unpublished data). Thus, this explanation can be excluded.
Second, the lack of telencephalic expression might be due to a positional effect, whereby endogenous negative cis-regulatory element(s) residing near the integration site repress the telencephalic Sox3 transgene expression. The effect of endogenous flanking regulatory element from the integration site on transgene expression regulation has been readily demonstrated in the Sr/+ line, where ectopic Sox3 was found in the developing gonad resulting in female to male sex reversal (Sutton et al., 2011). Yet, both the Sr/+ and Nr/+ embryos lack telencephalic Sox3 transgene expression at 10.5 to 12.5 dpc (N. Rogers, unpublished data) arguing against the case of a positional effect despite this explanation is still reserved.
Third, it is possible that endogenous telencephalic Sox3 expression was a result of the interaction between multiple positive and negative regulatory elements that share some redundant regulatory functions (given that Sox3 is important for CNS development, one may 89
imagine that evolution would indeed favour redundant regulation to ensure robust expression). The Sox3 transgene construct, although incorporating the minimal 3.5kb up- and downstream telencephalic regulatory sequences from Brunelli et al. (2003), may also contain other negative regulatory element(s) that override this positive regulation, and lack additional positive regulatory domain(s), which reside outside of the 34kb endogenous regulatory sequences incorporated that enable telencephalic expression. In order to further unravel the complex network of interaction in Sox3 expression regulation, a reporter expression analysis with subset(s) of sequences within the 34kb regulatory sequence from the Sox3 construct may be carried out together with complementation studies with further up- and downstream regulatory sequences including those long-range HCNEs identified (Ahituv et al., 2007; Pennacchio et al., 2006).
3.3.3 Transgenic Sox3 expression in the diencephalon
In comparison to the telencephalon, transgenic Sox3 was highly expressed within the diencephalon. This is in contrast to the endogenous SOX3, which was highly expressed in the telencephalon but was at a lower level in the diencephalon. In fact, the intensity of transgenic Sox3 was in an inverse relationship with that of endogenous SOX3. There are two possible causes for the relatively high level of transgene expression within the diencephalon. First, similar to the telencephalic transgene expression (discussed in 3.3.2), there cannot rule out the possibility of a positional effect whereby integration site residing transcription enhancing element might act on the transgene and upregulate its expression. However, this appears unlikely as the diencephalic transgenic Sox3 expression was apparent in both Sr/+ and Nr/+ embryos. Second, this might be the consequence of the absence of one or more repressive element(s), which downregulate the endogenous SOX3 expression within the diencephalon.
An interesting stripy transgene expression pattern was noted at the posterior limit of the SCO, near the di- and mesencephalic boundary. This stripy pattern was identical and consistently observed in all transgenic Nr/+ or Sr/+ embryos, suggesting this is under the control of regulatory sequences incorporated in the transgene construct rather than those residing at the integration site. It is well known that regulatory sequences for eukaryotes can be organised in modules, with different combinations of modules interacting together to allow dynamic temporal expression patterns within different spatial domains (Guenther et al., 90
2008; Kelly and Buckingham, 2000; Kirchhamer et al., 1996). The lack of transgenic Sox3 in the small subset of SCO cells at the posterior limit suggest that they may be under the regulation of different modular combination than those of anterior SCO cells.
Although not necessarily true in all cases, cells with similar specificity are often associated with identical modular combination (Guenther et al., 2008; Kelly and Buckingham, 2000; Kirchhamer et al., 1996). This thus poses the question of whether the posterior cells have a slightly different function and/or fate to that of the anterior SCO cells. The SCO is believed to be derived from a homogeneous group of cells at the prosomere 1 RP. However, a difference in the architecture and ultrastructure is apparent between the apically and the basally residing cells within the mature SCO (Rodriguez et al., 1998). The molecular mechanism underlying the derivation of these two cellular types is not well understood. However, it is possible that different modular combinations are employed to fine tune temporal gene expressions within different subset of SCO cells, which then allows differentiation of the two (or possibly more) SCO cell types. Stemming from this, the modular combination that allows the transgenic Sox3 expression in those posterior-most SCO cells may not be present in the regulatory sequence incorporated into the transgene construct to enable its expression. Further analysis by reporter mapping of these modular regulatory elements, followed by lineage tracing of cells under the regulation of the above identified modules through the use Cre mediated reporter may allow a better understanding of the molecular regulation that underlies the SCO cellular heterogeneity.
3.3.4 Sox3 has a role in SCO development
Endogenous SOX3 expression was detected in all SCO cells at 11.5 to 15.5 dpc, a developmental timeframe for peak SCO proliferation and differentiation. Furthermore, in contrast to the ubiquitous and comparably even SOX3 expression in early SCO precursors, a variable level of endogenous SOX3 signal was observed at 15.5 dpc, at which time the first terminally differentiated SCO cells arise. Previous literature found that SCO develops in a cell autonomous manner which most SCO cells are derived from local precursor cells from the RP of prosomere 1 (Rakic and Sidman, 1968). Thus, despite there was no correlation established between the SCO spatial domain, and the signal intensity of RF and SOX3 signal, 91
the fact that SOX3 is expressed during SCO proliferation and differentially regulated between SCO cells during differentiation suggests a role of SOX3 in both of these processes.
3.3.5 Sox3 overexpression leads to defective SCO development and CH
The presence of non-expanded 4V in Sox3 transgenic mice suffering from overt CH indicated a restricted or an obstructed CSF flow along the Aq. More importantly, SCO dysplasia was always observed with CH hallmarks in all Sox3 transgenic mice. These data are consistent with previous literatures where cases of non-communicating CH are highly associated with defective SCO development (Huh et al., 2009; Pérez-Fígares et al., 2001; Picketts, 2006). In contrast, a thinner SCO was observed in a mouse without LV or 3V expansion, typical CH hallmarks. Since overt CH was only found with low penetrance and that non-overt CH indicative of late onset was apparent in single hemizygous Sox3 transgenic mice, the presence of dysmorphic SCO without CH hallmark in a Nr/+ mouse is likely to be a case of late onset, non-overt CH, in which the mouse was analysed before CH pathological manifestations. Moreover, SCO dysmorphology was observed in all embryos from 13.5 dpc (described in 4.2.2 to 4.2.4) of the Sr/+; Nr/+ genotype, which suffered from overt CH with a near complete penetrance. This is in sharp contrast to the presence of a morphologically normal SCO in some non-CH individuals of either the Sr/+ or Nr/+ genotype, which displayed a low 18-30% overt CH penetrance. Together, these indicate that the defective SCO development is positively correlated with the penetrance of overt CH in Sox3 transgenic mice. In brief, the frequent co-existence and the direct correlation of dysmorphic SCO and CH in Sox3 transgenic mice, as well as the consistent presence of both endogenous and transgenic Sox3 in embryonic and adult SCO have together proposed a strong association between Sox3 overexpression and the dysplasic SCO development, which ultimately leads to CH.
3.3.6 Proper Sox3 dosage regulation is important during SCO development
The marked increase in the penetrance of overt CH in the double hemizygous Sr/+; Nr/+ mice in comparison to the single hemizygous Sr/+ and Nr/+ mice has clearly demonstrated that the frequency of CH is dependent on Sox3 dosage. Hence, this leads to the 92
hypothesis that the phenotype of defective SCO development may also be dependent on the dosage of Sox3. In fact, all Sr/+; Nr/+ embryos displayed SCO aplasia by 18.5 dpc (described in 4.2.2 to 4.2.4) while some Sr/+ or Nr/+ adult mice that did not suffer from CH were presented with functionally normal SCO. Interestingly, endogenous SOX3 expression was lower in the SCO than in laterally residing cells at 12.5 dpc, 15.5 dpc and 18.5 dpc. Although transgenic Sox3 was also present in both the SCO and cells flanking the SCO, the level of SCO SOX3 appeared to be comparable to those lateral residing cells in transgenic embryos. There is no doubt that the total level of SOX3 within the Sox3 transgenic dorsal diencephalon flanking the SCO may still be higher than that of the SCO. However, in the presence of a significantly upregulated level of SOX3, the magnitude of the fold-difference between the SCO and the cells flanking the SCO was perhaps too small to be detected using immunofluorescence technique which is only semi-quantitative. Nevertheless, this elevation of SOX3 expression in the transgenic SCO correlates well with observed Sox3 dosage dependency in SCO dysplasia. Thus, the maintenance of the SOX3 expression at or close to its endogenous level appears critical to allow normal SCO development. A requirement for Sox3 dosage regulation for CNS development has been previously reported in human. In fact, duplication of Sox3 containing chromosomal region has been identified in some XH patients (Woods et al., 2005), demonstrating that SOX3 overdosage also causes human CNS defect.
3.3.7 The role of endogenous SOX3 in mature SCO
SOX3 expression was detected at the SCO from 15.5 dpc into 18.5 dpc and even into adulthood. This implies that SOX3 may be required for the functional maintenance of the mature SCO. The persistent SOX3 expression at the mature SCO is in contrast to that in the VZ of both the telencephalon (Wood and Episkopou, 1999; N. Rogers, unpublished data) and the diencephalon, where endogenous Sox3 is widely expressed in neural progenitors within the VZ initially, but then is gradually downregulated as neurogenesis proceeds. Also, SOX3 expression is mutually exclusive to early neuronal marker and other differentiated neural markers in mouse telencephalon (N. Rogers, unpublished data) and chick NT (Bylund et al., 2003), suggesting it as a neural progenitor marker in these contexts. Moreover, some SOX3 positive cells was found to co-localise with SOX2, a marker for adult neural stem cell (Ellis et al., 2004; Ferri et al., 2004; N. Rogers, unpublished data), within the two adult neural stem cell niches, the subventricular zone of the CCx and the sub-granular zone of the Hpc (Ma et 93
al., 2009). Thus, the persistent SOX3 expression in the mature SCO indicates that it may possess some neural stem cell properties.
In support of the above, SOX2 and NES, a neural progenitor marker, were also detected in the adult mouse and rat SCO respectively (Chouaf-Lakhdar et al., 2003; Kuroda et al., 2005; Rodda et al., 2005, personal commincation, N. Rogers). To further this, proliferation study demonstrated that adult rat SCO is capable of self-renewal (Chouaf- Lakhdar et al., 2003). In addition, the SCO is one of all CVOs that express NES. Interestingly, in vitro cell cultures from three other CVOs, the organum vasculosum of the lamina terminalis, the median eminence and the subfornical organ, are capable of forming neurospheres which later underwent proliferation and neurogenesis (Bennett et al., 2009). When transplanted in vivo, these cells migrated towards the subventricular zone of CCx and differentiated into neurons with other resident neural stem cells (Bennett et al., 2010). Neurospheres formation is often a good indication of neural stem cells properties (Bennett et al., 2010; Eriksson et al., 2003). Thus, a neurosphere assay from a dissected adult SCO may be carried out to verify the above proposed neural stem cells property of the SCO. 94
Chapter4: Analysis of SCO development in Sox3 transgenic mice
4.1 Introduction
SCO is one of the earliest CNS organs to develop during embryogenesis (Rakic and Sidman, 1968). Thus, further analysis of SCO development in the Sox3 transgenic mouse model was required to determine the primary pathology of the CH phenotype. However, Sr/+ and Nr/+ mice had a low penetrance of overt CH. Thus, this deemed studies of the embryonic SCO development in Nr/+ and Sr/+ mice difficult, since it is not possible to determine whether a particular embryo will ultimately develop overt CH. As a result, further characterisation of SCO development was carried out in Sr/+; Nr/+ double hemizygote embryos, as this genotype resulted in a near complete penetrance of overt CH in adults. Moreover, analysis in Sr/+; Nr/+ double hemizygotes enabled further investigation (in a mouse model with higher Sox3 dosage) of whether the frequency of the SCO dysmorphology is dependent on Sox3 dosage and positively correlating with the penetrance of overt CH (as discussed in 3.3.5 and 3.3.6 above).
Prior to detailed cellular and molecular analyses, a morphological study of the SCO development in Sr/+; Nr/+ mice was carried out at a range of embryonic timepoints. The temporal mapping of SCO dysmorphology during development allowed appropriate SCO developmental stages to be chosen for further cellular and molecular analyses. cSox3 overexpression inhibits neuronal differentiation in chick NT (Bylund et al., 2003; Holmberg et al., 2008). Hence, an initial hypothesis of Sox3 acting as a SCO differentiation inhibitor was proposed. In order to test this hypothesis, a thorough interrogation of the SCO developmental cellular processes, namely proliferation, differentiation and cell death was performed in Sr/+; Nr/+ embryos following the morphological analysis.
4.2 Results 95
4.2.1 LV expansion and SCO aplasia in 18.5 dpc Sr/+; Nr/+ embryos
The near complete penetrance of overt CH suggested that non-overt CH was rare and that the CH had an earlier onset due to the higher penetrance in Sr/+; Nr/+ double hemizygotes in comparison to either single hemizygous Sr/+ or Nr/+ mice. Thus, Sr/+; Nr/+ mice were analysed at the earliest timepoint when ventricular expansion indicative of hydrocephalus can be detected as the first step of SCO dysmorphology characterisation. Histological study was carried out in Sr/+; Nr/+ embryos at 18.5 dpc, a timepoint when the volume of the LV is small with the ventricular walls closely positioned next to each other due to the birth of most telencephalic neurons (Guillemot, 2005), such that any LV dilatation can be readily detected. Detailed analysis revealed an expansion of the LV in all 18.5 dpc Sr/+; Nr/+ embryos (n=5) (Figure 4.1 A and B). Thus, this data confirmed an early onset of hydrocephalus which agrees with the near complete penetrance of overt CH in Sr/+; Nr/+ double hemizygotes.
As shown in prevous chapter, both the endogenous and transgenic Sox3 were robustly expressed in the developing and mature SCO and that a SCO dysplasia was highly associated with CH phenotypes in single hemizygous Sox3 transgenic adult mice. The presence of LV dilatation, a hallmark of hydrocephalus, in Sr/+; Nr/+ individuals at 18.5 dpc thus prompted the initial investigation of the SCO morphology of these embryos. In WT sagittal sections, the SCO contained densely packed elongated ependymals cells that lined the dorsal most aspect of the dienecphalon, with a distinctive PR to its anterior and a MR intercepting its AP axis (Figure 4.1 C). However, in Sr/+; Nr/+ embryos, elongated pseudostratified ependymal SCO cells could not be identified in the dorsal diencephalon (n=3). The PR or MR were also not evident (Figure 4.1 E). Somewhat densely packed elongated cells that morphologically resembled SCO cells were observed near the very posterior limit of the SCO region. Under high magnification, WT SCO displayed pseudostratifed columnar cells with their nuclei densely packed basally, leaving their cell bodies adjacent to the ventriculear area (Figure 4.1 D). However, nuclei of the SCO-like cells residing at the posterior limit of the SCO region in the Sr/+; Nr/+ embryos failed to localise basally and were found immediately dorsal to the ventricular area (Figure 4.1 F). Previous study found that nuclei of SCO cells migrate and localise basally during differentiation by 18.5 dpc (Rakic and Sidman, 1968). The lack of basal localisation in these cells thus indicate that they might be undifferentiated SCO remnant. WT Sr/+; Nr/+ A B
PG PC LV SCO PC Aq Aq LV
18.5 dpc LV LV Figure 4.1: Sr/+; Nr/+ mice displayed hallmarks of hydrocephalus and SCO aplasia at 18.5 dpc. A-B: LV dilatation indicating hydrocephalus was apparent in 18.5 dpc Sr/+; Nr/+ embryos. C-F: SCO elongated ependymal cells were absent from mice, although somewhat elongated SCO-like cells (hash) were 18.5 dpc Sr/+; Nr/+ observed at the posterior limit of the SCO (E). Nuclei of SCO cells were localised C D basally with a collection of cell bodies apically in the WT SCO (D). Howevever,
PG SCO-like ependymals cells failed to localise their nuclei basally and they were packed less densely then those of WT differentiated SCO cells at 18.5 dpc (F). (D)
SCO and (F) are boxed regions of (C) and (E) respectively. H & E stains (A and B). Nissl WT Aq stains (C-F). Arrowhead: PR. Doublet arrow: MR. Asterisk: missing PG. Equal SCO Aq sign: missing PR. Coronal sections (A-B). Sagittal sections (C-F). Top to bottom: dorsal to ventral. Left to right: anterior to posterior (C-F). Scale bar: 1mm (A and F B), 400µm (C and E), 100µm (D and F). E
# * Aq # Aq
Sr/+; Nr/+ Sr/+; = 96
Alternatively, these cells could be an attenuated cell type that is not found within the SCO region during normal development.
4.2.2 Normal morphology of the SCO at 12.5 dpc Sr/+; Nr/+ embryos
The SCO aplasia at 18.5 dpc in Sr/+; Nr/+ embryos demonstrated that the maldevelopment had an earlier embryonic origin. SCO morphology was thus analysed in an ascending chronological order in an attempt to identify the developmental timepoint at which the SCO morphological defect first appears. Thus, histology analysis of the 12.5 dpc Sr/+; Nr/+ SCO was first performed. According to previous literature, SCO primordium can be easily identified as the invaginated neuroepithelium ventral to the PC at 12.5 dpc (Rakic and Sidman, 1968). Like other non-SCO neuroepithelial cells at that timepoint, the cells within the WT SCO primordium were elongated and made up predominately of nuclei with minimal cell bodies. Numerous PC axonal tracts appeared as teardrop-like features residing immediately dorsal to the SCO primordium (Figure 4.2 A, B, E and G). Interestingly, no significant histological dysmorphology was discovered in 12.5 dpc Sr/+; Nr/+ SCO as invaginated SCO primordium was evident (n=5), with densely packed elongated ependymal cells (Figure 4.2 C, D, F and H). The abundance and shape of the PC appeared normal in four out of five embryos, while one embryo demonstrated slightly disorganised PC possibly due to biological variation. No overt developmental defects in other dorsal diencephalic features including the Pt, EpTh, DTh and VTh, to WT were evident in Sr/+; Nr/+ embryos (Figure 4.2 A-D).
4.2.3 SCO dyspalsia from 13.5 dpc in Sr/+; Nr/+ embryos
Since no significantly morphological defect was identified at 12.5 dpc Sr/+; Nr/+ SCO, histological analysis was then subsequently carried out in 13.5 dpc Sr/+; Nr/+ embryos. In order to give rise to the PG from the RP of prosomere 2, the neuroepithelium that is anterior to the SCO dorsally evaginates during development. This dorsally evaginated neuroepithelium then becomes the PG primordium while the invagination itself becomes the PR. As development progresses, the evaginated PG primordium eventually pinches off to form an enclosed lobule-like PG (Ko et al., 2005; Rakic and Sidman, 1968). Interestingly, in 97
13.5 dpc WT embryos, the SCO invagination became more prominent. Moreover, both the PG primordium and the PR evagination were clearly evident. The MR that intercepted the AP axis of the SCO was also apparent (Figure 4.3 A and C). Interestingly, when studied histologically, SCO dysplasia was already obvious in 13.5 dpc Sr/+; Nr/+ embryos (n=2). The SCO primordium was uneven and noticeably thinner although elongated SCO cellular morphology was maintained (Figure 4.3 A and B). At one of the thinnest points, the SCO was only a few nuclei thick between the ventral tip of the PC and the ventricular area (Figure 4.3 B). This was in contrast to the WT SCO which was eight to ten nuclei thick (Figure 4.3 A). The PC was grossly disorganised with reduction in both the abundance and size of their teardrop-like morphology (Figure 4.3 B). More apparently, the PG primordium and the evagination that forms the PR were no longer observed (Figure 4.3 D).
Subsequently at 14.5 dpc, the WT SCO primordium became thicker than that of 13.5 dpc and displayed a more prominent MR (Figure 4.3 E). However, the SCO dysplasia was exacerbated in 14.5 dpc Sr/+; Nr/+ embryos when compared to one day earlier (n=2). Elongated ependymal cells were no longer observed at the anterior and medial SCO spatial position (Figure 4.3 F). Instead, Sr/+; Nr/+ SCO cells displayed a similar density and morphology to that of its surrounding cells. The PC that was immediately dorsal to the anterior SCO region was absent. Only a small group of PC tract was found near the posterior limit of the SCO. Interestingly, SCO-like elongated and densely organised ependymal cells were found posterior to the small group of surviving PC (Figure 4.3 F).
4.2.4 Morphologically and functioally defective SCO from 15.5 dpc Sr/+; Nr embryos
The SCO morphological study was further pursued in 15.5 dpc Sr/+; Nr/+ embryos in order for a thorough characterisation of the defect. In 15.5 dpc WT embryos, SCO cells morphology was quite distinct as most neighbouring cells have differentiated and lost their elongated cell shape of the earlier neuroepithelium. SCO cells became more elongated and more densely organised into a pseudostratified layer. The SCO was well developed and both the MR and PR were clearly defined in WT embryos (Figure 4.4 A, C and D). In contrast, histological analysis of section series that accounts the entire dorsomedial region of Sr/+; Nr/+ embryonic brain has failed to identify any diencephalic RP feature resembling a SCO or PG (n=3). Sparsely packed non-elongated cells were present in the SCO region (Figure 4.4 B, WT Sr/+; Nr/+
A PC C EpTh EpTh PC SCO 3V Aq SCO DTh Aq DTh Figure 4.2: SCO was morphologically normal in 12.5 dpc Sr/+; Nr/+ embryos. A-H: WT 12.5 dpc SCO was ventrally invaginated (arrows) and was characterised by the presence of dorsally residing VTh VTh PC (A, E and G). Sr/+; Nr/+ SCO was also ventrally invaginated (arrows) and was morphologically indifferent to that of WT (C, F). Sr/+; Nr/+ SCO cellular density and localisation (H) were compa- rable to those of WT (G). The PG, EpTh, Pt, DTh and VTh were all found histologically normal in Sr/+; Nr/+ embryos (A-D). B D Note that sections (A) to (B) and (C) to (D) are in a medial to PC 12.5 dpc PC Pt EpTh EpTh lateral direction. (E) and (F) are boxed regions of (A) and (B) 3V Pt SCO SCO 3V respectively. (G) and (H) are boxed regions of (E) and (F) respec- DTh Aq Aq Aq tively. I: A transverse section from Schambra (2008) indicating the
DTh plane of sectioning for (A) to (D). Nissl stains. Sagittal sections. Top to bottom: dorsal to ventral. Left to right: anterior to posterior.
VTh Scale bar: 800µm (A-D), 200µm (E-F), 100µm (G-H).
VTh
12.5 dpc E WT Sr/+; Nr/+ G H I PC B, D WT A, C SCO Aq PC F PC
PC Sr/+; Nr/+ Sr/+; SCO Aq 13.5 dpc
A C PG SCO
PC WT Figure 4.3: SCO morphological defect was apparent in Sr/+; EpTh Aq SCO Nr/+ embryos from 13.5 dpc. A-B: WT SCO displayed thick SCO primordium (double-headed arrow) (A). In Sr/+; Nr/+ embryos, SCO primordium was thinner (double-headed arrow) B D with evidently less abundant and disorganised PC (B). C-D: The PG, PR (arrowhead) and an identifiable MR (doublet arrow) were * formed by 13.5 dpc in the WT embryos (C), while No PG and PR = were identified in Sr/+; Nr/+ embryos (D). E-F: SCO dysmor- phology was exacerbated by 14.5 dpc in Sr/+; Nr/+ embryos. In EpTh Sr/+; Nr/+ Sr/+; PC WT embryos, the SCO primordium was much thicker. MR also Aq became more prominent (doublet arrow) (E). In Sr/+; Nr/+ indi- SCO viduals, SCO was dysmorphic with little elongated ependymal SCO cells recognised at 14.5 dpc, and thus MR was no longer identified. Defective PC morphology was also present, with only some residual PC residing near the posterior limit of the SCO. WT Sr/+; Nr/+ SCO-like ependymal cells (plus sign) were found posterior to the group of residual PC (F). Nissl stains. Asterisk: absence of PG. E F Equal sign: absence of PR. Hash: absence of SCO cells. Sagittal sections. Top to bottom: dorsal to ventral. Left to right: anterior to posterior. Scale bar: 100µm (A-B), 200µm (C-D), 400µm (E-F).
+
PC PC
14.5 dpc Aq Aq # SCO Aq 15.5 dpc
WT Sr/+; Nr/+ C D A B PG
* PC PC PG WT Aq SCO PC 3V PC EpTh 3V Aq Pt SCO SCO # # Pt E F
Aq
Aq PC * PC
Sr/+; Nr/+ Sr/+; # EpTh = # Aq Aq
15.5 dpc G H I PC
Aq PC
WT M C, E Aq G, J J K L
PC Aq PC Sr/+; Nr/+ Sr/+; Aq Figure 4.4: Severe SCO dysplasia in Sr/+; Nr/+ embryo at 15.5 dpc. A-F: In contrast to the thick WT SCO primordium (A and C), Sr/+; Nr/+ embryos lacked SCO primordium (hash) at the SCO region, although some residual PC was found dorsal to the posterior limit of the SCO (B and E). High magnification showed that unlike WT SCO (D), cells ventral to the residual PC in Sr/+; Nr/+ mice (F) were not as elongated and densely packed. PG and PR were absent (asterisks and equal sign respectively), while MR was not found in Sr/+; Nr/+ individuals (B and E). (D) and (F) are boxed regions of (C) and (E) respectively. G-L: SCO remnant was observed at the very posterior limit in Sr/+; Nr/+ embryos (J-L). They were histologically comparable in both cell density and cell shape to those of WT (G-I). (H) and (K) are boxed regions of (G) and (J) respectivley, while (I) and (L) are boxed regions of (H) and (K) respectivley. Note that sections (C) to (G) and (E) to (J) are in a medial to lateral direction. M: A transverse section from Schambra (2008) indicating the plane of sectioning for (C), (E), (G) and (J). Nissl stains. Coronal sections (A and B). Sagittal sections (C-L). Top to bottom: dorsal to ventral. Left to right: anterior to posterior (C-L). Scale bar: 1mm (A and B), 200µm (C and E), 500µm (G and J), 100µm (D, F, H and K), 50µm (I and L). 98
E and F). Once again, these cells were morphological similar to their neighbouring tissue. No PR or MR was identified. Similar to 14.5 dpc Sr/+; Nr/+ embryos, a very small group of deformed and diminished PC were found residing at the posterior end of the SCO region. Dense elongated cells indicative of SCO remanent was found posterior to the small surviving group of PC (Figure 4.4 G, H, J and K). These cells of SCO remnant showed elongated cell morphology and were packed in a density that was comparable to those from WT. In fact, their morphology was not distinguishable from that of WT under high magnification (Figure 4.4 I and L).
In order to investigate the function of the dysplasic SCO at 15.5 dpc, RF immunofluorescence was performed. RF started to become detectable in some WT SCO cells from 15.5 dpc (Figure 3.20 A). Analysis of sagittal sections series has identified that RF was expressed in some SCO cells from the WT embryos at 15.5 dpc. In contrast, no RF signal was observed from the dysplasic SCO or the SCO remnant of stage-matched Sr/+; Nr/+ embryos (n=3) (Figure 4.5 B). In order to exclude delayed differentiation as one of the reasons for the lack of RF observed in the 15.5 dpc Sr/+; Nr/+ SCO, RF expression was further investigated at 18.5 dpc, by which the SCO is fully differentiated. Immunofluorescence analysis of the WT SCO at 18.5 dpc revealed that RF was highly and ubiquitously expressed (Figure 4.5 C and D). In contrast, Sr/+; Nr/+ embryos were devoid of any detectable RF signal in the SCO region including those SCO-like cells at the posterior region (n=3) (Fig 4.5 E and F). This indicate that the Sr/+; Nr/+ SCO lacked functional mature SCO cells at 18.5 dpc (3 days after RF expression is initiated in WT embryos) and argues against the case of a functional but developmentally delayed Sr/+; Nr/+ SCO at 15.5 dpc. Given the striking morphological defect and the absence of RF, it seems likely that SCO cells had already lost their WT morphology and function by15.5 dpc.
To summarize, in Sr/+; Nr/+ mice, the SCO was morphologically indifferent to that of WT at 12.5 dpc. SCO dysmorphology, namely thinner SCO primordium associated with diminished and disorganised PC became apparent from 13.5 dpc. An absence of the PG was also evident. Both the SCO and PC dysplasia have exercebated from 13.5 dpc. At 15.5 dpc, morphologically and functionally normal SCO cells were absent, despite some cells resembling SCO cell morphology were left at the posterior limit. The SCO defect in Sr/+; Nr/+ embryos was further demonstrated by SCO aplasia and the lack of RF production in any diencephalic region at 18.5 dpc embryos. It is interesting to note that the SCO remnant 99 observed at the posterior limit of the 15.5 dpc Sr/+; Nr/+ SCO did not appear to develop into mature SCO cells as no functional SCO cell was detected in Sr/+; Nr/+ embryos at 18.5 dpc.
4.2.5 Developmental defects in the diencephalon of Sr/+; Nr/+ embryos are restricted to the RP
In addition to the RP, transgene-derived SOX3 was widely expressed in the lateral regions of the dorsal diencephalon at 12.5 dpc. As a result, these lateral diencephalic features might also be defective. Thus, a histological analysis was performed to study the morphology of the dorsal diencephalic lateral regions during development. Series of coronal sections from 15.5 dpc WT and Sr/+; Nr/+ embryos were analysed in a caudal to rostral sequence. Habenuclar commissure (HbC) are axonal tracts that connect both lateral diencephalic habenular nuclei and form at the dorsal midline by 13.5 dpc (Funato et al., 2000; Wahlsten, 1981). In Sr/+; Nr/+ embryos, RP features anterior to the SCO, namely the PG (discussed in 4.2.3 and 4.2.4) (Figure 4.6 G’ and H’) and the HbC, were missing at the dorsal midline (n=3) (Figure 4.6 I’and J’). In fact, habenular nuclei were formed with HbC projected dorsomedially towards the midline as in WT (Figure 4.6 I, J, I’ and J’). However, in these Sr/+; Nr/+ embryos, HbC failed to meet and cross at the midline (Figure 4.6 I, K, I’ and K’). Intriguingly, lateral diencephalic features such as the Pt, EpTh, DTh and VTh were both histologically and morphologically normal (n=3) (Figure 4.6 A-K and A’-K’). In order to further investigate the formation of diencephalic dorsolateral structures, histological analyses were also performed at WT and Sr/+; Nr/+ embryos at 18.5 dpc. Similarly, the Pt, EpTh, DTh and VTh were present and morphologically normal in Sr/+; Nr/+ embryos (n=3) (Figure 4.7 A-F). Collectively, the restriction of the morphological defects to the dorsal diencephalic RP suggest that the development of dorsolateral diencephalic domains are less prone to an increased dosage of SOX3 in comparison to RP features.
4.2.6 Absence of apoptosis in SCO from WT and Sr/+; Nr/+ embryos
A histologically normal SCO was observed at 12.5 dpc Sr/+; Nr/+ embryos. By 13.5 dpc, the SCO primordium was thinner and by 14.5 dpc, a significant loss of SCO-like cells was apparent. The disappearance of SCO elongated ependymal cells suggested that the loss 15.5 dpc A
WT Sr/+; Nr/+ C D WT
PC RF SCO Aq PC O
Aq SCO Aq
B E F 18.5 dpc PC DAPI SCO PC Aq
3V SCO
Sr/+; Nr/+ Sr/+; Aq P PC + SOX3 # Aq RF PC
Figure 4.5: Sr/+; Nr/+ SCO was functionally defective at 15.5 dpc and 18.5 dpc. A-B: RF expression was evident in some SCO cells at 15.5 dpc in WT# embryos (A), while no RF was detected in Sr/+; Nr/+ SCO region or SCO remnant at the posterior limit (plus sign) (B). Note that SOX33V was also expressed by ventricular lining ependymal cells and SCO cells in both WT and Sr/+; Nr.+ embryos at 15.5 dpc. C-F: RF was ubiquitously expressed by mature WT SCO cells from 18.5 dpc embryos (C) but was not detected in the SCO region or from SCO-like cells residing at the SCO posterior limit (arrowheads) of Sr/+; Nr/+ embryos (D). (E) and (F) are DAPI stainings of (C) and (D) resepctively. Immunofluorescence. Hash: absence of SCO primordium. Sagittal sections. Top to bottom: dorsal to ventral. Left to right: anterior to posterior. Scale bar: 200µm (A-B), 100µm (C-F). WT Sr/+; Nr/+ WT Sr/+; Nr/+ F’ A SCO A’ SCO F PG PC * Di Di Pt Pt PC SCO # Pt Aq Aq Aq ChP ChP Aq Pt Tel Tel Tel DTh Mes Tel VTh Mes DTh Di Di VTh
3V Rb Rb Cer Cer Cer Rb
PG G’ SCO B’ SCO G B PC * Di Pt SCO # Pt Aq Pt Pt Tel Tel ChP Tel Aq Tel Aq Di ChP Tel Aq DTh DTh VTh Di Di VTh
Rb 3V Cer Rb Cer Cer Cer 3V C’ H PG H’ C PC SCO * # SCO PC Tel PC Tel # Pt 15.5 dpc Pt Pt Pt Tel Tel Tel Aq Tel Aq Aq ChP Di ChP Aq DTh DTh Di VTh VTh
Cer 3V Rb Rb 3V Cer D D’ I HbC I’ 3V HbC HN SCO # PC Tel EpTh Tel Tel PC 3V Tel HN Tel Pt EpTh Aq Tel Pt Tel DTh Tel Aq DTh Di Di VTh VTh
3V Rb Rb
3V E E’ J HbC J’ PC Tel Tel # HN EpTh 3V SCO Tel HbC PC Tel Tel Aq Pt Tel EpTh Tel DTh HN Tel Pt DTh Aq Di VTh Di VTh
3V 3V Rb
WT Sr/+; Nr/+ L K K’ HbC HbC
HN HN 15.5 dpc A A’
J J’ Figure 4.6: Dorsal diencephalic defect was restricted to the SCO and its anterior RP at 15.5 dpc Sr/+; Nr/+ embryos. A-H and A’-H’: SCO remnant once again was observed at caudal (posterior) sections in Sr/+; Nr/+ embryos (A’ and B’) but was absent from rostral (anterior) regions (hash) (C’-H’), while WT SCO was readily identified over numerous sections when analysed in a caudal to rostral direction (A-H). Similarly, Sr/+; Nr/+ embryos lacked PG (asterisk) (F’-H’) and PC in the most rostral sections (G’ and H’). Lateral diencephalic domains including the Pt, DTh and VTh appeared histologically normal. I-K and I’-K’: RP defect but not lateral defect was again observed in Sr/+; Nr/+ dorsal diencephalic region anterior to the SCO. HbC failed to cross the midline (arrowhead) (K’), despite the normal morphology of habenular nuclei (HN) (I’-K’). (K) and (K’) are boxed regions of (I) and (I’) respectively. Note that (A) to (J) and (A’) to (J’) are sections series in a caudal to rostral direction. L: A sagittal section from Schambra (2008) indicating the plane of sectioning for (A) to (J) and (A’) to (J’). Nissl stains. Tel: telencephalon. Di: Diencephalon. Rb: Rhomben- cephalon. Coronal sections. Top to bottom: dorsal to ventral. Scale bar: 1mm (A-J, A’-J’), 200µm (K and K’). WT Sr/+; Nr/+ A D Tel Pt Tel Pt EpTh Mes DTh Mes EpTh DTh VTh VTh
Rb Di Rb Di B E Tel PG * Tel Pt Mes EpTh Pt Mes 18.5 dpc EpTh DTh VTh DTh VTh Rb Di Rb Di C F Tel Pt Tel Pt EpTh EpTh Mes Mes DTh DTh VTh VTh Rb Di Rb Di
NOTE: This figure is included in the print copy of the thesis held in the University of Adelaide Library.
Figure 4.7: Dorsolateral diencephalic domains were histologically normal in Sr/+; Nr/+ embryos at 18.5 dpc. A-F: The Pt, EpTh, DTh and VTh from Sr/+; Nr/+ embryos were comparable to those of WT at 18.5 dpc. Note that (A) to (C) and (D) to (F) are section series in a lateral to medial direction. Nissl stains. G: A transverse section from Schambra (2008) indicating the plane of sectioning for (A) to (C) and (D) to (F). Nissl stains.Tel: telencephalon. Di: diencephalon, Mes: mesencephalon. Rb: Rhombencephalon. Sagittal sections. Top to bottom: dorsal to ventral. Left to right: anterior to posterior. Scale bar: 500µm. 100
of SCO-fated precursors might be the result of apoptosis. Due to the rapid SCO morphological deterioration between 12.5 dpc and 14.5 dpc, SCO apoptosis was investigated from 12.5 to 14.5 dpc in both WT and Sr/+; Nr/+ embryos. Cellular apoptosis induces DNA fragmentation, resulting in DNA breaks. The terminal dUTP nick ending labelling (TUNEL) assay allows the visualisation of apoptotic cells through the attachment of labelled terminal dUTP to the end of these DNA breaks. Thus, programmed cell death was initially investigated by TUNEL assay in Sr/+; Nr/+ and WT embryos using fluorescence labelled terminal dUTP. Interestingly, no TUNEL signal was found in both WT and Sr/+; Nr/+ embryos at 12.5 dpc (n=2) and 13.5 dpc (n=2) (Figure 4.8 A-D and K-N). Yet, TUNEL signals were widely observed in the positive controls, which were SCO neighbouring tissues that were treated with DNases to induce DNA breaks (Figure 4.8 E-J and O-T). Together, they supported that these TUNEL assays were valid and suggested that both the WT and Sr/+; Nr/+ SCO primordia lacked apoptotic cells at 12.5 dpc to 13.5 dpc. However, the lack of endogenous TUNEL signal, i.e. an area showing positive TUNEL signals without DNase- induced DNA breaks, in this assay posed a possibility that the TUNEL technique applied was not adequately sensitive to detect endogenous apoptosis within the SCO. Investigation of SCO programmed cell death was thus further pursued using on alternative technique.
Immunodetection of activated Caspase-3 has been widely used as a technique for the investigation of apoptosis. During apoptosis, the proenzyme Caspase-3 is proteolytically cleaved such that it becomes activated. Activated Caspase-3 is then required to further cleave the ADP ribosome polymerase for downstream effect of apoptosis (Nicholson et al., 1995; Schlegel et al., 1996; Urase et al., 1998). Thus, the SCO cellular apoptosis was interrogated in Sr/+; Nr/+ and WT embryos at 12.5 to 14.5 dpc through immunodetection of activated Caspase-3. Previous publication found that the trigeminal ganglia (TrG) is positive for endogenous activated Caspase-3 at 12.5 dpc during murine CNS development (Urase et al., 1998). The TrG hence, served as an endogenous positive control in this study. Intriguingly, immunodetection of activated Caspase-3 analysis has discovered a lack of developmental apoptosis at 12.5 dpc and 13.5 dpc WT SCO (Figure 4.9 A-C and Figure 4.10 A-C). Similarly, Sr/+; Nr/+ SCO from 12.5 dpc (n=1) and 13.5 dpc (n=1) were devoid of any activated Caspase-3 signal, indicating the absence of apoptosis (Figure 4.9 D-F and Figure 4.10 D-F). Although it was unlikely that the increased apoptosis at 14.5 dpc would account for the loss of SCO cells in Sr/+; Nr/+ embryos as most SCO cells are “lost” by this stage, further analysis of activated Caspase-3 was carried out to confirm this was true. Again, activated 101
Caspase-3 positive apoptotic cell could not be identified in the WT SCO nor the SCO region and posterior SCO remnant of Sr/+; Nr/+ embryos at 14.5 dpc (n=1) (Figure 4.11 A-F). In contrast to the SCO, cells within the TrG demonstrated specific activated Caspase-3 signals from both 12.5 dpc, 13.5 dpc and 14.5 dpc WT and Sr/+; Nr/+ embryos (Figure 4.9 G-H, Figure 4.10 G-H and Figure 4.11 G-H). Some of these activated Caspase-3 signals were closely associated with chromatin condensation, a morphology common in apoptotic cells, when observed with DAPI nuclear staining (Figure 4.9 G-H, Figure 4.10 G-H and Figure 4.11 G-H).
Collectively, the data from both the TUNEL assay and the activated Caspase-3 analysis suggest that the WT SCO was not undergoing developmental apoptosis and that SOX3 overdosage did not elevate SCO apoptosis between 12.5 dpc and 14.5 dpc during embryogenesis. As a result, programmed cell death is unlikely to account for the loss of cells with SCO morphology in Sr/+; Nr/+ individuals observed from 13.5 dpc onwards.
4.2.7 Overproliferation of the SCO primordium of 12.5 dpc Sr/+; Nr/+ embryos
SCO proliferation peaks during 10.0 dpc to13.0 dpc (Rakic and Sidman, 1968), at a time when transgenic Sox3 was detected. Thus, the effect of SOX3 in SCO proliferation was investigated at 12.5 dpc, a timepoint when an evaginated SCO primordium can be firstly observed. Two strategies were initially employed to study this cellular mechanism, namely BrdU incorporation and Ki67 immunofluorescence. Ki67 is a nuclear protein associated with ribosomal RNA synthesis and can be found in cells during active cell cycle, i.e. G1, S, G2 and M phases. Thus, it is highly associated with proliferation. Moreover, Ki67 has a short half-life (60-90 minutes) and has been widely used as a proliferative marker (Bullwinkel et al., 2006; Nabi et al., 2008; Rahmanzadeh et al., 2007). BrdU, on the other hand, is a thymidine analogue which can be incorporated into DNA during replication. Actively proliferating cells that have undergone S phase then can be detected with BrdU antibody.
Initial study with Ki67 found a ubiquitous expression in all SCO cells in both WT and Sr/+; Nr/+ embryos, although there were a few apically residing cells displaying a more intense Ki67 signal (n=1) (Figure 4.12 A-F). This indicated that all SCO cells were within the active cell cycle. This was expected as previous study found that SCO does not subside from Sr/+; Nr/+ WT M K Sr/+; Nr/+ WT C A PC PC TUNEL SCO SCO SCO PC TUNEL SCO PC Aq Aq SCO L N Aq Aq PC SCO B D PC DAPI SCO SCO SCO SCO Aq PC Aq DAPI PC O R Aq TUNEL Aq H E TUNEL P S F I +DNase DAPI +DNase DAPI Q T J G
TUNEL+DAPI TUNEL+DAPI
12.5 dpc 12.5 13.5 dpc 13.5 Figure 4.8: No apoptotic signal was observed in WT and Sr/+; Nr/+ SCO at 12.5 dpc and 13.5 dpc by TUNEL assay: A-D and K-N: There was no TUNEL signal indicative of apoptosis identified in the SCO of either WT or Sr/+; Nr/+ embryos at 12.5 and 13.5 dpc. (C) and (D) are DAPI stainings of (A) and (B) respectively, while (L) and (N) are DAPI stainings of (K) and (M) respectively. E-J and O-T: Positive TUNEL signals were observed from tissue immediately lateral to SCO that was treated with DNase to induce DNA breakage (E, H, O and R). (F) and (I) are DAPI stainings of (E) and (H) respectively, while (P) and (S) are DAPI stainings of (O) and (R) respectively. (G) is a merge of (E) and (F), (J) is a merge of (H) and (I), (Q) is a merge of (O) and (P) and (T) is a merge of (R) and (S). Note the non-specific TUNEL signals due to speckles (arrowheads). Immunofluorescence. Sagittal sections. Top to bottom: dorsal to ventral. Left to right: anterior to posterior. Scale bar: 100µm (A-D and K-N), 50µm (E-J and O-T). WT Sr/+; Nr/+ A D Caspase-3 Activated
PC PC
SCO SCO Aq Aq B E Background
PC PC
SCO SCO Aq Aq C F DAPI 12.5 dpc PC PC
SCO SCO
Aq Aq
SCO Activated Activated Caspase-3 Background DAPI Caspase-3 + DAPI G H I J WT
K L M N Sr/+; Nr/+ Sr/+;
TrG Figure 4.9: Absence of activated Caspase-3 positive, apoptotic cells in both the WT and Sr/+; Nr/+ SCO at 12.5 dpc embryos. A-F: Activated Caspase-3 was not detected in 12.5 dpc SCO from either WT or Sr/+; Nr/+ embryos (A and C). There were non-specific signals from speckles and autofluores- cence haematoma in activated Caspase-3 fluorescence (yellow arrowheads) (A and C), which were also apparent in an unrelated fluorescence (background) (B and E). (B) and (E) are unrelated fluorescence signals of (A) and (C) respectively, while (C) and (F) are DAPI stainings of (A) and (C) respectively. G-N: Activated Caspase-3 was readily detected in the TrG (positive control) of both WT and Sr/+; Nr/+ embryos (G and K). Some activated Casapase-3 signals were detected in nuceli displaying the character- istic apoptotic figure, chormatin condensation (white arrowheads) (I, J, M and N). (H) and (L) are urelated fluorescence signals of (G) and (K) respectively, while (I) and (M) are DAPI stainings of (G) and (K) respectively. (J) is a merge of (G) and (I), while (N) is a merge of (K) and (M). Immunofluores- cence. Sagittal sections. Top to bottom: dorsal to ventral. Left to right: anterior to posterior. Scale bar: 100µm (A-F), 50µm (G-N). WT Sr/+; Nr/+ A D Caspase-3 Activated
PC PC
SCO SCO Aq Aq B E Background
PC PC
SCO Aq SCO Aq C F DAPI 13.5 dpc
PC PC
SCO Aq SCO Aq
SCO Activated Activated Caspase-3 Background DAPI Caspase-3 + DAPI G H I J WT
K L M N Sr/+; Nr/+ Sr/+;
TrG Figure 4.10: Absence of activated-Caspase 3 positive, apoptotic SCO cells in the SCO of 13.5 dpc WT and Sr/+; Nr/+ embryos. A-F: Activated Caspase-3 was absent from 13.5 dpc SCO in WT or Sr/+; Nr/+ embryos (A and C). Non-specific signals from speckles and autofluorescence haematoma were observed in activated Caspase-3 fluorescence (yellow arrowheads) (A and C). Identical non-specific signals were also identified in an unrelated fluorescence (background) (B and E). (B) and (E) are unre- lated fluorescence signals of (A) and (C) respectively, while (C) and (F) are DAPI stainings of (A) and (C) respectively. G-N: Activated Caspase-3 was readily detected in the TrG (positive control) in both WT and Sr/+; Nr/+ embryos (G and K). Some activated Casapase-3 signals were detected in nuceli display- ing the characteristic apoptotic figure, chormatin condensation (white arrowheads) (I, J, M and N). (H) and (L) are urelated fluorescence signals of (G) and (K) respectively, while (I) and (M) are DAPI stain- ings of (G) and (K) respectively. (J) is a merge of (G) and (I), while (N) is a merge of (K) and (M). Immunofluorescence. Sagittal sections. Top to bottom: dorsal to ventral. Left to right: anterior to poste- rior. Scale bar: 100µm (A-F), 50µm (G-N). WT Sr/+; Nr/+ A D Caspase-3 Activated
PC PC + SCO # Aq Aq B E Background
PC PC + SCO # Aq Aq C F
PC DAPI PC +
14.5 dpc SCO # Aq Aq
SCO Activatd Activated Caspase-3 Background DAPI Caspase-3 + DAPI G H I J WT
K L M N Sr/+; Nr/+ Sr/+;
TrG Figure 4.11: Absence of activated Caspase-3 positive, apototic cells in the WT SCO and the SCO region or the SCO remnant of Sr/+; Nr/+ embryos at 14.5 dpc. A-F: Activated Caspase-3 was not detected from the WT SCO (A) and the SCO region or SCO remnant (plus sign) of Sr/+; Nr/+ (D) embryos at 14.5 dpc. Non-specific signals from speckles and autofluorescence haematoma were observed in activated Caspase-3 fluorescence (yellow arrowheads) (A and D). Identical non-specific signals were also identified in an unrelated fluorescence (background) (B and E). (B) and (E) are unrelated fluorescence signals of (A) and (C) respectively, while (C) and (F) are DAPI stainings of (A) and (C) respectively. G-N: Activated Caspase-3 was readily detected in the TrG (positive control) in both WT and Sr/+; Nr/+ embryos (G and K). Some activated Casapase-3 signals were detected in nuceli displaying the characteristic apoptotic figure, chormatin condensation (white arrowheads) (I, J, M and N). (H) and (L) are urelated fluorescence of (G) and (K) respectively, while (I) and (M) are DAPI stainings of (G) and (K) respectively. (J) is a merge of (G) and (I), while (N) is a merge of (K) and (M). Immunofluorescence. Hash: absence of SCO cells. Sagittal sections. Top to bottom: dorsal to ventral. Left to right: anterior to posterior. Scale bar: 100µm (A-F), 50µm (G-N). Sr/+; Nr/+ WT O D A Aq Aq SCO PC SCO SOX3 PC E B Aq PC Aq Ki67 SCO SCO PC C F Aq PC DAPI Aq SCO SCO PC G I Aq Aq SCO BrdU PC PC (A-F, andK-N), 50µm(G-J). bottom: dorsal toventral.Leftright: anteriortoposterior. Scalebar:100µm from haematoma(arrowheads). Immunofluorescence.Sagittalsections. Top to erating indexofthe cation ofBrdUpositivecells indicatedtherewasa1.5-foldincreaseinthe prolif- and M).(L)(N)are DAPI stainingsof(K)and(M)respectively. O:Quantifi- displayed highnumberof cellsundergoing DNA synthesisbyBrdUlabelling(K DAPI stainingsof(G)and(H)respectively. K-N:the VZ ofdorsalMGE localised basallyinthe WT (G),butnotinthe (I) incomparisontothe WT (G)byBrdUlabelling.Mostpositivecellswere G-J: Sr/+;Nr/+SCOhadanincreasenumberofcellsundergoing DNA synthesis respectively, while(C)and(F)areDAPIstainingsof (A) and(D)respectively. WT andSr/+;Nr/+SCO(BE).(B)(E)areKi67signalsof(A) and(D) ating cellsat12.5dpc. A-F: A ubiquitousKi67expression waspresentinboth Figure 4.12:Sr/+;Nr/+SCOhada1.5-foldincrease inthenumber ofprolifer- SCO SCO H J Aq Aq SCO Sr/+; Nr/+SCOat12.5dpc.Notethe non-specificsignals DAPI PC PC SCO M K LV LV VZ VZ Sr/+; Nr/+SCO(I).(H)and(J)are BrdU MGE N L LV LV VZ VZ
DAPI 12.5 dpc 12.5 102
peak proliferation until 13.0 dpc (Rakic and Sidman, 1968). The general signal of Ki67 although might be higher in Sr/+; Nr/+ SCO, a precise measurement of its immunofluorescence intensity was deemed difficult due to the ubiquitous expression. BrdU technique was subsequently employed, since it only labels cells that had undergone DNA synthesis, but not other phases of the cell cycle, given a brief timeframe. The detection of only a subset of cells undergoing the cell cycle i.e. cells at S-phase, allows the quantification of any difference that is too subtle to be detected with Ki67 during peak SCO proliferation.
The exact SCO cell cycle time at 12.5 dpc is unknown. However, previous study performed had assumed a cell generation time of 24 hours (Rakic and Sidman, 1968) and this was thus implied in the current experiment. In order to ensure specific labelling of cells that are in active S phase at the time of BrdU administration, embryos were harvested shortly (2 hours) after intraperitoneal BrdU injection into the gestating mouse. WT and Sr/+; Nr/+ embryos from the same litter were paired for analysis to compensate for any difference in BrdU labelling efficiency among litters. Immunofluorescence technique was used to detect BrdU that were incorporated into proliferating cells. A quantitative method was then employed to analyse SCO proliferation on tissue sections from WT and Sr/+; Nr/+ embryos. BrdU positive cells were scored as a ratio to the total number of SCO cells. In order to further compensate for differences in BrdU detection efficiency between tissue sections, proliferative index of the MGE was used as an internal reference. As discussed in 3.2.2, The VZ of the dorsal MGE is devoid of transgenic Sox3 expression at 12.5 dpc. However, the VZ of the MGE was found to have abundant BrdU positive proliferating cells at 12.5 dpc (Figure 4.12 K-N). Thus, the ratio of BrdU positive cells in the MGE was scored in a similar fashion to that of the SCO, i.e. the number of BrdU positive cells in the MGE to the total number of MGE cells. Subsequently, the ratio of BrdU positive cells in the SCO of each section was normalised to the ratio of BrdU positive cells in the MGE from the same section.
Detailed quantitative analysis found that the 12.5 dpc Sr/+; Nr/+ SCO had a 1.5-fold increase in the number of BrdU-positive cells in comparison to the WT (n=4) (Figure 4.12 O). This was consistent with the qualitative observation as immunofluorescence images demonstrated that the BrdU positive cells were markedly higher in density within the Sr/+; Nr/+ SCO when compared to the WT (Figure 4.12 G-J). Moreover, most of the BrdU positive cells resided basally within the WT SCO (Figure 4.12 G). This was in contrast to those of Sr/+; Nr/+ embryos, where BrdU positive cells were found throughout the entire 103
SCO primordium (Figure 4.12 I). In addition, the BrdU intensity appeared to be generally higher in the SCO from Sr/+; Nr/+ embryos, despite that the signal intensity has not been measured quantitatively. Collectively, these data indicate that Sr/+; Nr/+ SCO had a higher proliferative index to that of the WT.
4.2.8 Loss of SCO progenitors from 12.5 dpc in Sr/+; Nr/+ embryos
Since an increased dosage of cSox3 inhibits neural progenitor differentiation in chick NT (Bylund et al., 2003; Holmberg et al., 2008), it was originally hypothesised that SOX3 overexpression might have a similar effect on SCO progenitor proliferation. However, the analysis of SCO differentiation has been impeded by the lack of immature SCO marker up to the stage of this investigation. Hence, prior to further study, a detailed expression analysis was performed using the online brain gene expression atlas (Allen atlas, http://www.brain- map.org/) in order to identify potential developmental marker(s) for the SCO. The brain atlas has a comprehensive developmental gene expression database that is made up of series of in situ hybridisation on mouse CNS sections (Lein et al., 2007). First, a subset of genes was selected as they are expressed exclusively within the developing SCO or RP of the dorsal diencephalon during embryogenesis. Then, their levels of gene expression were carefully compared across different stages between 11.5 dpc to 18.5 dpc (given that the data was available from the brain atlas). Finally, gene(s) were selected as potential immature SCO marker if they were highly expressed within an immature SCO but not in a mature SCO. Interestingly, the gene fibrinogen C domain containing protein 1 (Fibcd1) was identified as a potential candidate SCO progenitor marker. Fibcd1 encodes a transmembrane receptor that binds to acetylated components and elicits endocytosis upon ligand binding (Schlosser et al., 2009; Thomsen et al., 2009). From the atlas, a ubiquitous and high level of Fibcd1 transcript is present in the SCO within the developing CNS at 11.5 dpc and 13.5 dpc. This expression becomes patchy and less intense within the SCO by 15.5 dpc. At this stage, Fibcd1 transcript is no longer restricted to the SCO, but also in the dorsal mesencephalon. At 18.5 dpc, Fibcd1 transcript is expressed in the medial habenular nucleus, a component of the EpTh, and the mesencephalon, but not in the mature SCO (Lein et al., 2007).
To verify the data observed from the brain atlas (Lein et al., 2007), Fibcd1 in situ hybridisation was performed on WT 12.5 dpc, 15.5 dpc and 18.5 dpc SCO. As expected, 104
ubiquitous and strong Fibcd1 transcript expression was apparent at most 12.5 dpc SCO (n=2), except where a stripy expression pattern was observed at its posterior limit (Figure 4.13 A and E). Although Fibcd1 transcript was detectable at 15.5 dpc SCO (n=2), the signal was less intense and was patchy (Figure 4.13 B). By 18.5 dpc, Fibcd1 transcript was not detectable within the SCO (n=2) (Figure 4.13 C), while the positive control SCO staining of purkinje cell protein 4 like 1 (pcp4l1) at its neighbouring section was evident (Figure 4.13 D). Similar to the observation in brain atlas (Lein et al., 2007) and in contrast to the absence of expression in SCO, Fibcd1 transcript was apparent in the medial habenular nucleus, a structure anterior to the SCO, at 18.5 dpc (Figure 4.13 C). Together, these indicated that Fibcd1 is downregulated in mature SCO cells and could therefore serve as a marker for early SCO progenitors.
In order to investigate whether the overdosage of Sox3 inhibits SCO differentiation, expression of Fibcd1 at Sr/+; Nr/+ SCO was studied. At 12.5 dpc, Fibcd1 expression was present at the SCO of Sr/+; Nr/+ embryos. However, unlike the strong and ubiquitous expression in most of the WT SCO, Fibcd1 expression was patchy and lower in intensity throughout the SCO (n=2) (Figure 4.13 E-F). Fibcd1 transcript was still expressed at a moderate level at 14.5 dpc WT SCO (Figure 4.13 G). In contrast, Fibcd1 was not detectable in most of the Sr/+; Nr/+ dysplasic SCO where ependymal SCO cells were absent, (n=2) despite its evident expression at the posterior SCO remnant (Figure 4.13 H). Collectively, these results indicated that SCO progenitors were progressively lost from the early SCO primordium of Sr/+; Nr/+ embryos probably prior to 12.5 dpc. Hence, it is unlikely that Sox3 overdosage inhibits SCO differentiation.
4.2.9 Downregulation of diencephalic midline markers in 12.5 dpc Sr/+; Nr/+ embryos
In an attempt to unravel the underlying molecular mechanism(s) for the above described cellular defect in Sr/+; Nr/+ SCO development, expression analysis was performed. Previous literature has demonstrated the importance of Msx1 in SCO development. Within the SCO, Wnt1 expression is dependent on Msx1 and a lack of either Msx1 or Wnt1 also led to defective SCO development and CH (Bach et al., 2003; Louvi and Wassef, 2000). Moreover, 15.5 dpc Msx1 null mice displayed a dysmorphic SCO with some morphologically normal cells at its posterior, similar to the case in Sr/+; Nr/+ embryos (Bach et al., 2003). 105
Thus, the expression of Msx1 and Wnt1 were studied in the diencephalic RP at 12.5 dpc, the last stage at which normal SCO morphology was observed in Sr/+; Nr/+ embryos.
In agreement to previously reported results (Bach et al., 2003; Shimamura et al., 1994), both Msx1 and Wnt1 were highly but specifically expressed by cells at the RP within the 12.5 dpc WT diencephalon, including the SCO (Figure 4.14 A, B , E and G). This RP specific expression was not restricted to the diencephalon but also extended posteriorly into the mesencephalon (Figure 4.14 A, B, E and G). No significant difference in the level of Msx1 and Wnt1 expression was found in the Sr/+; Nr/+ RP posterior to the SCO. Interestingly, a significant downregulation or absence of Wnt1 and Msx1 transcripts was observed in the SCO as well as the RP anterior to the SCO (n=3) (Figure 4.14 C, D, F and H). More detailed analysis found that both Msx1 and Wnt1 were present at the very posterior SCO limit, albeit with lower intensity. However, a marked decrease or complete absence of both transcripts was detected in majority of the SCO, as well as any rostral RP (Figure 4.14 F and H). The lack or reduced expression in SCO could be further demonstrated when comparing the dorsally residing PC from the WT, where there was a clear PC outline contrasted by the colorimetric development of the abundant Wnt1 or Msx1 transcripts in the SCO, to that of Sr/+; Nr/+, where the PC outline was ambiguous (Figure 4.14 F and H). In summary, the expression of Msx1 and its target gene, Wnt1 were downregulated in the SCO cells and the diencephalic RP that was anterior to the SCO posterior limit in Sr/+; Nr/+ embryos, indicating that the molecular identity of diencephalic RP was compromised.
4.2.10 Characterisation of ChP in Sox3 transgenic mice
There are four ChP within the CNS, with one ChP residing in each brain ventricle. Together, these ChP are responsible for 70-80% of the total CSF production within the CNS (Brown et al., 2004; Lowery and Sive, 2009; Milhorat, 1975; Pérez-Fígares et al., 2001). Thus, proper ChP development is important for a normal and regulated CSF production (Dziegielewska et al., 2001). Moreover, hyperactive ChP and ChP hyperplasia, which ultimately cause CSF overproduction, have been reported to be the underlying cause of CH in some mouse models (Baas et al., 2006; Banizs et al., 2005). Thus, ChP development in Sox3 transgenic mice was also analysed. ChP is derived from two different lineages, a mesenchymal origin that forms the structural component, and a ChP epithelium that overlays Fibcd1 Fibcd1 A E G PC
PC PC Aq SCO WT
12.5 dpc Aq SCO
Aq SCO F H B PC + PC # PC Aq Sr/+; Nr/+ Sr/+; Aq SCO 12.5 dpc 14.5 dpc 15.5 dpc Aq SCO
Figure 4.13: Loss of SCO progenitors in the SCO from both 12.5 and 14.5 dpc Sr/+; Nr/+ embryos. A-E: Fibcd1 was ubiquitoursly expressed C at high level in early WT SCO at 12.5 dpc, except at the posterior limit where a stripy expression pattern was observed (yellow arrowheads) (A PC and E). Lower and patchy (black arrowheads) Fibcd1 expression was detected in the WT SCO at 15.5 dpc (B). FIbcd1 transcript was no longer
18.5 dpc present in the mature WT SCO at 18.5 dpc, although a moderate level of MHN SCO Aq transcript was evident at the MHN (C). Pcp4l1 was highly expressed in the neighbouring section of (C). F: Less intense Fibcd1 transcript expres- sion was present in a patchy pattern (black arrowheads) in Sr/+; Nr/+ 12.5 Pcp4l1 dpc SCO (F) when compared to the WT (E). G-H: At 14.5 dpc Sr/+; Nr/+ D embryos, there lacked Fibcd1 expression in the dysplasic SCO, where elongated SCO cells were absent (hash), although Fibcd1 expression was maintained in the posterior SCO remnant (plus sign) (H). In situ hybridisa- tion. MHN: medial habenular nucleus. Arrows: points SCO invagination. PC Doublet arrow: MR. Sagittal sections. Top to bottom: dorsal to ventral. 18.5 dpc SCO Left to right: anterior to posterior. Scale bar: 100µm (A-C), 200µm (E-H). Aq WT Sr/+; Nr/+ A B
Figure 4.14: Reduced or absence of Wnt1 and Msx1 expression at the SCO primordium and the RP anterior to the SCO in 12.5 dpc Sr/+; Nr/+ embryos. A, C, E and G: Dien- Msx1 cephalic expression of Msx1 and Wnt1 were restricted to the dorsal midline in WT (A, C, E 12.5 dpc and G). The SCO posterior limit was indicated by an arrowhead. (E) and (G) are boxed regions of (A) and (C) respectively. B, D, F and H: Msx1 and Wnt1 dorsal midline expression C D were downregulated or absent in the RP anterior to the SCO posterior limit (arrowhead) of Sr/+; Nr/+ embryos (B, D, F and H). (F) and (H) are boxed regions of (B) and (D) respec- tively. In situ Hybridisation. Di: diencephalon. Mes: mesencephalon. Sagittal orientation. Top to bottom: dorsal to ventral. Left to right: anterior to posterior. Scale bar: 500µm. Wnt1
Msx1 Wnt1 E G PC PC
WT SCO SCO Mes Mes 12.5 dpc Di Di F PC H PC SCO
Mes SCO
Sr/+; Nr/+ Sr/+; Di Mes Di Nr/+ WT A D LV LV SOX3 Me Me B E LV LV
Nr/+ WT EGFP LV M Q 4V 4V Me Me Me SOX3 F C LV LV DAPI Me N R 4V Me Me Me 4V EGFP G J Me Me Me 4V SOX3 3V O S 3V 4V Me DAPI 4V K H Me M Me P T 4V EGFP Me SOX3+DAPI 3V 3V 3V 4V Me
L I 12.5 dpc 12.5 Me Me DAPI
3V
3V 12.5 dpc 12.5 Figure 4.15: Expression of endogenous and transgenic Sox3 at 12.5 dpc ChP. A-F: Endogenous SOX3 was expressed at a varying low level in the LV ChP. Some ChP cells displayed stronger endogenous SOX3 expression (yellow arrowheads) than others (A). EGFP and hence, transgenic Sox3 was present at Nr/+ LV ChP (D and E). Most LV ChP cells from Nr/+ embryos demonstrated a high level of SOX3 expression (D). (B) and (E) are EGFP and (C) and (F) are DAPI stainings of (A) and (D) respectively. G-L: Endogenous SOX3 was not detected in 3V ChP (G). Ectopic EGFP, indicative of transgenic Sox3 (J and K) was present at Nr/+ 3V ChP. Note that autofluorescence from haematoma (white arrows) (G and J). (H) and (K) are EGFP and (I) and (L) are DAPI stainings of (G) and (J) respectively. M-T: Both endogenous (M) and transgenic Sox3 (Q and R) were not detected in WT or Nr/+ 4V ChP. Note the autofluo- rescence from mesenchymal cells (white arrowheads) (M and Q) and non-specific speckles (yellow arrows) (R). (N) and (R) are EGFP and (O) and (S) are DAPI stainings of (M) and (Q) respectively. (P) is a merge of (M) and (O), while (T) is a merge of (Q) and (S). Immunofluorescence. Me: ChP mesenchyme. White box with broken lines: ChP epithelium. Yellow line: boarder of neuroepithelium/mesenchyme or ChP epithelium/mesenchyme. Coronal sections (A-L). Sagittal sections (M-T). Top to bottom: dorsal to ventral. Left to right: anterior to posterior (M-T). Scale bar: 100µm (A-L), 50µm (M-T). SOX3 EGFP DAPI SOX3+DAPI A B C D
Me Me Me WT Me LV E F G H
Me Me Me Me Nr/+
I J K L WT Me Me Me Me 3V Me P M Me N Me O Me
Me Me Me Me Nr/+
Q R S T WT
Me Me Me Me 4V
U V W X Nr/+ Me Me Me Me
12.5 dpc Figure 4.16: Both endogenous and transgenic Sox3 was not detected in the LV, 3V and 4V ChP at 15.5 dpc. A-H: There was no endogenous SOX3 (A) or EGFP indicative of transgenic Sox3 (E and F) detected in the LV ChP of WT or Nr/+ embryos. (B) and (F) are EGFP and (C) and (G) are DAPI stainings of (A) and (E) respectively. (D) is a merged of (A) and (C) while (H) is a merge of (E) and (G). I-P: WT (I) and Nr/+ (M and N) 3V ChP also displayed a lack of endogenous and transgenic Sox3 expression. (J) and (N) are EGFP and (K) and (O) are DAPI stainings of (I) and (M) respectively. (L) is a merge of (I) and (K), while (P) is a merge of (M) and (O). Q-X: Both endogenous (Q) and transgenic Sox3 (U and V) were not detected in the WT or Nr/+ 4V ChP. (R) and (V) are EGFP and (S) and (W) are DAPI stainings of (Q) and (U) respectively. (T) is a merge of (Q) and (S), while (X) is a merge of (U) and (W). Note the autofluorescence from mesenchymal cells (white arrowheads) (A, D, E, H, I, L, M, P, Q, T, U and X). Immunofluorescence. Me: ChP mesenchyme. Yellow line: boarder of ChP epithelium/ChP mesenchyme. White line: boarder of ChP epithelium/ventricular area. White broken line: boarder of neuroepithelium/ChP mesenchyme. Sagittal sections. Top to bottom: dorsal to ventral. Left to right: anterior to posterior. Scale bar: 50µm. WT Sr/+; Nr/+ 15.5 dpc G J A B C CCx CCx CCx
LV DTh LGE DTh WT
3V DTh 3V LV D E F CCx H K LV 3V
LGE 15.5 dpc Sr/+; Nr/+ Sr/+; DTh
Figure 4.17: Both the LV and 3V ChP were histologically normal in 15.5 dpc Sr/+; Nr/+ embryos. A-L: Both LV and 3V ChP from Sr/+; Nr/+ embryos displayed normal histological morphology (D, E, I and K) when compared to those of the WT (A, B, G and H). There was no significant difference observed in the size, number or density of cells from Sr/+; Nr/+ LV and 3V ChP. (E and K) to that of WT (B and H). Upon high magnification, no irregularity in cell shape was detected in any ChP cell from I L Sr/+; Nr/+ embryos (F and L). (B), (E), (H) and (K) are the boxed region of (A), (D), (G) and (I) respectively, while (C), (F), (I) and (L) are the boxed region of (B), (E), (H) and (K) respectively. Nissl stains. Sagittal sections. Top to bottom: dorsal to ventral. Left to right: anterior to posterior. Scale bar: 200µm (A, D, G and J), 50µm (B and E), 100µm (H and K), 25µm (C, F, I and L). 106
the mesenchyme and is continuous with its neighbouring neuroepithelium. ChP development initiates around mid-gestation in mouse and by 12.5 dpc, the ChP epithelium of all four ventricles has already been defined from its surrounding neuroepithelium (Sturrock, 1979). In mouse, further differentiation of the ChP epithelium into mature ChP initiates at 13.5 dpc in the 4V. This is then followed by the ChP from the LV at 14.5 dpc and that of the 3V at 15.5 dpc (Dziegielewska et al., 2001; Johansson et al., 2005; Sturrock, 1979). Thus, for the initial stage of ChP development characterisation, both endogenous and transgenic Sox3 expression were analysed at 12.5 dpc and 15.5 dpc within the single hemizygous Nr/+ embryos through the analysis of SOX3 and EGFP protein expression.
Endogenous SOX3 was not detected within the 3V and 4V ChP epithelium from WT embryos at 12.5 dpc and 15.5 dpc (Figure 4.15 G and M and Figure 4.16 I and Q). However, a very low level of endogenous SOX3 expression was present at the LV ChP was detected from WT embryos at 12.5 dpc but not at 15.5 dpc (Figure 4.15 A and Figure 4.16 A). Analysis also revealed a lack of EGFP and hence, transgenic Sox3 expression in the 4V ChP from 12.5 dpc (n=3) and 15.5 dpc (n=3) Nr/+ embryos (Figure 4.15 N-T and 4.16 R-X). Surprisingly, EGFP indicative of transgenic Sox3 was highly expressed at both the LV and 3V ChP at12.5 dpc embryos (n=3) (Figure 4.15 B-F and H-L). This transgenic Sox3 expression was transient, however, since it was no longer detected in the 3V and LV ChP from Nr/+ embryos at 15.5 dpc (n=3) (Figure 4.16 B-H and J-P). Due to the presence of transient expression of transgenic Sox3 between 12.5 dpc to 15.5 dpc, there poses a possibility that LV and 3V ChP development might be affected. Morphological study was thus carried out on both the LV and 3V ChP at 15.5 dpc, a time point when the differentiation of both ChP have initiated, in Sr/+; Nr/+ double hemizygous embryos. In contrast to a severely dysmorphic SCO at 15.5 dpc, both LV and 3V ChP were present with normal morphology at this timepoint (n=3). There was no significant difference in the size, density and number of 3V and LV ChP cells (Figure 4.16 A, B, D, E, G, H, Jand K). Moreover, no irregularity in ChP cell shape was identified (Figure 4.16 C, F and I, L).
Together, these data indicate that unlike the SCO, endogenous SOX3 was not expressed consistently throughout ChP development. Despite a transient expression of transgenic Sox3 in LV and 3V ChP epithelium at 12.5 dpc, no apparent morphological difference was identified in these ChP from 15.5 dpc Sr/+; Nr/+ embryos. This is in great 107
contrast to the SCO, which was severely defective in both its morphology and secretory function at this stage.
4.3 Discussion
4.3.1 Defective SCO development is the primary cause of CH
The regulation of CSF homeostasis is a complicated process which requires a balance between a steady rate of CSF production, a dynamic CSF flow and an efficient CSF reabsorption (Dziegielewska et al., 2001; Pérez-Fígares et al., 2001). Thus, the CH in Sox3 transgenic mice is caused by the deregulation of one or more of the above three processes. Although it was established in the previous chapter that the SCO is causative to the CH phenotype in Sox3 transgenic mice, the presence of other CH-related defect(s), which may exacerbate the CH severity, cannot be excluded. CSF is mostly synthesised by ChP within the CNS. Upon production, the CSF flows in a rostrocaudal direction through all brain ventricles into the subarachnoid space and the spinal lumen for reabsorption (Pérez-Fígares et al., 2001; Picketts, 2006). The presence of Aq blockade and non-expanded 4V indicate that CH causative domain(s) were rostral to the hindbrain. This has excluded retarded CSF reabsorption as the primary cause of CH in Sox3 transgenic mice.
Numerous studies have shown that proper cilia function is essential to maintain the CSF laminar flow as loss of cilia formation or motility has proven to be detrimental to CSF homeostasis (Brody et al., 2000; Chen et al., 1998; Davy and Robinson, 2003; Ibanez-Tallon et al., 2002; Ibanez-Tallon et al., 2004; Kobayashi et al., 2002; Lang et al., 2006; Lechtreck et al., 2008; Lechtreck and Witman, 2007; Sapiro et al., 2002). Most ependymal cells are born between 14.0 dpc to 16.0 dpc from radial glia cells residing at the VZ lining all ventricles. Although there was no transgene expression within the telencephalic VZ, transgenic Sox3 was apparent in those of the diencephalon. This has prompted the question of whether a perturbation in the function of diencephalic ependymal cells has contributed to the cause of CH. However, ependymal differentiation and cilia formation do not take place until the first week of postnatal stage (Banizs et al., 2005; Spassky et al., 2005; Wilson et al., 2009). Furthermore, ependymal denudation, which is often associated with CH mouse models 108 caused by defective ependymal development (Batiz et al., 2006b; Nechiporuk et al., 2007; Tullio et al., 2001), was not observed in CH affected Sox3 transgenic mice. Collectively, the malformation of ependymal cells and their cilia are unlikely to account for the early onset of hydrocephalus, i.e. from 18.5 dpc, observed in Sr/+; Nr/+ mice.
The ChP are responsible for 70-80% of the total CSF production (Pérez-Fígares et al., 2001). The lack of endogenous or transgenic Sox3 expression in the 4V ChP together with the conclusion of the causative region to be rostral to the hindbrain as concluded above, indicate that the CH in Sox3 transgenic mice was unlikely to be caused by the hyperactivity of the ChP of the 4V. In contrast, the presence of endogenous and/or transgene-derived SOX3 in the ChP of the LV and 3V raises the possibility that these regions might be causative of the CH phenotype. Previous literature has associated macroscopic ChP defects such as decreased convolutions, cytoplasmic expansion and shortened cell shape with CH (Baas et al., 2006; Dziegielewska et al., 2001; Wodarczyk et al., 2009). This is in contrast to the Sox3 transgenic mice, which had no macroscopic dysmorphology observed at the ChP of both the LV and 3V. In a few CH mouse models, ChP was detected with microscopic but not macroscopic dysmorphology (Banizs et al., 2005; Lindeman et al., 1998). However, analysis of microscopic ChP defects, such as the lack of cellular polarity, cilia formation or motility, has not been performed in Sox3 transgenic mice. As a result, a causative role of defective ChP in CH development cannot be confidently excluded. Yet, the early severe SCO dysmorphology correlates well with the early onset and the high penetrance of CH in Sr/+; Nr/+ double hemizygous mice. This argues that the defective SCO development as the primary cause of CH and that any microscopic ChP defect(s) would have an additive effect in the development of CH in Sox3 transgenic mice. To further elucidate the association between the ChP and the CH of Sox3 transgenic mice, electron microscopy study of the ChP cilia formation and motility, as well as analysis of the ChP epithelium polarisation through detection of its polarised proteins, acetylated-α-tubulin and IFT88, may be performed in the future.
4.3.3 SCO dysplasia was a consequence of loss of SCO cellular identity
A morphologically normal SCO primordium was present during early embryogenesis of Sox3 transgenic embryos, indicating that induction of SCO precursors occurred normally. However, these SCO precursors did not differentiate into normal SCO cells, nor retain their 109
precursor cell state as development progressed. In fact, a loss of elongated SCO precursors cells was evident as early as 13.5 dpc in Sox3 transgenic mice, shortly after peak proliferation (Rakic and Sidman, 1968), with a minute amount of SCO elongated ependymal cells remaining by 15.5 dpc. Two of the most direct explanations for the loss of early SCO cells in the presence of Sox3 overdosage are the downregulation of cellular proliferation and/or elevation of apoptosis in the early SCO cells. However, an increased proliferation and a lack of programmed cell death were observed in the SCO of Sox3 transgenic embryos and conclude these explanations not plausible. In fact, based on these observations, one may expect there would be more cells within the SCO primordium of the Sox3 transgenic embryos. Hence, this poses a question of what did this “larger mass” of cells transformed into. Three possibilities are explored in an attempt to answer this question. First, SCO precursors may undergo precocious differentiation. This is however, very unlikely as a morphologically normal SCO was not found at any stage after 12.5 dpc in Sox3 transgenic embryos. Furthermore, no RF expression was detected at the SCO of Sr/+; Nr/+ embryos at 15.5 dpc. Second, Sox3 has been shown to inhibit neurogenesis in chick spinal cord (Bylund et al., 2003; Holmberg et al., 2008). Perhaps SCO differentiation was also blocked by the increase dosage of Sox3 in transgenic mice. This is also unlikely as one would expect an accumulation of SCO precursors at later stage(s) of embryogenesis should a blockage of SCO differentiation be present. Expression analysis of the early SCO marker Fibcd1 failed to identify SCO cells within most SCO domain of 14.5 dpc embryos. Thus, the remaining possibility is that SCO precursors lost their identity. This is indeed supported by the apparent loss of early SCO cells observed by the early SCO marker Fibcd1, such that some SCO precursors have taken up an alternative identity. Consistent with this possibility, the disappearance of elongated early SCO cells was accompanied the appearance of flattened, sparsely packed cells resembling neuroepithelial cell from SCO neighbouring regions.
4.3.2 Morphological defect was not restricted to SCO but was RP sepcific
Sox3 transgene expression was widely detected within the diencephalon of Sox3 transgenic embryos. Not surprisingly, developmental defects associated with Sox3 expression were not restricted to the SCO. Prominent dysmorphology was in fact identified in a number of dorsal midline structures, i.e. PC, HbC and PG, which are discussed separately below. 110
Loss of PC
Proper projection of commissural axons requires a dynamic but yet coordinated complex signalling network that first attracts the midline crossing neurons to project their axons towards the midline. Upon reaching the midline, repulsive signals then are required to send those axons away from the midline contralaterally and to prevent them from re-crossing the midline again (Kaprielian et al., 2000; Kaprielian et al., 2001; Lindwall et al., 2007). Previous studies have established that midline glial structures are essential to provide these signals to pathfinding neurons within the forebrain (Lindwall et al., 2007).
The PC is a collection of axonal bundles projected from lateral residing nuclei across the midline dorsally to the SCO. During chick embryonic development, the magnocellular nucleus is the first nuclei population that projects their axons contralaterally across the midline dorsal to the SCO primordium, forming the initial PC bundles. As development progresses, a few other pretectal nuclei also send their axons across the midline dorsal to the SCO, leading to the growth of the PC axonal bundles (Cobos et al., 2001; Figdor and Stern, 1993; Hoyo-Becerra et al., 2010). Due to the juxtapositioning of the PC to the SCO and the coexistence of SCO and PC dysmorphology in numerous mouse models (Bach et al., 2003; Dietrich et al., 2009; Estivill-Torrus et al., 2001; Fernandez-Llebrez et al., 2004; Louvi and Wassef, 2000), it has been proposed that the SCO is required for instructing PC midline crossing. Moreover, it has been demonstrated previously that SCO explants are capable of either attracting or repelling axonal growth from magnocellular or pretectal explants, depending on their embryonic stages, when co-cultured in vitro (Hoyo-Becerra et al., 2010; Stanic et al., 2010).
Consistent with the above mentioned studies, there was a progressive deterioration of the PC in Sox3 transgenic mice that correlated temporally with SCO dysplasia. This close relationship was further demonstrated in 14.5 dpc and 15.5 dpc Sox3 transgenic embryos by the co-survival of a small group of PC residing in proximity to the SCO remnant at the posterior limit. These observations in Sox3 transgenic mice provide additional support to the proposal that the SCO is required for proper PC formation. Similar to the SCO primordium, a morphologically normal PC was present before 12.5 dpc, suggesting there were functional signals from the SCO precursors to direct proper PC projection to cross the midline at and prior to this stage. The loss of these initial PC fibres indicates that perhaps the SCO is not only required to instruct PC formation, but also for the maintenance of the PC formed at the 111
midline. This is supported by studies which showed that secretions from the mature SCO cells can influence neuronal aggregation and neurite outgrowth in vitro (El Bitar et al., 1999; Monnerie et al., 1995; Monnerie et al., 1996; Monnerie et al., 1997; Monnerie et al., 1998). Perhaps, these SCO secretions have a role to play in the maintenance of the PC once they have crossed the midline.
HbC failed to cross midline
The habenular nuclei are a pair of lateral nuclei residing at the dorsal diencephalon anterior to the PG. Their major function is to serve as a signal relay centre between the basal ganglia and the limbic system in motor control (Hikosaka et al., 2008). The HbC is an axonal tract projected from the two lateral habenular nuclei into the prosomere 1/prosomere 2 boundary of the dorsal midline. The HbC allows communication between the two lateral nuclei. The exact molecular mechanism for the formation of HbC is still unknown, however one would expect that it would also require local RP signalling in order to cross the midline (Kaprielian et al., 2000; Kaprielian et al., 2001). Indeed, mediodorsally projections of habenular commissure were observed in Sox3 transgenic embryos, but there lacked any successful midline crossing. The presence of mediodorsally projected HbC in Sr/+; Nr/+ embryos are in contrast to the PC where no collected axonal projections were observed. This is perhaps due to the fact that lateral domain(s) other than the RP may also be required for habenular axonal pathfinding, since an in vitro study has demonstrated that tissue explants from prosomere 1 or habenular nucleus are capable of stimulating or inhibiting habenular axonal growth (Funato et al., 2000). Nevertheless, the lack of midline crossing of the HbC suggests a possible RP defect within prosomere 1 and prosomere 2 of the diencephalon of Sox3 transgenic embryos.
Absence of PG development
PG is a prosomere 2 RP derived endocrine gland that is mainly responsible for melatonin production in response to light for circadian cycle regulation (Arendt, 1998; Blackshaw and Snyder, 1997; Furukawa et al., 1999; Møller and Baeres, 2002; Nishida et al., 2003). Similar to the SCO, normal morphological PG primordium was not observed in Sr/+; Nr/+ embryos at any stage of embryogenesis after 13.5 dpc. Interestingly, PG dysmorphology has been associated with SCO dysplasia in previous literatures, since 112 dysplasic PG was identified together with dysmorphic SCO in Pax6 small eye mutant and Wnt1 driven ectopic En1 CH mouse models, both of which displayed general RP defects (Estivill-Torrus et al., 2001; Louvi and Wassef, 2000). Similarly, the loss of both Msx1 and Msx2 within the RP has been shown to cause RP defect whereby the development of both PG and SCO were missing (Bach et al., 2003).
General RP defect
In summary, the presence of multiple RP associated dysmorphologies strongly suggests a general RP defect rather than a SCO specific defect in the Sox3 transgenic mouse model. Since the SCO derives from the RP, the loss of SCO cellular identity is likely to be secondary to a general loss of RP cellular identity. Previous investigations in mouse and chick NT demonstrated that RP precursors are required to withdraw from cellular proliferation in order for proper RP differentiation, during a time when the lateral neural tissues that flank the RP are still highly active in cell cycle progression (Baek et al., 2006; Kahane and Kalcheim, 1998; Sah et al., 1995). As a result, the increased proliferation observed in Sox3 transgenic SCO primordium suggests that these cells have deviated in their RP fate. However, it is worth noting that, unlike the Pax6 small eye mutant and Rfx4-v3 null mice (Blackshear et al., 2003; Estivill-Torrus et al., 2001), which do not have the RP specified even at very early embryonic stages, Sox3 transgenic mouse model did appear to at least have some proper RP specification before 12.5 dpc. This is supported by the following two pieces of evidence. First, a morphological normal SCO primordium with properly formed PC was present by 12.5 dpc in Sox3 transgenic embryos, suggesting a functional RP at that stage. Second, Fibcd1 specifically labels the SCO primordium within the RP at 12.5 dpc. Although, the level of Fibcd1 expression was lower in Sox3 transgenic embryos, either by a general downregulation in all cells and/or the complete loss of expression of some cells, the successful induction of Fibcd1 expression indeed suggest there was at least some degree of proper RP patterning along the AP axis. Thus, it is likely that the RP was specified and perhaps patterned along AP axis with some success before the later loss of identity.
Intriguingly, all diencephalic lateral domains were morphologically normal and were in sharp contrast to the general dysmorphology of the RP. However, both the Sr and Nr Sox3 transgene were present in the RP and the lateral domains within the dorsal diencephalon at 12.5 dpc. This raises a question of how can a general elevation of Sox3 cause a specific 113 developmental defect in the RP? In conjunction with the lower dosage of endogenous Sox3 within the developing SCO and the Sox3 dosage dependency in SCO dysplasia and CH, the following hypothesis is proposed: a significantly lower level of Sox3 is required to specifically maintain RP identity and to allow further differentiation of its derivatives. If the level of Sox3 were elevated, perhaps above a threshold or to a level where the levels Sox3 abundance between the RP and its lateral domains are similar, RP cells would adopt an alternative (lateral, non-RP) identity and lose its ability to response to signals to RP specific differentiation (Figure 4.18).
In addition to the diencephalon Sox3 may also be responsible for telencephalic RP maintenance. A commissural midline projection defect was noted in the corpus callosum and dorsal hippocampal commissure in Sox3 null mice. Fasciculation of corpus callosum and hippocampal commissure were observed juxtalaterally to the midline in these mice (Rizzoti et al., 2004). Moreover, corpus callosum defect was also identified in a Sox3 associated XH patient (Woods et al., 2005). Together, these underline a possible role of SOX3 in telencephalic RP development. Recent findings indicated cerebral axons require the signalling guidance from a telencephalic midline structure, the glial wedge, to successfully cross the midline (Paul et al., 2007; Richards et al., 2004; Shu and Richards, 2001). Intriguingly, a preliminary observation found both endogenous and transgenic Sox3 were expressed at the glia wedge primordium of Nr/+ embryos.
In conclusion, the results presented in this chapter suggest a correct dosage of Sox3 is required to maintain RP identity to allow normal SCO development and the proper formation and maintenance of PC. Based on currently available circumstantial evidence, Sox3 may also implicated in the RP development within the telencephalon. Further analysis in Sox3 null and transgenic mice may allow the elucidation of whether the glia wedge is also regulated by Sox3 in a similar mechanism to that of the SCO, and perhaps can provide a plausible explanation of the corpus callosum phenotype observed in some XH patients.
4.3.4 Sox3 regulates Msx1 and Wnt1 for RP development
The significant downregulation of both Msx1 and Wnt1 within the 12.5 dpc anterior RP was well correlated spatially with the Sox3 transgene expression, which was present in the A
Low Sox3 Low proliferation
RP cells High Sox3 High proliferation
lateral neural cells
B
High Sox3 High proliferation
lateral neural cells? High Sox3 High proliferation
lateral neural cells
Figure 4.18: Hypothesis of the Sox3 dosage dependent RP specification. A: In a WT embryo, Sox3 expression is low within the RP (in yellow), allowing RP specification and withdrawal from cell cycle progression. The high level of Sox3 expression in RP neighbouring tissues thus specifies their lateral identities. B: In a Sox3 transgenic embryo, the elevated level of Sox3 expression leads the RP precur- sors to specify as its lateral tissues, resulting in the loss of RP identity. Moreover, the mis-specified RP cells fail to downregulate from cell cycle progression. Coronal view. Top to bottom: dorsal to ventral. 114
RP anterior to the SCO posterior limit including the SCO. Thus, Msx1 and Wnt1 may be one of those earliest genes that were downregulated in the RP of the Sox3 transgenic mouse model upon Sox3 overexpression. One may argue that the lack of Msx1 and Wnt1 expression might be a secondary effect of the loss of RP identity. However, both the presence of a normal SCO morphology and the expression of Fibcd1 SCO specification marker in Sr/+; Nr/+ embryos indicated that most if not all the RP cells were still of its correct identity at 12.5 dpc. Thus, this supports that the downregulation of Msx1 and Wnt1 expression was likely to be a close downstream effect due to Sox3 overexpression rather than a consequence secondary to the deviated RP cell fate in Sox3 transgenic mice.
Previous literature demonstrated that Msx1 null mice only displayed RP defect limited to the prosomere 1 due to the functional redundancy of Msx1 and Msx2. In fact, Msx1 and Msx2 double null mutant failed to develop functional RP along the entire cranium (Bach et al., 2003). Since the RP defect extended anteriorly across prosomere 1 into prosomere 2 and prosomere 3 in Sox3 transgenic embryos, it is very likely that Msx2 may also be downregulated by Sox3 overexpression. Future analysis is required to verify this postulation. Moreover, the repression of both Msx1 and Wnt1 expression in Sox3 transgenic embryos further demonstrated the dependency of Wnt1 expression regulation on Msx1 observed in Bach et al. (2003). In contrast, a subtle reduction of developmental cellular death within the prosomere 1 RP of Msx1 null mice was detected 11.5 dpc (Bach et al., 2003). There was no developmental cell death detected in either WT or Sox3 transgenic at 12.5 dpc. However, apoptotic death in the prosomere 1 of 11.5 dpc Sox3 transgenic embryos was not studied and hence, this possibility cannot be excluded.
Expression of the Sox3 transgene was observed within the dorsal neural folds at 8.5 dpc during neurulation (P. Thomas, personal communication). Transgenic Sox3 was also detected at 11.5 dpc in the diencephalic RP. Thus, it is possible that the downregulation of Msx1 and Wnt1 expression may initiate prior to 12.5 dpc in the diencephalic RP (or even at the diencephalic dorsal neural folds briefly before NT fusion) of Sox3 transgenic embryos. If this were the case, it suggests that perhaps both Msx1 and Wnt1 are not important for early RP specification. In fact, this correlates well with the phenotypes observed in the Wnt1 loss- of-function mouse model. Despite Wnt1 is expressed at the dorsal neural fold and maintained within the early dorsal midline and RP cells, there was no depletion of RP cells at 12.5 dpc Wnt1 mutant (Shimamura et al., 1994). Thus, perhaps Sox3 upregulation perturbs gene 115
expressions important for late RP development from12.5 dpc onwards but not critical for early RP or dorsal NT development.
The loss of Wnt1 within the developing RP appears to disrupt the expression of adhesion molecules. In Wnt1 mutant, Cdh1 was upregulated, while that of Ctnna2 was downregulated within the RP. Ctnna2 have been shown to regulate Cdh1, which encodes an epithelial cell adhesion molecule, and Cdh2, which encodes a neural cell adhesion molecule, to enable intercellular interaction for cellular and tissue sorting during CNS development (Drees et al., 2005; Hirano et al., 1992; Lien et al., 2006; Stepniak et al., 2009). Thus, the loss of RP identity in Sox3 transgenic mice was perhaps the result of the ultimate cellular consequence, a perturbed intercellular interaction, and hence failure of cellular and tissue sorting during RP development.
Dorsal NT induction begins before complete NT fusion. RP is induced by diffusible Bmp signals, Bmp4 and Bmp7 from neighbouring epidermal ectoderm at both the dorsal horns of neural folds during neurulation (Liem et al., 1995). Bmp induction is believed to induce the expression of the master RP regulator, the transcription factor Lmx1a, at both dorsal tips of the neural folds. The expression of Lmx1a was found at the entire dorsal NT in mouse during and shortly after neurulation (Failli et al., 2002). Interestingly, Dreher mouse that lacks the function of Lmx1a fails RP development (Millonig et al., 2000), and at least in the chick spinal cord, ectopic Lmx1a was sufficient to activate downstream expressions of RP markers Gdf7, Wnt1, Msx1 and Msx2 within the dorsal tip of the closing neural fold. (Chizhikov and Millen, 2004b). Here, the lack of both Msx1 and Wnt1 in Sox3 transgenic mice has placed Sox3 upstream of both of Wnt1 and Msx1, and downstream or on par with Lmx1a. Thus, despite successful RP induction and early development, a Sox3 dosage is required to be kept low within the developing RP to permit the proper expression and function of Msx1 and Wnt1 and allow further RP development (further discussed in chapter 6.4).
Sox3 acts as a transcriptional activator in the inhibition of proneural gene expression in chick spinal cord (Bylund et al., 2003). If Sox3 acts similarly in the developing RP of Sox3 transgenic embryos, one would expect that the repressing effect of Msx1 and Wnt1 expression by Sox3 overdosage is likely to be an indirect and through the upregulation of some unknown Msx1 and Wnt1 repressors. Alternatively, Sox3 may exert this repressive effect through simply protein-protein interaction. In Xenopus DV patterning, xSox3 inhibits Wnt signalling 116 through physical interaction with its signalling molecule β-catenin, and represses the transcription of its downstream effectors (Zorn et al., 1999). Likewise, SOX3 represses β- catenin induced transcriptional activity in human embryonic kidney cells, despite it is not known whether hSOX3 exerts this effect through a protein-protein interaction as that in the Xenopus (Wong et al., 2007). Thus, it is yet to be elucidated whether Sox3 represses Wnt signalling through physical interaction with β-catenin within the developing RP of the mouse diencephalon. At last, Sox3 may act as a transcriptional repressor to inhibit the transcription of Wnt1 and/or Msx1 within the mouse diencephalic RP. Biochemical studies on the binding partners and SOX3 recognition and binding sequences during RP development will thus allow further verification of the two latter speculations. 117
Chapter 5: Molecular function of Sox3 in SCO development
5.1 Introduction
Results in the previous chapters indicated that overdosage of Sox3 in the diencephalic RP leads to a loss of RP identity, including the SCO primordium, and disruption of Msx1 and Wnt1 expression. Since Sox3 has been demonstrated to be a transcriptional activator (Bylund et al., 2003), Sox3 transcriptional regulation of Wnt1 and Msx1 was likely to be indirect. Loss of midline marker expression suggested that the RP cells that normally form the SCO may adopt alternate cell fate(s) with a non-SCO molecular profile. It was reasoned that, a comprehensive analysis of the gene expression profile in the SCO of Sox3 transgenic mice would provide insight into the molecular fate of SCO progenitors and Sox3 target genes in this region. Thus, microarray analysis was performed to compare the expression profile of the SCO primordia from Sr/+; Nr/+ and WT embryos. In order to maximise the potential for identification of differentially expressed genes directly due to Sox3 dosage, SCO tissue from 12.5 dpc, the last timepoint when the SCO primordium was morphologically normal with some SCO cells correctly specified, was studied.
5.2 The design of the microarray experiment
In order to isolate the SCO primordium, microdissection was performed. Briefly, the PC was identified under the dissecting microscope. The SCO primordium was then trimmed around the PC both rostrocaudally and laterally (Figure 5.1). SCO primordia were pooled into groups of four for each genotype (Figure 5.2). There were two rationales behind this sample pooling. First, due the continuous nature of the SCO with the surrounding neuroepithelium, the SCO primordium collected was non-homogenous in terms of “contamination” by neighbouring tissues. Sample pooling was thus employed to “average” out the influence of any non-SCO specific differentially expressed genes from individual tissue samples. In 118 addition, due to the minute tissue size, pooling of samples was required to achieve sufficient RNA for direct labelling prior to microarray analysis, thereby avoiding the possible confounding effect of RNA overamplification.
A microarray power atlas, which estimates the sample size for the required statistical power based on a study with similar parameters to the current microarray experiment, suggested more than five biological replicates were required for the identification of 10% differentially expressed gene with 95% confidence (Gadbury et al., 2004; Page et al., 2006). This meant a total collection of twenty SCO primordia (in 5 groups of four) for each genotype. However, the frequency of Sr/+; Nr/+ embryos was less than 16% per litter, which is much lower than the expected 25% based on the Mendelian ratio, possibly due to early embryonic death associated with exencephaly (discussed in 5.3.6). Thus, the sample size was adjusted to the minimum of three (Figure 5.2), which allows a 0.6% and a 3% discovery of true differentially expressed genes at a confidence of 95% and 90% confidence respectively (Gadbury et al., 2004; Page et al., 2006).
One of the most important factors for a good quality microarray is the provision of good quality RNA. Thus, the integrity of the RNA isolated from the pooled samples was assessed in terms of RNA integrity number (RIN). Only RNA with a RIN ≥8 out of a total 10 (high quality) was proceeded to the Adelaide Microarray Centre, where it was amplified, processed and hybridised to array chips. The Affymetrix GeneChip® Gene 1.0 ST array system for mouse was chosen for the experiment for the following reasons: 1) this system has incorporated an RNA amplification step which enables reasonable amplification of RNA with an abundance as low as 100ng. This was critical due to the minute tissue size of the current study; 2) the Affymetrix platform utilises random primers for RNA amplification and allows the detection of transcripts without polyadenylation signals; 3) this system interrogates the transcript abundance for each gene using a probe set consisting an average of 26 probes, hence, allows the detection of differential expression in splice variants; and 4) the analytical software and expertise for the Affymetrix platform were readily available and easily accessed. In brief, RNA was amplified, labelled (see Appendix IV for a flowchart of the microarray experiment), and then hybridised to a microarray chip. Hybridisation signal intensity, which is in direct proportion to the amount of given RNA was read by a scanner and converted into a number in digital form. A B C E Mes Di Tel Di Di Tel Rhom Mes Mes Brightfield Tel
D F
Tel Di Di Mes Mes Live EGFP Live Tel
G I K
PC PC PC Brightfield
H J L
PC PC PC Live EGFP Live Figure 5.1: SCO microdissection for microarray and qRT-PCR tissue collection. A-B: A straight cut was made dorsal to the eye of a 12.5 dpc embryo to harvest a cup-like tissue (boxed) (A) and the rest of the embryos (B). C-D: (C) is the dorsal view of the boxed region in (A). Strong live EGFP signal was present in dorsomedial diencephalon (arrowhead) (D). Further incisions were made to yield the enclosed tissue. E-F: Dorsal view of the enclosed tissue in (C) and (D) respectively. Strong live EGFP signal in dorsomedial diencephalon (arrowhead) (F). G-H: Respective ventral view of the tissue from (E) and (F) after it was cut at the ventral midline and placed flat like an “open book”. PC was readily observed as a cluster of dorsal midline residing, fibrous like structure that was more transparent than the surrounding tissue (G). A clear EGFP signal boundary was apparent at the posterior end of the PC (arrowhead) (H). Tissue was further trimmed into the boxed region. I-J: The ventral view of the boxed region of (G) and (H) respectively. Tissue was trimmed only to surrond immediate PC both anteroposteriorly and laterally, as indicated by the boxed region. K-L: Tissue was ready to be collected for further processing. (K) and (L) are the ventral view of the boxed region from (I) and (J) respectively. (D), (F), (H), (J), and (L) are live EGFP signals of (C), (E), (G), (I) and (K) respectively. Di: diencephalon. Mes: mesencephalon. Rhom: rhombencephalon. Tel: telencephalon. Grey line: dorsal midline (C-L). Lateral view (A-B). Darkfield (A-B). Left to right: anterior to posterior. Top to bottom: dorsal to ventral (A-B). Sr/+; Nr/+ embryo shown. Scale bar: 1mm (A-B), 500µm (C-F), 200µm (G-L). WT Sr/+; Nr/+
Biological replicate 1 Biological replicate 2 Biological replicate 3 Biological replicate 1 Biological replicate 2 Biological replicate 3
RNA RNA RNA RNA RNA RNA
Chip 1 Chip 2 Chip 3 Chip 4 Chip 5 Chip 6
Compare
Figure 5.2: Sematic representation of the microarray setup. SCO primorida were harvested from WT and Sr/+; Nr.+ embryos at 53-55 somite stages. For each genotype, there were three biological replicates with each replicate pooled from four SCO primordia. RNA was isolated from each pooled sample for microarray hybridisation. Gene expression were compared between all Sr/+; Nr/+ biological replicates and all WT biological replicates. 119
This chapter describes the statistical analysis of the microarray data. This is followed by the microarray validation in which candidate gene(s) identified from microarray were subjected to quantitative verification by qRT-PCR and qualitative verification using in situ hybridisation.
5.3 Results
5.3.1 Microarray data processing and quality control
Microarray raw data readout files (.CEL files) were obtained from the Adelaide Microarray centre. Data was first processed for background subtraction, probe specific background correction and normalisation for non-sample related variances in the Partek® software followed by the statistical study, analysis of variance (ANOVA). However, no significantly regulated gene was identified. Hence, independent data processing and analysis were repeated with the Affy package (Release 2.8) software from the Bioconductor project (Ihaka and Gentleman, 1996) by similar normalisation algorithms. This normalised data was then statistically analysed with linear models for microarray analysis (Limma). The discussion below outlines the details on data processing and the reasons to perform these two independent analyses. Please note that all microarray graphical representations in this result chapter were generated by the Affy package (Release 2.8) from the Bioconductor project (Ihaka and Gentleman, 1996), since only post-normalisation MA plots were able to be generated by the Partek® software, and they were included in Appendix V.
Normalisation is an important step in microarray analysis, since it removes data variation contributed by factors not due to biological differences, e.g. discrepancies in the sensitivity of signal detection. Thus, the data was first normalised to remove these variations. Subsequently, a RNA degradation plot, a plot demonstrating the mean intensity of each probe within the average 26 probes for each gene, was generated from the normalised data. The RNA degradation plot suggests that high intensity was observed in probes spanning most of the medial region within a gene. In contrast, both the 5’ and 3’ end displayed much lower intensity, since degradation starts from the flanking ends of a transcript (Figure 5.3). No unexpected pattern of low or high intensity was observed, indicating all microarray RNA 120
samples were of good quality. Moreover, all samples displayed an acceptable level of mean intensity across probe sets (Figure 5.3). This was in agreement with the RIN assessment prior to the microarray experiment, which found that the RNA used were of high quality.
Prior to statistically analysis, graphical analysis was performed to assess the quality of the microarray data. The box plot was carried out to study the overall range of relative intensity for each sample. A slight variation in intensity range was observed between samples before normalisation (Figure 5.4 A). However, normalisation has successfully removed these technical variations, since the intensity range amongst all samples was comparable in post- normalised data (Figure 5.4 B). Moreover, MA plots with a fan-like shape on the left and a short tail on the right were apparent in the pre-normalised data (Figure 5.5 A). These features of the pre-normalisation MA plots are typical observations in microarray data as genes with lower intensity are more prone to non-biological related variation, which then leads to a higher fold-change in expression. In comparison, the post-normalisation MA plot demonstrated a reduced fan shape on the left and a more elongated comet tail on the right, suggesting a more restricted range of intensity ratio, i.e. all dots were closer to zero at y-axis (assumed most gene expressions did not change between samples) (Figure 5.5 B). This transformation of post-normalisation MA plot indicates that the microarray data was of good quality.
5.3.2 Microarray statistical analysis
A brief screening of relative intensities in all six samples, i.e. three replicates each for WT and Sr/+; Nr/+, found that the relative intensity of most genes were consistent across WT or Sr/+; Nr/+ pooled samples, except for the telencephalic-specific genes, Neurod6, Tbr1 and Eomes, which were abundant in Sr/+; Nr/+ replicate 3 but the other two Sr/+; Nr/+ replicates (Figure 5.6). Further investigation of these genes on the online brain gene expression atlas (Allen atlas, http://www.brain-map.org/) indicated that they are highly expressed in the telencephalon but absent from the dorsal diencephalon at 12.5 dpc (Lein et al., 2007). These suggested that the Sr/+; Nr/+ replicate 3 sample was contaminated by telencephalic material, possibly during tissue collection. Therefore, both WT and Sr/+; Nr/+ replicate 3 were excluded from further analysis. WT replicate 1 WT replicate 2 WT replicate 3 Sr/+; Nr/+ replicate 1 Sr/+; Nr/+ replicate 2 Sr/+; Nr/+ replicate 3
Figure 5.3 RNA degration plot of all microarray samples indicated good RNA quality. RNA degration was measured by the mean intensity (y-axis) of the ~26 different probes within a probe set (x-axis). All samples displayed high intensity from probe 5 towards probe 20, which often span the sequences in the middle of a trancript. Low intensity was observed at the two flanking ends, which are regions where degradation starts. A Pre-normalisation box plot B Post-normalisation box plot ) ) 2 2 Intensity (log Intensity Intensity (log Intensity
WT WT WT Sr/+; Nr/+ Sr/+; Nr/+ Sr/+; Nr/+ WT WT WT Sr/+; Nr/+ Sr/+; Nr/+ Sr/+; Nr/+ replicate 1 replicate 2 replicate 3 replicate 1 replicate 2 replicate 3 replicate 1 replicate 2 replicate 3 replicate 1 replicate 2 replicate 3
Figure 5.4: Intensity box plot of pre- and post- normalised microarray data. A: There was slight variation in the range of signal intensity (y-axis) between different samples (x-axis) before normalisation. B: Comparable level of intensity was observed from samples after normalisation. For each sample, the line spans the intensity range from its minimum to its maximum. The box represents the interquartile range (IQR), i.e. from 1st to 3rd quartile. The black horizontal line represents the median intensity. (A) MA plot (pre-normalisation) (B) MA plot (post-normalisation) M (Intensity ratio) M (Intensity M (Intensity ratio) M (Intensity
A (Average intensity, log2) A (Average intensity, log2)
Figure 5.5: Pair-wise pre- and post-normalisation MA plots. A-B: Each dot represents a gene with its ratio of intensity between the two comparing samples (y-axis) and the average intensity (x-axis). The median intensity and interquantile range (IQR), i.e. range from 1st to 3rd quantile, are indicated in respective grey boxes. The blue line repsents the line at postion zero along the y-axis. The red line represents the trend of the data. There was a wider spread of intensity ratios and a slight skewed trend (red line), i.e. away from the zero (blue) line, in pre-normalised data (A). In post-normalisation MA-plot, the intensity ratios were much more collected and the trend (red line) was much closer to zero (blue) line. WT1: WT replicate 1. WT2: WT replicate 2. WT3: WT replicate 3. Sr/+; Nr/+1: Sr/+; Nr/+ replicate 1. Sr/+; Nr/+2: Sr/+; Nr/+ replicate 2. Sr/+; Nr/+3: Sr/+; Nr/+ replicate 3. Relative Intensity Reference Probeset ID Gene symbol WT Sr/+; Nr/+ Gene assignment sequence Replicate 1 Replicate 2 Replicate 3 Replicate 1 Replicate 2 Replicate 3 10544936 Neurod6 NM_009717 3.94881 3.81068 3.72646 3.33057 3.77351 7.20025 NM_009717 // Neurod6 // neurogenic differentiation 6 // 6 B3|6 29.0 cM // 11922 10589994 Eomes NM_010136 4.44706 4.95856 4.79823 4.62272 4.58133 6.9903 NM_010136 // Eomes // eomesodermin homolog (Xenopus laevis) // 9 F3|9 67.0 cM // 10472313 Tbr1 NM_009322 4.66529 4.60745 4.57868 4.05805 4.04543 6.96634 NM_009322 // Tbr1 // T-box brain gene 1 // 2 C1.3|2 32.0 cM // 21375 /// ENSMUST
Figure 5.6: Detection of telencephalic specific genes in the Sr/+; Nr/+ replicate 3 sample. Neurod6, Eomes and Tbr1 were at a relative basal level of 3.5 to 4 in all three WT replicates and both replicate 1 and 2 of Sr/+; Nr/+ samples. However, their relative levels were up to about 7 in Sr/+; Nr/+ replicate 3 (highlighted in yellow). 121
Previous studies have suggested ANOVA analysis as one of the most reliable and easy to use statistic analytical methods for microarray data analysis (Jeanmougin et al., 2010). Thus, ANOVA analysis was performed to determine the statistical significance of differentially expressed genes. A step-up P value was calculated on a false discovery rate (FDR) of 10%. ANOVA analysis indicated there was no significant difference in any gene expression (step up p value ≤ 0.05). Moreover, Sox3, which served as a positive control, although indicated a 1.6-fold increase in Sr/+; Nr/+ samples (Figure 5.7), it did not yield a significant p value. The ten lowest p values ranged from 0.1 to 0.6, with the lowest p value at 0.11 from the gene H3 histone, family 3A (H3f3a). These ten genes also showed a relatively small fold-change, i.e. less than 2, which could be attributed by non-sample related variances (Figure 5.7). This lack of significance in differential gene expression was somewhat expected as previous prediction model has established that a minimal of 3 replicates are required to achieve significance (Gadbury et al., 2004; Page et al., 2006). Limma has been recommended as a more powerful analytical tool for experiment with small samples (Page et al., 2006). Thus, a Limma analysed was also performed. However, due to the limited sample size (n=2), Limma analysis also demonstrated an overall similar significance level with a comparable list of genes to that of the ANOVA analysis, and failed to produce any differential expression with confidence over 95% (see Appendix VI for gene lists). Thus, instead of focusing on statistical analysis, an alternative approach that combines the fold-change, expression pattern and biological function of each gene was employed.
5.3.3 Candidate genes selection
Due to the vast amount of microarray data, candidate gene selection has to be specific and tailored towards the direction for further investigation. In this project, there were three main goals to achieve in further study of the candidate genes. The first goal was to verify the change in candidate gene(s) expression using alternative technique(s) in order to confer the validity of the microarray. The second goal was to further unravel the molecular pathway(s) that was regulated by Sox3 during SCO development. The last goal was to elucidate the molecular profile of alternatively fated SCO cells in Sox3 transgenic embryos. As a result, the candidate gene selection process described below was optimised according to these three goals. 122
An average fold-change for each gene in Sr/+; Nr/+ SCO was deduced through the ANOVA algorithm after pooling the data from all analysing biological replicates. A subset of genes was selected for the fold-change in modulus that is ≥2. This was to minimise the chance of identifying differential gene expression merely due to the influence of background variance. Intriguingly, there were markedly more genes that were downregulated (128) than upregulated (51) in the Sr/+; Nr/+ SCO within this subset of genes (see Appendix VII for the list of these genes). Next, the expression pattern of these genes during embryonic CNS development was investigated using the online brain gene expression atlas (Allen atlas, http://www.brain-map.org/) (Lein et al., 2007). Only genes that demonstrated an endogenous dorsal diencephalic expression between 11.5 to 18.5 dpc were included for further analysis. An interesting trend was observed. A diencephalic RP and/or a SCO restricted expression was seen in most genes that were downregulated in the Sr/+; Nr/+ SCO. In contrast, the expression of upregulated genes were often only found at the Pt but not the RP of the prosomere 1. The Pt is a diencephalic subunit that has lateral domains flanking the prosomere 1 RP and a dorsoventrally compacted dorsal midline domain residing juxtadorsally to the RP (termed dorsal midline Pt) (Figure 5.8). This has suggested that perhaps the SCO (or the RP) might have adopted the fate of the laterally residing Pt cells. In order to further to explore this hypothesis, genes with restricted expression in either the SCO or the Pt were prioritised for further analysis. A functional evaluation based on previous literature was also carried out to select genes known to have a role in dorsal diencephalic development. A summary of the selected candidate genes is illustrated in Figure 5.9.
Overall, the selected candidate genes were grouped into three categories: 1) very high fold-change with restricted expression in either in the SCO or the Pt; 2) genes previously implicated in SCO and RP development and/or relevant to Wnt and Bmp signalling (as both pathways are essential for RP development); and 3) genes with a SCO specific expression along the RP. For category 1, the three most up-regulated and down-regulated genes were selected. As these genes displayed highest fold-changes, they were considered to be a prudent starting point for microarray validation. The three most downregulated genes were Flavin- containing monooxygenase (Fmo1), A Disintegrin-like and Metalloprotease domain with ThromboSpondin type I motif, 3 (Adamts3) and 3-hydroxy-3-methylglutaryl-CoA synthase 2 (Hmgcs2). Previous studies indicated that the expression of these genes are restricted to the SCO along the RP and that they were absent in regions lateral to the RP (Finger et al., 2010; Le Goff et al., 2006; Lein et al., 2007). On the other hand, the three most upregulated genes Fold-Change (Sr/+; Stepup p value (Sr/+; Gene Symbol Gene assignment Reference sequence Nr/+ vs. WT) Nr/+ vs. WT))
ENSMUST00000081026 // H3f3a // H3 histone, H3f3a 1.21951 0.111646 ENSMUST00000081026 family 3A // 1 D2.3 // 15078 /// ENSM
NM_008289 // Hsd11b2 // hydroxysteroid 11- Hsd11b2 1.77996 0.113059 NM_008289 beta dehydrogenase 2 // 8 D3|8 50.8 cM
NM_198886 // Zbtb12 // zinc finger and BTB Zbtb12 1.54537 0.491756 NM_198886 domain containing 12 // 17 B1 // 1937
ENSMUST00000037268 // 1700106J16Rik // RIKEN 1700106J16Rik -2.02532 0.57222 ENSMUST00000037268 cDNA 1700106J16 gene // 11 C // 742
NM_008235 // Hes1 // hairy and enhancer of split Hes1 -1.95362 0.57222 NM_008235 1 (Drosophila) // 16 B2|16 27.0
NM_027444 // Bbx // bobby sox homolog Bbx -1.71745 0.57222 NM_027444 (Drosophila) // 16 B5 // 70508 /// ENSMUST
NM_025736 // Ttc35 // tetratricopeptide repeat Ttc35 -1.68715 0.57222 NM_025736 domain 35 // 15 B3.2 // 66736 ///
NM_178788 // Dctd // dCMP deaminase // 8 B1.2 Dctd -1.60442 0.57222 NM_178788 // 320685 /// ENSMUST00000033966 /
BC025128 // Lrrc51 // leucine rich repeat Lrrc51 -1.55759 0.57222 BC025128 containing 51 // 7 F1 // 69358 /// ENS
NM_009237 // Sox3 // SRY-box containing gene 3 Sox3 1.63689 0.765684 NM_009237 // X A7.3-B|X 24.5 cM // 20675 //
Figure 5.7: No differential regulation was observed in any gene from the microarray with 95% confidence. The lowest p value (0.111646, H3f3a) by ANOVA analysis did not indicate significance. The ten lowest p values are boxed and arranged in an ascending order. The fold-change was also relatively small amongst these genes. As a positive control, Sox3 indicated an increase of 1.6-fold (highlighted in yellow), its p value also did not indicate significance. dorsal midline Pt
RP
MZ VZ VZ MZ
Figure 5.8: Coronal section demonstrating the spatial relationship between the prosomere 1 diencephalic RP and the Pt at 12.5 dpc. The Pt is shaded in colour, with the lateral domains shaded in pink and the dorsal midline domain shaded in light blue. Each lateral domain of the Pt is further divided into the VZ and the MZ (broken lines). The RP is enclosed in solid red lines. Top to bottom: dosal to ventral. Scale bar: 200µm. Position (ranked Candidate ascendingly on Fold-change Dorsal diencepalic expression pattern Rationale(s) genes fold-change, out of 28853 entries) SCO: present at 11.5, 12.5, 13.5, 15.5 and 18.5 dpc (ISH, Lein et al. Fmo1 1 -6.84214 High fold-change 2007)
SCO: present at 17.5 dpc (Immunofluorescence, Le Goff et al. Adamts3 2 -6.56237 High fold-change 2006)
Hmgcs2 3 -5.43513 SCO: present at 14.5 dpc (ISH,Finger et al. 2010) High fold-change
SCO: absent at 11.5 dpc, low at 13.5 dpc, present at 15.5 and 18.5 High fold-change; Bmp family member Bmp5 8 -4.56046 dpc (ISH, Lein et al. 2007); Diencephalic dorsal midline: present at (Furuta et al. 1997) 10.5, 11.5 and 13.5 dpc (ISH, Furuta et al. 1997)
SCO: strong and specific at 11.5 dpc, strong at 13.5, 14.5 and 15.5 dpc (ISH, Visel et al. 2004; Lein et al. 2007; and Finger et al. 2010), High fold-change; Wnt family member (Zakin Wnt2b 17 -3.70574 weak at 18.5 dpc (ISH, Lein et al. 2007); Diencephalic dorsal et al. 1998) midline: present at 9.5 dpc (ISH, Zakin et al. 1998)
Figure 5.9 continues into next page... Position (ranked ascendingly on Candidate fold-change, out Fold-change Dorsal diencepalic expression pattern Rationale(s) genes of 28853 entries) SCO: low at 11.5 dpc, highest at 12.5, and13.5dpc (ISH, Lein et al. 2007), at 14.5 and 15.5 dpc (ISH, Visel et al. 2004; Lein et al. 2007; and High fold-change; agonist of Wnt signalling (Kamata et Rspo3 20 -3.6154 Finger et al. 2010 ), low at posterior SCO at 16.5, 17.5 and 18.5 dpc al. 2004; and Nam et al. 2006;) (ISH, Lein et al. 2007)
SCO: realtively specific to SCO at 11.5, 12.5, 14.5 and 16.5 dpc (ISH and High fold-change; antagonist of Wnt signalling (Hocevar Dab2 34 -3.04962 Immunofluorescence, Visel et al. 2004; Cheung et al. 2008; and Finger et al. 2003; and Jiang et al. 2008, 2009) et al. 2010)
High fold-change; Bmp family member; required for SCO: present at 11.5, 12.5, 13.5, 15.5 and 18.5 dpc (ISH, Bach et al. diencphalic dorsal midline development and implicated Bmp6 35 -3.04961 2003; and Lein et al. 2007); Dorsal diencephalic midline: present at in SCO development (Furuta et al. 1997; and Bach et al. 11.5 and 13.5 dpc (ISH, Furuta et al. 1997) 2003)
SCO: present at high level at 11.5 and 12.5 dpc (ISH, Bach et al. 2003; and Lein et al. 2007), present at 14.5 dpc (ISH, Finger et al. 2010; and Required for SCO and diencephalic dorsal midline Wnt1 55 -2.68569 Visel et al. 2004), present at low level at 15.5 and 18.5 dpc (ISH, Lein et development (Shimamura et al. 1994; and Bach et al. al. 2007); Diencephalic dorsal midline: present at 10.5 dpc (ISH, Bally- 2003) Cuif et al. 1992; and Shimamura et al. 1994)
SCO: present at 11.5, 12.5, 13.5, 14.5, 15.5 and 18.5 dpc (ISH, Bach et Required for SCO and diencephalic dorsal midline Msx1 58 -2.63885 al. 2003; Gray et al. 2004; Visel et al. 2004; and Shimogori et al. 2010) development (Bach et al. 2003)
Figure 5.9 continues into next page... Position (ranked ascendingly on Candidate fold-change, out Fold-change Dorsal diencepalic expression pattern Rationale(s) genes of 28853 entries)
SCO: present only specific in SCO at 11.5 12.5 and 14.5 dpc (ISH, Visel et al. 2004; Lein et al. Fibcd1 60 -2.61436 2007; and Finger et al. 2010), lower expression at 15.5 dpc SCO and absent at 18.5 dpc (ISH, Lein Specific to SCO; SCO ealry marker et al. 2007)
Exclusive to RP cells and is required for RP identity SCO: present at 12.5 dpc (Louvi and Wassef 2000); Diencephalic dorsal midline: present at 10.5 Gdf7 157 -1.91216 maintenance and dorsal neural tube patterning (Lee et al. and 12.5 dpc (Monuki et al. 2001) 1998; and Lee et al. 2000)
SCO: absent at, 11.5, 13.5, 15.5 and 18.5 dpc. Lateral to SCO: at pretectum MZ at 11.5 dpc (ISH, Lein et al. 2007; and Shigomori et al. 2010); at dorsal midline pretectum (a thin layer dorsal to High fold-change; its expression is lateral to but excludes Lhx5 28849 2.94516 SCO), MZ of pretectum, tectum, tegmentum at 12.5, 13.5, 14.5, 15.5, 16.5 and 18.5 dpc (ISH, dorsal diencephalic midline Visel et al. 2004; Lein et al. 2007; Vue et al. 2007; Abellan et al. 2010; Finger et al. 2010; and Shimogori et al. 2010)
SCO: absent at, 11.5, 13.5, 15.5 and 18.5 dpc. (ISH, Lein et al. 2007); Lateral to SCO: present at pretectum MZ at 11.5 dpc (ISH, Lein et al. 2007), at dorsal midline pretectum (a thin layer dorsal High fold-change; its expression is lateral to but excludes Gata3 28852 3.70456 to SCO), MZ of pretectum, tectum, tegmentum at 12.5, 13.5, 14.5, 15.5 and 18.5 dpc (ISH, dorsal diencephalic midline Nardelli et al. 1999; Visel et al. 2004; Lein et al. 2007; and Finger et al. 2010).
SCO: absent at, 11.5, 13.5, 15.5 and 18.5 dpc. Lateral to SCO: at P1 MZ of diencephalic walls at 11.5 dpc, at dorsal midline pretectum (a thin layer dorsal to SCO), at MZ of pretectum, tectum, High fold-change; its expression is lateral to but excludes Sox14 28853 4.51502 tegmentum at 12.5, 13.5, 14.5, 15.5 and 18.5 dpc (ISH, Visel et al. 2004; Hirata et al. 2006; Lein dorsal diencephalic midline et al. 2007; Finger et al. 2010; and Hashimoto-Torii et al. 2003)
Figure 5.9: Table of candidate genes identified using the alternative approach. The expression pattern of these genes within the developing dorsal diencephalic development was interrogated. Rationales for selection as a candidate gene are also stated on the right. 123
were LIM homeobox protein 5 (Lhx5), GATA binding protein 3 (Gata3) and Sox14. These genes are highly expressed in the dorsal midline Pt and at the MZ of lateral Pt (Abellán et al.; Finger et al., 2010; Hashimoto-Torii et al., 2003; Hirata et al., 2006b; Lein et al., 2007; Nardelli et al., 1999; Shimogori et al., 2010; Visel et al., 2004; Vue et al., 2009), but are absent from the SCO or RP. The correlation of these specific expression patterns with three of the most upregulated and downregulated genes further strengthened the previous proposal that Sr/+; Nr/+ SCO tissue might take on a molecular profile similar to that of its neighbouring regions.
The category 2 genes were RP specific genes that are implicated in the signalling pathways required for RP development, i.e. Wnt and Bmp signalling pathways. The genes selected for this category were Bmp5, Wnt2b, Rspo3, Bmp6, Msx1, Wnt1 and Gdf7 (Figure 5.9). In accordance with previous in situ experiment, both Msx1 and Wnt1 were found to be downregulated (Bach et al., 2003; Louvi and Wassef, 2000). Bmp and Wnt family signalling pathways have been shown to function in RP induction, maintenance and RP dependent dorsal CNS patterning in both chick and mouse (Chizhikov and Millen, 2004c; Chizhikov and Millen, 2005; Hébert et al., 2002; Lee et al., 2000a; Lee et al., 1998; Megason and McMahon, 2002). Moreover, Bmp6 expression was lost in the SCO of Msx1 null mutant embryos (Bach et al., 2003). Thus, the Bmp family members Bmp6 and Bmp5, which also display RP restricted expression (Furuta et al., 1997; Zakin et al., 1998), were selected. Another Bmp family member, Gdf7, is the only gene identified so far that is exclusive to definitive RP cells (Krispin et al., 2010a; Lee et al., 1998), i.e. not expressed in early RP and hence, NC progenitors. Thus, Gdf7 was included in the selection despite it was only downregulated with a borderline fold-change in modulus of 1.9. In addition, Wnt2b, a Wnt family member (Zakin et al., 1998) and Rspo3, a Wnt signalling pathway agonist (Kamata et al., 2004; Nam et al., 2006), were also selected.
Category 3 genes are those with an expression domain restricted to the SCO primordium, rather than the entire RP. One would expect these genes may be important for SCO specification or function rather than for RP development in general. In accordance to previous in situ hybridisation finding (Figure 4.13), Fibcd1 expression was reduced 2.6-fold in Sr/+; Nr/+ SCO. Disabled homolog 2 (Drosophila) (Dab2) encodes a membrane- associated signalling adaptor molecule. Dab2 RP expression is restricted to the SCO along the RP from 11.5 to 16.5 dpc (Cheung et al., 2008; Lein et al., 2007). Dab2 transcripts are 124
also present at 16.5 dpc in the PG but not the adult SCO. Dab2 expression was downregulated 3-fold by the microarray analysis. Interestingly, DAB2 is a Wnt signalling antagonist whose transcription activity is regulated by Wnt signalling (Jiang et al., 2009; Jiang et al., 2008; Railo et al., 2009). Thus, Dab2 was selected for further evaluation because this might add to our knowledge of the cellular identity at 12.5 dpc Sr/+; Nr/+ SCO, as well as complement the Wnt signalling investigation as outlined for category 2 genes.
5.3.4 Validation of transcriptional profiles in microarray through qRT-PCR
In order to validate the expression changes in Sr/+; Nr/+ SCO identified by the microarray, qRT-PCR was carried out with the above selected genes using RNA extracted from 12.5 dpc embryos. SCO microdissection technique identical to that of the microarray was used to obtain material for qRT-PCR. Analysis was firstly performed on the two biological replicates from the microarray. Further repeats were also carried out in three extra pooled biological replicates for Fmo1, Adamts3, Msx1, Wnt1 and Sox14. Due to the limited availability of desired embryos, only two of three extra pooled biological replicates were included for Hmgcd2, Bmp5, Wnt2b, Rspo3, Dab2, Fibcd1, Gdf7, Lhx5 and Gata3, while only one of three extra pooled biological replicate was included for Bmp6. Succinate dehydrogenase complex, subunit A, flavoprotein (Fp) (Sdha) served as the reference gene in these qRT-PCR with the Glyceraldehyde-3-phosphate dehydrogenase (Gapdh) as another endogenous control. Both Sdha and Gapdh expression did not alter significantly between and amongst WT and Sr/+; Nr/+ pooled samples, indicating that they were appropriate endogenous control genes for the prosomere 1 RP tissue. On the other hand, Sox3 and EGFP served as positive controls and were also incorporated into the qRT-PCR in both microarray samples and the three extra biological replicates. As expected, the expression of both EGFP and Sox3 were significantly elevated (n=5) (Figure 5.10-5.14). In fact, Sox3 transcript was upregulated 2.5-fold in Sr/+; Nr/+ prosomere 1 RP in comparison to that of the WT. Together, these conferred the validity of the qRT-PCR experiments. Hence, the result of candidate genes validation by qRT-PCR is described below. * p=0.0500
* p=0.0142
* p=0.0006 * p=0.0001 * p=0.0060
Figure 5.10: qRT-PCR verification of the three most downregulated genes in the Sr/+; Nr/+ prosomere 1 RP by the microarray. The expression of these genes were significantly downregluated in the Sr/+; Nr/+ SCO or prosomere 1 RP. Please note that these genes display a RP specific expression within the dorsal prosomere 1. *p=0.0500
*p=0.0307
*p=0.0126
*p=0.0142
Figure 5.11: qRT-PCR verificaiton of the three most upregulated genes in the Sr/+; Nr/+ prosomere 1 RP by the microarray. Gata3 and Lhx5 expression were significantly increased in the Sr/+; Nr/+ prosomere 1 RP or SCO. No significance was obatined for Sox14, despite its expression appeared to be elevated in the Sr/+; Nr/+ prosomere 1 RP or the SCO. Note that all three genes are only expressed within the Pt but not the RP of prosomere 1. *p=0.0500
*p=0.0142
*p=0.0159 *p=0.0138 *p=0.0011 *p=0.0266 *p=0.0012
Figure 5.12: qRT-PCR validation of genes that are implicated in Wnt and Bmp signalling pathway and were downregulated in the Sr/+; Nr/+ prosom- ere 1 RP by the microarray. Bmp5, Wnt2b, Rspo3, Bmp6 and Msx1 were all signficantly downregulated in the prosomere 1 RP of Sr/; Nr/+ embryos. The trend of Wnt1 differential expression in Sr/+; Nr/+ embryos was not consistent and hence, did not yield any significance. Gdf7 data is not presented in the above graph as it could not successfully amplified in the qRT-PCR experiments. *p=0.0500
*p=0.0142
*p=0.0016
Figure 5.13: qRT-PCR validation of SCO specific genes that displayed a decrease in abundance in the Sr/+; Nr/+ prosomere 1 RP by the microarray. Fibcd1 was significantly downregulated in the Sr/+; Nr/+ prosomere 1 RP. However, Dab2 expression was not signficantly different in the prosomere 1 RP of Sr/+; Nr/+ embryos when compared to the WT. 125
Category 1: genes with high fold-change
qRT-PCR analysis of the three most downregulated genes (Fmo1, Adamts3 and Hmgcs2) revealed consistent and statistically significant downregulated expression (approximately 5-fold) in the Sr/+; Nr/+ 12.5 dpc SCO. Their differential expression agreed with the microarray, although the average fold-changes were slightly less than the 5- to 7-fold detected from the microarray (Figure 5.10). Similarly, two of the three most upregulated genes, Lhx5 and Gata3, also displayed a similar trend to that indicated by the microarray data. Both genes were significantly elevated (5- and 6-fold respectively) in Sr/+; Nr/+ embryos, which was slightly higher than the 3.7- and 4-fold observed, respectively, from the microarray (Figure 5.11). Conversely, Sox14 displayed an average 3.9-fold increase from qRT-PCR (Figure 5.11). Its expression was consistently upregulated across all five biological replicates, although its fold-changes were widely ranging from 1.2- to almost 8-fold across samples. This might due to the fact that Sox14 expression is more prone to influences from sample to sample variations (discussed in 5.4.1). This idea is further supported by its insignificant but low p value of 0.0975, which suggests that Sox14 may reach statistical significance if a larger sample size was utilised.
Category 2: Genes implicated in Wnt and Bmp6 signalling pathways
qRT-PCR was also performed to validate the altered expression of genes implicated in Wnt and Bmp signalling. Interestingly, Bmp5 and Rspo3 were detected with an average of 5- and 4-fold decrease respectively, while both Wnt2b and Bmp6 showed a 3-fold downregulation by qRT-PCR (Figure 5.12). As described in chapter 4 (Figure 4.14), Wnt1, and Msx1 were downregulated in the SCO and the RP anterior to the SCO. Consistent with these results, Msx1 was found to be downregulated 2-fold by qRT-PCR in the Sr/+; Nr/+ SCO, which was comparable to the fold-change detected in the microarray. Intriguingly, this was not true for Wnt1. Wnt1 expression analysis by qRT-PCR failed to demonstrate a consistent trend in its differential expression in Sr/+; Nr/+ SCO. In fact, qRT-PCR found that Wnt1 expression was downregulated in Sr/+; Nr/+ SCO by 2- and 10-fold in the two microarray samples. While the three additional replicates displayed a 10-fold decrease, an opposing 7-fold increase and no change respectively. Although a markedly larger fold-change in modulus was detected in the two microarray samples by qRT-PCR than that of the microarray (approximately 2.6-fold), a consistent trend (both were downregulation) was 126 detected by the two techniques, demonstrating that Wnt1 was not a false positive in the microarray experiment. However, the inconsistent trend of Wnt1 expression regulation in the extra three replicates deemed the qRT-PCR validation of Wnt1 failed. Nevertheless, Wnt1 expression was reduced in the SCO and its anterior RP at 12.5 dpc Sr/+; Nr/+ embryos, as evident by in situ hybridisation experiment (Figure 4.14). The reason(s) behind the considerable variation in fold-change differences are not known. Yet, with the exception of Wnt1 and Gdf7, all genes from category 2 were significantly downregulated in Sr/+; Nr/+ SCO in qRT-PCR with comparable fold-changes to those of the microarray.
Lastly, despite repeated attempts using multiple primer pairs, a robust qRT-PCR assay for Gdf7 was unable to be developed. Gdf7 is highly expressed in 12.5 dpc SCO according to previous literatures (Louvi and Wassef, 2000; Monuki et al., 2001). Moreover, from the post- normalised microarray data (see Appendix VI), the signal intensity and hence, the relative expression of Gdf7 in the WT embryos (approximately 6) was well above background (approximately 3), suggesting that Gdf7 was at least moderately expressed within the SCO or the prosomere1 RP of 12.5 dpc embryos. Thus, the failure of its detection by qRT-PCR was likely to be a technical problem that is yet to be resolved.
Category 3: Genes with SCO specific expression
In accordance with both the microarray and earlier performed in situ hybridisation (Figure 4.13), an average of 4-fold decrease in Fibcd1 expression was observed in Sr/+; Nr/+ SCO with significance, although this fold-change was slightly higher than the 2.6-fold change detected in the microarray. In contrast, Sr/+; Nr/+ SCO Dab2 expression had a 1.25-fold decrease from the two microarray samples. This did not correlate with the microarray, which indicated a 3-fold decrease in expression. These data suggested that Dab2 was a false positive from the microarray experiment. This finding was further supported by the two extra biological replicates, which displayed a 1.25-fold decrease and a 2-fold increase in Sr/+; Nr/+ SCO, leading to an average of 1.12-fold increase and a p value of 0.7355. Thus, Dab2 was not differentially regulated in the Sr/+; Nr/+ SCO.
In summary, qRT-PCR has successfully validated the microarray-derived transcriptional profiles of Sr/+; Nr/+ SCO. Most selected candidate genes displayed differential gene expression with consistency across the two techniques. These genes are Fmo1, Adamts2, Hmgcs2, Lhx5 and Gata3, Bmp5, Rspo3, Msx1, Bmp6, Wnt2b and Fibcd1. WT Sr/+; Nr/+ WT Sr/+; Nr/+ E E’ M A A’ ] ]
] ] * MZ Pt * * Pt Pt VZ NOTE: Pt MZ * 3V VZ 3V This figure is included in the print copy of the thesis 3V 3V held in the University of Adelaide Library.
F B B’ ] F’ ] ] ] Pt * * Pt * Aq * Aq Aq Pt Aq Pt
3V 3V
Lhx5 G Gata3 C C’ ] ] G’ ] ] * * * * Pt Aq Pt Aq Aq Aq Pt Pt
D D’ H ] H’ ] ] ] * Pt * * Pt * Aq Pt Aq Pt Aq Aq
12.5 dpc 12.5 dpc Figure 5.14: Continues into next page... WT Sr/+; Nr/+ N
I ] I’ * ] NOTE: Pt MZ * VZ Pt This figure is included in 3V the print copy of the thesis 3V held in the University of Adelaide Library.
J J’ ] ] * Pt * Pt Aq Aq Figure 5.14: The expression domains of Lhx5, Sox14 and Gata3 were altered 3V and expanded at the Pt of 12.5 dpc Sr/+; Nr/+ embryos. A-L: In WT embryos, , and were highly expressed at the MZ of the Pt, as well as the 3V Lhx5 Sox14 Gata3 dorsal midline Pt (above the RP) (bracketed). The expression of Lhx5, Sox14 and at the dorsal midline Pt were relatively more collected in sections anterior Gata3 K K’ Gata3 ] ] to the SCO (A, B, E, F, I and J) than those above the SCO (C, D, G, H, K and L), * * due to the intermingling of dorsal midline Pt and PC. A’-L’: In Sr/+; Nr/+ Pt embryos, the expression of Lhx5, Sox14 and Gata 3 at the dorsal midline Pt Aq Pt Aq (bracketed) were obviously expanded. A double-layer expression pattern (arrow) and bifurcation at the ventral-most border of the lateral domains (arrowhead) were observed in some sections (A’, B’, E’, F’, I’ and J’). Note that (A) to (D), (A’) to (D’), (E) to (H), (E’) to (H’), (I) to (J) and (I‘) to (J’) are section series in an L L’ anterior to posterior direction. M-N: A sagittal section from Schambra (2008) indicating the plane of sectioning for (A) to (L) and (A’) to (L’). Note the section- ] ] ing planes were color-coded to the respective sections. In situ hybridisation. * * Asterisk: RP. Coronal sections. Top to bottom: dorsal to ventral. Scale bar: Pt Pt Aq Aq 200µm.
12.5 dpc WT Sr/+ Nr/+ A B C Mes Mes Mes Di Di Rb Rb Rb Di Tel Tel Tel 12.5 dpc
Sr/+; Nr/+ Sr/+; Nr/+ exen1 Sr/+; Nr/+ exen2 D E F Mes Mes Mes Di Rb Di Rb Di Rb Tel Tel Tel
Figure 5.15: Exencephaly with low penetrance was observed in Sr/+; Nr/+ embryos at 12.5 dpc. A-D: The anatomy of the cranium was comparable between the WT and Sr/+; Nr/+, Sr/+ and Nr/+ embryos that had undergone successful fusion. E: Sr/+; Nr/+ embryo with exencephaly that exposes the entire cranium, i.e. from proencephalon, including the telencephalon (arrowheads) and diencephalon, to the rhombencephalon. F: Sr/+; Nr/+ embryo with exencephaly only within the diencephalon, mesencephalon and rhombencephalon, but not the telencephalon (arrowheads). A-E: Darkfield. F: Brightfield. Tel: telencephalon. Di: diencephalon. Mes: mesencephalon. Rb: Rhombencephalon. Top to bottom: dorsal to ventral. Left to right: anterior to posterior. Scale bar: 500µm. 127
Although a slightly varied fold-change, i.e. less than ±2 difference where fold-changes > 2.5, was detected between the qRT-PCR and microarray for Fmo1, Adamts3, Lhx5, Gata3 and Fibcd1, their trends of changes were significantly conserved between the two techniques.
5.3.5 The expression domains of pretectal markers are expanded
To further validate the microarray result and to investigate the hypothesis that the SCO adopts a fate that is similar to its neighbouring regions, in situ hybridisation expression analysis of Gata3 and Lhx5 in the prosomere 1 RP of Sr/+; Nr/+ embryos were carried out (performed by Michael Morris). In addition, in situ hybridisation expression analysis of Sox14, which displayed a consistent trend of downregulation by qRT-PCR but did not reach statistical significance, was also performed. Sox14 is a transcriptional repressor that belongs to the SoxB2 subgroup. It has been previously shown that Sox21, the other SoxB2 member, confers an opposing effect on the transcriptional activity of target genes regulated by Sox3 within the developing chick NT (Bylund et al., 2003; Sandberg et al., 2005). Thus, its expression pattern was also studied along with those of Lhx5 and Gata3.
Expression of Gata3, Lhx5 and Sox14 were analysed on tissue sections from WT and Sr/+; Nr/+ 12.5 dpc embryos. Interestingly, the expression of these genes in WT were most prominent in the MZ (see Figure 5.8 for details on Pt subdivisions) of the Pt lateral domains, as well as its dorsal midline domain (Figure 5.14 A-L). This is in agreement with previous literature, which found that Gata3, Lhx5 and Sox14 are specific markers for the MZ but not the VZ of the caudal Pt (Abellán et al.; Finger et al., 2010; Hashimoto-Torii et al., 2003; Hirata et al., 2006a; Lein et al., 2007; Nardelli et al., 1999; Shimogori et al., 2010; Visel et al., 2004; Vue et al., 2009). Comparison of WT and Sr/+; Nr/+ Gata3, Lhx5 and Sox14 in situ hybridisation data revealed that there was no detectable difference in signal intensity within the lateral and dorsal medial Pt. However, the expression domain was clearly different in the Sr/+; Nr/+ embryos. In fact, a bifurcation was observed at the ventral-most boarder of the lateral domains in the anterior sections from Sr/+; Nr/+ embryos (n=3) (Figure 5.14 A’-L’). Moreover, an expansion of the expression domain at the dorsal midline Pt was apparent in the region dorsal to the SCO and in the RP anterior to the SCO (n=3) (Figure 5.14 A’-L’). All three genes displayed a very restricted thin layer of expression at the dorsal midline Pt within the WT embryos. However, the expression domain of these genes appeared as a double-layer 128 at the dorsal midline Pt of Sr/+; Nr/+ embryos (Figure 5.14 A’-L’). Interestingly, no Gata3, Lhx5 and Sox14 ectopic expression was observed in the region next to the ventricular area within the diencephalic RP (n=3) (Figure 5.14 A’-L’), indicating that these cells did not take up a cellular identity of the dorsal midline Pt or the MZ pretectal cells.
5.3.6 Low penetrance of exencephaly observed in Sr/+ and Sr/+; Nr/+ embryos
During generation of double hemizygotes Sr/+; Nr/+ embryos, an exencephaly phenotype was detected with low penetrance. Exencephaly is a condition where growth of neural tissue continues despite a failure of NT closure. This results in CNS development outside of the skull vault (Greene and Copp, 2009). Exencephaly was observed from 10.5 dpc to as late as 15.5 dpc. Due to a limited sample size, exencephaly was quantified at 12.5 dpc only. Exencephaly was detected in 6 out of 38 (16%) Sr/+; Nr/+ embryos at 12.5 dpc and was not observed in Nr/+ embryos at any stage. In addition, 1 out of 53 Sr/+ embryos (2%) had exencephaly at 12.5 dpc. The markedly higher frequency of exencephaly in Sr/+; Nr/+ embryos in comparison to either Sr/+ or Nr/+ embryos indicated that this phenotype, similar to the CH, was also sensitive to Sox3 dosage.
There was no significant difference observed in the gross cranial morphology of WT, Sr/+; Nr/+, Sr/+ and Nr/+ 12.5 dpc embryos that have undergone successful NT closure (Figure 5.15 A-D). In contrast, Sr/+; Nr/+ embryos that displayed exencephaly consistently presented with exposed brain tissue from the diencephalon to the rhombencephalon (Figure 5.15 E-F). In some cases, the telencephalic midline was successfully closed (Figure 5.15 F). Overall, exencephaly was observed in Sr/+; Nr/+ embryos with incomplete penetrance and varying level of severity. This lack of proper NT enclosure indicates that in addition to the CH, a neurulation defect was also present in Sr/+; Nr/+ embryos.
5.4 Discussion
5.4.1 Some candidate genes displayed variable fold-changes between the microarray and qRT-PCR experiments 129
It is important to note that for Fmo1, Adamts3, Lhx5, Gata3 and Fibcd1, the fold- change detected by the microarray varied slightly to that by the qRT-PCR experiment. There are a few reasons behind these variations in fold-changes. First, there were two biological replicates used in the microarray, while qRT-PCR validated the microarray with additional biological replicates. This thus, might introduce biological variation amongst different embryos, which contributed to these discrepancies. Moreover, SCO dissections were non- homogeneous, which might lead to expression variation depending on the level of tissue homogeneity. Third, previous literature has demonstrated that microarray is less sensitive in transcript detection in comparison to qRT-PCR (Cue et al., 2009). Hence, there poses a possibility that the microarray might not detect the transcript as accuratly as qRT-PCR, contributing to the variations observed.
5.4.1 Bmp and Wnt RP expression was diminished transgenic embryos
Expression downregulation has been detected in a number of genes implicated in Bmp or Wnt signalling pathways at the developing prosomere 1 RP of Sox3 transgenic embryos by the microarray and qRT-PCR experiments. Bmp and Wnt are two signalling pathways that have important roles in a wide variety of biological processes, including CNS specification and differentiation, haematopoiesis, bone morphogenesis, craniofacial development and oncogenesis. Both Bmp and Wnt signalling pathway can be activated by a large repertoire of ligands, which have varying affinities to their common receptors. Moreover, Bmp and Wnt signalling are able to functionally interact with each other and/or with other signalling pathways or factors, e.g. Notch signalling, dependent on cellular contexts. Together, these properties confer the capability of Bmp and Wnt signalling to exert temporally dynamic regulation across a broad range of biological processes (Chen and Panchision, 2007; Freese et al.; Fuerer et al., 2008; Gordon and Nusse, 2006; Grigoryan et al., 2008).
Bmp signalling is activated by the binding of a Bmp family ligands to either a Bmp type I or type II receptor. Upon ligand binding, the Bmp receptor triggers the canonical signalling by inducing formation and nuclear translocation of the SMAD transcription factors protein complex. SMAD complex then partners with its co-activator, CBP/p300, in order to initiate the transcription of downstream target genes. In neurogenesis promotion, Bmp signalling can also activate mitogen associated protein (MAP) kinase pathway and 130
mammalian target of rapamycin (mTOR) pathway to activate STAT proteins for further recruitment of CBP/p300 co-activator (Figure 5.16) (Chen and Panchision, 2007).
Similarly, canonical Wnt signalling is activated via the nuclear translocation of the co-transcriptional activator, β-catenin. In the absence of Wnt ligand, the β-catenin degradation complex, which is made up of adenomatous polyposis coli (APC), glycogen synthase kinase (GSK3) and Axin proteins, constantly phosphorylates cytoplasmic β-catenin and signals it for degradation. Without β-catenin in the nucleus, the transactivating activity of the T cell-specific transcription factor (TCF) protein is repressed by the binding of the transcriptional repressor, Groucho. Wnt signalling is activated when Wnt ligand binds to the G-protein coupled receptor, frizzled, together with the co-receptor, low-density lipoprotein receptor (LDLR)-related protein (LRP) receptor. Upon binding, the ligand-receptor complex then phosphorylates the cytoplasmic protein Dishevelled (Dsh), which inhibits the action of β-catenin degradation complex through phosphorylation. This then allows accumulation and nuclear translocation of β-catenin. Upon translocation, β-catenin binds to TCF to activate the transcription of Wnt induced target genes (Figure 5.17) (Gordon and Nusse, 2006).
Both Bmp and Wnt signalling pathways have been previously shown to be essential for dorsal NT patterning and development (Chizhikov and Millen, 2004a; Chizhikov and Millen, 2004c; Chizhikov and Millen, 2005; Lee et al., 2000a; Lee et al., 1998; Liem et al., 1997; Liem et al., 1995; Shimamura et al., 1994). Thus, it is perhaps not surprising that a number of Bmp and Wnt signalling related genes were identified to be differentially regulated in the Sox3 transgenic SCO. In situ hybridisation experiments presented in chapter 4 (Figure 4.14) showed reduced RP expression of Wnt1 and Msx1 in the Sr/+; Nr/+ embryos. Thus, the downregulation of Msx1 and Wnt1 in the microarray, to some degree, served as positive controls for the study. Surprisingly, qRT-PCR validated Msx1 but failed to verify the Wnt1 downregulation observed in the microarray and in situ hybridisation experiment. Unlike Sox14, there was no consistent trend of fold-change in Wnt1 expression. While this issue remains unresolved, it is possible that the variability in qRT-PCR expression related to the slightly different developmental windows used for the qRT-PCR and microarray analyses, since SCO tissue was collected from embryos between 53 to 55 somites-stages for the microarray but between 52 to 56 somites-stages for the qRT-PCR experiment. Perhaps, Wnt1 expression is very dynamically regulated during this developmental window such that its expression level is drastically different across a timeframe more than three somites-stages. Figure 5.16: Schematic overview of Bmp signalling. The type I and type II Bmp receptors undergo dimerisation upon ligand binding. Activated dimerised receptor triggers the formation of the SMAD complex, which is then translo-
cated into the nucleus to initiate downstream target genes transcription when NOTE: partnered with its coactivator, CBP/300. In the promotion of neurogenesis, This figure is included in the print copy of Bmp signalling also activates mTOR signalling, which then recruits STAT the thesis held in the University of Adelaide Library. proteins to complex with SCBP/p300 coactivator. Bmp signaling elicits a wide range of biological consequences, e.g. upregulates caspase activation for apoptosis, and increases the expression of Msx1 for dorsalisation and Wnt1 for proliferation. BMPR: Bmp receptor. Figure from Chen and Panchision (2007)
NOTE: This figure is included in the print copy of the thesis held in the University of Adelaide Library.
Figure 5.17: Schematic overview of Wnt signalling. A: In the absence of Wnt ligand, β-cantenin is constantly phosphorylated for degradation by its destruction complex, i.e. APC/GSK3/Axin. B: Upon ligand binding to Frizzled and LRP receptors, Dsh is phosphorylated (P). Phosphorylated Dsh in return, phosphorylates and inhibits β-cantenin destruction complex (P). This allows the accumulation and nuclear translocation of β-cantenin. Within the nucleus, β-cantenin partners with TCF to activate the transcription of Wnt target genes. Figure from Gordon and Nusse (2006). 131
Bmp ligands
Intriguingly, Bmp5, Bmp6 and Gdf7, which are all ligands of Bmp signalling pathway, were downregulated in Sr/+; Nr/+ embryos by the microarray analysis. qRT-PCR verification further validated this microarray result to be significant for both Bmp5 and Bmp6. Similar to Bmp4 and Bmp7, Bmp5 has been shown to induce dorsal NT neuronal differentiation in chick explants when overexpressed in vitro (Liem et al., 1997). Moreover, Bmp5 is important for proper mid-hindbrain organiser functioning in CNS development (Tilleman et al., 2010). In addition, Bmp5 is capable of inducing dendritic growth in vitro (Beck et al., 2001), and promoting early proliferation but later differentiation of chondrogenesis in vivo and in vitro (Mailhot et al., 2008). Less is known about the function of Bmp6 during CNS development. However, in vitro study in human ES cells demonstrated the ability of Bmp6 to induce stem cell proliferation when heterodimerised with Bmp2 (Valera et al., 2010). Moreover, its expression was downregulated in the P1 RP in the Msx1 null mice, demonstrating its possible function in RP and dorsal NT development (Bach et al., 2003).
Another Bmp member, Gdf7, is also important for chick dorsal NT patterning. Loss of Gdf7 function or Gdf7 induced RP ablation results in a lack of NT patterning, implicating a yet to be determined function of Gdf7 in RP development (Lee et al., 2000a; Lee et al., 1998). Interestingly, Gdf7 mutant mice displayed CH in addition to the loss of NT patterning (Lee et al., 1998). Moreover, Gdf7 is required for RP repulsion of spinal cord commissural axons as well as dorsal telencephalic ChP development, which are processes highly dependent on a normal RP formation (Butler and Dodd, 2003; Currle et al., 2005). Unfortunately, qRT-PCR validation of the Gdf7 downregulation in Sox3 transgenic SCO was not successful due to experimental technicalities. Since the Gdf7 mode of action in RP development is still unknown, it is thus interesting to further pursue the qRT-PCR validation and hence, elucidate its regulating molecular mechanism in RP development.
Wnt ligands
Both the microarray and qRT-PCR indicated a downregulation of Wnt2b and Rspo3 in Sox3 transgenic SCO. Unlike Bmp signalling, Wnt signalling although is required for RP development, but there is no evidence of Wnt signalling in dorsal NT induction. The fact that Bmp signalling is upstream of Wnt signalling during RP development is readily demonstrated 132
in the experiment within the chick NT where the expression of dorsal NT markers were induced by exogenous Bmp signalling but not Wnt signalling (Liem et al., 1995). Interestingly, Wnt2b RP expression is restricted to the diencephalon and mesencephalon (Zakin et al., 1998). This is in contrast to the expression of Wnt1, which is expressed along the RP of the entire neuraxis (Shimamura et al., 1994). There is limited knowledge of Wnt2b function in dorsal NT development. Nevertheless, study on other biological processes showed that Wnt2b appears to have paradoxical cellular influences on development depending on cellular context, since it inhibits neuronal differentiation in the developing retina (Kubo et al., 2005) but promotes specification of endothelium progenitors in lung (Gross, 2000).
Lastly, the R-spondin family was recently demonstrated to bind to the LRP receptor and synergistically activate Wnt signalling with Wnt ligands (Nam et al., 2006). Rspo3 activates Wnt signalling and has been shown to promote endothelial proliferation and sprouting in angiogenesis in vitro (Kazanskaya et al., 2008), as well as enable endocytosis of clathrin-associated ligand-receptor complex in the signal transduction of Wnt signalling during Xenopus gastrulation (Ohkawara et al., 2011). Interestingly, in vitro analysis demonstrated that Rspo3 is able to amplify Wnt1 induced signalling (Kim et al., 2008). Thus, overexpression of Sox3 in the RP did not only inhibit Wnt1 expression for Wnt signalling, but also the expression of Wnt1 synergistic partner, Rspo3.
In summary, the identification of a number of differentially expressed Wnt and Bmp ligands, i.e. Bmp5, Bmp6, Gdf7, Wnt2b, Wnt1 and Rspo3, indicates perturbation of several aspects of Bmp and Wnt signalling in the RP result from Sox3 overexpression. However, the cellular consequence of the putative alteration in Wnt or Bmp signalling on RP development is not clear, since both signalling pathways have cellular-context dependent function. Moreover, both Bmp and Wnt ligands are diffusible factors that can function locally in dorsal NT and/or RP development or instruct neighbouring dorsal NT patterning upon diffusion. (Chizhikov and Millen, 2004c; Chizhikov and Millen, 2005; Hébert et al., 2002; Lee et al., 1998). In fact, this dual effect on RP development is readily demonstrated by Wnt1, where the lack of RP Wnt1 function caused the misexpression of adhesion molecules within the RP of a Wnt1 loss-of-function mouse model (a function in RP development) (Shimamura et al., 1994), while RP secreted Wnt1 was capable of instructing commissural migration of pathfinding axons within the NT (a function of RP dependent dorsal NT development) (Bovolenta et al., 2006; Liu et al., 2005). As a result, downregulation of the above identified targets, which 133
includes Wnt1, may have a dual effect in NT development, as it can first lead to a disrupted RP development, and second a defective dorsal NT patterning due to the failure of proper RP function. In support of this, a failure of dorsal diencephalic patterning was also identified in Sox3 transgenic embryos (discussed in 5.4.4). Further study to elucidate a possible association between the identified Wnt and Bmp molecules and the observed patterning defect may provide a more complete understanding of how the RP development and dorsal NT patterning are affected by Sox3 overexpression.
5.4.2 Fmo1, Adamts3, Hmgcs2 – a possible function in the synthesis of SCO secretions
Both microarray and qRT-PCR demonstrated that Fmo1, Adamts3 and Hmgcs2 were downregulated in the Sox3 transgenic SCO at 12.5 dpc, suggesting their role in SCO development or function. Moreover, these three genes are all expressed in the 12.5 dpc SCO, and for Fmo1 and Hmgcs2 particularly, the expression is present along RP within the diencephalon (Finger et al., 2010; Le Goff et al., 2006; Lein et al., 2007). However, none of these three genes has been implicated in dorsal diencephalic development.
Fmo1 encodes an enzyme that metabolises sulphur and nitrogen containing drugs and pesticides and it plays an essential role in biological detoxification (Krueger and Williams, 2005). In addition to the developing SCO, Fmo1 is highly expressed in human and mouse liver and kidney, and has also been found in adult mouse lung, ChP and cerebrum. (Dolphin et al., 1996; Janmohamed et al., 2004; Koukouritaki et al., 2002). In vitro functional analysis through the use of recombinant human Fmo1 demonstrated that hFMO1 is capable of metabolising carbon chain through oxygenation, in particular methamphetamine analogue through N-oxidation (Lee et al., 2009; Zhang et al., 2007). In agreement with this in vitro result, Fmo1 loss-of-function mouse model displayed an enhanced response to the drug imipramine, an anti-depressant that is metabolised by N-oxidation (Hernandez et al., 2009).
Adamts3 encodes a metalloproteinase that is highly expressed in mouse cartilages, bones and musculoskeletal tissue, and is well characterised to have a role in connective tissue synthesis by processing fibrillar procollagen into collagen through cleavage of its aminopropeptide. Col1a1 encodes a type I collagen which can be processed by ADAMTS3 as demonstrated by an in vitro study (Fernandes et al., 2001; Le Goff et al., 2006). Interestingly, 134
Col1a1 is expressed at the 13.5 dpc diencephalic RP (Lein et al., 2007). Moreover, its expression was downregulated by 1.9 fold in the 12.5 dpc Sr/+; Nr/+ prosomere 1 RP by the microarray experiment. Together, these prompt speculation that ADAMTS3 may process COL1A1 during SCO genesis. However, it is also possible that ADAMTS3 may function in a manner not related to collagen processing within the developing SCO. Further elucidation of the mode of action of ADAMTS3 is required to understand its molecular function in the SCO during development.
Hmgcs2 encodes a rate-limiting enzyme that catalyses the second last condensation step in the production of ketone bodies from fatty acids and amino acids during metabolism (Dashti and Ontko, 1979; Hegardt, 1999). In addition, HMGCS2 heterodimerises and enhances the activity of a ligand activated transcription factor encoded by peroxisome proliferator activated-receptor alpha (Pparα) (Kostiuk et al., 2010), a gene that has been implicated in fatty acid metabolism and ketogenesis of liver and brain astrocytes (Cullingford et al., 2002a; Cullingford et al., 2002b). Hence, it would be interesting to investigate whether Hmgcs2 is also involved in ketogenesis within the developing SCO in future experiments.
It is interestingly to note that Fmo1, Adamts3 and Hmgcs2 all encode a protein that is involved in metabolism. SCO is a brain gland responsible for the synthesis of a range of glycoproteins (Rodriguez et al., 1998) that are important during NT development, e.g. for axonal migration (Hoyo-Becerra et al., 2010; Monnerie et al., 1995; Monnerie et al., 1996; Monnerie et al., 1997; Monnerie et al., 1998). Thus, these enzymes may participate in the production of SCO secretions, either through its known function or through an unknown or de novo mechanism. Alternatively, the possibility of these genes having a direct action on SCO development cannot be excluded. Nevertheless, the above propositions are purely speculative and further investigation is required to elucidate the function of these genes in prosomere 1 RP development.
5.4.3 Fibcd1– a possible endocytic activity in SCO development
Microarray, qRT-PCR and in situ hybridisation analysis have demonstrated that Fibcd1 was downregulated in Sox3 transgenic SCO. Fibcd1 encodes a transmembrane endocytic receptor that specifically binds acetylated oligosaccharides and is located 135
predominantly at the brush boarder of enterocytes. One of the ligands of Fibcd1 is chitin, a pathogenic biopolymer that provides the rigidity for fungus and insects and is capable of triggering mammalian immune response (Reese et al., 2007; Schlosser et al., 2009; Thomsen et al., 2009). The mammalian CNS is immuno-privileged and the SCO is the only CVO that possesses an intact BBB (Duvernoy and Risold, 2007). Thus, foreign molecules and immune cells should not have access to the SCO. A recent study indicated a functional BBB in mouse embryos was present from 15.5 dpc (Daneman et al., 2010). Moreover, Fibcd1 is only present in an immature SCO and but not at 18.5 dpc and postnatal SCO. As a result, the function of Fibcd1 within the developing SCO is unlikely to be associated with the chitin-induced immunity. The endocytic property of Fibcd1 encoding receptor suggests that it may interact or respond to endocytosis of external signals during SCO development. In fact, it has been well established that the adult SCO is innervated by serotonergic and other peptidergic fibres (Rodriguez et al., 1998). Perhaps, the developing SCO is also innervated similarly, with the transient expression of Fibcd1 to allow the transmission of incoming signals for the SCO development. Alternatively, it may serve as another yet unidentified molecule that aids in the endocytosis of the receptor for known signalling pathways, e.g. Bmp and Wnt signalling.
5.4.4 Disrupted dorsal diencephalic patterning due to defective RP development
Quantification techniques, i.e. microarray and/or qRT-PCR indicated Gata3, Lhx5 and Sox14 were upregulated by Sox3 overdosage. All of these genes share an overlapping expression pattern at the MZ of the Pt and the dorsal midline Pt. Although their functions in Pt development are not currently known, Gata3, Lhx5 and Sox14 appear to implicate in the promotion of neural differentiation and specification in the development of other CNS domains (El Wakil et al., 2006; Zhao et al., 1999). Hence, this suggests that Gata3, Lhx5 and Sox14 may exert similar function to allow the specification and differentiation of pretectal neurons. Current evidence that supports the function of Gata3, Lhx5 and Sox14 in neural differentiation is discussed below.
Gata3 belongs to the GATA transcription factor family, which its members play an important role in haematopoiesis. This is particularly true for Gata3, which is required for thymocyte development and T-cell differentiation during immune response (Clevers, 2006; Ho and Pai, 2007). Although Gata3 is widely expressed across the developing mouse 136
hindbrain and the dorsal midline Pt and the MZ of Pt (Nardelli et al., 1999), its function in these regions of the CNS is still unknown. However, another GATA family member, Gata2 appears to arrest cell cycle progression and stimulate neural progenitor differentiation through interaction with Notch signalling in the chicken spinal cord (El Wakil et al., 2006), while loss of Gata2 impedes neurogenesis in Gata2 null mice (Nardelli et al., 1999). Interestingly, the expression of Gata3 is dependent on Gata2 (Nardelli et al., 1999), suggesting a possibility that Gata3 may share a similar to Gata2 in CNS development.
Lhx5 encodes a LIM homeobox transcription factor that is expressed in the MZ of the Pt as well as the developing VTh and telencephalon (Abellán et al., 2010; Mastick and Andrews, 2001; Nakagawa and O'Leary, 2001). Lhx5 null mice had defects in the differentiation of hippocampal and Caja-Retzius cells in the forebrain, Perkinje cells in the Cer and GABAnergic interneurons in the spinal cord (Miquelajauregui et al., 2010; Pillai et al., 2007; Zhao et al., 2007; Zhao et al., 1999). Similar to the cellular function of Gata2 in chick spinal cord (El Wakil et al., 2006), Lhx5 is required to allow cell cycle exit followed by differentiation in the mouse Hpc (Zhao et al., 1999). Interestingly, Lhx5 counteracted Wnt signalling through upregulation of Wnt antagonists, secreted frizzled-related protein 1a (Sfrp1a) and secreted frizzled-related protein 5 (Sfrp5), to promote zebrafish forebrain development (Peng and Westerfield, 2006)
In vivo and in vitro experiments in the chick DTh and spinal cord have established that Sox14 expression is regulated by Shh signalling in a dose-dependent manner (Hargrave et al., 2000; Hashimoto-Torii et al., 2003). In agreement with this, subsequent in vitro analysis in human hepatocyte and glioblastoma-astrocytoma cell lines found that Shh signalling is able to induce Sox14 expression through upregulation of Foxa2 (Popovic et al., 2010). In fact, Sox14 labels and specifies a subset of postmitotic neurons with the DTh (Hargrave et al., 2000; Hashimoto-Torii et al., 2003). This role of Sox14 in neuronal specification is indeed supported by the fact Sox14 is capable of mutually repressing the transcriptional activity of gastrulation brain homeobox 2 (Gbx2), a transcriptional factor that specifies the subset of neurons residing next to Sox14 within the developing DTh when overexpressed in the chick NT (Hashimoto-Torii et al., 2003).
Sox3 transgenic embryos displayed abnormal Gata3, Lhx5 and Sox14 expressing domains within the MZ of the lateral Pt domains, i.e. bifurcation at the ventral border. This suggests the presence of a NT patterning defect within the Sox3 transgenic embryos. The RP 137
has been well established to be important for dorsal CNS patterning in both the spinal cord and diencephalon. Genetic ablation of RP inhibits the DV specification of spinal cord interneurons. In prosomere 1, the loss of RP identity in Msx1 null embryos led to ectopic Pax6 expression across the dorsal midline and hence, the expansion of Pax6 positive neural progenitors (Bach et al., 2003; Chizhikov and Millen, 2005; Lee et al., 2000a; Lee et al., 1998). Thus, this disrupted diencephalic RP patterning in Sox3 transgenic embryos is likely to be secondary to the loss of normal RP formation. However, given the presence of Sox3 transgene in the developing Pt at 12.5 dpc, one may argue the defective dorsal NT patterning might be a primary defect of the Pt development. Yet, both the presence of Pt MZ markers Sox14, Lhx5, Gata3 and a morphologically normal Pt at 15.5 dpc in Sox3 transgenic embryos do not support this proposition.
In addition to the Pt lateral domains, the Gata3, Lhx5 and Sox14 positive domain at the dorsal midline Pt has expanded towards the RP. There are two possibilities behind this observed defect. First, this may be a direct effect of Sox3 overexpression such that the cells at the dorsal aspect of the RP have been alternatively fated to take on an identity of their neighbouring tissue (dorsal midline Pt). Alternatively, similar to the lateral Pt domains, this defect may be secondary to the loss of RP function. The growth of dorsal midline Pt may be normally inhibited by RP secreted factors. In the absence of a normal RP in the Sox3 transgenic embryos, the dorsal midline Pt may expand the growth of its dorsal midline domain towards the RP. In order to confidently verify these two possibilities, a lineage analysis using the Cre-mediated reporter system may be used to investigate whether these expanded Gata3, Lhx5 and Sox14 positive cells are derived from the precursors of the dorsal midline Pt or the RP.
One may propose that Gata3, Lhx5 and Sox14 would not show an upregulated expression in the microarray and qRT-PCR experiment if the latter possibility proposed in the above paragraph is true, since these techniques analyse the transcriptional profile of the dissected SCO primordium. However, it is important to note that the microdissected tissues used in the microarray and qRT-PCR experiments included the thin layer of dorsal midline Pt cells that are immediately dorsal to the SCO primordium. Thus, an increase in the dorsal midline Pt cells could be translated as an elevated expression of Pt specific genes. This also highlights the limitations of microarray and qRT-PCR analyses which do not provide spatial 138
information relating to gene expression, and the importance of incorporating a qualitative technique for microarray validation.
Intriguingly, cells residing at the ventral aspect of the RP (cells next to the ventricular area within the RP) did not display an ectopic expression of either Gata3, Lhx5 or Sox14, indicating that they did not take on a cellular identity of the dorsal midline Pt or the MZ of the lateral Pt domains. These RP cells are laterally adjoined to cells residing in the VZ of the lateral Pt domains. It is thus possible that they have adopted the transcriptional profile of the cells in the VZ of the lateral Pt instead. Ascl1 and Neurog2 are two genes specifically expressed at the VZ of the lateral Pt at 12.5 dpc (Lein et al., 2007). In fact, ASCL1 forms a heterodimer with NEUROG2 to promote neurogenesis and neural differentiation both in vitro and in vivo (Henke et al., 2009; Kim et al., 2009; Pattyn et al., 2006; Puelles et al., 2006). Also, Ascl1 promotes differentiation of GABAergic neurons within the developing Pt (Puelles et al., 2006). An in situ hybridisation analysis of Ascl1 and Neurog2 (upregulated for 1.7-fold and 1-fold respectively from the microarray) in the RP can be performed to further elucidate whether these alternatively-fated RP cells have taken up a similar transcriptional profile or cellular fate to those VZ residing pretectal neural progenitors.
5.4.5 Failure of cranial NT fusion might be associated with disrupted dorsal NT tube patterning
The formation of a successfully enclosed cranium relies on the interaction between neurulation initiation and NT closure. Neurulation initiation is dependent on Shh signalling emanating from the notochord and prechordal plate (Bassuk and Kibar, 2009; Vieira et al., 2010). Upon neurulation initiation, proper NT closure requires the architectural change of the neuroepithelium through cellular convergent extension, neural furrowing and folding, followed by fusion initiation and complete fusion (Bassuk and Kibar, 2009; Colas and Schoenwolf, 2001; Copp and Greene, 2010; Copp et al., 2003; Pyrgaki et al., 2010; Vieira et al., 2010). Since there was normal NT closure and hence, neurulation observed within the spinal cord and telencephalic vesicles, neurulation initiation is likely to be occurring normally in Sox3 transgenic mice. 139
Dorsal NT fusion of the mesencephalon and rhombencephalon relies on the fusion initiation at closure 2 (see Figure 1.6 for sites of closure initiation), which upon successful initiation, migrates posteriorly towards the mes- and rhombencephalon an anteriorly into the diencephalon (Greene and Copp, 2009). The consistent unfused di-, mes- and rhombencephalic tissue strongly suggests the fusion initiation at closure 2 rather than the migration of the fusion from closure 2 has failed. In contrast, telencephalic fusion relies on the fusion initiation at closure 3, which resides at the rostral limit of the telencephalon and migrates posteriorly into the telencephalon (Greene and Copp, 2009). The failure of telencephalic fusion in some embryos indicates that either fusion initiation and/or migration from closure 3 were also affected. Certainly, given the limited data on the Sox3 transgenic exencephalic phenotype, it cannot be excluded that the lack of NT fusion in both the diencephalon and telencephalon was indeed caused by an prior defect in the architectural change within the neuroectoderm, i.e. cellular convergent extension, neural furrowing and folding, which are all critical for later neural fusion initiation.
Upon neural induction, the neural plate undergoes active proliferation to generate a thickened neuroectoderm. In order to generate neural progenitors in preparation for neurogenesis, this high level of proliferation is maintained in all cells even after NT closure. However, cells within the dorsal neural fold (or dorsal NT) initiate cell cycle withdrawal prior to NT fusion, such that it allows the specification of these dorsal NT cells into an early RP (Baek et al., 2006). Failure to downregulate cell cycle progression within the dorsal neural fold or NT has previously shown to result in exencephaly in a few mouse models (Baek et al., 2006; Harrington et al., 2009; Ishibashi et al., 1995; Kim et al., 2007; Sah et al., 1995). An example of these is the mice that are null for p53, a well known tumour suppressor gene required for cell cycle arrest, which displayed exencephaly with low penetrance (Sah et al., 1995).
Endogenous Sox3 is expressed along the entire neuraxis, with a lower level of expression at the dorsal neural fold during neurulation (Collignon et al., 1996; Wood and Episkopou, 1999). Thus, it appears that this pattern of lower Sox3 expression within the dorsal NT originates long before 12.5 dpc, the earliest timepoint presented in this thesis that shows a lower level of SOX3 in the SCO when compared to its laterally flanking tissues. Interestingly, transgenic Sox3 expression was observed in the dorsal neural folds at 8.5 dpc during neurulation (P. Thomas, personal communication), indicating the presence of Sox3 140
overdosage. Thus, perhaps similar to the RP development, a low level of Sox3 expression within the dorsal neural fold is permissive to NT closure during neurulation. Sox3 overdosage has been shown to increase cellular proliferation in the diencephalic RP and result in a loss of RP identity in previous chapters. Together, it leads to the speculation that upregulation of Sox3 may elevate cellular proliferation within the dorsal neural fold and NT, which results in a failure of dorsal NT fusion. 141
Chapter 6: Discussion
6.1 The knowledge of Sox3 function in mammlian CNS development was limited
SOX3 mutation has been implicated in XH with or without mental retardation in a number of families (Laumonnier et al., 2002; Solomon et al., 2004; Woods et al., 2005). Moreover, SOX3 duplication was associated with two identified XH families. Together, these human genetics studies strongly indicate the importance of Sox3 dosage regulation during CNS development (Solomon et al., 2002; Woods et al., 2005). This is further supported by Sox3 null mouse models which displayed hypothalamic-pituitary defects, similar to human XH patients (Rizzoti et al., 2004). However, despite Sox3 is widely and highly expressed during CNS development (Collignon et al., 1996; Solomon et al., 2004; Stevanovic et al., 1993; Wood and Episkopou, 1999), only mild defect was identified in other CNS regions of Sox3 null mice, possibly due to functional redundancy by other SoxB1 members. This has impeded further investigation of SOX3 function in mammalian CNS development. An overexpression study within the chick spinal cord suggested that Sox3 functions as neural differentiation inhibitor through its transcriptional activation activity (Bylund et al., 2003). Similar neural differentiation inhibition activity was observed in Xenopus overexpression assays, although Sox3 acts as a transcriptional repressor in that respect (Rogers et al., 2009). However, these Sox3 properties need not necessarily translate into the endogenous function of Sox3 during mammalian CNS development, since they were both from non-mammalian model organisms with exogenous Sox3 expression.
The transgenic Sox3 mouse model studied in this thesis utilises an endogenous Sox3 regulatory sequence in an attempt to upregulate Sox3 dosage within its endogenous spatial and temporal CNS domains. It provides a model for the XH patients associated with duplicated Sox3 chromosomal region (Solomon et al., 2004; Woods et al., 2005), who also have an elevated dosage of Sox3 under endogenous regulation. Moreover, it was established as an in vivo mammalian model organism for the understanding of the molecular and cellular function of Sox3 in the CNS. 142
6.2 Diencephalic RP development was sensitive to overdosage but not underdosage of Sox3
Elevation of Sox3 expression within the dorsal neural developing RP led to the loss of RP identity associated with increased cellular proliferation. In Sox3 transgenic mice, SOX3 proteins were observed distributing across the entire nuclei without aggregation, a figure commonly associated with protein toxicity. Moreover, there was no SCO cellular death detected, indicating the RP defect was unlikely due to toxicity induced by over accumulation of SOX3, but rather due to an increase in dosage effect (overdosage) in Sox3 transgenic mice.
In contrast, several lines of evidence indicate that loss of RP identity was not associated with Sox3 loss-of-function (underdosage). First, Sox3 null or heterozygous female mice do not display CH. Second, a morphologically normal SCO with SOX2 expression was detected in the adult SCO of these mice (N. Rogers, personal communication). Third, the loss of SOX3 function in XH patients with SOX3 mutation, namely polyalanine expansion, did not result in CH. These suggest that, at least within the diencephalon, Sox3 is dispensable for RP development, since it may be functional redundant with Sox2 such that the function of Sox3 is substituted by Sox2 in Sox3 null embryos. In fact, functional redundancy between SOXB1 members was readily reported in CNS and pharyngeal development in Sox3 null mice, lens induction and spinal cord neural differentiation in chick, as well as neural differentiation in Xenopus (Archer et al., 2010; Bylund et al., 2003; Kamachi et al., 1995; Kamachi et al., 1998; Rizzoti et al., 2004; Rizzoti and Lovell-Badge, 2007). Hence, it would be interesting to study mouse model(s) with multiple knock-out of SoxB1 to further elucidate these overlapping function(s) between SoxB1 members during RP development.
6.3 Sox3 dosage regulation is important for proper RP specification
Upon neurulation, dorsal NT and later RP cells withdraw from cell cycle progression to allow RP specification and further differentiation (Baek et al., 2006; Chizhikov and Millen, 2004b; Kim et al., 2007). This downregulated level of proliferation correlates with a lower 143
level of SOX3 within the dorsal NT and later RP observed in previous literatures (Collignon et al., 1996; Wood and Episkopou, 1999). Interestingly, in Sox3 transgenic mice, elevation of RP SOX3 to a level similar to its lateral Pt region has upregulated RP proliferation and is in agreement with previous literature where overexpression of Sox3 promotes neural fate uptake and expands the neural plate of an early Xenopus embryo through upregulation of 1.4-fold in cell cycle progression (Rogers et al., 2009). Furthermore, a reduced level of proliferation was detected at the hypothalamic region of the Sox3 heterozygous mice (Rizzoti et al., 2004). However, without further investigation, it is not known whether Sox3 elevated the diencephalic RP proliferation through direct upregulation of cell-cycle related genes that are essential for cellular proliferation or by maintenance of cellular population(s) that are in active cell cycle progression.
In addition to elevated cellular proliferation, Sox3 overdosage led to the loss of RP identity, suggesting that a low level of Sox3 has a permissive role for proper RP specification. Interestingly, cellular identity appears to couple with cellular proliferation during dorsal NT/RP development in chick. Lmx1a, the master gene that induces dorsal NT formation upon epidermal ectoderm Bmp signalling, withdraws the dorsal NT/RP cells from cellular proliferation upon expression. Loss of Lmx1a function within the mouse dorsal NT has led to an increased cellular proliferation followed by a loss of dorsal NT specification (Chizhikov and Millen, 2004b). Moreover, Lmx1a null mice also displayed the absence of dorsal NT or later RP development from 9.5 dpc (Millonig et al., 2000). Together, these indicate that normal RP specification requires the proper regulation of Sox3 dosage and is coupled to proliferation.
Differing levels of SOX3 expression diffrence were found across different regions of the developing CNS, e.g VZ residing cells within the LGE showed higher endogenous SOX3 signal intensity that those residing at the VZ of MGE. Thus, it would be interesting to investigate whether this SOX3 dosage dependent neural specification is specific to the RP or whether SOX3 employs a similar mode of action in other domains of the developing CNS. Moreover, SOX3 expression was switched off in other domains of the RP but not in the differentiated SCO during late embryogenesis, postnatal stages and into adulthood. One would presume that a differentiatied SCO has already been specified. Thus, its role at later stages is unlikely to be associated with RP specification. Since high level of Sox3 is associated with increased cellular proliferation in RP development earlier, would the low 144
level of SOX3 expression simply persist after SCO differentiation to maintain a progneitor property with a low level of cellular proliferation rather than to specify to the RP? A previous study demonstrated that adult rat SCO cells are active in the cell cycle (Chouaf-Lakhdar et al. 2003), suggesting the SCO as a site of neural stem cells residence (discussion on SCO as site of neural stem cells in 3.3.7). In contrast, a thymidine pulse labelling experiment of P21 mice failed to reveal any evidence of SCO proliferation (Rakic and Sidman, 1968). This may be explained by the very slow turnover of neural stem cells such that the pulse exposure of labelled thymidine might have missed the S phase of the cell cycle resulting in the lack of label incorporation.
6.4 Sox3 activity was repressed to upregulate Bmp and Wnt expression in RP development
The downregulation of the expression of multiple Bmp and Wnt ligands in Sox3 transgenic SCO indicate a possible disruption of these signalling pathways. It is not possible to speculate the hierarchy of the genes within the Bmp and Wnt signalling pathways simply based on their fold-changes from the microarray. However, Msx1 loss-of-function has led to a failure of Wnt1 expression (Bach et al., 2003), indicating the dependence of Wnt1 on Msx1. In addition, both the microarray and the qRT-PCR have showed a consistent downregulation of genes encoding metabolic enzymes, i.e. Fmo1, Adamts3 and Hmgcs2. Hence, these genes may be responsible for eliciting the ultimate cellular consequences for SCO development. Together, the arguments above place Fmo1, Adamts3 and Hmgcs2 downstream of Bmp and Wnt signalling pathways. However, it is yet to be elucidated that whether Bmp and Wnt pathways directly regulate the expression of Fmo1, Adamts3 and Hmgcs1, or whether the expression of these genes are induced as a result of RP specification by Bmp and Wnt signalling.
During neural development, Wnt1 establishes a dorsal to ventral mitogenic gradient with its highest density dorsally. Ectopic expression of Wnt1 in chicken NT has demonstrated that Wnt1 acts as a mitogen and upregulate proliferation within the chick dorsal NT (Chesnutt et al., 2004; Megason and McMahon, 2002). However, the loss of Wnt1 within the developing RP led to disrupted NT patterning with minimal reduction in cellular proliferation 145
in mouse NT (Muroyama et al., 2002; Zechner et al., 2007), suggesting Wnt1 may have a dual function in promoting cellular proliferation and patterning within the dorsal NT. On the other hand, Chesnutt et al. (2004) argued that Wnt1 only acts as a mitogen by demonstrating that the patterning ability of Wnt1 within the dorsal NT is coupled to its ability to promote proliferation (Chesnutt et al., 2004). In contrast to Chesnutt et al. (2004), the RP of Sox3 transgenic mice had a loss of Wnt1 expression that coincided with an increase in proliferation and a loss of RP identity. Hence, in consistent with the proposed dual function by Zechner et al. (2006) and Muroyama et al. (2002) (Muroyama et al., 2002; Zechner et al., 2007), the evidence presented in this thesis suggests that Wnt1 is at least associated with RP specification, if not the whole dorsal NT patterning, and its function is unlikely to be related to its previously proposed mitogenic effect. It is well documented that Wnt signalling can exert varying effect dependent on cellular context (Chen and Panchision, 2007; Freese et al.; Fuerer et al., 2008; Gordon and Nusse, 2006; Grigoryan et al., 2008). Thus, perhaps the observed discrepancy in Wnt1 mode of action is due its dynamic function in different cellular context.
6.5 Hes1 – a direct target of Sox3 in RP development?
Although it is well established that Sox3 encodes a transcription factor, its target genes are poorly defined. It is yet to be elucidated whether any of the differentially regulated genes identified from the microarray are Sox3 direct targets. However, preliminary promoter sequence analysis found no evidence of Sox3 binding site on both Wnt1 and Msx1. The question of what are the direct target(s) of Sox3, thus, still stands.
A role in dosage dependent patterning of the dorsal NT or RP has been recently reported for the Notch signalling effector gene hairy and enhancer of split 1 (Drosophila) (Hes1), a basic helix loop helix transcription factor. In contrast to Sox3, Hes1 expression is high in the RP but low in its lateral flanking region (Baek et al., 2006). The developing RP of Hes1 null embryos had an increase in cellular proliferation. Although affected embryos still displayed some RP Wnt1 expression indicative of proper dorsal NT and RP induction, the level of Wnt1 expression was extremely low. In contrast, marker expression of laterally flanking cells mostly appeared normal (Baek et al., 2006). Similar to Sox3 transgenic mice, Hes1 null embryos failed to perform NT closure from 8.5 dpc (Ishibashi et al., 1995). Hes1 is 146
a transcription repressor which was previously shown to represses neural differentiation within the developing telencephalon (Ishibashi et al., 1994; Nakamura et al., 2000). In fact, during telencephalic neurogenesis, Hes1 expression is downregulated, presumably to release the repression on neuronal markers. Since, the RP is an organising centre that is mostly made up of neural ependymal cells and is devoid of neurons. The function of Hes1 within the developing RP thus, appears to inhibit neural differentiation, in particular neurogenesis and to allow an uptake of a non-neuronal fate of the RP cells (Baek et al., 2006; Ohtsuka et al., 2001).
The similarities of Hes1 null mice phenotype to those observed in our Sox3 transgenic mouse model prompts the speculation that perhaps Sox3 may be regulating Hes1 in the developing RP. Previous literature has readily demonstrated Hes1 to be a direct target of another SoxB1 member, Sox1, which promotes neurogenesis. Sox1 directly binds to Hes1 regulatory region and represses its expression in vitro (Kan et al., 2004). Moreover, according to the microarray presented in this thesis, Hes1 was downregulated (1.9-fold) in the presence of Sox3 overdosage (see Appendix VI for microarray gene lists). To gain further support, recently study using chick NT found that neurogenesis inhibition by Notch signalling was dependent on Sox3 (Holmberg et al., 2008). Perhaps, Sox3 acts similarly to its other SoxB1 member and directly represses the transcription activity of Hes1, which Hes1 then acts as a transcriptional repressor for genes further downstream, i.e. genes involved in Wnt and Bmp signalling (Figure 6.1 B).
The speculation made above proposes that Sox3 may work as a transcriptional repressor, which contradicts to the observation as a transcriptional activator previously in chick NT (Bylund et al., 2003). However, the conclusion from Bylund et al. (2003) was derived from an overexpression of an artificial fusion protein with a Sox3 HMG domain upstream of the herpes simplex virus transactivation domain, VP16 (Bylund et al., 2003). Moreover, SoxB1 transcription factor acts in a cellular context dependent manner through interaction with different binding partners (De Santa Barbara et al., 1998; Remenyi et al., 2003; Rodda et al., 2005; Yuen et al., 2002). Perhaps Sox3 has different mode of action, i.e. a transcriptional repressor or activator, depending on the cellular context and the availability of partnering proteins. In fact, the transcriptional repressing activity of Sox3 has been observed in organiser establishment during early zebrafish embryogenesis (Shih et al., 2010), Xenopus body axis formation (Zhang et al., 2003) and neurogenesis (Rogers et al., 2009). 147
6.6 The role of Sox3 in RP neural differentiation
Inhibition of neural differentiation is important to allow proper neural ependymal differentiation of the RP cells, since ectopic neuron generation was demonstrated to disrupt RP differentiation (Baek et al., 2006). Sox3 has been thought to be a neurogenesis inhibitor in chick and Xenopus NT (Bylund et al., 2003; Holmberg et al., 2008; Rogers et al., 2009). Interestingly, upregulation of Sox3 causes a loss of RP precursor identity that ultimately led to the failure of ependymal differentiation in the Sox3 transgenic mice. It is not known whether those Sox3 overexpressing, alternatively-fated RP cells would take up a neuronal identity. However, one would expect Sox3 to have other function in addition to neurogenesis inhibition if its overexpression leads to a change of cell fate in RP neural progenitors, which initiate differentiation prior to neurogenesis. In fact, neural differentiation study in zebrafish has found that although Sox3 inhibits neurogenesis, its function is required for neurogenesis initiation (Dee et al., 2008). Altogether, these suggest that perhaps Sox3 is important to maintain the pool of progenitor cells that is destined to become neurons, such that upon a high level of Sox3 expression, neural progenitors commit to a neuronal fate although remaining undifferentiated in order to undergo cell cycle progression for progenitor pool expansion.
In summary of the above, a preliminary model is proposed in an attempt to explain the relationship between the Sox3 dosage-dependent RP patterning and its downstream target genes, as well as the role of Sox3 in neural differentiation (Figure 6.1). In brief, Sox3 is ubiquitously expressed at a high level throughout the neuroectoderm. Upon neurulation, Bmp4 and Bmp7 signalling instruct the specification of the dorsal NT. The induction of dorsal NT identity triggers the downregulation of Sox3 and the withdrawal of cell cycle progression (either through Sox3 repression or by a Sox3 independent pathway). The downregulation of Sox3 then allows these dorsal NT cells to remain as neural progenitors such that they are able to maintain and be responsive to signalling pathways that instruct further RP differentiation, i.e. Wnt and Bmp signalling. The complex interactive network of Bmp and Wnt signalling downstream of Sox3 then further regulates other genes for the ultimate effector functions during RP development (Figure 6.1). Interestingly, Lmx1a is a master mediator that is expressed within the dorsal NT and early RP upon dorsal NT induction. Once expressed, Lmx1a induces the expression of other dorsal NT markers for further RP development. Unlike Sox3 transgenic mice where the development of the early RP 148
appeared normal, the loss of Lmx1a function led to the failure of early RP formation in mouse and chicken (Chizhikov and Millen, 2004b; Millonig et al., 2000). This has placed Lmx1a upstream of Sox3 such that Sox3 may be directly repressed by Lmx1a or by other molecules that are upstream of Lmx1a (Figure 6.1).
In contrast to the RP, a high level of Sox3 persists from the neuroectoderm stage to direct the cells in the non-RP domains to become neuronal-fated neural progenitors, which are under active cellular proliferation (Figure 6.1). As a result, in the presence of Sox3 transgene, the high level of Sox3 has alternatively fated the RP neural progenitors into a neuronal fate and hence, they behave like neuronal progenitors, which are no longer responsive to RP differentiation instruction, but instead they later differentiate as neurons. Thus, analysis that verifies the neuronal identity of these alternatively-fated RP cells in Sox3 transgenic mice at later embryonic stages may be carried out to gain further support for the above proposed model.
6.7 Sox3 in earlier dorsal NT development
The correct induction and maintenance of the dorsal NT is an important first step for NT fusion, followed by NC migration and further RP development. Since transgenic Sox3 was found in the dorsal NT from 8.5 dpc (P. Thomas, personal communication), the above described process could be perturbed by elevation of Sox3 dosage. A brief discussion on each of these processes is presented below.
6.7.1 Dorsal NT induction
The presence of a correctly induced RP in Sox3 transgenic mice, suggests that dorsal NT induction was independent of Sox3 dosage. In fact, it appears that RP development was dependent on Sox3 dosage only when a definitive RP is formed. Upon induction from epidermal ectoderm, the early dorsal NT is comprised with both NC and RP cells. Initially these dorsal NT cells all express RP markers of Bmp4, Lmx1a, Msx1 and Wnt1. Later, migrating NC cells switch off RP marker expression and gradually move out of the dorsal NT, leaving the dorsal NT with cells making up the definitive RP (Chizhikov and Millen, 2004b; Bmp4, Bmp7 ? (epidermal ectoderm)
? Lmx1a Sox3
Bmps and Wnts (Bmp5, Bmp6 Hes1 Wnt1, Wnt2b, Rspo3) ? (3) Msx1, Msx2? Rfx4_v3, Rfx3 proliferation (1) (4) (2) non- Sox3 Fmo1, Adamts3 neuronal (3) Hmgcs2 fate Hes1? proliferation (4) neuronal fate
Figure 6.1: The proposed model of Sox3 in diencephalic RP specification. In normal diencephalic RP (yellow) development, Sox3 expression is repressed either by Lmx1a or other molecules upstream of Lmx1a upon dorsal NT induction by Bmp4 or Bmp7 from neighbouring epidermal ectoderm. The repression of Sox3 allows the expression of molecules implicated in Bmp and Wnt signalling pathways through the regulation of other mediators. Here, Hes1 (red) is proposed to be one of these mediators and that it is repressed by Sox3. Hes 1 is a Notch signalling effector gene that is highly expressed within the RP. Hes1 expression is required for the down- stream expression of Bmp and Wnt signalling molecules. Interaction between the Bmp and Wnt signalling pathways with other transcription factors such as Msx1, Msx2, Rfx3 and Rfx4, may allow the expression of effector genes (1), i.e. Fmo1, Adamts3, and Hmgcs2, that are responsible for the ultimate cellular consequences. It is important to note that Fmo1, Adamts3 and Hmgcs2 may not be directly regulated by Bmp and Wnt pathways, but may be indcued by other molecules that are activated as a result of the RP development instructed by Bmp and Wnt singalling (2). In addition, RP cells withdraw cellular proliferation upon downregulation of Sox3. Again, this withdrawal from cell cycle progression may be a direct repression of cell-cycle related genes due to Sox3 repression (3), or it may be a result of cell-fate specification that is brought about by the downregulation of Sox3 (4). In non-RP lateral tissue (white), Sox3 is highly expressed, which leads to a high level of cellular proliferation and the repression of Hes1 expression. This ultimately results the neural progenitors to take up a neuronal fate. Coronal view. Top to bottom: dorsal to ventral. 149
Copp, 2005; Copp et al., 2003). In mouse embryos, the definitive RP marker, Gdf7, is not expressed until 12.5 dpc (Louvi and Wassef, 2000), suggesting further RP differentiation initiated from 12.5 dpc. This corresponds to the last appearance of a morphologically normal diencephalic RP or SCO in Sox3 transgenic mice, correlating with the failure of the definitive RP to further differentiate into ependymal cell fate.
6.7.2 NT closure
NT fusion failure was observed occasionally in Sox3 transgenic mice. It is not known whether this defect is underpinned by a similar molecular mechanism as the later RP development. However, it is worth noting that NT fusion is more resistant to Sox3 dosage changes than SCO development as all Sr/+; Nr/+ embryos observed displayed a RP defect, while only 16% of them were affected by exencephaly. NT closure requires the complex interaction between the neuroectoderm and the epidermal ectoderm residing at the very dorsal tip of the neural folds (Copp and Greene, 2010; Copp et al., 2003; Greene and Copp, 2009; Pyrgaki et al., 2010). Hence further cellular and molecular analysis of the dorsal neural folds in early Sox3 transgenic embryos during NT fusion may allow a more thorough understanding this defect.
6.7.3 Cranial NC development
Cranial NC cells migration is one of the very important steps in craniofacial morphogenesis. Cranial NC cells begin migrate out of the dorsal NT before NT fusion and persist to as late as 16 somite-stage (~9.0 dpc). Diencephalic NC cells mainly give rise to cranial mesenchyme that migrates towards the region between the diencephalon and the eye. (Santagati and Rijli, 2003; Serbedzija et al., 1992). As a result, craniofacial NC formation or migration may be affected in Sox3 transgenic mice. Although there was no detrimental craniofacial defect identified. However, a flatter upper and mid-face with significantly shorter nose were observed in some single hemizygous and double hemizygous mice. The diencephalic NC cells gives rise to the premaxillary bone, the frontal bone and the nasal bone, which make up the upper facial structure (Santagati and Rijli, 2003). Thus, the flatter face could be due to abnormally developed bones as a consequence of defective diencephalic NC 150
cells formation or migration. Alternatively, the possibility that this craniofacial difference was merely a secondary effect of CH cannot be excluded.
6.8 Similarities and differences in overdosage of Sox3 between mice and human
SOX3 duplication was found in two XH families with patients displaying short stature (dwarfism) due to growth hormone deficiency (Solomon et al., 2002; Woods et al., 2005). However, Sox3 overdosage in mouse did not result in growth hormone deficiency, which is always observed in patients with XH. Some Sox3 transgenic mice although appeared to have growth retardation, however this was likely to be secondary to CH as they only appeared to be smaller to their WT counterparts after P7 and/or after a clear CH-induced dome-shaped cranium was identified. Preliminary analysis found that the Sox3 transgene is present during hypothalamic development (P.-S. Cheah, unpublished data). This indicates that hypothalamic-pituitary development perhaps is less sensitive to Sox3 dosage. In addition, the penetrance of the dwarfism phenotype may be dependent on other modifier genes, since it is well known that the difference in mouse genetic background greatly affected the phenotypes displayed (Laronda and Jameson, 2011; Raverot et al., 2005; Specht and Schoepfer, 2001; Threadgill et al., 1997).
Here, Sox3 overdosage mouse model displayed abnormal RP development, which ultimately leads to CH. In humans, an underlying genetic cause was estimated to account for up to 40% of hydrocephalus detected, and 15% of these cases with genetic causes were identified to be X-linked (Zhang et al., 2006). However, the heterogeneity and multietiology properties of CH have deemed gene linkage analysis difficult. Currently, only one gene, L1CAM, has been linked to CH in humans (Pérez-Fígares et al., 2001; Zhang et al., 2006). The Sox3 transgenic model has demonstrated the overdosage of Sox3 as a cause of CH and hence, is a candidate gene for X-linked CH cases in humans. Interestingly, a couple of genetic markers residing at chromosomal location Xq27.3, which is near the SOX3 locus, were identified to be associated with hydrocephalus in an affected family (Strain et al., 1994). In contrast to the Sox3 transgenic mouse model, the function of Sox3 is dispensable for SCO development and does not cause CH, as demonstrated by the Sox3 null mice. This is possibly 151
due the functional redundancy between SoxB1 members. However, it remains possible that SCO development in humans may rely on Sox3 function. In fact, hydrocephalus was identified in one patient from a family with a deletion of Xq26.3-27.3, which includes Sox3 (Wolff et al., 1997).
Moreover, in Sox3 transgenic mice, defects were also observed in RP associated or dependent processes, i.e. NT closure, midline crossing of commissural axons and dorsal midline patterning, and the development of other RP derivates, i.e. the PG. Thus, SOX3 overdosage may also account for one or all of these defects in human. Interestingly, spina bifida, which results from a failure of NT closure in the spinal cord, was linked to a SOX3- containing X-chromosomal duplication (Hol et al., 2000). In mouse, Sox3 is endogenously expressed along the neuraxis of the developing mouse, but is downregulated along the dorsal NT. Thus, it is logical to speculate that perhaps the overdosage of SOX3 within the dorsal spinal cord in humans leads to the failure of NT fusion. In addition, the lack of midline crossing by the corpus callosum of the Sox3 null mice suggest that telencephalic RP development and/or functions may rely on Sox3 function, such that the RP function perhaps is compromised within the telencephalon in these mice (Rizzoti et al., 2004). Interestingly, corpus callosum dysgenesis was also observed in XH patient associated with SOX3 duplications (Woods et al., 2005).
6.9 Normal SCO development and CSF homeostasis in human
Although the SCO has been demonstrated to be required for CSF homeostasis maintenance and its defective function has been established as a causative role in mouse CH (Huh et al., 2009; Pérez-Fígares et al., 2001; Vio et al., 2000), the link between SCO dysfunction and CH in humans is debatable (Castañeyra-Perdomo et al., 2004; Rodriguez et al., 2001). Previous histological study has identified SCO dysplasia in hydrocephalic human foetuses (Castaneyra-Perdomo et al., 1994). Here, the Sox3 transgenic mice display a severe lack of embryonic SCO development and function with a prenatal onset of hydrocephalus. This further adds to the expanding list of mouse models with SCO defect as the primary cause of CH. Moreover, it supports previous mouse models that the SCO regulation of CSF 152 homeostasis originates from embryogenesis (Bach et al., 2003; Blackshear et al., 2003; Vio et al., 2008). Thus, a similar role in CSF regulation is possible for the embryonic human SCO such that a defective embryonic SCO development may account for CH in human.
6.10 Future directions
This thesis demonstrates that a Sox3 repression is critical for the maintenance of RP identity and the development of a definitive RP. A number of questions have been raised from these findings which need to be addressed by future experiments. First, it is speculated that functional redundancy is responsible for the lack of SCO dysgenesis and CH in Sox3 null mice. Thus, double or triple SoxB1 mutants may help to unravel this complex level of functional redundancy and to elucidate the exact role of Sox3 in RP specification. To circumvent the early embryonic lethally displayed by Sox2 null mouse (Avilion et al., 2003), a Wnt induced Cre recombinase system can be used (Danielian et al., 1998), since Wnt1 is highly expressed by the dorsal NT during early neurulation and is later restricted to the RP (Hollyday et al., 1995; Megason and McMahon, 2002; Shimamura et al., 1994). Comparison of the single and multiple conditional knockdowns of Sox1, Sox2 and Sox3 within the RP may allow the elucidation of any redundant functions or genetic interaction between these SoxB1 members.
Second, Hes1 is speculated to be the mediator downstream of Sox3 during diencephalic RP development. In fact, the microarray experiment in this thesis demonstrated that Hes1 was downregulated for 1.9-fold in the diencephalic RP in the presence of Sox3 overexpression. Hence, Hes1 expression analysis by in situ hybridisation and qRT-PCR can be performed in Sox3 transgenic mice to validate the microarray experiment. Moreover, Hes1 is a direct target of Sox1 as demonstrated in previous literatures (Kan et al., 2004). As a result, bioinformatics analysis of the Hes1 regulatory sequences followed by chromatin immuneprecipitation experiment with Sox3 antibody can be performed to elucidate if Hes1 is directly regulated by Sox3 during RP development. In order to unravel more potential direct targets of Sox3, identical techniques can also be used to investigate whether any of the validated candidate genes identified by the microarray experiment is a direct target of Sox3. 153
Third, it is demonstrated in this thesis that Sox3 overdosage directs the RP progenitors into an alternative fate. It is speculated that these RP cells have taken up a cell fate that is similar to cells in neighbouring tissues, i.e. the Pt. To further explore this possibility, expression analysis of pretectal neuronal markers Ascl1, Neurog2, Sox14, Lhx5 and Gata3 at the RP region can be performed in Sox3 transgenic embryos after 15.5 dpc, when the loss of RP cells is morphologically apparent, by in situ hybridisation.
At last, NT fusion failure is apparent in some Sox3 transgenic embryos. To investigate whether this defect is due to an increased proliferation and identity compromisation of the dorsal neural folds, similar to that of the RP during later embryogenesis, cell-cycle progression study by BrdU analysis and dorsal NT markers analysis by in situ hybridisation can be carried out in Sox3 transgenic embryos at 8.5 dpc, a timepoint when neurulation initiates. Moreover, some Sox3 transgenic mice were observed with a flatter face. In order to explore whether this is a direct consequence of craniofacial defect due to abnormal NC cells development and not a secondary effect of CH, skeletal preparations can be performed on postnatal Sox3 transgenic mice. If craniofacial defect were identified, expression analysis of NC cells markers, e.g. Sox9, Foxd3 and Snail2, in the premigratory NC cells at the diencephalic dorsal neural fold at 8.5 dpc, a timepoint when diencephalic NC migration initiates, may help to gain some molecular insight on how Sox3 overdosage affects NC cells migration.
6.11 Concluding remarks
This thesis has analysed the Sox3 transgenic mouse model and outlined a requirement of Sox3 dosage regulation in RP development and its relationship with CH and a range of RP associated defects. Moreover, it has demonstrated that Sox3 overdosage leads to the loss of RP cellular identity, possibly through the loss of expression in genes implicated in Wnt and Bmp signalling, and is coupled to an increase in cellular proliferation. Most importantly, it suggests that a low level of Sox3 is permissive for the proper specification of the non-neural RP cells. Currently, there is no association of SOX3 in any CH or RP development in humans. RP associated defects often lead to severe NT dysmorphology and are often embryonic lethal (Harris and Juriloff, 2007; Harris and Juriloff, 2010). Together with the multietiology and heterogeneity of CH, it is not surprising that a genetic linkage of Sox3 with these conditions 154 is underrepresented in humans. Hence, the findings in this thesis have put SOX3 forward as a potential candidate gene responsible for human CH and NT defects, which may aid the process in genetic screening in affected patients. 155
Appendix I: The sequences of Btbd11 last exon
Features:
3’ UTR
Deleted sequence
Note: the sequence of the sense strand is shown in a 5’ to 3’ direction
1741 TGCAAAAAGCATCAATACTGACAACTGTGTGGATATCTACAGCCACGCCAAGTTCCTTGG 1801 GGTGACAGAGCTCTCGGCGTACTGTGAAGGCTACTTTCTTAAAAATATGATGGTCCTTAT 1861 TGAAAACGAAGCGTTCAAGCAGCTCCTGTACGACAAAAATGGCGAAGGGGCGGGCCAGGA 1921 TGTGCTTCAGGACCTACAGAGGACGTTGGCCATTAGGATTCAGTCTATCCACCTGTCGTC 1981 TTCCAAGGGTTCCGTGGTATGAGACGCCCCACGAGGCAATGGTCCCCAGGAGCAAGGTTT 2041 CTGGCTCCCTTCTGCATTGGCACCACAAAAACACAAAAATTGTTCCTGGTGTCTGTCTGT 2101 CTGTACTGGGCCAAGGAGCTGCCTCCACTGCCTCAGCTACTCTCCGGTCAGTGAACTTGG 2161 CTTCCGTGTGCTCCTGAGCGAAGAACCGTCCACGCGTCTGCAGGTTAGATTGTAGCAGAG 2221 GCCAGAACCTGCAGGCCAACTCACAGCAACCAAGCCCCCTCCAGTGCCCCGGAGCCATGG 2281 GCACCTGCTGCCAAGGACCATGATCTAGCAGACGGACCAAGGCTTAGAGATCTCTGCTTC 2341 CTGGTTGACTGATGAAAACCTCAGTAGTATGACCTTGAAAAGCATCACATCACCCACGAA 2401 TGACCTTGGAGGCTGTACTTGCCCAGGCATCAGCCCTGGTGGCTAGTTGTTGATGCCATT 2461 GCCCGGCATGAGGAAGTCACATGGCCAATTTTTAGGGGTGGGAACTTTTTAAAAAAAATG 2521 TTCATGTCTAGCAAACACAGAGCAGTCTCGTAGTTTCGAATATATTTCTAAAAGCTTAAC 2581 AGTTCATTTTCGACTGAATTGTGTTTAGGGAATCTGTGTTTGGGGTTCTTTTCATTTTTT 2641 TTTAAGATAGTGAGTTGCAGGTCCGTGCCGTGTAAACTTGCATCTGACAAATTCCTGCTG 2701 ATACTATACAAGGTCACATATTCCTGCCTTAGTGATTCGATCATGGTTCCACAAAAGAGG 2761 TGCACTGCCAATTATCTCTAATTATCTCGAATCTAATAGTTGACAATAACCTGCATGCTG 2821 ATTGTGGTTTGGGGGTGAGACTAGGGCACCCCCACATAACAAAACATGCATTTGCATGGA 2881 CCAAACCTGTACAACATCATGTTTGTGCTTTGTTTTCTTTTGCCAAAGTGAGATATGTTC 2941 TCTCGCTTCTCGCTAAAAGCCCACTTGGTGAAACAAACATATCTGTTTCTTTTCCTGATC 3001 ACGAATGGCTCTGGGGATACTTTTGCAGAAACGTGTTCATTTCTGATAACTGAGTGGGGG 3061 TGGGGGAGCACGTCAGAGAATCTACTATTAATCTCAGAGCTTTATCTGTATTCTCAGAAG 3121 ATGTTCATAGTTTGTTACTAGCCAGAGAGTATGACTATGTTAGATTATTTTTAAAGATGA 3181 AAAGTGTAAGATAGGCTTGACACTCTCTGGATTCTGAGAGTGTGCAGTCTGCAATTATCT 3241 ATGATGAAAGTTTATATTGACAGGGTTGTGTTTACTCTGCCATATAATATATATACATAG 3301 ATGTATATATATATGCAAAAGAGATTGGTAAAGTGCCACAATACTCCAAATAATTCTTGA 3361 GTTGCGGCCCTTTTTTTCTTTCCTTGTTCTTTATTTTTAATCCTAGATACATTAAGGTAA 3421 AAGCTAGGAGAATGGCTTTTGTCTTTTGGGATGGCCAAGTTTTGTTGATAACCCCATTTC 3481 CTAGCATGATTTTTGGGAAGTTACGCCAAACCCTAGAGTCTACAGGAACTTGTATCCTTC 3541 GCCTGTTGTCCTACTCCCTCAACACCTGTCAAGGACCCACGTCAAGTCTCGTTTCATGGG 3601 TGTGGTGTGCAAGGTACCCAGTCTGAGTCTATCTGCAGCATCAACTCCTTTGTGATGCAG 3661 GCCCCCGAGGCCTCTCTCTTCCATGTAGTCATCCATGAGCCTACCCAGCCCTACCATCAA 3721 ATGATTATGGGAGCAACTGCTTGAGGTGACAACCAAGTCACAGACACAAGAGTGACACAC 3781 TGTCGCTTGCCTTCAGGGATGCTGCTCTCTAACCCAGCTGGTTCTCTGTGGATGTCGAAC 3841 ACTCTCTGGACCTCCAGGTGTTGATCTGTACATTGGTAACTCCCACCTAGACCTCTGTTC 3901 AGTTTGGTGTGCTTTCTCTGCCATTAAAAGAAATTATTTCCCACA
156
Appendix II: The sequences of the Sox3 transgene integration site in Nr mouse model
Features:
Sox3 ORF
Sox3 3’UTR
IRES
EGFP
SV40 polyAdelylaton signals
FRT site
Chromosome 10 sequence: 85,122,881-85,122,902bp (NCBIm37)
Chromosome 10 sequence: 85,122,667-85,122,751bp (NCBIm37)
Chromosome 10 deleted sequence: 85,122,868-85,122,880bp (NCBIm37)
Note: the sense strand of the Sox3 transgene is shown in a 5’ to 3’ direction
85,122,881bp 85,122,902bp 1 AGAGAGCAGCATCCCTGAAGGCTTTCCTCTAAACCATAGTCCAGCCAGGATGCAAAAGCC 61 TTGGAATAAGGGTCCGCACAAAGAAGTGCTCCTGGCACAAGTATTTCTCTATATTACAAT 121 CTTTACATACATAAAATAGGGCCAAGAGCATACCTACTCACAAGGCTGTTTAGAAGAGAA 181 AATAAGATCGTGTTGCTGAAGCACTTCAGAAAGAAGGCTATTATCCTGATTTGTTCCATT 241 CAAAGGGAAGAAGAGCTGAGTGTCCTAGAATGTAGTATAGTTGTTCTATTGACCTACTCA 301 AGTCCCTGGCCCCCAATGCTTATGCATCACAGAAACAAGGCAGTTTGGTCAGCTCAAACC 361 TCAGAAAATAAAAACAGAAAGAACAGGGTTGGAAATCATCAGTCTGTAAAATCTCCAGCC 421 ACGTTTTTGCTTTTGTACCATCCTCAACCCCCTCCAACACTCTATAGCTACTCTAAATTG 481 GCTTTTCACTTGGATTCCCCTACTTATCTAATCTGTGCCCCAAATCCCCATGATCCTATT 541 GACCACCGCATATTTGTATCAATTTTGTAGGCTTCACCATGTATTTTTCGCTTGTGCTAA 601 GTGCTTATATACATTAAAGCATTATCCTTATAACAGCCCTGTTAAGTAGCTATTATTGCA 661 TCTAATATTTTACGGATGAGGAAACCAATTCAAAGAAGTAACTTGCCCAAGGTCTCACAG 721 CTGTGGCAGAGCTGGGATTCCAAAGCCTGTGCTGGGAAGAGCAAACATATCCGGCTGAAA 781 ATGAGAGTGGAGGTTGAGATTTACAGATAGTAAATTTGCAGGGAGATGGAGGTTAGGCCA 841 GGATGGTACCCTCTTGTTAGGTGAGGTTGGTCAGAAATGCTATCTGGCTGGCATTTCCTT 901 AATATGATCTGCAGAGGTAACAAAAGGCAGGCTGCATGAGCAAGAAGGGGAAAGTGGATA 961 CATAAATTATTTCATGTTTCCACACATATGGCTAAACTTTTCTAGGCCAAGGCAATAAAT 1021 TTGTAATATTATCATCACCTCCTGCTGGATGGCAGTCATGTTTTGTTATCCAGAGCCTGA 1081 GCATGTATAACTGCATGCGCAGTCTAATGTACTGGTGACACAGGAGAAGAGAGTCAATGT 157
1141 GCATGAGGCCCTGAGAGCCATGTCACTACTGTCTGTGGTTGGTCAAGGTTACTATTTTGC 1201 CTTTTGGACATACTATCAATTATCCTGCCAACAGAGATACTGGACCTTTAGTTTTCTTAT 1261 AATATAAGAAACTGACTTTAATTAAACCATCAGTAACTTGAACCTAATTATTTCAGACAG 1321 TATTTAATTAATGTTCTATGGATACCTAATATAATTAAGGTTCCTTTAATAAGACCTTCT 1381 GTGACATTAAAATATATGACCTTTATCTTCCTTATCCACCACATTCTGACAACTCAGATT 1441 TGTCATTTTCACTAAAAGGTATAGTTTGATTAACTAGAACAGGCTATAAAATTAATCTGG 1501 CAAATTACACTTGTTTGTTTCTTTTTTTTCTTCCCCCTCAAGGCCTGAGAAGTTAAGTCT 1561 TCTAATTAGCAGTTCAGATCCAAGTGTTCATACCACCGTTTTACTTTTAATTTGCCAAAG 1621 TGTTCCTTTTGTACATGAATTTTTGTGACCAAAGCAGATAACAACTATCCTTTATTAATC 1681 ATGGTGAAAGGTTCAATCTTTAAGATATGAAAAAGAATGAGCTGTTCGTTACTTTAACAA 1741 TGACAGTGATTGTCTTAAGATGTCCTGCCTTTAATAAGGGGCTTCAGTTTCTAGTATTAA 1801 CAAGGCAATCAAAATATTAAATCCCTATAGGAACTGGCAGCAAACAGAAGGCTCCTCCAA 1861 ACGTGACTCACGACAGGCAGGGTTTGTGTATTTCTCGCCTCCAGTTCAGAAGCTCCCAGT 1921 GACCCAGTTTCTAATTGTCTTCACAACCTTCCACCCTCACCCTGCAAATACACAAGGCCT 1981 GATTCAAGGAAAAGGGAAGCAGCTATCAGTCTCTCGCCTCCTTTGCAATCTGCTGAGAAA 2041 CTATCTTGCTTTGGGGTTAGGGGCTTTACCTTGGAAGCGGAGAGCTTGCCTAGTATATCC 2101 AGAGCCCTGGATTCAGTAAGTATGGAGAAAGGTATGCTAAAGAGAAATACAATTTGAGTG 2161 GGAATCCTTAAGGTCCCTAACTCACATTCAAGACCACCAAAGACCAGTCGTATTTTAGTA 2221 TTTTAGCATCCAGGGCATCAGAGTACTGCTTCCTCTCAGGGTCAGGGGGAGATCTGTTAG 2281 ATCTCTCTCTCTCTCTCTCTCTCTCTCTCTCTCTCTCTCTCTCTCTCTGTGTGTGTGTGT 2341 GTGTGTGTGTGTGTGAGAGAGAGAGAGAGAGAGAGAGAGAGAGAGAGAGGGAGAGGGAGA 2401 GGGAGAGGGAGAGGGAGAGGGAGAGGGAGAGGGAGAGGGAGAGGGAGAGGGAGAGGGAGA 2461 GGGAGAGAGAGAATTGAGAAGGGTAGAAGTAGTAGAGATGGAAAGAAATGAGAGGGTGAG 2521 GGGTGCGTGGACACTATGAGGTTTGCAGGGAGCCTCTCCTGACTGTTGCTAAGGGAGAGC 2581 TCCCAATATAAAGGACAGTGTCTGCCACTGTTGTGGTGCTCAACAGATCCCTCTGCAGTA 2641 AAGGGAGTCTTCTAACTCCACTGCAATCCACATGGTAACCAGAGCGTAGTGGACACTTGA 2701 CACAAGCCTAGAGAAACTGCTTTCGATTTTAATTAACTTGAAGTAATGTTAACAAATTTA 2761 AAGCCTCACATGTGACGAATGGTTACCTTACTGAACAAAGGCATCGTTTAACAGCTCATC 2821 TCCTTCACTTTACACAAGGAGAAAATGTGCAGAATAGACAGCCAGTTTGGTAGATGACTC 2881 CCCTTGCTGAGGAAAATCATGTCTGGGGACTAGACTGCAAATCCTAAATTGTAGCTCTTT 2941 GGACACCTTTCCCAGTTCATACTACTTCCCTTCCCAGCAGCTCTGAAGTAATTCTCTAAG 3001 GGATAATCTTCACATATACTGACACTGTCATTTCTCACTGGTCCAATGACTGATAAAACT 3061 ACTGTGCAATGTGAGGCGGAAACAGAAAAGTCAAGAAATACGAACTCCACATGCCACCTC 3121 CATAGGGACAAACACACACGCACACATGTACACACCCACTCCCTACACACACATACACAC 3181 ACACACGCACACGCACACGCACACACACACACACACACACACACACACAGAAAGACAGAC 3241 AGAGACTGACAGAGATGGAGACAAAGACAGAGAGAAAGACAGCATGACAGACAGCGTTTC 3301 AAACGGAGCCACTGTTCTGGAGCAGTGTTAATTGACTGTGAGATCAGAGCAATCTGAAGT 3361 CAAATTCCAGAGGAGCACACTTGCACAAATCTGTGATCCCAGCACTTGGCAGGCTAAGGC 3421 AAAAGAATTGCAAGTCTGAGAAAAGCTTGCATTCCATAGATAGATGCTGTCTAAAACTAA 3481 AATAAAATAACTAGTTCAACACCAGCCCAGGATATATAAGATCCTGCTACAAAAACTAAA 3541 CAAAAGAAAATGTGATTCTTCTCTATATTTAAGATCTAGGGAGCTCACTGAATGTGGACT 3601 TTTATCTCACAATTTGCTTTTTAACTTTAATTACTAAGTTTGTTTTGCTGTGCTAGGGAA 3661 TGGGTAAGCATTCTACACTAAGTTGCATCCTTGGTCCTCAGTCCATCATAATAGCATACA 3721 GACCTCTACAATGAAACTGACAAAATATACTGAATCCTAAATGCTTTCAACTTTCTGCAG 3781 ATTAGCCGCTGAAAACAGAGCTCTTGGAAATAGGAAGTTTAGAAAATGAATAAGAGCTGG 3841 AGAAAACAAACAAACCACCAGGACAAATCTGGTGATATTTTAGAATATTTGTTATCACTC 3901 CTGCAATTACATGAAAAACTCTATTTCCTGACAGCGATACATTTGAATTAACAAAATCGT 3961 TGCAGGATGAACTTACATGTAATAAAGACATTTCTAGCCAGGCGTGGTGGTTGGTATGCA 4021 ACCATAATTCCTACCCTGGAGGCTGGGAAGGAACTTGCAATAATAAATTCAAGGTAAACT 4081 AGGCCTCACCAAAGGAGCCTCTTTTCTTCATTTTAATTTCTAATACCCTGATATGAAACA 4141 CAAATTATTTCCTAAAACAGAAAATAATTCTATTTCAATACATGAATGTCTTGAATTAAA 4201 GGTTCACTATCGATCTATTGTTTTACTAAGCAAATAGAATTTTTTTGCATACATTTTTCT 4261 CAATTTCATAAATGTGGATACATGGCCACAATAAAAAAAAAAAACTAACCATCTTAACAT 4321 AGAAATAAGGTGGGAATGGAAAAACAGACTTGGATTATGAATTTTAAAAAAAAAATAGAA 4381 TAATTGCAGCCAGTATGTCATGACAGATACAAATCGGGCATCGTTTTAATCAAACTGAAT 4441 TTAGATGATAGACTTTCATGGAAAACCTAATTTACTTTGAATTCCCCGGTCATGACTTGT 4501 GCACGCTTTACACGACAAGGTTAAAGAGATTACATCTGGGTTCCCGTCAGCAGCACCTGC 4561 TGGTTAGTTCTCCGCACGCTGATTTAATGAGAGGTCAAGCGCTGTCTGTCCAGAGGACTT 4621 CTGTATCAGGCAGGTGATGGGAAGAAGTCTTCATTTCCGTGGCAGTTGGTGCTGGGGCTT 4681 TCACAACTGATGTATCATCATCCCAGCATTTACAAACTATCTTGGCATTTGTCACCATAC 4741 TAAACTTAGAAGCAAAAAAAAACACAATCAAAGCTTTTTTTTTTCCAGTACAAAGAAGGA 158
4801 AGAAATGTTCTTCTTTTTGTGGCTTTTTCAAAACTGACTTACATGTATTTCTTAAAAGTA 4861 CAGTCATTGTAGAAAAGTATATGTCTATGTGACACAATGCCATACGCAATTACATAAAGT 4921 AATGTCCAGATGGCAGGAGAATTCATTAACAACTTCTTATGCTTCATAAACAGGGGAAGA 4981 CAATACAGAAATCATTCATCCTAAAAATTCTAATCCAATTTTAAAAATAATTTAAATTCA 5041 TTAAAACAGTTTCTATTTTAATTCACAAGTTGTTCAGAGAAGAGCTTAAAGATGTTCCTG 5101 TAATTATTGTAACAGGAAGGGCTTTGAGCCATTAAAATAATCTCTGGCAATTCAGTGTGA 5161 TTCCCCTTATTTATGCAGCACGCTTTTCTGAGTGTGCCTAAGTCCCGCTTTGCCTTTACA 5221 GAAATGGTCACTTCTAAGCAGTGTATCACAAAGAGGCTCAAATTCCAAAATCAATCAAGC 5281 CATTTTGCCAAAGACGGAGGATTTACAATCGAGTAATCTGTAATCATATAGATGTTTTTA 5341 TGATAATTATCTTCCTTTTGAGCACACATTTGTGACCAAGACTATGGGGGGGGGGGGGAC 5401 GAGGTAGATATAGTTTGTGGTATTTTATTAGGATGAGACTGAGAAATCCCAAGCATTTTA 5461 TTTTTATATTAAAACCCACTTTTGCCCAGGCCCTGCTCATACATCACTGTACAATACTGA 5521 CAACTGGACACAGCTCTTGACAAATCACAGTTTAAATATTCATTTTGAGTATGCATAAAA 5581 ATTATAGTAAGAAGATGGGCAATTTTCTCAACAGGCCTATCATAATCATGAGAAGAACTG 5641 ATTATTCGTGGTAGGGGATACAGGATGAACAGGTGTTTGGGCTCAAGAGAATACATTAAT 5701 TTTTCCAATTTCAACAACTGAGATCTTTACTAAAGATGATAGAAGTTTTCTTTTGCCTGT 5761 GCCTTTGAGAGAAGCAGATATGAGTATAGTAGGCTAAGCTGAGATGTTCTGAAGACACTT 5821 CATACAGATGGTTTATCTACTTCAAAATGTATTTTTAGCCTGTGATTCTTTGTTCACTAG 5881 TGTGCATCAGCTTCTATACAATCTCCATTTCTTTCTCACAGAGAGCTGTCACCAGACTAC 5941 CACACCAAAGTAAAGGTTCTAACTCTCCTGGCATCTGGCACAACCCTCTAACTGCTTGAG 6001 ATAAGTCCAAAGGGAAGTCCATTCAGCCACAAGCAGTGCAAAGGCATGTGATAGGTAATA 6061 ACTTAGAAGTAAGCTGTATTGACAATCCACCCAATGTTTAAATCAAACATTTAAAACTAG 6121 TTCAAAAGTCAGACCAGAGAAGTTTAATTTCAGGACAATTTCTTTGAAAGCAAAGAAGCC 6181 ACCTCGCTAAGTCAATTTCTGACAGAATGACAGTTTCTCCCGCCCCACATTCACAGCAGT 6241 GAGTGGATACAGGGGTCAACATGACAAACAGCATGATGCTAAAATGTTCCACAGTCTGAA 6301 GCATGCCTTGCCCAATTTTGAAAATATGTGAAAAGATAATAACAGAAATTTCCCACTTTT 6361 TTTTTGCTTCCAACATGCATGTTGTGTACTGACAAGAAAGGGAGAGGGGCTTGTGGGTAT 6421 TGCATAGATTTCTAAAATTATGCCCTTTATGATCCTTCCTGTACCTCTGCGATAGAAGCA 6481 TTAAGACTTGGTTGTACGGACTGGAGAGATGGCTCAGCTGTTAAAGACTAGGCTCACAAC 6541 CAAACCCACAGAACTCCGTTGTACAGCATATTATTAACTGGGCAGATTTCCTTACTGCAG 6601 AAGTACATAGTTTGGCAGACTTTGCTTTAGCAGTCTCCACATTCAAACAATCAATTGCCT 6661 GTTGTAATTACCGCAGGGCATTAAAAAAGAAAAACAGATGAACATTATGTATCAATATCA 6721 ACCCATCAGAATATAAAACTCAAACTATAAACTATTTTCTTGTCCTGCAGCTGCTCAAAT 6781 GTGACATGTGCCCTTCTTTATAGAAGCCTGCAGATCAGAACGTGAGTCCTGGCTCTCAAC 6841 CACTTAAGAAGGCACTGCCTTAAAGAGGTGAGATGTGGAGGGCAGCGAAATGGCCCAAAG 6901 AGCAATAAATAACCTCCCAAAGAAATGGACCTTGGACAGAGGCTTCAGTATGTACTCTGA 6961 GATGCTCATTTCCCCAAGATCACATGTGACTTTTCCCTAAAGCATTTGCCTAGAACCCTT 7021 TGGGTCGCATGGCAAAATCCTAGGATCAGAAATATTGTGACTCATTGTTAGCTTCTGGAT 7081 TCCTTTTTGAAATACAGGCTGCCCCTTCTCAGCTCAGGCCAAGAACCAGAAAAGGTTGAA 7141 TTCCAGGAAGTGAAAACTGCTTCCCTGCCAAAGCCATCCACTCTTCAGGTCCAGGTGACC 7201 TCCCAGGCATTGTGCTGGGGAGCAGCTGGTCAAGACTTCCCACTCCAGCCTCCTCCTCCC 7261 CTCTCCAACACCATCCCCTCTGTCCTCCAATCTGAAGAGGCAACTGCTTTCTGCAAACCC 7321 CACCACAAGGTGCTGAGGACATAGAGGCTCTGACATGAAGAAGCTCACATTCTTTTCACC 7381 AGGAAGCAGGTCAGAGCATTACACAGTGGCCAGGGCTGACAAAGAGTTCAGTACAGGAGG 7441 CTCTGCCCACACAGAAGAGATAGCCAGTAGTAGCTCTGCTCCCAAAGAAGATACACACTA 7501 TCATCTGTGCCTTGAAGTACCTAAAAATGGAGCCTCCCCACAGATTGAAGTAATAAGGGA 7561 AAAGGGAATTGGCAACAGCAAAGCATGAGAGAGCAGGTAGGAGGTGGTGCAACAGGTGGA 7621 TGGAAGTCCATACCCCCACCTCCATCCCAGCTTCCTAAACGTGCCAACAGAATGGCAGAG 7681 GCTTAGCAGAGAAGCAGGCTCCTGCTAAACACAAAGCATGGTGTCTCTTTTGTTTAGGAT 7741 GTTTTCATAGAGCTTAAGGCAGTGTGTCATACTTTGAAGCAGTGATAGCCAAACAATTGG 7801 GTTTCTGCAGTGGAATCTCTTTACCCCACCACCAAAACAAAAGCCTCTACAAAAATCCCA 7861 AGAAGAAAAGAGACAATCAGAGCTGCTCCTGGTGAAGACTTCCAAAGGCCTTTCTGAAGA 7921 CCCCAAAAGCTCCGTAGGCTATTCTGAGGAATTTCTTCCCATTACACCCTGAGACCCACA 7981 AAACAGCAACTACATCGTCCTCATATTCTTTTCTGATACTACCACACTGACCAATCAATA 8041 GCTTGAACTAAAGTGATCTAAATGGAAAAGCAATGTAAGAGATCAGAAAAGAAAAAGAGG 8101 TATACCAGAAAGCGGCTGGTTGTATCAAGGAACATTTAAGAATAAGTGACCAGGGGCTGG 8161 TGAGATAGCTCAGTGGGTAAGAGCACCCGACTGCTCTTCCAAAGGTCTGAAGTTCAAATC 8221 CCAGCAACCACATGGTGGCTCACAACCATCTGTAACAAGATCTGACGCCCTCTTCTGGAG 8281 TGTCTGAAGACAGCTACAGTGTACTTACATATAATAAATAAATAAATCTTAAAAAAAAAA 8341 AAGAATAAGTGACCTATCTCATGCAGTGACTGAAGGAACAGAAAAGTTGGGATGGAAAAT 8401 GAAAAATGCTTCTCTCAAGTACCTAGAAGCCTGGTGCAGAGATAGGGAAAAGCGATCATT 159
8461 GAGACCCTCTGTAATTAAAGCAAGATTTATGAAAAGTCAACATAGGCAACAACACAAAAA 8521 ATGCACACAACAAAGGCTGCTTGGCAGACATAACACTAGATCTAAAACATCAAGGTCCTC 8581 TCGTTTTACTGAGAGGGAAGGAAAGCAACATAGGCCTCAGATAAATCACGAAGACACTTA 8641 TATGCTCCATCCAAAGTTTTATCTTTGGTTGCAGATGTAGTTCAGTTCATAGACTGCTTA 8701 CCTAGCATGTATGGGACCCTGGCTTTGATTCTCAGCACAGAATAAGTGGACATAGTGACA 8761 CATGGCTGTAATCCAAACATTTTACCCAGGCCCAACAACCTGCAAAACTAACTGTATTCT 8821 TCATCATTCATTCATAAATTCATAGTCCCTATTCTATGCTGAGCACTCATATACAGTTAG 8881 AATGCATGAGAAACAGGTCATATCCTTAATCTTCTGCTATTTATATGCTGTCTGACTATG 8941 GGACCAGCAATAAACAGCATTTGGGAAGTAGAGGAAGCTCCATAGACCATGTTGAAGAAT 9001 TTCTTCCCATTACACCTAGAGACCCACAAAACAGCAACTACATCTTCCTTATGTTCTTCT 9061 TTCTAACACTGTCACGCTGACTAATGAATAGCTTTGGGGAATCTAATCCAAACAGCATTT 9121 GGGAGGTAGAGGAAAGGATCAGAAAGTCAGGGTTATCCTTGGCTACATAGAAAGTTTGAG 9181 TCAAGCCTGGTATACATATGTATGAACATACATAAGTATATACGTCCATATGTAGAAGTG 9241 GAGAGCCAAGGAATAATTCAAAATAGGTACTCAAAAAGGGTTGGATATTAGAAAGCCTAG 9301 AATTTAAGAAGTTATTTTCATAAGCCAACTATTTTCATAAACCAGAGATTGTGAGACGAT 9361 ACCTTTGGGGTTAGAAAGCAAGCCATGGACTGAAGAGCTATAAATGAGTGAGGACTTGAT 9421 TGTTCGCCTCTGGTTATTCTTCATCTTTCCAACAGGAAGCAGAGTATCACCTTCGCCTGG 9481 GGCTTCATGGTGGCATTAGTCAGATATTCAGTAAGCTCAGGAGCTGCAGGAAAGGAAGAA 9541 GAGAAGTGAGAGGAGGGAATGAGGAAGAAGAGAAAGAACCAGAAGGGTAGAAGGAAGAGT 9601 AAGGGAGAATAGAAGGGCATAGGGAAATGAGGAAAGGGGAAGAGGAGGGAAGTAGAGACA 9661 GATACATCTACCAGCTAATTAACCTGAAGCCAGCCTGGCCTGGAGGGGTGTCCAGGGTCT 9721 TTCATTTGCCTTTCTTCTTGGAAAGCAATTCTAACTTTTCCAAGTCTGCTCTCTTGCTGT 9781 CTTCTTGTAGTTACTATTGTGAAGGTCACGGCCTTCCTTTACTAATGCCTCCAAACTACC 9841 TATATGTTATACACTACCTATATGTTATACCTATTATCTACTCTATGGCTACAATGGACA 9901 AAAGCCAATGCTTTCTCTGTAATGAAGGGGCTACCTTACACACTTGATGAGTTTGCTACA 9961 GGGTTCCTCAGGAAGGAATTTACTCAACCAACCATACCAGTGTGCCTACTTCAGTATCTG 10021 GGTCTCAGGGTCACTAGCAAAACTTGCTGATTCCAATGACATCAGTACCCTTCATCAGAG 10081 CCATTCCCCCCACTCCCCAACATTAGCTGTTCAGCATGAAGAAGGGTTTCCCTAATTACA 10141 CTCACTGCCTCCACTCCAAGTGTGACTCACTTGGGGACATACATCTACAAGGCTGGGTTG 10201 TTACAAAGATTAAATGAGATCACATATATAAATAATGCTTTAGCAGCCATTCCATCATAA 10261 TAGCCATCATTATTATCATCAGCATTTATTCATTCAACCAAGCAGGCACTCAATAAATAT 10321 TGATTGACTATATGAATGGATGAGTTAAAAAAAAAAAGTTTCTTCAAGCAGGCCAAGTTA 10381 ACAAAAGGATGCAGGCAACTTCATACTTTCCAAGCCTGGCTACACATAGTCATACCTACT 10441 ATCTGTTTAGGGCTCTTCCACTTATGGAGCCCTTTTACAAAAGGTTTATTTAAAAATTGT 10501 ACACGATTCCTATGCCCAGCAGACATCAACACCCTTACCCTATTTTTCCCACAGTACTTA 10561 TCATGTCTTCTCAGATGCTATATTTTAATTTTGTTTTGTTTACCTGTTTATTGCTGGTCC 10621 CACAGTCAAACAGCATATAAGCTATAGAAAATTAAGGATATTGCCTGTTTCTAATGCATT 10681 CTAAGTGCATGTGAATGTTCAGCATAGAATAAGGACTACGAATGAATGAATGAATGATGA 10741 AAATACAGTTAGTTTTGCAAGCTGTTAGCTCTGGGTAAAATGTTGGGTTGATTGGGAGTC 10801 AGGCTACATTTGAAAAGGATCAGTAAAGGCTGAGGTTAGAAATAAAATGGACCTGTTTAC 10861 CACTGGCCTTAGTCTAAAAAGACCAAGGTGCATGTAAGAATCTCTCTTATGGAACACAGA 10921 TTTATTTTATACTTCTGCTCTCAGGAGTACCCAGTTCTTCATTTACATATGGTATTTTAG 10981 AGTACACAGAAAGCTGTTTCTCTATCAGCCTCTGAGATGGTATCATTTCTAGTTTGCTCT 11041 TAGTTGCTATAATGAACCCTGATCAAGACTGACTTGGGGGGACAGGGTTTATTTATTTCA 11101 TCTTAAAGCTTTCAATCTATCCTGGAAGAAAGGAAGGAAGGTGGGAACTTAAGGCAGGAA 11161 CCTGAAGGCAGGAGTTGATGCAGAGACCAAGGAGGAGCCCTGCTTACTGGCTTGCTCCTC 11221 ATAGCTTGCTCAGCCTGCCATTTTGTTTTGTTTTGGTTTTGTTTTGTTTTGTTTTGTTTT 11281 GTTTTGTTTTGTTTTGTTTAAATGCAAGCCAGGATCACCTGCTCAGGGATGGCATGACCC 11341 ATGGTAGGCTGGGCCCGCCCTTGTTCTGTTTTGAATTGATTTTCTTTATGAGGTGAAAGA 11401 TACAAATCTAATTTCATTCTTCTACTGCTAGAAATCCAGTTTTGCCATCACCATTTGTTG 11461 ACTAGGCTATTTTCTCATCCATAGTATGTGTTTGCTTTCTTTGTCAAAAAATAAGCATAG 11521 CTTTGCCATTTATTCCCAGGTCCTCTATTCTGTTCCATTAGTCTTTCTGTCTATTTTTAT 11581 GTCAATGTAGGATGTCTTTACCACTACAATGCTATAGTATAACATAAGGTGCAGTATCAT 11641 TCTTACTCCTCATTATTGTTGTAACTATTCATGGTGTTTTGTAGTCTATCATGAATTCCC 11701 ATATTATTTTTTCTAAGCCTGTGAAATATGACATGGGAATTTTGATAGGTATTGCATTAA 11761 TTCTGTACATTAATTTTGGTGGTATAAAACATGGACTGTGGTCTCTTTTGCCGATAGTAA 11821 CTGAACAAAATAAAGGAGGGGGTGGAATCAAGTTCTGGGGAGATGGCTTAGCAGGCAAAC 11881 CCTTGCCACACAAGCAAGAAGACCTGAGTTCAGAATCCCAGCACCCATGTAAATGACGTC 11941 TATACAGCAATCCAGTGCTCAGGAGGCAGAGATAAGGTATCCACATAAGAAGCTGGCTGG 12001 CTAGACTAGCTGAAGCTGCAAGCTATAGATTCAAATGAGACACTGTGCCTTAATATTTAA 12061 AGTGGGGAACTACCAAGAAGATACTCGATGTTAAGTTCCCACCTTCCTTCCAGGATAGAC 160
12121 TGAAAGCTTTAAGATGAAATAAATAAACCCTTTCCTCCCAAGTCAGTAACACACATGTGC 12181 ACACATGTACATGTTTATACATACATGATCACTTATATACATGCACATACAAAACATAAA 12241 CAAAACCTCCAACAACCTGTTAAAAATGGGGTCTCCTTTCTCATCAAGTGCTTTCTGGAG 12301 ATCACGTGGTTTCAAATTTGCTGAATCTCACATCCCCAATGTCAAATGGACATTGTGGCT 12361 AGGGAAAATTGTAAATCCCTGCAAAATCAACTAGGGTAGAAACCCAGTTATGCTTGAGAT 12421 AGGAAAGAAAAATAAAAACAACCATAAAGAATGAAAGACATGAATGAAGGAGCTGTCTAG 12481 GCATAAAGATTATCTACATTTTATGGTTGAAAAACTGAGGGATTGGAAGTGATTCCTTAA 12541 AGTAGCTAGATGATCAGGTGTGGTGGTCCACATCTGTAACCTCAGAACTTAAGAAACCCG 12601 TGTATTCCCTGAGAAGTGCCTCAGCATAGCTTGGGTAGGCAAGAACGAGCTAAGTGATCA 12661 CCTGGATGCCATGGACTCTAACCTGGACCACCTGCAGACCATGATGACAAGCCATGGCTT 12721 CAGTGTGAACACCAGCACCCTGCTTGGCCTGTGCAGCGCCTCGGTGACCATGTGTGACAT 12781 GAGTCTGCCTGACCTGGACAGTAGCCTGGTCGGCGTTCAGGAGCTTCTATCTCCACAAGA 12841 GCCTCCCAGGCCTATTGAGGCAAAGAAGAGTAACCCCGACTCAGGAAAGCAGTGGGTGCA 12901 CTGCAGGGCTCAGCCTCTGTTCCTGCTGGACCCTAATGCTGTGAACTTTGAGCTGTGGGA 12961 GAGCTCCTACTTTTCTGAAGGAGATGACTACATGAACAGTCCCACCATCTCGCTTCTGGT 13021 GGGCATTGAACCCCACAAAACCAAGGACCCCACTGTCTCCTAGAAGCTTGAATAGTGCTA 13081 CTGAATAGTCAAACACACTCCACCCCTGGGAGAAATATTCCACAGCCAAAACATTCCAAC 13141 CCCCAAATCTGGCCATGTCCCCACCCTATCTTTTATATATATATATTTTTAATTAGGTAT 13201 TTTCCTCATTTACATTTACAATGCTACCCAAAAGTCCCCCATACCCTCCCCCCAACTCCC 13261 CTACCCACCCACTCCCACTTTTTGGCCCTGGCGTTCCCCTGTACTGGGGCATATAAAGTT 13321 TGCAAGTCCAATGGGCCTCTCTTTCCAGTGATGGCTGACTAGGCCATCTTTTGATACATA 13381 TGCAGCTAGAGTCAAGAGCTCTGGGGGGTACTGGTTAGTTCATAATGTTGTTCCACCTAT 13441 AGGGTTGCAGTTCCCTTTAGCTCCTTGGGTACTTTCTCTAGCTCCTCCATTGGGGGCCCT 13501 GTAATCCATCCAATAGCTGACTGTGAGCATCCACTTCTGTGTTTGCTAGGCCCCGGCATA 13561 GTCTCACAAGAGACAGCTATATCTGGGTCCTTTCAGCAAAATCTTGTTAGTGTATGCAAT 13621 GGTGTCAGTGTTTGGACGCTGATTATGGGATGGATCCTGGATACGGCAGTGTCTAGATGG 13681 TCCATCCTTTCGTCTCAGCTCCAAACTTTGTCTCTGTAACTCCTTCCATGGGTGTTTTGT 13741 TCCCAATTCTAAGAAGTCCCCACCCTGGGGGCTGGAGGGATGGCTCAGCGGTTAAGAGCA 13801 CTGACTGCTCTTCCAGAGGTTCTGAGTTCAATTCCCAGCAACCACATGGTGGCTCACAAC 13861 CATCTGTAATGAGATCTGATGCCCTCTTCTGGTGTATCTGAAGACAGCTACTGTGTACTC 13921 ACATACATAAAATAAATAAATAAATAATCTTTAAAAAAAAAAAAAAAAAAAAAGAAGTCC 13981 CCACCCTATCTTAAAACCCTTCAAGAATATGGAGATGGTTTAGCGAGTTCCTGAAAAGGC 14041 CAAATATAATTAATAATAACAGTCATTTAAAGACTCTGGGGATTGATTAGAAGGCTACAA 14101 CAAATTATAAAGGTCTTGTTCAACACAGTCTACAGAACCTCAGTGGTGTATGGATCATTC 14161 TCTACAGTAGAAGACTGGGGTATTCAAACTACAGTGACACCTCCATTATCCCAACAGCTG 14221 CACTTTATACAGACCATTGTACAGATTCATACATGTTGAAGATCCATTAAGTGAAACTCT 14281 GAGCACAAATTTATCAATGTTTCCTCCTAACCCCTAGCATTAATATGAAGTGAGAGGACA 14341 GAAATTATTTAGACTCTGGATATTATTATTATTATTATTATTATTATCAGAACTGTCTCA 14401 TGGTTCTAATCTCCAGTTTCAGCATCCCTTCACCTATTAGCCAAGAATGTAGGCTTGTCC 14461 TCCTGATTCTCACTTCTCTGCTTTGTTAGAACAGCAAGGCCTAAGTTATAAGAAGCCCTA 14521 ATGGACATGCTCATTTGGAAACCAGTTTTGCTTACTCTGAGACCCAAATCATTGAGGGCG 14581 CCAATCAGGTAGAACTCTGTTCAAAGTAGAAGGTGAAACACCTTTAGGGGCATATATAGT 14641 CTCTAATGAAGATGGATTGGTAACTGGAGACCTTTTTAATAAAAGAAATGGCCAGGCGGT 14701 GGTGGCGCATGCCTTTAATCCCAACACTTGGGAGGCAGAGGCAGGCAGATTTCTGAGTTC 14761 AAGGCCAGCCTGGTCTACAGAGTGAGTTCCAGGACAGCCAGGGCTACACAGAGAAACCCT 14821 GTCTTGAAAAACCAAAAAATGAAATAAAATAAAATAAAATAACCCATTCCTCCACAGAAA 14881 TAAAGATTTTATCAGAGTATAGAGTTAGCAAAGAGAACGTGCACATATGTGTGGCCACGT 14941 GCCCTTTACTTCACTGCCTCCAAGACAAAATTCAGTATGATGGCCAGAGAACCAGCATAA 15001 GGGTGAAAGAGGAAATATTAGCAGGTCAGCCATGACCATGCCAGTACTAAGCTGCTATGT 15061 CTCATTCATTCTCCAAGTCTCATAGGCAATTCCAGAACTACACTTTGTGTGTGTGTGTGT 15121 GTGTGTGTGTGTGTGTGTGTGTGTGTGTGACCTCTTTTTTTATTAGATATTTTCTTATTT 15181 ACATTGCAACTGTTATTCCCTTTTCTGACTTCACCTCTGAAAACCCCCTATCCCCTCCTC 15241 CCTCTCCTTGTTCCCTAACCCACCCACTTCCCTGTCCTGGCATTCCCTATACTGGGGCAT 15301 AGAACCTTCACAGGACCAAGGGCCTCTCCTCTCATTGATTCCCGACAAGGCCAGAGCCAT 15361 TGGCCCCTCCACATGTACTCTTTTTGGTGGCTTTTGTTTCTTTGGTTTGGGATAGGGTCT 15421 GTGTAGCTCAAGCTGGCCTCCAGCGCTTCAGCCTCCCTCTCAAGTAGCCATGTGTACTAT 15481 CAAGCTGCAGTTGTGTAACAGCACACTCAGCTATCAAATCACAGATTTAGACTAGAGCTC 15541 TGAGGAGTTAAAGAAGACCTCCTAAGGATCCAGGCAGGATAGTAGGTGTGCTGGGCTGTT 15601 GACCAATTCTCTGTGTTTAATTTGTATCCATTGAGCCTACTTACACTTTCAGCAAGAAGT 15661 TTAACAGTCGTCAATTCATGCTAGATTTTGTGTTTCAGATGCAATGTCACAGACAAGCCA 15721 TAAAACTATTTTTAATGTGAACTAAAGTCTCTCTGACTTATTAGAAATGATTTTTATAAT 161
15781 ATTCTAGGTCTTTCTCAAAGCCCTGTAAAGAACGATAAAGTTTTAATCTCTCCTAGTGAT 15841 TACTGCTACTTTTTTTTCATTCCACTTGGCCAGTTGTGCCTATGTCTGAGGTGTTGGAGA 15901 AAGCATAAAAGCTGTGAAATTTGAACAAGGCCTCTTCTCTCATTCAACAGGACTCCTGAT 15961 AAAATCGACAGAACAAGATTATTTTCTATTGGGCTCTGAAAGAAAAAAAAATTAAATTTT 16021 ACATTTTTAAGCAAAAGTAAAAGCTGTAGAATCTAACTGTTCTTAGTTGTATGTATTGAG 16081 CAGCAGGCCTGCCTCTTTCATAGAGTTATTCTATGTCAGAATTAAACAGCAGAACTGTTT 16141 GTCACAATCCAAAGCACGTTGGTTCAGTGAGTACTCTCCAGAGAGCCTTGACACCAACAA 16201 TCTCAAGCAACGATGCCAAACCAAGAAAGAGTGTCATACATGATTAGTTCAAATTTTTAA 16261 AATTCCTTTCAAAAAGGACTTCTTCAGATACTACAAGTAATTCCACAAACATCTACGCGG 16321 AGCCCTGTAAAGGCTAAGGAGAGTGGCACACGTTAAAAAGCGTGTTCAATTCTACTTGGT 16381 AGCTAAGAGGTCCCATGTAAGAAACAAGGCAGCTCGTAACTGAACTTGCTGTGTTAAGCG 16441 GATGTCAGGCACTTTCAGCTGTACTGGGAATCAAACCCCACAATGGCAAAGCTGCCTTTC 16501 AGTGAGATCAATAAACTGGTCCTACTCATCAGTTAACAACTGCTCATCCTGAAATGTAAG 16561 CCCAGATTTAGTACACCTCTGCAGTTGGGTCTCCTATTGGAACACAGAACTGAAAGCAAA 16621 CAACTGATAACAGCTTTAGGAAGTCTCTGAAACTGACCAGAGCCACCCTGCTAGAAAAAG 16681 CAGAGCAAAGCTACAAAGAAAACTCTGAGACCAGACCAGCTGCCTACAAAAAGCAGAAAC 16741 CTGCAGATATGTGCCTGGGAGAGGTTAAAATCATCTGAGTTACTTGGAAAGGACTCCGTC 16801 CAGCCTGTTGAGCTGCCTGCAGGCTCTGCATTGTGCTCCAGATCTCCAGCTTTGCGAGCT 16861 CTCACACGGAGGTGGGCTTCTGCTCCTATAAGTAATCCCTCAGTTTCATCTAGAACCCCA 16921 ATAAAAATCATTGATAAACTAAAAGGGGGAGCAAATATTGGGGATACAGTTTAATTTCTT 16981 GACTAGGATTTCGAAAGCCCTAGGTTTCATTTTCAGCACCACAGAATAAAGACAGGCTCA 17041 TCGATATATGCCTGTAATCCTAGCAGTATAGGAGGTGAAAATGGGAAGTACAGATATCCC 17101 TGTTCATTCTCAACTACATAGCAAGTTTAGGACAAGCCTAGGATATATAAGACCTTGTAG 17161 GAAAAAAAAAGGGGGGGGGGAATGGAAGCAGGGACGGATGGAAAGATGGGAAGAAATGGT 17221 GGAGGGTAAGCAAGTTTGCAGTTAGTGATGAATTAGGCTTTAATGCTTAATTCTTAGAGA 17281 TTTACTAGAAGCCAATTAAATGTCAGTCCAGGCCATAAATGGAACCTGGGCAAGTCAAGG 17341 GAGGTACCTAGGACCCAACACGAAAGCAGGCAAAAGCCCTGCTGCATCATCCTGAACAAA 17401 ACCCAAGCAAATATGTAGCCAGATCTCTGGCTACTAAAAGACCACAGTAATTTAACACAT 17461 CTAAACCCCGAGTCTTGACTATTGAACACATCAAACAACTATATTGATTTTTAAGATTCT 17521 ATCTAATCCAGGAGTACACTTTCAGGACCTTAACTATCTATACTAGTTACCTCGCTAGCT 17581 TTAATCCTGGCTTAGGAATTCACTCTCTTGCTACAATCTCAGACCTCTAGGGCCAGATGA 17641 AAAATAAAAATCGTAACCAAGCAAAAATAAGGATGATGATAATGATGACGAGTATGGTGA 17701 TTGATGATGATGAGTGTACTTTAGATTTCTACTTCAGCATCACCTCAGAGAAGCTTTTCC 17761 TGATCACCCCAAAATATCAACTCCCATCACTCTCCACTCCCTCGTCTGCTTTGACCCCCA 17821 CACAGTTGCTGATTCACGGATGCACTGGTGACTGCTATTCCTGTCCCCACCAGTGTGCAA 17881 GTGTCATGAGTGTTAGGACTTTGTTTAATCCACTGTGGTAACTAGAGGAATTAAAGTGTG 17941 CCATGCTATAGTAGATCTATAATAAAAAAAATACATCAACTATATAGTGATCACAATTGT 18001 ATATGTCCCTGTGGTCCTGGAGCCTATCTATGGTCTAGTTATCATCTAAGCAAGACCTTA 18061 CTGGTGCAATGGAAATCACCTGCACAATAATTGTGTCCTGGCACATTTCTTCTCAGACAT 18121 TAGAAACACTTCGTTTTCTTCATTTTTAAAACATGGAGATGCTCAGATTCTACTTCTGTC 18181 AAAAAGAAAGAGAAGTAGCTCTTACTATTTGTCTTCTGAATATGTTGTCCCTTTGCAAAG 18241 CTGTTGTCAAAACCCAGGCTTATTTGCCATGGGGTCTGGTACTTTTCAAAACTCCAATTC 18301 AGTTGCCCACATTACATGCAGGATTCCACCTGAAATAAGTGTAGGTGTTGTAATACACAG 18361 ACCAGACTCAAGAAAATACTTGAACAGTGCCTAGTCTGACAGCTTCTAGGCTCTGATGAG 18421 AAAGTACACTTTGCTAAATTCGGCCACTCCACCTCATCTACAGAAGGTAGCAAGTACTCC 18481 TGCAGCCAGTATCTATACAGCTCTGCGGAGAAGGACACTATCATCTGCAAGAGCAAGAGG 18541 AATGCTCTCCCCATTGCATACATTCTGCATTATAGGAGCCCAGAAGTAGGGAAGGAAAGT 18601 AAAGCAATTTGATTGTTAAGGAAGATTCCCTTCATCCAGCTGCCAAATAAATCAGATATG 18661 CTCTCTTCTCTTTGGGGGTGAAAAAAAAAACGCTTTTTAGCAGCTGTGTTCTGACATCCT 18721 TTTGAACAGAATAACAAAGAAGCAAAAGGTAAACATAATAATGTACTTGATATGAAAAAT 18781 ACGGTGGGAAGCCATGAAATAGTAGCATCCTTCTTTAAGGAGGGATGCATGCTGAAAGCA 18841 GGGGCACAGGGAACATTAAGGTGGCAATTTCTGGGGTTGTTCTTTTGGAAACACATAGCG 18901 ATCCTACTGGCTCTACTCCCTGCCTCCCAAATTGGCCAGCAATCTCTCTTTAAATTTGTT 18961 TTGAAAGACAAAGGGAATGGATTCTGCCAGCCCAGTGACAAACCCACAGCCATAGTGTTA 19021 TCTTATTTAAATGACTTTACAGTCCTGTAAAAATATGTCCTTATCTTGTGAATACATTGG 19081 TTATTTTGTGTTGTAGTTAAAAATGACCTAAAGGCATTTTCTAAGCATCATGGTAAACAG 19141 GGAGATCATAAACAAACCATGTTTAGCATTGCTGCTGCAGTTTCAGTATTTCAGATTTCC 19201 ATAGCAATTTAGAAAGATCAAATCATTGGCATTCATTTTCAGAAGAAGTAAAGCCTTCTG 19261 AGCCTACAAAGCTAATTGAATGGTTTGAGAACATACGTTTACAAGACGCCTGGGCAACTA 19321 TCACTCAATGTTTTCTTTATTCTCACTTTTAATATGTTGATGGTTTTATTTTACAACATA 19381 AACACCCAAATAAGAAAAATTCTTCTTCCATGCTAGATAATTTCACCAAAATGGTGACGG 162
19441 CAATATTTTTTTAAATGAAAATTTATCTAATGGCACGTAACCAATAGGGCTAGAAATTTT 19501 CAAGATTATATTCTTAAAGCACAGAGCTGTATGATTCAAATTGTGATCTAAAGGCCACTG 19561 CAAAATTACGAAGAGATGGAGTGAGCTCTGGGAACAATTAGAAACAATCCTGAAAGCTGC 19621 TCATTTATATCCCTCCCTGTTATCACCTCGTCTGGGTCTTTCAATGATGGACTAAGAAAT 19681 GAAGGGGGGAGCAGGGTTTTCAATATGTCTTTTCTTTGAGAAACAATAAAAACATTAAAC 19741 TTTATACATTGAAGTACCTCTGCACCGAATGGTAGCCCTTTCATGGGAGTCCAGCTTGGG 19801 TTTTGTCTGGGTGGGGAAGTAGGGAGGGAGACTGTGAATGCCCCCAGAGTCATTTAAAAT 19861 ATATAGTATTAGCTAATTGATGAGAGGACTAAAAACAATAGTAAGTAGGGAGCCTGGCTA 19921 CGAGCCAGCAGCAAGGCTTTATGGTGAAAGCCAGTTTACATTAGAGGCCTGTTTTTCCCC 19981 TTTCCTTATATGAAATGAACAGCTGTAATTGAAAAGCAAGGAGCTCTATGGTTTATAGAA 20041 ATAGGTGTTGTGAAGAGCCTTGGCCGTCTAATTTTCTTCCTATTGCAGGGCAGGCTTCCA 20101 TTGTCTGTTTTTTAGAGCTAATTTCTTTTCTGATCAAAAGGCCCACGTCGTTGAAAAGGT 20161 AGTAATGACTATTCATTTGGCTGAAGCCTCCATTCTCCAGACTGCAAGCATTATTTAGAA 20221 ATGGCTATTTCTTTCTAAAAGGAAAAGTTCATTTATGTCCTCTATAATGTCAATTAAAGT 20281 AGAATATATTTTAATTATTTATCCTCAGGAGGAACCTTTTAGGAACTTAATTTAGTATGC 20341 CTATGCTATGCAGACACTACAATATGCGAATGTGGAATGATATCTAGGTACATGGGTTTT 20401 ATTTTATTTTACCACTAATGAATAGTTTATTTGCAAGTAATTAACAAGTTAAATCGGCAC 20461 AATTAGTGTCGGATTTCTTTAATAATATAACTCATAAAAGGATAGTCATTGGAACTAGAA 20521 AATAAGGTTGGTTGTTGACTTTGTAAAGAAACCATTTCACAAAAGGGGTAGTTTTTTCCC 20581 CATTAGAAAGGGTATACTGAGACGTTCCTCGCGATGCACTTTCAAACGAAAGCTAGGCTG 20641 TGCGGGCAGAGAAAATATGTTCAGAAACGAGGGGAGATGGTGTTAATTACACAAACATCA 20701 CATTAGCACAGCATCGAGGTCCCTTAATCTGTCGATGCTTTCTGGGGAAATGGAATTGGG 20761 GTGACAGCGGCCTAGAAGTCAAAAAAAAAAATGTGGACAAACCAGAACGGGCTTTGTCGC 20821 GAGGTCTGCCCCTCAACTCCAGTTTGCTGGATTGCAAGTCGACCGGATGGGCAGCACACT 20881 AACTCTTCCCAGTCCACCCCCACAGTCGCGCCCCGGGAGCTGTGCAAGCCGGTGGCCGAG 20941 ACAGTGGCACTGGGGTGTTCGACAAGCTTGCTGGGCTAGCGCTCTCTCAGAGAGAAAGCC 21001 GCGGTGGGAAAGCGCAAGCCTGCAAGGATAAGCAGGGTCACTGAGGGCCCCCGGGGGCGT 21061 ACCACTCCCAATCAGTCCAGATCCTTGCACAGGACCTTCCTCTTCCCTGGGAGCAGACTT 21121 CTCTGTGGATCTGCTAGGAGTTTGGGGTAAGTGAAATCCCGTTTGCTCAGTCTCAAGTCT 21181 TCTCTCAGACTGTTTCCAGTTTGCAATGCAGAGTGAAAAGGGCAAGATCTCACTCACTGT 21241 TGCATACTATACAAACAACTTAGCACATAATAAATGTAGGGTGGAAACTTAATGTAACCA 21301 AAGGAAATGCTGGGCCAAGAAGGAAAATAGAAAATCTGCGCATCGAATTCCATAGCTCTC 21361 GACTTTTTTAGAAAATGAAAACGTTTAGTAAAAAGAGAGAAGGAAAACATGGGATTAAGG 21421 ACAGCTGAATCCTGCCTTGAGATTTATCCCCACCCCGACCCCCACCCCCGGAAAGCTCTA 21481 CCCAAACCCTCTCAAAGCTAACAGGCATCGCCCGCGGAAGGAGCAGGTGAACCTGCACAG 21541 AGGCAGGGGAGATGCCTAGTCTGCATTGATTTTTCCAGGGAAGGGTAGTCAGGTCTAGTG 21601 GACTCCTACTCACCCCCACCCCGGCGGAATTCTAAATCTGTGCAGATAACAATTTTAAGA 21661 TCTTTTCCCTGGTCCCATACTCCTATGAATCACCTTGATTCTGCTTAAGCGAGGGCCATG 21721 CCTTTGCATCAGCAAAGGAGATGGACTCCCTTTCCTGGCAAGCACAGTGGGAGGGCAGCC 21781 ACCCCCCATCCGGGAGCCAGGTTTCTGGGGGTTGCTCTGCGGGTGTGTGTGTGAGTTGGT 21841 TGATCTGAATTACAGAGCTCCTGCAGAGTGGTGGGAGGGGGTGGGGGTGAGGAGGAGACA 21901 AACAGGGCAATGACAGTGGCCCTGGAAGCCCTCTGAAATGGGACCCTGTTCCTGGGCAGT 21961 GCTTTCCCCCCTCTGGGAAGAAGGCAATGGATATGGGTTGTGAAGTCAACAAGAGGGCGC 22021 TGAAGGCAAAGTTTGTCAAAATCCTCAGATTTGCTGCACACCTAGGGAAAAGAGAGCGGG 22081 TAATTAATAATACCTTAACACCCTTGTTAGTAAGAGTCCCTCTTACAGATGAGAGCCCTC 22141 AAGTGTGGGGAGGCCAAAGAGACCCCAGCATCTCTACAACAACTAAGTAAATGAATTGAA 22201 TTCAGAAACTTGTTTCAAAAGTGCTCACCCCAAAAATATCTACAGCCAAAAAGCAGTAGA 22261 ATAAATTGTGCCACCACACAGTTTGTCACCTCCAGAACAATGCTATTCCATGTCTCAAAG 22321 CTCCTACAAGTGTTTCTTTAGAGAGCATCAACTAGGTTGTTTAAAGCCTGTATCCACCAA 22381 CCTAGGATCTTCAGGTCCCTCTCTGACCTTAAAGACAGGAAAGCTGTCCCTTCACACACA 22441 CACACACACCCCCACACACACACATACACACACACATACACACTGCTGTGTGGGGTTTGG 22501 GATGGGAGGGGTCCCACATAGTGCTAGTGGAATGGGAAAGTTATGTCTTTTCCACCAGTA 22561 GGAGTGACCTTTCTGACCTTTTTACAAAGGATGGTTCTAAAATAGATTCAAGCTTCTTGA 22621 TCAAGGCACTCCCTCTGTAGCCTGAATACAGTCTTGGATGTACCAGGGTAAATGGTCAAG 22681 CAATGCTCTCAATGAAGAGACAGAATTGACTTTTAAAATGTAAGGCTTTTGTCTGCTGGA 22741 AAAGGTTTCCAAAAGAAGTTGTTTTCATGTCCTCAAGCCAGTATTTAGCCCTAGGCACCA 22801 ACATACTTGGAAAACCTTAGAGCACTGTGGGCCAGCTACCACGGATGGCACTACTCTTGG 22861 GCCTGAACTAGCCTGAACAAACTTCAAACACTGCCATCCTCTCCACGTGCCAGCAGGTAC 22921 CTACACAGATCTAACGCTCCCTATCAGCTCTAAAGCACAATACCAGAAGTTCTCTGCAGT 22981 TCTGAGGCCCTTTGGAACACCTGATTCTCCCGAAGCAACAGGAAGGAGGGAGATAGGGGA 23041 TAAACTATATACCTGGAGTTCGCTGGGTTAGACCGCCAAATCACTGGAGTTTTACACCAG 163
23101 CTTCCTAAAGATAGGTCCTGTTCTGCCCTGCAGCCAAGCAGCTGTTTGTCTTTCCGGGGT 23161 TGAGAAGACCTGGAAGACCCTCAGGGAGAGAAGGGGGCTCCTCTGGGTGAAGTTTCCAAA 23221 TGAAGCCAACAAACGGCGCTATTGTCCCGCCATGGGTCCTAGGAGCCCCTGGCTCGACCC 23281 CAATCGGAGGGAGTCTTAACGCTCCTTGGAGCGGGGGTGGGGCATGGGGGGCGGGCAGGG 23341 AGGGGTGGAGGGGGAAGCACTTGGCCGGTTTCCTAGTCTGGGATCCAGAGCTCTTGGTAT 23401 TGCCCACCGAATCCCAGAGGCCTCTAAGGCTAAGAGCTCTGCTGAGCGCATTCTTTCCTT 23461 CCTGTGTCCCGATGTATTTTGCTTTTCGTTGTTGTTGTTAACAGATACAGAGAATACAGT 23521 TTGAAAGACTCTACAGTTCAATACCTTGCCATCCGGAAAGCCTCTGCGGAGACAGAAGCC 23581 ACTAAGATTAAGGAAGTACCCGCTCAGATACGTCTACAGACCCCCCCAATAAAAGGAATT 23641 TGGCATGACTGGTGTCTTAGTCTTCAGTGCGGGGAGGGTCTGGAAAGCTGGTTTCCTAGC 23701 CGCCCTGAATATCCTCAGGTCCACAGTGTTAACTCTGCTTGTCTACTGAAATAATAAATT 23761 GTAAGAAGAAAGTTTCCTATGAAGTGTTTTGAAGGGAATTCGGTGAAACACCTGTTTGGC 23821 TCTGTCTAGCCCAGCCCAGAGCCACCACACAGAGATCAGTTTAATCACAGTCACCGCAGA 23881 GAAGCCAAGAGGGTGAGAGAAGGTGTCTGATCCTACCAAAAGCCCTCAGCGAATCTGGCA 23941 GGAGGATGTAATGTGAAAATGAGCTCACTTAAACCAGGACAAGGTGTGATCAAATGACAA 24001 ACAGCTAATCTGCTTCGGGGAGGAGCAGGGACCCAGACCCACATCCTCTGTCTTGGGAAA 24061 TATGCAGCCCAGGCACTGTCACCACAACGGCTTACTTGGGTCCTAAAACTTGGAAAGAAT 24121 CCTGACATCTGTCAGGCTACTGGACGCCAGGCTGCGAGTCGGTTGGGTAGTTTCCTAAAC 24181 GGCTCTTCACTTAGCAACAGCCTCAGGGGCCTGGTGGAGGGGGGGGGGGCGACGAGAGAG 24241 GGCGTGAAGAGGGTGCTGGGAGGGCAAGGGCCAGCGCACATTGGTTTTTTTAGGGGACCA 24301 TGGGTGTGCCTGAGAAAATCTTGAGTCTGTCAGCCAATTTACCAAGAAAACGAGTGCAGG 24361 TGGGGCCTGCTGAAACATTCCCTGTTCTGAAACATGTTTTCCAGCTCAGCCGTGCGTGTG 24421 CGTGCACACTTACTAACATAGATCCCCACACAGTCTCCAGAGACTTTAAACCCCCTACAC 24481 CCACCTAATTTACTGATATCCTTGGCAGGAAACGGCTCAGACCAGGTAGAACACAAACTT 24541 AAAAGGAAGGAACCAGGACCACAAACTTTCTGGAACTGATGCCGGCTAGAAAGCAAGATT 24601 CGGGTGCCAACGGACCCTGTGGCCACAAATGGACAGAACTGGTCTGCGCCCTGCAAACGT 24661 TGGCGGCTTACTTTTTAATTAGTGCATGAGAGAAAGAGACAGAAGCCGGGAGTACAGGCA 24721 TCCCAAATCACCTCATTTAGTATTATCGCTGAAAAGCTAGACGCCTGGAATTCTCTCTCT 24781 CTCTCTCTCTCTCTCTCTCTCTCTCTCTCTCTCTCTCTCTCTCTCTCTCTCACACACACA 24841 CACACACACACACACACACACTTAGGGCATGTGCTTTTCTTTAGGTTTCCCACCTTTCCA 24901 CCAAGCTAAAGGAGGAAGCGGGAATCCAGTACTGCGCGGACACCTTATTTCGGTTTCTGC 24961 CACATCTGCCCAGATGGCTTCCTATTCCCTCCAGCAGCCTCCGAGGGCTTCAAGAGTTTT 25021 CAAACGCCATGCCACTTGCTTGAGCTGGCAGCGGAGAATGTCCGGGTTCCCTCTCCGGGC 25081 AGCGGAGGTTTCTCTAGATGGTAGTCCTTGGCAGATAACCCAACAAAGTGCCTGAAGGTC 25141 CCACTCTGCTTCTCCAGCCTTCGGTCAGAAAGGCTGTAAAACAAGAAAGGAGAAGCAGAG 25201 TCTGTTAGCACAGCCCGCTTGGGGGGGGGGGGGAGCGCGGCTGCGTGCCGAGAGCCGCTC 25261 GCAGCGGTCGTAGCAGGTGCGCGGTCGGAGGAGGGGCGGTGCACCCGGAGCAAGGCCTAG 25321 GCTCAGCTCCGGCCACCCCCCCCCAAACTTTTCTTTTACTGCGAGAGGCTCCGTGTTGCC 25381 ATGAAAACGGATTTTCCAAGGGGCCCGACCCTCCCCCTCAGGAGCATCGGACTGGAAAAC 25441 CCGAAATGGCCAGACAGCAGAGACAAAAAGGGGCCTCTAAGCAGGGGCCCAGAGCGAAAA 25501 AGTGAAGGCTCGTTCGCCCTTTGCATCCTAAACCCGCTCCCCAGGAGCTACGTTCGAGCG 25561 CCTCCCCCGGGCACAGGCCGAGGCCACTCCCCACTCGCACTGGCAAACCCCCCATGGAGC 25621 CTCACAGCCAGGTGCGCTAGCAGTCCCGCTCCCCGGCCCCCGCCCCTCTCTCGGTCCTGC 25681 ACGCAGTGGCCCTCGGGGTGGAGGGTTCAGGCAACCGGCTACCATCAATTCCAAGGAAGG 25741 CTTCTCAGTGGCTCAGCGTAGCGCACCTCCGTTATACACTCAACGTCCCAAGCATGCCTG 25801 GACTGCTAGAGGTCCCCAGTGTCCTTAGCGCTGGCAACAGCAGGGTTTGGAGGTCCTCCG 25861 TGCAGACGGCGGGCGGGGAGCTTTCCAGGTGGGCCATGGGATGGGGCGCTGGTCAGCCTA 25921 GTCGAGTCTGCAAAGTTGTTTCCGAACCCCGCGGGCGCTCCTCTGCCCCTCCTCCCCCCA 25981 GTCTCGCTTGCTCGCGCGCCCTCTTCCCCCCTCCTCCCTCACCTCCTCGGGTTTCTCGCT 26041 CTTTCAAACTTTTTGAGACCCTAATTGGTGGTCTCAGAGCGGGTCTCGTGGACTCCCGCC 26101 TCCTGAAAAAGCTCTCATCACGTCACTCGACCCAAGAGGGGTCTGGAGGGTGAACGAAAG 26161 GGCTCCCCGAACTTTTTTTTCCAAGCTGGGCCACAGGGGGCTCGGTGTTGATTGGCCAGG 26221 ACTCATCACGGCGAGCCTGTCAATCACGAGGCCCACGGGTGGTGAGGGGGCGGACCGAGC 26281 CCCAACCCTCCCGGATCTGAGCAGGTATATAAGGGACCGGGCAGGCTTCCCGGGCAAGCT 26341 GCGAATGATGCGACCAGCTCGAGAGAACGCATCAGGTGAGAGAAGCCCGCGAGTTCCCGC 26401 CGACTTTGCGCGGAGCCCTTCGGCTAGCCTGCCCTTCCCGCCTGAGCTGCCGGCCCGCCG 26461 GCCCCTGAGCACCACTCCGACGGAGTCCCCCGGCCTTTTCACGGTGGCCGCTCCAGCCCC 26521 CGGAGCGCCGTCTCCTCCCGCCACGCTGGCGCACCTCCTTCCCGCCCCGGCGATGTACAG 26581 CCTGCTGGAGACTGAACTCAAGAACCCCGTGGGGCCGCCCACCCCAGCCGCGGGCACCGG 26641 CGTCCCCGCAGCTCCCGGCGCTGCAGGCAAGAGTGGCGCGAACCCAGCCGGCGGAGCGAA 26701 CGCAGGCAACGGGGGCAGCGGGGGCGCGAACGGCGGCGGTGGTGGTGGTGGCGGCGGGGG 164
26761 CAGCGACCAGGACCGCGTCAAGCGACCCATGAACGCGTTCATGGTGTGGTCCCGCGGGCA 26821 GCGGCGCAAGATGGCCCTGGAGAACCCCAAGATGCACAACTCCGAGATCAGCAAGCGCTT 26881 GGGCGCCGACTGGAAACTGCTGACCGATGCGGAGAAGCGGCCGTTCATCGACGAGGCCAA 26941 GCGACTGCGTGCGGTGCACATGAAGGAGTACCCGGACTACAAGTACCGGCCCCGCCGCAA 27001 GACCAAGACGCTGCTCAAGAAGGACAAGTACTCGCTGCCCGGCGGCCTCCTGCCCCCGGG 27061 CGCCGCCGCAGCCGCCGCCGCTGCCGCCGCAGCCGCCGCCGCCAGCAGCCCGGTGGGCGT 27121 GGGCCAGCGCCTGGACACGTACACGCACGTGAACGGCTGGGCCAACGGCGCCTACTCGCT 27181 CGTGCAGGAGCAGCTGGGCTACGCGCAGCCCCCGAGCATGAGCAGCCCGCCGCCGCCACC 27241 CGCGCTGCCTCAGATGCACCGCTACGACATGGCCGGCCTGCAGTACAGCCCCATGATGCC 27301 ACCCGGCGCCCAGAGCTACATGAACGCCGCCGCCGCCGCTGCCGCCGCCTCGGGCTACGG 27361 GGGCATGGCGCCCTCCGCCGCGGCCGCCGCCGCCGCCGCCTACGGGCAGCAGCCCGCCAC 27421 CGCTGCCGCCGCGGCCGCCGCCGCCGCCGCCATGAGCCTGGGCCCCATGGGCTCCGTGGT 27481 GAAGTCGGAGCCCAGCTCTCCGCCGCCCGCCATCGCTTCGCACTCGCAGCGCGCGTGCCT 27541 CGGCGACCTGCGCGACATGATCAGCATGTACCTGCCACCTGGCGGGGACGCGGCCGACGC 27601 CGCTTCTCCGCTCCCAGGCGGCCGGCTGCACGGCGTGCACCAGCACTACCAGGGCGCCGG 27661 GACTGCGGTCAATGGAACGGTGCCGCTGACCCACATCTGAGCTCCGGCCCGCGCTCCCAG 27721 CTCTTCGCCCCCACCCCGCCCCACACCCCCCACGTTGGGACGCCTTGGCCCCTCTCCCTC 27781 CCCCCCCCCTAACGTTACTGGCCGAAGCCGCTTGGAATAAGGCCGGTGTGCGTTTGTCTA 27841 TATGTTATTTTCCACCATATTGCCGTCTTTTGGCAATGTGAGGGCCCGGAAACCTGGCCC 27901 TGTCTTCTTGACGAGCATTCCTAGGGGTCTTTCCCCTCTCGCCAAAGGAATGCAAGGTCT 27961 GTTGAATGTCGTGAAGGAAGCAGTTCCTCTGGAAGCTTCTTGAAGACAAACAACGTCTGT 28021 AGCGACCCTTTGCAGGCAGCGGAACCCCCCACCTGGCGACAGGTGCCTCTGCGGCCAAAA 28081 GCCACGTGTATAAGATACACCTGCAAAGGCGGCACAACCCCAGTGCCACGTTGTGAGTTG 28141 GATAGTTGTGGAAAGAGTCAAATGGCTCTCCTCAAGCGTATTCAACAAGGGGCTGAAGGA 28201 TGCCCAGAAGGTACCCCATTGTATGGGATCTGATCTGGGGCCTCGGTGCACATGCTTTAC 28261 ATGTGTTTAGTCGAGGTTAAAAAAACGTCTAGGCCCCCCGAACCACGGGGACGTGGTTTT 28321 CCTTTGAAAAACACGATGATAATATGGCCACAACCATGGTGAGCAAGGGCGAGGAGCTGT 28381 TCACCGGGGTGGTGCCCATCCTGGTCGAGCTGGACGGCGACGTAAACGGCCACAAGTTCA 28441 GCGTGTCCGGCGAGGGCGAGGGCGATGCCACCTACGGCAAGCTGACCCTGAAGTTCATCT 28501 GCACCACCGGCAAGCTGCCCGTGCCCTGGCCCACCCTCGTGACCACCCTGACCTACGGCG 28561 TGCAGTGCTTCAGCCGCTACCCCGACCACATGAAGCAGCACGACTTCTTCAAGTCCGCCA 28621 TGCCCGAAGGCTACGTCCAGGAGCGCACCATCTTCTTCAAGGACGACGGCAACTACAAGA 28681 CCCGCGCCGAGGTGAAGTTCGAGGGCGACACCCTGGTGAACCGCATCGAGCTGAAGGGCA 28741 TCGACTTCAAGGAGGACGGCAACATCCTGGGGCACAAGCTGGAGTACAACTACAACAGCC 28801 ACAACGTCTATATCATGGCCGACAAGCAGAAGAACGGCATCAAGGTGAACTTCAAGATCC 28861 GCCACAACATCGAGGACGGCAGCGTGCAGCTCGCCGACCACTACCAGCAGAACACCCCCA 28921 TCGGCGACGGCCCCGTGCTGCTGCCCGACAACCACTACCTGAGCACCCAGTCCGCCCTGA 28981 GCAAAGACCCCAACGAGAAGCGCGATCACATGGTCCTGCTGGAGTTCGTGACCGCCGCCG 29041 GGATCACTCTCGGCATGGACGAGCTGTACAAGTAAAGCGGCCGCGACTCTAGATCATAAT 29101 CAGCCATACCACATTTGTAGAGGTTTTACTTGCTTTAAAAAACCTCCCACACCTCCCCCT 29161 GAACCTGAAACATAAAATGAATGCAATTGTTGTTGTTAACTTGTTTATTGCAGCTTATAA 29221 TGGTTACAAATAAAGCAATAGCATCACAAATTTCACAAATAAAGCATTTTTTTCACTGCA 29281 TTCTAGTTGTGGTTTGTCCAAACTCATCAATGTATCTTAAGTGAAGTTCCTATACTTTCT 29341 AGAGAATAGGAACTTCCCAATATGCTGGAGTCCTGATAATTTTGCGGGTCCGAAATTTGA 29401 AGTGCGGGTTGGGAATGGGGATAGTCTTAGTTTTATAACTGAGTTTGCAGACCTGGGCGG 29461 GGGCGGGGGGGGGAGCATGATGTAAAATGATTTCCTTTAAGCTATTGGAGATCCAAGCTA 29521 CAACCGCCCCCAATACTCTCAAGTGGGAACTAGTCTCAAGTGGAGAAAACCTCTGTCTTC 29581 TGGGAGCACCTACCCAGAGGCAGCCCACGCTACCTGAGCCGCACACCCCACGCGCAGTCT 29641 TGCACTTGTTGATTGCGGTGGTGAGTGCCCGGGATCGAGTGTCTGCTTCCGCAGCGATGC 29701 AGAGCGGGAGGCAGACACCTGCGGCTGCGGAAGACTTACCAAAAGCTGAGCACTTTTACC 29761 TCTGTGCACTGGGGCGGGAAGATACTCGCTGAGTTGAAATATTTACAAAACGAAAGCCCT 29821 TCGGAAATCTCAATGATCGTCTTTAAAATAAATGGTTACAGACGTCCCCCTTAAGGAAAT 29881 GAGTATTTCCACCGCGGTAGCTAACGCGGGAGAGTGAGCGCTCAACACCGAAAGATTGGA 29941 TTTTAAAAGTTAGATGGTAGATGGGTCAGAGTTGTACAACTCCTCCCAAGGCTTGCTTTG 30001 AACTTGTCAAAGCAGTTTACTTACTTGTTTTCTAGAAAGAGGCTCCCTCTCATGTCGAGC 30061 CTAGAGCTTGGTTTTTACTTCCCTTTTCCTACTTTACTCCATCCCATAACCACTGACCAT 30121 TCCATGGTACCACCCCTAAGGGTCCTTCGAGTCGTATGGAGCCCCAGCGTAGCTGTTTGG 30181 AGCCTAGGAAGGAAACAAATAGAACCCGTGTTTTTAACCCCCTCCTCAGCTGTTTGAGGT 30241 GATTTCCTTAATCCCTTTAATCCTGTTACCTCGCTATCCCAGAATAACTAAACTCGGCTT 30301 TTCCAATAAGCTTGGCCTGGTGCACTGTTTTTATCCTCATCCACTCCTGACCACAGGTTT 30361 CTTTGCTTGAGGTCTGTTGGGACCATAGGCGACAGTCCTCTGTAAATTGGTCCGAATTCC 165
30421 ACGTTTCCAGCCAGCTTCCTTTGTGTGGACCAGCCATTCACCTGCGGGTCTGGGCGGGCC 30481 CCAGGGGGAGGGGCGGGGGTTTCTGAGGCTTCGGGGGTCGTGCCCGCTTTTTACACATCA 30541 CTGCAGATCCCGGAGACTAAAGCCACCCCGCCTCTCCAAGAAAGGTCGGGTCTCCCTGGG 30601 AAGAAGTTTGCACAGCCATCCCCACTCCCGACACACACCCATCCCCCACTGTCCAGCATG 30661 AGGTGGGAGGGGGCCAGGGTTGATTTACAGATGCAGTCCCATCGCCCTCTTCGCAGGGCT 30721 CTTCCTGGGCCTGCTGTGTTCAGACTGGGCCTCTTCCCTCCTCTCCTCATCCTAAAGCTA 30781 TAGACCTCAAGCCCCTCTTGACCGATGGGAGCTTCCATCCTTTACTGCAGTGATTGGAGC 30841 AAAGCGGCAGGCTGCCTGGTAGGCGAGACTCTGCCCCCATCCGCCCCCCCCCCCCCGCCA 30901 AAGTCCCAGTAGGAAGCTTGTCATACCTACCACCATATTCCAATTTTTCGGTCTTTGGCC 30961 TTAAATAAAACATCTGCCTAAGGTTATTCTAGATTTTGCTCTGAAATGACTCCAGGAAAG 31021 GGTGGGGTACAAACAAGCAATGCCCCCCACTGAGGGGGCGGTGGAAGAGGGGGAATAAAC 31081 TCTGACACTGCTTCTTTGGAAGAGCTCATTGTGCTGCTGCCTTCTTAGAAGTGCCAGGGT 31141 CACCCTTTCAGCATGTCGCACATACTGCACTCTCTCTCTCTAAATATCCATGGCTTTGTG 31201 CTCCACTTTGTCTGATTTCCAAATTGCCTTTAAGAGAAGATTTCATCAATCAGTACATTT 31261 GGAGGATTGAAATACAGATGAAAGATAATGTTTTTAAACTTCAGCAATAGGGAAGAGAAC 31321 CAGACTGAAGTGGGATGAAGTTTCCGTTCTCTAACCCGGAGCGCTCACACCAGAATCTGT 31381 CCTCCTCTACACTGCTCTGTCTACCAGGGACTGCTCCTTAATCTGTTTTTATAGAGTTGC 31441 GTTTCTGATAAGGCTTAATTCAAAACAGCAACAAAATCAACACGTGCCCAGACTCCAATC 31501 CCTTCGGCCTTCTCTGGCAGGTTTTGTCACCAAAAGGGGAAATGACAGTGTTCAAGGTTG 31561 CCTAGCTAGTTCCAAATACAGCCTAGTGGAGAACCAGATTCTCAGGTCCCTGGGAAGGAT 31621 TCATTTTGCTCTGCCTGCCACAAAAGGAAGACAGAAGCCAAGGAAAGCTTAGCCCCAGTT 31681 GCCTCTGCAGCCAAAATCCACCAAGACCTGTGGAGAAAAGAGGTAATTCTGGAGATAGTA 31741 GGCTCTGCTTGGGCCATGGCATGTACGAGGAAAACAGATCCTTACCGACACAAAAGACCA 31801 ACGTCTTTTAGTTTCCCTGCCAGGACCACAGCTATTTTTAGTGGGAATTATCCTGCTAGA 31861 GATTTGCTTTCCTGTCCTGACACTACATCATTTTCCTTATTAATTCCCCTCCTAGACGTC 31921 CACATGGGTGATGCTGGGTATTCAAAGTGCAGGCAAAGAATGCTAAGCAACCTTGTCTGT 31981 CATAGCTGTTCTTAATACTAGTTGTCATAGCTGTTCTTAATACTAGTTATAACCGCCTCA 32041 CATGTCCAGGAACCAGGAAAGAGTAAGTATGATTTTTTTTTTTAAAAAATGGCTGCTGTG 32101 TAATTATCACAAACATTTATGGAGTAATTTAAATAACACACTAAACCATCCGAATAGCAA 32161 GAGAACAATATGTCAGAACTCTTATTTAGCAAGTATGACCTTGACTATTTAAAAATAAGT 32221 ATATATGACAGGAAACATGCAAAAAGGGCAGTAGTCTGAATACTGGCAAGTCTTCTTTAT 32281 ACTCTTTAATTTTCTCTAATATGATTATATTTCTGTTATCAGGAAAAAAGGCAGGTTTAA 32341 AATATATGCCTTCCCCATTATGAACACTGAAATGCGTCTCTTGTAAAAGAAAGTTCAAAC 32401 CTTTCCTTCAGCTGTTTTGGGGGATAGGAGGGGGACGGGGAGCCCCAGGAGGAATAATCC 32461 CCTAGCTTTTCTCTAGGGCTGATGCTGCAGTTAATAGAGGCAGGAGGAGCTCACCAGAGC 32521 AACCCCCTGCCAGGTCCGCTCCTCAGAAAGCAGGAACAGAGGCAGATTAGTGACTGGCTC 32581 ATCATTCTTCAGACTCACTCACTGGAGAAAAGGACCTCCTTTGGAGTCATGCCTAACAGT 32641 CTTGGATTGCAAGGAGCAGGGGCCACCAGGGTCCCCCAAGATCTATGGTCACAGGTGGTC 32701 CAGCAGCTCGTGCTGTTTTCTTGGGAGAAGCAAAGAGCAAGTCTGGGGATGGGATAGTGA 32761 TTGGAGGAAGAGGGGACTAGGTGAGCAGCATTTTCTTCCTGTCAGCCGCCACCATCTTCA 32821 TCCCTCCCTCCATCTCTGCCTGAGCTCCCACCTGTTCTCTGTGCTTTCATTCTGCTGCCT 32881 CTGTCCCAAACTAGCTCCCTAGCATATGCGCATAAATTCCATTATGCTATTTCCCTTTGC 32941 AAATAAGTTTCTGCTAGTTTCCAGGACAGAGATTTTAGGGGGAAAATCCTAACTATTGCA 33001 AAGTTTTGTGTGATCAGTTTCTCTCTACCTTTCTGAGCTTAACTACTTTTATTTTACATA 33061 TTTTCACCTTGGTCCCATTTCCTTCAGTCCATCACATGAAGCACAGATTTTCTTACAACT 33121 AATAAAACCTACTACCACAAATGTATGTTCCTCCCCACTGGGGACAGGGTGGCGAGGTGG 33181 CTTTTGAGTTTGATGTTTCCCTGTGCTTGGAAAATTCTTACACTCTTCCCTCAGACTGTC 33241 CAGCATTCTTCAGAAAAATCTCTCAGTTCAAGTATCACCTCTTCTAGGAAGCTTTCCATG 33301 ATGCTTGCATTTTAAATATGTCCCAGTTCCTTGTTATGAGTTCTCTAAGCACTTAGTGTC 33361 TCTCCTAAGCATTGGTGTGTGTGTGTGTGTGTGTGTGTGTGTGTGTGTGTGTGTGTCTTA 33421 GGGTTTTACTGCTGTGAACAGACACCAAGACCAAAGTAATTCTTACAAAGACAATATGTA 33481 ATTGGGGCTGGTTTACAAGTTTAGAGGTTGAGTACATTAACATTAAGGTGAGATCATGGC 33541 AGCCACAGACAGGCATGGTGCAGGCAGAGTTATGGCAAACCAAAGACTGACTACCAGGCA 33601 TCTAAGATGAGGGTCTTAAAGCCCACACCCACAAGGATACACCCACTCCAACAAGGCCAC 33661 ACCTACTCCAACAGTGCCACACTTCCTACTAGTGCCATTCCCTGGAACAAGCGTACACAA 33721 AGCACCACAGTAGTCATTGTATATTTGTTTGTGTGATGCTGTCTGCTGTTTCTATCCAAG 33781 TGTGTAAGTTCCATGAGGGCAAAGACTGTGTCTAGTTTGCACACCAAAACACAGCATGAT 33841 GTTTCTCTCATTGTAGTCTCTCAAGAATATATATGAAGTCTTACTGCATATTTAAGAAAT 33901 GGTAGTTCTGTGGGCTTGAAGCTGATGTGTGCTAGATCTCAATGTGATAGCCATACTGGT 33961 GTATACATGTAATCCCTGTATAGAAGATGAGGCAGAAGGATCAGGAGTTCAAGGCCAGTC 34021 TTGGCTCCATAAGAGAACTGGCTCTAAAAAGTTGTCCTCTGACCTCCACACATGCATGTT 166
34081 GGCATGCACTCACCCCACCACGCACAATGACACAAAATAAATACTTTTAAAGCAGATAAA 34141 TTAGACTTTATCAGAATACGAGATGTAAGGTTTTTGTTTGTTTGTTTGTTTGTTTTGACT 34201 GATTGGCATAATTTCAGAGGAAAATATCAAGAAAGTGCAAAGAAAAGCAACCCACAGAAT 34261 GGAAGAATGTGTGTGTGTGTGTGTGTGTGTGTGTGCGTGCACACACACACACAAGGCATT 34321 TGTTCAGGGTTTATATTTAGACCATGTAGAGACTCCTAGAATTCAACAATCATAAGACAA 34381 CTACTTTAAAACAGGCAAAGCGACTTGAATATGCTTTTTTTTTTCAAAAGAATGTATACA 34441 AATGGACACTAACACATGGACAGATGCGTGGCACCACTAGTCCATAGGGAAATGTAAACT 34501 GACAAGCTACCACTTCAACACACACATGGTTATAAAAATAGTTTACAAATAGACAACACA 34561 AGTCTTGCTGACAATGTGGAGAAACTGTAGTCCTTACATTGTACTGATGAAAGTGGGAAA 34621 AGAAAATCTTGTAAACAGACTGGCAGTTCCAGGGGTTAAAACAAACTTCACATAGCAGTC 34681 AGCAATTCCACAGGGAGGAAAAGTTAGAAAGGGGGAATACATATGTCTGTACAAAACATA 34741 TATATCCATATTCTATCATTATTATTCTTATTTCCCCAATGGGGTAAACAATGTACATAT 34801 TTATTTGCTGATTAGGAGATAAATCTAGTGTGGGTTTTTAGGAGAGAGAGAGAGAATTGA 34861 AAACAACAAACGGAATGTCAGAGTCAAAGGAACCCAAACGTCAGTTAAAGCAGCTCCTGA 34921 TGATCAAATATGGAATCATCTGAACAATAAAATGATTAAAATAATACTGAACTATAATCT 34981 AAAGTATAAACATTTCTAGGTGCCCATACTGATAGAAATCAAAACAACTAAATAAACAGG 35041 TGACAAGATGAGTGCAGGCAGTGAAGGACAATTCTAAGTAATGGTTACCACTAGTCTGAC 35101 TTCAGGGAGGTGGAGCGTAGCCCCCTAACTCCTAAATTCAGGCTCATGCTTTGACTACTT 35161 CTCAAAGAATAGAATAAGGGAAGGGAGGACGTGTAACATTGTACTAAAAAAGTATAACAA 35221 ACATAACCTCCACCATACTGTCAGAGTCAATATCAATTATTGACAGATAACATACTGTTC 35281 TTCATGAAATGCACTGGAAACCGAGAACCATACTTCTGTTGTCTAACCTCCCCAAAATTC 35341 ATAATCCCAAGCTAATCCTAGTAAAACATTAAATCTTTTTTATAGATTTATTTATTTGTA 35401 TGTGAGTACACTGTAGCTGTCTTCAGACACACCAGAAGAGGGCATCGGATCCCATTGCAC 35461 ATGGTTGTGGGCCACCATGTGGTTGCTGGGAACTGAACTTAGGACCTCTGGAAGAGCAGT 35521 CATTGCTCTTAACCACTGGGCCATCTCTCCAGCCCAAAACATCAAATCTTTTGTTGGTTT 35581 TGTTTTGTTTTGTTTTGTTTTGTTTTGTTTTGTTTTGTTTTTGAAACAGAGTCTCACTAT 35641 GTAGCCCTAACTGTTCAGAAACTCACTATATAAGGCAGGTTGGCCTCAAACTCACAAAGA 35701 TCCTCTGCCTCTCAAGTGCTGGTAAATATCAGATCTTAATTCTAAAAAATGCCTGGTCAA 35761 TACTCCTCCAATTTGTTAAAGTCATCAAAAATAGGTAAATAAGCTGTCACATTTAAGACT 35821 TACCTTGGGGGCTAGAGAGATGGCTCAGCGGTTAAGGGCACACTGACTTGCCCTTAACTA 35881 AATCTTAAAAAAAAAAAAAAAAAAGACTTACCTAAAGAAATAAGACTACTACTAAATGTA 35941 TAAGGTATCGATTTTATTAGTAATAATATGTTAATATTGGTTTGTTAATTATTACAAATA 36001 AGAGCTGGGAGTATAGCTCAGTTGCTAGGGTACTTTCCTAACTTGAGCAAAGTCCTGAGT 36061 TCTGTCCCCAGTACTGAATAAACAAGACATGGTGGAATGCATCTGTAAACCCAGCATTGC 36121 AGAGGTTGCAGTAAGAGGATCAGAAGTTTAGGGTCATCTTCAACTACACATAGGAAGTTT 36181 AAGACTATCTTGGGATACATGAGACCCTGTCTCAAAGACAGAAAAAGATAGGTAAAGGAT 36241 AGAGTCAGGGAAAAAAAATGAAATACTTTAATGAAAGTTGTTAATTATAAGGAAATGTTC 36301 AAATATGGCAACTTCTCACCCAATCTCCCCAATTTTCATATAAATTCAAAACTGTTCCAG 36361 GAAGTAAAGTTTATTTTTAAAAAAGAATAGGAATAAATTTCTGGTACACACACACACACA 36421 CACACACACACACACAAAATGAATGACAGGTGAACCTTGAAAGCATGCTAAGTGAAAGGG 36481 TACAACGACAAAATGCTATATGCTGTATGATTATCTACTGACAAGATCTCACTCTGAAGT 36541 CTAGGCTAGCCTGGAAGTCACTAATGTTGCCCCAGCAGGCATCAGACCATGAAAACCCTC 36601 TTGCCTCAGTTTCCCAAATGCTACGACCACAGGCTTGAGGCATCATACTCACCTCTGTAT 36661 TACTTCATTTAGAGGACTGCAGAATTAGCTTAATCTTTAGAGTCAGAGAATTGATTACTG 36721 TTGCCCAGAGCTAGAGGGACTATTAGAGTCAGACATCCAGTTTAGCATCTTTATTTGGCA 36781 CATAAGGCATTAGAGACCAGAATGAAACAAGGAAACCAAAATAACATAGCTTGCTAATGC 36841 CAGAGTAAATGAACTAGAACTTCCACATTTAATCACTTTCACCGTAAGTAGCATCCCAGT 36901 TGGAGAATAGATGGAAAAGGGAAGGCCATTTCCAGAATGTTGCTAAAAGCCTAGAGTTCT 36961 ATCCATTTCCAGTTTGATTCTTGTGGTGGCAGAAGAAATCGTTTTGAAATACTTGAAATA 37021 CGTTTTAAAAGTGAGTGCACTTATGTTAAAGAAAATTCTCCTTTGCAAAAACAAGCTGCT 37081 GTTGGAGGGAAAATAATGAGGATGTAGTTCAGTGGTAGCATGCTTGCTATACACTCTTTT 37141 ACATGTGGACTTTATAAGTCCTAATGAGCGAAAGTGAATATGGATAATGCTTACTAAAAA 37201 GTATTCGAAACATTTGATGAACTTTAATATTAATTTAAGTATAAAATTATAAACATTTCT 37261 GTGGTCCTACAAGTTTAAACCAAGCGACAGTGTGTCACTCTTGTGTCTGTGACTTGGTTG 37321 TCACCTCAAGCAGTTGCTCCCATAATCATTTGATGGTAGGGCTGGGTAGGCTCATGGATG 37381 ACTACATGGAAGAGAGAGGCCTCGGGGGCCTGCATCACAAAGGAGTTGATGCTGCAGATA 37441 GACTCAGACTGGGTACCTTGCACACCACACCCATGAAACGAGACTTGACGTGGGTCCTTG 38501 ACAGGTGTTG
85,122,667bp 85,122,751bp 167
Appendix III: SOX3 antibody specificity
Please see next page for figures. A
NOTE: This figure is included in the print copy of LV theLV thesis held in the University of Adelaide Library. LV LV
The SOX3 antibody used is specific to SOX3. A: SOX3 antibody staining on a 12.5 dpc WT tissue section at the plane of the telencephalon showed a level SOX3 expression at the VZ. B: Identical antibody staining at a comparable section from 12.5 dpc Sox3 knockout (KO) mice showed a lack of positive signal, indicating the antibody used is specific to SOX3. Immunofluorescence. Coronal sections. Top to bottom: dorsal to ventral. Scale bar: 4006µm. Figure from N. Rogers (unpublished data). 168
Appendix IV: An outline of the microarray experiemnt
Please see next page for the flowchart.
NOTE: This figure is included in the print copy of the thesis held in the University of Adelaide Library.
A schematic flowchart showing the process for Affymetrix GeneChip® Gene 1.0 ST Array System for mouse. Figure from Data Sheet, Affymetrix GeneChip® Gene 1.0 ST Array System (2007). 169
Appendix V: The pairwise MA plots generated from Partek® analysis
Please see following pages for MA plots. M (Intensity ratio) WT replicate 1vs.WT replicate 2 WT A (Average intensity, log 2 )
M (Intensity ratio) WT replicate 1vs.WT replicate 3 WT A (Average intensity, log 2 ) M (Intensity ratio) WT replicate 2vs.WT replicate 3 WT A (Average intensity, log 2 )
M (Intensity ratio) WT replicate 1vs.WT Sr/+;replicate Nr/+ 1 A (Average intensity, log 2 ) M (Intensity ratio) WT replicate 1vs.WT Sr/+; replicate Nr/+ 2 A (Average intensity, log 2 )
M (Intensity ratio) WT replicate 1vs.WT Sr/+; replicate Nr/+ 3 A (Average intensity, log 2 ) M (Intensity ratio) WT replicate 2vs.WT Sr/+; replicate Nr/+ 1 A (Average intensity, log 2 )
M (Intensity ratio) WT replicate 2vs.WT Sr/+; replicate Nr/+ 2 A (Average intensity, log 2 ) M (Intensity ratio) WT replicate 2vs.WT Sr/+; replicate Nr/+ 3 A (Average intensity, log 2 )
M (Intensity ratio) WT replicate 3vs.WT Sr/+; replicate Nr/+ 1 A (Average intensity, log 2 ) M (Intensity ratio) WT replicate 3vs.WT Sr/+; replicate Nr/+ 2 A (Average intensity, log 2 )
M (Intensity ratio) WT replicate 3vs.WT Sr/+; replicate Nr/+ 3 A (Average intensity, log 2 ) M (Intensity ratio) Sr/+; Nr/+replicate 1vs. Sr/+; Nr/+replicate 2 A (Average intensity, log 2 )
M (Intensity ratio) Sr/+; Nr/+replicate 1vs. Sr/+; Nr/+replicate 3 A (Average intensity, log 2 ) M (Intensity ratio) Sr/+; Nr/+replicate 2vs. Sr/+; Nr/+replicate 3 A (Average intensity, log 2 ) 170
Appendix VI: Microarray CEL files and gene lists
Due to the vast amount of data, a digital copy of the microarray CEL files and gene lists generated from statistical analyses are recorded in the DVD and enclosed below.
NOTE: The data DVD is included as file 03Appendix6. Appropriate software is required to access the data 171
Appendix VII: Microarray – a table of genes with a fold-change in modulus of 2 or above
Fold-Change(Sr/+; Direction of Gene symbol Referece sequence Probeset ID Nr/+ vs. WT) fold change
Fmo1 NM_010231 10359571 -6.84214 down Adamts3 NM_001081401 10531177 -6.56237 down Hmgcs2 NM_008256 10494643 -5.43513 down Cubn NM_001081084 10480155 -5.37619 down Cpne4 NM_028719 10588380 -5.20237 down Adamts3 NM_001081401 10531193 -4.87371 down Pla2g7 NM_013737 10445293 -4.72728 down Bmp5 NM_007555 10587231 -4.56046 down Adamts3 NM_001081401 10531191 -4.25281 down Adamts3 NM_001081401 10531183 -4.18264 down Trpm3 NM_001035244 10462039 -4.17463 down Adamts3 NM_001081401 10531179 -4.12323 down Slc39a12 NM_001012305 10469389 -3.96219 down Alcam NM_009655 10439895 -3.825 down Npr3 NM_008728 10427796 -3.80201 down Kcnc2 NM_001025581 10366391 -3.7406 down Wnt2b NM_009520 10500938 -3.70574 down Adamts3 NM_001081401 10531197 -3.69273 down Adamts3 NM_001081401 10531203 -3.65927 down Rspo3 NM_028351 10368495 -3.6154 down Vtn NM_011707 10379190 -3.57591 down Edaradd NM_133643 10407782 -3.56781 down Col8a1 NM_007739 10440091 -3.54137 down Lum NM_008524 10365983 -3.50909 down Rbp4 NM_011255 10467319 -3.46993 down Adamts3 NM_001081401 10531195 -3.46308 down Adamts3 NM_001081401 10531187 -3.44871 down Cdh20 NM_011800 10348999 -3.44208 down Nebl BC119802 10480288 -3.36191 down Prl8a2 NM_010088 10404316 -3.27293 down Adamts3 NM_001081401 10531175 -3.23938 down Calml4 NM_138304 10586118 -3.15671 down Adamts3 NM_001081401 10531173 -3.08065 down Dab2 NM_023118 10422728 -3.04962 down 172
Bmp6 NM_007556 10404686 -3.04961 down Adamts3 NM_001081401 10531189 -3.02595 down Adamts3 NM_001081401 10531166 -3.01937 down Trhr NM_013696 10423953 -2.98524 down Ptgds2 NM_019455 10545101 -2.97438 down Adamts3 NM_001081401 10531181 -2.96932 down --- 10466886 -2.91756 down Hpgd NM_008278 10571840 -2.86651 down --- 10462086 -2.85107 down Pygl NM_133198 10400844 -2.84676 down Dpy19l3 NM_178704 10562500 -2.83306 down Plekhg1 NM_001033253 10367673 -2.79579 down Adamts9 NM_175314 10546450 -2.77026 down Gja1 NM_010288 10363173 -2.74464 down Adcy8 NM_009623 10429029 -2.74253 down Pdk1 NM_172665 10472846 -2.72742 down Pdgfra NM_011058 10522503 -2.71736 down Leprel1 NM_173379 10438753 -2.7065 down Ociad2 NM_026950 10530615 -2.70531 down Wnt1 NM_021279 10426637 -2.68569 down Dcn NM_007833 10365974 -2.65362 down Tmem132c ENSMUST00000051260 10525932 -2.64671 down Epha7 NM_010141 10503659 -2.6403 down Msx1 NM_010835 10529651 -2.63885 down Adamts3 NM_001081401 10531185 -2.62397 down Fibcd1 NM_178887 10481566 -2.61436 down Lhfpl3 NM_001081231 10520043 -2.60769 down Clybl NM_029556 10417167 -2.6044 down Neil3 NM_146208 10578690 -2.55883 down Slc38a11 NM_177074 10483199 -2.55297 down Lamb1-1 NM_008482 10395163 -2.55199 down Dpy19l1 NM_172920 10591816 -2.51811 down Ddn NM_001013741 10432278 -2.49425 down Ferd3l NM_033522 10395317 -2.45067 down EG328479 BC130025 10422500 -2.44537 down Adamts3 NM_001081401 10531201 -2.44415 down Tom1l1 NM_028011 10389816 -2.4368 down 3632451O06Rik BC023359 10419416 -2.39512 down Thrb NM_009380 10412882 -2.39392 down Glis3 NM_175459 10466888 -2.37862 down Dock5 NM_177780 10421046 -2.37549 down Nfia NM_001122952 10506050 -2.37378 down Rnf152 NM_178779 10357003 -2.35876 down Pcp4l1 NM_025557 10360053 -2.34276 down Aard NM_175503 10424082 -2.32311 down Kcna5 NM_145983 10548043 -2.31818 down 173
Adamts9 NM_175314 10546454 -2.29917 down Wnt9a NM_139298 10376490 -2.29762 down BC057079 NM_001081184 10505791 -2.29193 down Luzp2 NM_178705 10553537 -2.28685 down Pde6c NM_033614 10462887 -2.24522 down Galnt3 NM_015736 10483249 -2.2266 down Slc38a4 NM_027052 10431915 -2.22215 down Igfbp5 NM_010518 10355500 -2.21419 down Cck NM_031161 10597817 -2.19177 down 2610301F02Rik ENSMUST00000049544 10484207 -2.16949 down Adamts1 NM_009621 10440522 -2.16874 down Bche NM_009738 10498710 -2.15475 down Kdr NM_010612 10530692 -2.15107 down F2rl1 NM_007974 10411226 -2.14802 down --- 10461160 -2.14497 down Scube2 NM_020052 10566822 -2.1399 down Ptbp1 NM_001077363 10407390 -2.13823 down Garnl3 NM_178888 10481772 -2.13791 down Ppargc1a NM_008904 10529979 -2.12402 down Sntb1 NM_016667 10428698 -2.12253 down Mrc1 NM_008625 10469358 -2.11701 down Hs3st3a1 NM_178870 10376956 -2.10449 down Vwa5a NM_172767 10584435 -2.0888 down Lrig3 NM_177152 10366746 -2.08579 down Slc2a3 NM_011401 10547641 -2.08537 down Ass1 NM_007494 10363541 -2.0842 down Mobkl2b ENSMUST00000098156 10512022 -2.07802 down Ass1 NM_007494 10471154 -2.07614 down Gsbs NM_011153 10538519 -2.0747 down Enpp1 NM_008813 10368289 -2.07141 down Fgfr2 NM_010207 10568436 -2.06418 down Cx3cl1 NM_009142 10574220 -2.06329 down Pkp4 NM_026361 10472212 -2.06107 down Cp NM_001042611 10490989 -2.05207 down ENSMUSG00000072690 AK143453 10542592 -2.04267 down A830059I20Rik NM_021427 10554960 -2.03637 down Ecm2 NM_001012324 10405033 -2.03397 down Hmox1 NM_010442 10572897 -2.02918 down Lrp1b NM_053011 10482336 -2.02774 down OTTMUSG00000008561 NM_001085549 10507099 -2.02644 down 1700106J16Rik ENSMUST00000037268 10380244 -2.02532 down F13a1 NM_028784 10408693 -2.0244 down --- 10362426 -2.02155 down Fbn1 NM_007993 10487040 -2.01378 down Acot2 NM_134188 10397145 -2.01255 down Sfrp1 NM_013834 10570957 -2.00344 down 174
Suclg2 NM_011507 10546538 -2.00297 down Efcab1 NM_025769 10433782 -2.0005 down Gad1 NM_008077 10472707 2.00427 up Dpf1 NM_013874 10551741 2.00867 up Tpbpa NM_009411 10409881 2.02101 up Fbxo44 NM_173401 10518484 2.02716 up Pdzd4 NM_001029868 10605104 2.02915 up Tcfap2a NM_011547 10408798 2.04714 up Vat1 NM_012037 10391454 2.04984 up Cxx1b NM_001018063 10604633 2.05565 up --- 10433428 2.05678 up
Thsd4 NM_001040426 10594199 2.06697 up Atp1a3 NM_144921 10560919 2.07286 up Klhl35 ENSMUST00000037359 10555225 2.07436 up Lypd1 NM_145100 10357339 2.07778 up Emx2 NM_010132 10464391 2.0797 up Hcn4 NM_001081192 10585851 2.10161 up S1pr1 NM_007901 10501586 2.10231 up Slc16a14 NM_027921 10356240 2.11015 up Nxph1 NM_008751 10536401 2.14784 up Samd14 NM_146025 10380489 2.15317 up D430041B17Rik NM_172737 10559681 2.1598 up Sst NM_009215 10438730 2.17213 up Mpped2 ENSMUST00000016530 10474361 2.1729 up Rcan2 NM_030598 10445325 2.17611 up Pnoc NM_010932 10420853 2.18349 up Jarid1d NM_011419 10607972 2.19478 up Sez6l2 NM_144926 10557535 2.20153 up Sv2a NM_022030 10494372 2.20704 up Rgs6 NM_015812 10397030 2.2535 up Accn2 NM_009597 10426782 2.26344 up 2700081O15Rik BC096638 10461071 2.27446 up Brunol5 NM_176954 10371188 2.28125 up Cnih2 NM_009920 10464917 2.28676 up ENSMUSG00000066088 AY512949 10506452 2.32644 up Grik3 NM_001081097 10508052 2.33842 up Adarb2 NM_052977 10403396 2.34108 up Rpp25 NM_133982 10585703 2.37565 up Lhx1 NM_008498 10389283 2.38849 up Cbfa2t3 NM_009824 10582429 2.41547 up Nxph1 ENSMUST00000101672 10536405 2.44633 up Scrt1 NM_130893 10429944 2.50283 up Tmem130 NM_177735 10535577 2.52956 up --- 10423917 2.58033 up
Gata2 NM_008090 10539873 2.64899 up Slc32a1 NM_009508 10478124 2.6746 up 175
Cacna2d3 NM_009785 10418300 2.685 up Dub2a NM_001001559 10566205 2.7048 up Accn4 NM_183022 10347650 2.76526 up Lhx5 NM_008499 10525069 2.94516 up Nrxn3 NM_172544 10397575 3.01117 up Gad2 NM_008078 10469672 3.21428 up Gata3 NM_008091 10479981 3.70456 up Sox14 BC055767 10596023 4.51502 up 176
Appendix VIII: Plasmid maps
Please see following pages for plasmid maps. The plasmid map of pYX-Asc-Fibcd1. The plasmid map of pGEMTEasy-Gata3. The plasmid map of pGEMTEasy-Lhx5. The plasmid map of pGEMT-Msx1. The plasmid map of pGEMTEasy-Sox14. The plasmid map of pBluescript KS(+)-Wnt1. I
References
Abellán, A., Vernier, B., Rétaux, S., Medina, L., 2010. Similarities and differences in the forebrain expression of Lhx1 and Lhx5 between chicken and mouse: Insights for understanding telencephalic development and evolution. The Journal of Comparative Neurology. 518, 3512-3528. Abramoff, M. D., Magelhaes, P. J., Ram, S. J., 2004. Image Processing with ImageJ. Biophotonics International. 11, 36-42. Ahituv, N., Zhu, Y., Visel, A., Holt, A., Afzal, V., Pennacchio, L. A., Rubin, E. M., 2007. Deletion of ultraconserved elements yields viable mice. PLoS Biol. 5, e234. Albrecht, A. N., Kornak, U., Boddrich, A., Suring, K., Robinson, P. N., Stiege, A. C., Lurz, R., Stricker, S., Wanker, E. E., Mundlos, S., 2004. A molecular pathogenesis for transcription factor associated poly-alanine tract expansions. Hum Mol Genet. 13, 2351-9. Altschul, S. F., Madden, T. L., Schaffer, A. A., Zhang, J., Zhang, Z., Miller, W., Lipman, D. J., 1997. Gapped BLAST and PSI-BLAST: a new generation of protein database search programs. Nucleic Acids Res. 25, 3389-402. Ang, S. L., Jin, O., Rhinn, M., Daigle, N., Stevenson, L., Rossant, J., 1996. A targeted mouse Otx2 mutation leads to severe defects in gastrulation and formation of axial mesoderm and to deletion of rostral brain. Development. 122, 243-252. Aota, S., Nakajima, N., Sakamoto, R., Watanabe, S., Ibaraki, N., Okazaki, K., 2003. Pax6 autoregulation mediated by direct interaction of Pax6 protein with the head surface ectoderm-specific enhancer of the mouse Pax6 gene. Dev Biol. 257, 1-13. Archer, T. C., Jin, J., Casey, E. S., 2010. Interaction of Sox1, Sox2, Sox3 and Oct4 during primary neurogenesis. Developmental Biology. 350, 429-440. Arendt, J., 1998. Melatonin and the pineal gland: influence on mammalian seasonal and circadian physiology. Rev Reprod. 3, 13-22. Avilion, A. A., Nicolis, S. K., Pevny, L. H., Perez, L., Vivian, N., Lovell-Badge, R., 2003. Multipotent cell lineages in early mouse development depend on SOX2 function. Genes Dev. 17, 126-140. Baas, D., Meiniel, A., Benadiba, C., Bonnafe, E., Meiniel, O., Reith, W., Durand, B., 2006. A deficiency in RFX3 causes hydrocephalus associated with abnormal differentiation of ependymal cells. Eur J Neurosci. 24, 1020-30. Bach, A., Lallemand, Y., Nicola, M.-A., Ramos, C., Mathis, L., Maufras, M., Robert, B., 2003. Msx1 is required for dorsal diencephalon patterning. Development. 130, 4025- 4036. Baek, J. H., Hatakeyama, J., Sakamoto, S., Ohtsuka, T., Kageyama, R., 2006. Persistent and high levels of Hes1 expression regulate boundary formation in the developing central nervous system. Development. 133, 2467-2476. Balslev, Y., Saunders, N. R., MØllgard, K., 1997. Ontogenetic development of diffusional restriction to protein at the pial surface of the rat brain: an electron microscopical study. Journal of Neurocytology. 26, 133-148. Bani-Yaghoub, M., Tremblay, R. G., Lei, J. X., Zhang, D., Zurakowski, B., Sandhu, J. K., Smith, B., Ribecco-Lutkiewicz, M., Kennedy, J., Walker, P. R., Sikorska, M., 2006. Role of Sox2 in the development of the mouse neocortex. Developmental Biology. 295, 52-66. Banizs, B., Pike, M. M., Millican, C. L., Ferguson, W. B., Komlosi, P., Sheetz, J., Bell, P. D., Schwiebert, E. M., Yoder, B. K., 2005. Dysfunctional cilia lead to altered ependyma II
and choroid plexus function, and result in the formation of hydrocephalus. Development. 132, 5329-5339. Banks, A. S., Li, J., McKeag, L., Hribal, M. L., Kashiwada, M., Accili, D., Rothman, P. B., 2005. Deletion of SOCS7 leads to enhanced insulin action and enlarged islets of Langerhans. The Journal of Clinical Investigation. 115, 2462-2471. Bassuk, A. G., Kibar, Z., 2009. Genetic Basis of Neural Tube Defects. Seminars in Pediatric Neurology. 16, 101-110. Batiz, L., Oliver, C., Alvarez, M., Rodriguez, S., Rodriguez, E., 2006a. Molecular mechanisms underlying neuroepithelial/ependymal denudation in the hydrocephalic hyh mutant: spatial and temporal expression of alpha-SNAP and N-cadherin. Cerebrospinal Fluid Research. 3, S16. Batiz, L. F., Paez, P., Jimenez, A. J., Rodriguez, S., Wagner, C., Perez-Figares, J. M., Rodriguez, E. M., 2006b. Heterogeneous expression of hydrocephalic phenotype in the hyh mice carrying a point mutation in [alpha]-SNAP. Neurobiology of Disease. 23, 152-168. Beck, H. N., Drahushuk, K., Jacoby, D. B., Higgins, D., Lein, P. J., 2001. Bone morphogenetic protein-5 (BMP-5) promotes dendritic growth in cultured sympathetic neurons. BMC Neurosci. 2, 12. Beddington, R. S., 1994. Induction of a second neural axis by the mouse node. Development. 120, 613-620. Beddington, R. S., Robertson, E. J., 1999. Axis development and early asymmetry in mammals. Cell. 96, 195-209. Belo, J. A., Bouwmeester, T., Leyns, L., Kertesz, N., Gallo, M., Follettie, M., De Robertis, E. M., 1997. Cerberus-like is a secreted factor with neuralizing activity expressed in the anterior primitive endoderm of the mouse gastrula. Mechanisms of Development. 68, 45-57. Bennett, L., Yang, M., Enikolopov, G., Iacovitti, L., 2009. Circumventricular organs: A novel site of neural stem cells in the adult brain. Molecular and Cellular Neuroscience. 41, 337-347. Bennett, L. B., Cai, J., Enikolopov, G., Iacovitti, L., 2010. Heterotopically transplanted CVO neural stem cells generate neurons and migrate with SVZ cells in the adult mouse brain. Neurosci Lett. 475, 1-6. Benowitz, L. I., Routtenberg, A., 1987. A membrane phosphoprotein associated with neural development, axonal regeneration, phospholipid metabolism, and synaptic plasticity. Trends in Neurosciences. 10, 527-532. Bergstrom, D. E., Young, M., Albrecht, K. H., Eicher, E. M., 2000. Related function of mouse SOX3, SOX9, and SRY HMG domains assayed by male sex determination. Genesis. 28, 111-24. Blackshaw, S., Snyder, S. H., 1997. Developmental Expression Pattern of Phototransduction Components in Mammalian Pineal Implies a Light-Sensing Function. The Journal of Neuroscience. 17, 8074-8082. Blackshear, P. J., Graves, J. P., Stumpo, D. J., Cobos, I., Rubenstein, J. L., Zeldin, D. C., 2003. Graded phenotypic response to partial and complete deficiency of a brain- specific transcript variant of the winged helix transcription factor RFX4. Development. 130, 4539-52. Blatt, E. N., Yan, X. H., Wuerffel, M. K., Hamilos, D. L., Brody, S. L., 1999. Forkhead Transcription Factor HFH-4 Expression Is Temporally Related to Ciliogenesis. Am. J. Respir. Cell Mol. Biol. 21, 168-176. Bovolenta, P., Rodriguez, J., Esteve, P., 2006. Frizzled/RYK mediated signalling in axon guidance. Development. 133, 4399-4408. III
Brody, S. L., Yan, X. H., Wuerffel, M. K., Song, S.-K., Shapiro, S. D., 2000. Ciliogenesis and Left-Right Axis Defects in Forkhead Factor HFH-4-Null Mice. Am. J. Respir. Cell Mol. Biol. 23, 45-51. Brown, P. D., Davies, S. L., Speake, T., Millar, I. D., 2004. Molecular mechanisms of cerebrospinal fluid production. Neuroscience. 129, 957-70. Brunelli, S., Silva Casey, E., Bell, D., Harland, R., Lovell-Badge, R., 2003. Expression of Sox3 throughout the developing central nervous system is dependent on the combined action of discrete, evolutionarily conserved regulatory elements. Genesis. 36, 12-24. Bruni, J. E., 1998. Ependymal development, proliferation, and functions: A review. Microscopy Research and Technique. 41, 2-13. Bullwinkel, J., Baron-Lühr, B., Lüdemann, A., Wohlenberg, C., Gerdes, J., Scholzen, T., 2006. Ki-67 protein is associated with ribosomal RNA transcription in quiescent and proliferating cells. Journal of Cellular Physiology. 206, 624-635. Burkitt Wright, E. M., Perveen, R., Clayton, P. E., Hall, C. M., Costa, T., Procter, A. M., Giblin, C. A., Donnai, D., Black, G. C., 2009. X-linked isolated growth hormone deficiency: expanding the phenotypic spectrum of SOX3 polyalanine tract expansions. Clin Dysmorphol. 18, 218-21. Butler, S. J., Dodd, J., 2003. A Role for BMP Heterodimers in Roof Plate-Mediated Repulsion of Commissural Axons. Neuron. 38, 389-401. Bylund, M., Andersson, E., Novitch, B. G., Muhr, J., 2003. Vertebrate neurogenesis is counteracted by Sox1-3 activity. Nat Neurosci. 6, 1162-8. Callen, D. F., Baker, E. G., Lane, S. A., 1990. Re-evaluation of GM2346 from a del(16)(q22) to t(4;16)(q35;q22.1). Clin Genet. 38, 466-8. Camus, A., Perea-Gomez, A., Moreau, A., Collignon, J., 2006. Absence of Nodal signaling promotes precocious neural differentiation in the mouse embryo. Developmental Biology. 295, 743-755. Castañeyra-Perdomo, A., Carmona-Calero, E., Perez_Gonzalez, H., Martinez-Peña y Valenzuela, I., Plaza-Moreno, P., Ormazabal-Ramos, C., Gonzalez-Marrero, I., Trujillano-Dorado, A., Ferres-Torres, R., 2004. Ontogenic development of the human subcommissural organ. European Journal of Anatomy. 8, 107-120. Castaneyra-Perdomo, A., Meyer, G., Carmona-Calero, E., Banuelos-Pineda, J., Mendez- Medina, R., Ormazabal-Ramos, C., Ferres-Torres, R., 1994. Alterations of the subcommissural organ in the hydrocephalic human fetal brain. Brain Res Dev Brain Res. 79, 316-20. Castaneyra-Perdomo, A., Meyer, G., Ferres-Torres, R., 1985. The early development of the human subcommissural organ. J Anat. 143, 195-200. Chae, T. H., Kim, S., Marz, K. E., Hanson, P. I., Walsh, C. A., 2004. The hyh mutation uncovers roles for [alpha]Snap in apical protein localization and control of neural cell fate. Nat Genet. 36, 264-270. Chandler, K. J., Chandler, R. L., Mortlock, D. P., 2009. Identification of an ancient Bmp4 mesoderm enhancer located 46 kb from the promoter. Dev Biol. 327, 590-602. Chandler, R. L., Chandler, K. J., McFarland, K. A., Mortlock, D. P., 2007. Bmp2 transcription in osteoblast progenitors is regulated by a distant 3' enhancer located 156.3 kilobases from the promoter. Mol Cell Biol. 27, 2934-51. Chao, A. T., Jones, W. M., Bejsovec, A., 2007. The HMG-box transcription factor SoxNeuro acts with Tcf to control Wg/Wnt signaling activity. Development. 134, 989-97. Chen, H.-L., Panchision, D. M., 2007. Concise Review: Bone Morphogenetic Protein Pleiotropism in Neural Stem Cells and Their Derivatives—Alternative Pathways, Convergent Signals. STEM CELLS. 25, 63-68. IV
Chen, J., Knowles, H. J., Hebert, J. L., Hackett, B. P., 1998. Mutation of the mouse hepatocyte nuclear factor/forkhead homologue 4 gene results in an absence of cilia and random left-right asymmetry. J Clin Invest. 102, 1077-82. Chesnutt, C., Burrus, L. W., Brown, A. M. C., Niswander, L., 2004. Coordinate regulation of neural tube patterning and proliferation by TGF[beta] and WNT activity. Developmental Biology. 274, 334-347. Cheung, K.-K., Mok, S., Rezaie, P., Chan, W., 2008. Dynamic expression of Dab2 in the mouse embryonic central nervous system. BMC Developmental Biology. 8, 76. Chew, J.-L., Loh, Y.-H., Zhang, W., Chen, X., Tam, W.-L., Yeap, L.-S., Li, P., Ang, Y.-S., Lim, B., Robson, P., Ng, H.-H., 2005. Reciprocal Transcriptional Regulation of Pou5f1 and Sox2 via the Oct4/Sox2 Complex in Embryonic Stem Cells. Mol. Cell. Biol. 25, 6031-6046. Chizhikov, V. V., Millen, K. J., 2004a. Control of Roof Plate Development and Signaling by Lmx1b in the Caudal Vertebrate CNS. J. Neurosci. 24, 5694-5703. Chizhikov, V. V., Millen, K. J., 2004b. Control of roof plate formation by Lmx1a in the developing spinal cord. Development. 131, 2693-2705. Chizhikov, V. V., Millen, K. J., 2004c. Mechanisms of roof plate formation in the vertebrate CNS. Nat Rev Neurosci. 5, 808-812. Chizhikov, V. V., Millen, K. J., 2005. Roof plate-dependent patterning of the vertebrate dorsal central nervous system. Dev Biol. 277, 287-95. Chouaf-Lakhdar, L., Fevre-Montange, M., Brisson, C., Strazielle, N., Gamrani, H., Didier- Bazes, M., 2003. Proliferative activity and nestin expression in periventricular cells of the adult rat brain. Neuroreport. 14, 633-6. Cifuentes, M., López-Avalos, M. D., Pérez, J., Grondona, J. M., Fernández-Llebrez, P., 1996. Identification of a high molecular weight polypeptide in the subcommissural organ of the chick embryo. Cell and Tissue Research. 286, 543-546. Clevers, H., 2006. Wnt/β-Catenin Signaling in Development and Disease. Cell. 127, 469-480. Cobos, I., Shimamura, K., Rubenstein, J. L., Martinez, S., Puelles, L., 2001. Fate map of the avian anterior forebrain at the four-somite stage, based on the analysis of quail-chick chimeras. Dev Biol. 239, 46-67. Cohen, N. R., Taylor, J. S. H., Scott, L. B., Guillery, R. W., Soriano, P., Furley, A. J. W., 1998. Errors in corticospinal axon guidance in mice lacking the neural cell adhesion molecule L1. Current Biology. 8, 26-33. Colas, J.-F., Schoenwolf, G. C., 2001. Towards a cellular and molecular understanding of neurulation. Developmental Dynamics. 221, 117-145. Collignon, J., Sockanathan, S., Hacker, A., Cohen-Tannoudji, M., Norris, D., Rastan, S., Stevanovic, M., Goodfellow, P. N., Lovell-Badge, R., 1996. A comparison of the properties of Sox-3 with Sry and two related genes, Sox-1 and Sox-2. Development. 122, 509-20. Connor, F., Cary, P. D., Read, C. M., Preston, N. S., Driscoll, P. C., Denny, P., Crane- Robinson, C., Ashworth, A., 1994. DNA binding and bending properties of the post- meiotically expressed Sry-related protein Sox-5. Nucleic Acids Res. 22, 3339-46. Copp, A. J., 2005. Neurulation in the cranial region--normal and abnormal. J Anat. 207, 623- 35. Copp, A. J., Greene, N. D., 2010. Genetics and development of neural tube defects. J Pathol. 220, 217-30. Copp, A. J., Greene, N. D., Murdoch, J. N., 2003. The genetic basis of mammalian neurulation. Nat Rev Genet. 4, 784-93. Cremer, H., Lange, R., Christoph, A., Plomann, M., Vopper, G., Roes, J., Brown, R., Baldwin, S., Kraemer, P., Scheff, S., Barthels, D., Rajewsky, K., Wille, W., 1994. V
Inactivation of the N-CAM gene in mice results in size reduction of the olfactory bulb and deficits in spatial learning. Nature. 367, 455-459. Creveaux, I., Gobron, S., Meiniel, R., Dastugue, B., Meiniel, A., 1998. Complex expression pattern of the SCO-spondin gene in the bovine subcommissural organ: toward an explanation for Reissner's fiber complexity? Molecular Brain Research. 55, 45-53. Cue, D., Lei, M. G., Luong, T. T., Kuechenmeister, L., Dunman, P. M., O'Donnell, S., Rowe, S., O'Gara, J. P., Lee, C. Y., 2009. Rbf promotes biofilm formation by Staphylococcus aureus via repression of icaR, a negative regulator of icaADBC. J Bacteriol. 191, 6363-73. Cullingford, T. E., Dolphin, C. T., Sato, H., 2002a. The peroxisome proliferator-activated receptor [alpha]-selective activator ciprofibrate upregulates expression of genes encoding fatty acid oxidation and ketogenesis enzymes in rat brain. Neuropharmacology. 42, 724-730. Cullingford, T. E., Eagles, D. A., Sato, H., 2002b. The ketogenic diet upregulates expression of the gene encoding the key ketogenic enzyme mitochondrial 3-hydroxy-3- methylglutaryl-CoA synthase in rat brain. Epilepsy Res. 49, 99-107. Currle, D. S., Cheng, X., Hsu, C.-m., Monuki, E. S., 2005. Direct and indirect roles of CNS dorsal midline cells in choroid plexus epithelia formation. Development. 132, 3549- 3559. Dahme, M., Bartsch, U., Martini, R., Anliker, B., Schachner, M., Mantei, N., 1997. Disruption of the mouse L1 gene leads to malformations of the nervous system. Nat Genet. 17, 346-9. Daneman, R., Zhou, L., Kebede, A. A., Barres, B. A., 2010. Pericytes are required for blood- brain barrier integrity during embryogenesis. Nature. 468, 562-566. Danielian, P. S., Muccino, D., Rowitch, D. H., Michael, S. K., McMahon, A. P., 1998. Modification of gene activity in mouse embryos in utero by a tamoxifen-inducible form of Cre recombinase. Curr Biol. 8, 1323-6. Dashti, N., Ontko, J. A., 1979. Rate-limiting function of 3-hydroxy-3-methylglutaryl- coenzyme A synthase in ketogenesis. Biochem Med. 22, 365-74. Dattani, M. T., Martinez-Barbera, J.-P., Thomas, P. Q., Brickman, J. M., Gupta, R., Martensson, I.-L., Toresson, H., Fox, M., Wales, J. K. H., Hindmarsh, P. C., Krauss, S., Beddington, R. S. P., Robinson, I. C. A. F., 1998. Mutations in the homeobox gene HESX1/Hesx1 associated with septo-optic dysplasia in human and mouse. Nat Genet. 19, 125-133. Davidson, B. P., Kinder, S. J., Steiner, K., Schoenwolf, G. C., Tam, P. P. L., 1999. Impact of Node Ablation on the Morphogenesis of the Body Axis and the Lateral Asymmetry of the Mouse Embryo during Early Organogenesis. Developmental Biology. 211, 11-26. Davy, B. E., Robinson, M. L., 2003. Congenital hydrocephalus in hy3 mice is caused by a frameshift mutation in Hydin, a large novel gene. Hum. Mol. Genet. 12, 1163-1170. De Martino, S. P., Errington, F., Ashworth, A., Jowett, T., Austin, C. A., 1999. <i>sox30:</i> a novel zebrafish <i>sox</i> gene expressed in a restricted manner at the midbrain-hindbrain boundary during neurogenesis. Development Genes and Evolution. 209, 357-362. De Santa Barbara, P., Bonneaud, N., Boizet, B., Desclozeaux, M., Moniot, B., Sudbeck, P., Scherer, G., Poulat, F., Berta, P., 1998. Direct Interaction of SRY-Related Protein SOX9 and Steroidogenic Factor 1 Regulates Transcription of the Human Anti- Mullerian Hormone Gene. Mol. Cell. Biol. 18, 6653-6665. Dee, C. T., Hirst, C. S., Shih, Y. H., Tripathi, V. B., Patient, R. K., Scotting, P. J., 2008. Sox3 regulates both neural fate and differentiation in the zebrafish ectoderm. Dev Biol. 320, 289-301. VI
Del Bigio, M. R., 2010. Neuropathology and structural changes in hydrocephalus. Developmental Disabilities Research Reviews. 16, 16-22. Del Bigio, M. R., Zhang, Y. W., 1998. Cell Death, Axonal Damage, and Cell Birth in the Immature Rat Brain Following Induction of Hydrocephalus. Experimental Neurology. 154, 157-169. del Brío, M. A., Riera, P., Muñoz, R. I., Montecinos, H., Rodríguez, E. M., 2000. The metencephalic floor plate of chick embryos expresses two secretory glycoproteins homologous with the two glycoproteins secreted by the subcommissural organ. Histochemistry and Cell Biology. 113, 415-426. Denny, P., Swift, S., Connor, F., Ashworth, A., 1992. An SRY-related gene expressed during spermatogenesis in the mouse encodes a sequence-specific DNA-binding protein. Embo J. 11, 3705-12. Detrick, R. J., Dickey, D., Kintner, C. R., 1990. The effects of N-cadherin misexpression on morphogenesis in xenopus embryos. Neuron. 4, 493-506. Didier, R., Dastugue, B., Meiniel, A., 1995. The secretory material of the subcommissural organ of the chick embryo. Characterization of a specific polypeptide by two- dimensional electrophoresis. Int J Dev Biol. 39, 493-9. Dietrich, P., Shanmugasundaram, R., E, S., Dragatsis, I., 2009. Congenital hydrocephalus associated with abnormal subcommissural organ in mice lacking huntingtin in Wnt1 cell lineages. Hum. Mol. Genet. 18, 142-150. DiLeone, R. J., Marcus, G. A., Johnson, M. D., Kingsley, D. M., 2000. Efficient studies of long-distance Bmp5 gene regulation using bacterial artificial chromosomes. Proceedings of the National Academy of Sciences of the United States of America. 97, 1612-1617. Dolphin, C. T., Cullingford, T. E., Shcphard, E. A., Smith, R. L., Phillips, I. R., 1996. Differential Developmental and Tissue-Specific Regulation of Expression of the Genes Encoding Three Members of the Flavin-Containing Monooxygenase Family of Man, FMO1, FMO3 and FMO4. European Journal of Biochemistry. 235, 683-689. Dominguez-Pinos, M. D., Paez, P., Jimenez, A. J., Weil, B., Arraez, M. A., Perez-Figares, J. M., Rodriguez, E. M., 2005. Ependymal denudation and alterations of the subventricular zone occur in human fetuses with a moderate communicating hydrocephalus. J Neuropathol Exp Neurol. 64, 595-604. Drees, F., Pokutta, S., Yamada, S., Nelson, W. J., Weis, W. I., 2005. [alpha]-Catenin Is a Molecular Switch that Binds E-Cadherin-[beta]-Catenin and Regulates Actin- Filament Assembly. Cell. 123, 903-915. Duvernoy, H. M., Risold, P.-Y., 2007. The circumventricular organs: An atlas of comparative anatomy and vascularization. Brain Research Reviews. 56, 119-147. Dy, P., Penzo-Mandez, A., Wang, H., Pedraza, C. E., Macklin, W. B., Lefebvre, V., 2008. The three SoxC proteins Sox4, Sox11 and Sox12 exhibit overlapping expression patterns and molecular properties. Nucleic Acids Research. 36, 3101-3117. Dziegielewska, K. M., Ek, J., Habgood, M. D., Saunders, N. R., 2001. Development of the choroid plexus. Microscopy Research and Technique. 52, 5-20. El Bitar, F., Dastugue, B., Meiniel, A., 1999. Neuroblastoma B104 cell line as a model for analysis of neurite outgrowth and neuronal aggregation induced by Reissner's fiber material. Cell and Tissue Research. 298, 233-242. El Wakil, A., Francius, C. d., Wolff, A., Pleau-Varet, J., Nardelli, J., 2006. The GATA2 transcription factor negatively regulates the proliferation of neuronal progenitors. Development. 133, 2155-2165. Ellis, P., Fagan, B. M., Magness, S. T., Hutton, S., Taranova, O., Hayashi, S., McMahon, A., Rao, M., Pevny, L., 2004. SOX2, a Persistent Marker for Multipotential Neural Stem VII
Cells Derived from Embryonic Stem Cells, the Embryo or the Adult. Developmental Neuroscience. 26, 148-165. Engelhardt, B., Sorokin, L., 2009. The blood–brain and the blood–cerebrospinal fluid barriers: function and dysfunction. Seminars in Immunopathology. 31, 497-511. Eriksson, C., Bjorklund, A., Wictorin, K., 2003. Neuronal differentiation following transplantation of expanded mouse neurosphere cultures derived from different embryonic forebrain regions. Experimental Neurology. 184, 615-635. Estivill-Torrus, G., Pearson, H., van Heyningen, V., Price, D. J., Rashbass, P., 2002. Pax6 is required to regulate the cell cycle and the rate of progression from symmetrical to asymmetrical division in mammalian cortical progenitors. Development. 129, 455- 466. Estivill-Torrus, G., Vitalis, T., Fernandez-Llebrez, P., Price, D. J., 2001. The transcription factor Pax6 is required for development of the diencephalic dorsal midline secretory radial glia that form the subcommissural organ. Mechanisms of Development. 109, 215-224. Failli, V., Bachy, I., Rétaux, S., 2002. Expression of the LIM-homeodomain gene Lmx1a (dreher) during development of the mouse nervous system. Mechanisms of Development. 118, 225-228. Fernandes, R. J., Hirohata, S., Engle, J. M., Colige, A., Cohn, D. H., Eyre, D. R., Apte, S. S., 2001. Procollagen II Amino Propeptide Processing by ADAMTS-3. Journal of Biological Chemistry. 276, 31502-31509. Fernandez-Llebrez, P., Grondona, J. M., Perez, J., Lopez-Aranda, M. F., Estivill-Torrus, G., Llebrez-Zayas, P. F., Soriano, E., Ramos, C., Lallemand, Y., Bach, A., Robert, B., 2004. Msx1-deficient mice fail to form prosomere 1 derivatives, subcommissural organ, and posterior commissure and develop hydrocephalus. J Neuropathol Exp Neurol. 63, 574-86. Fernandez-Llebrez, P., Lopez-Avalos, M. D., Mota, M. D., Cifuentes, M., Andrades, J. A., Grondona, J. M., Perez, J., Perez-Figares, J. M., Rodriguez, E. M., 1996. Secretory glycoproteins of the roof and floor plates and their rostral derivatives, the subcommissural and flexural organs, in the developing central nervous system of vertebrates. An immunocytochemical study. Int J Dev Biol. Suppl 1, 151S-152S. Fernández-Llebrez, P., Pérez, J., Nadales, A., Pérez-Fígares, J., Rodríguez, E., 1987. Vascular and leptomeningeal projections of the subcommissural organ in reptiles. Histochemistry and Cell Biology. 87, 607-614. Ferri, A. L. M., Cavallaro, M., Braida, D., Di Cristofano, A., Canta, A., Vezzani, A., Ottolenghi, S., Pandolfi, P. P., Sala, M., DeBiasi, S., Nicolis, S. K., 2004. Sox2 deficiency causes neurodegeneration and impaired neurogenesis in the adult mouse brain. Development. 131, 3805-3819. Figdor, M. C., Stern, C. D., 1993. Segmental organization of embryonic diencephalon. Nature. 363, 630-4. Finger, J. H., Smith, C. M., Hayamizu, T. F., McCright, I. J., Eppig, J. T., Kadin, J. A., Richardson, J. E., Ringwald, M., 2010. The mouse Gene Expression Database (GXD): 2011 update. Nucleic Acids Res. 39, D835-41. Fischer, M., Skowron, M., Berthold, F., 2005. Reliable transcript quantification by real-time reverse transcriptase-polymerase chain reaction in primary neuroblastoma using normalization to averaged expression levels of the control genes HPRT1 and SDHA. J Mol Diagn. 7, 89-96. Fliegauf, M., Benzing, T., Omran, H., 2007. When cilia go bad: cilia defects and ciliopathies. Nat Rev Mol Cell Biol. 8, 880-893. VIII
Fransen, E., Lemmon, V., Van Camp, G., Vits, L., Coucke, P., Willems, P. J., 1995. CRASH syndrome: clinical spectrum of corpus callosum hypoplasia, retardation, adducted thumbs, spastic paraparesis and hydrocephalus due to mutations in one single gene, L1. Eur J Hum Genet. 3, 273-84. Fransen, E., Vits, L., Camp, G. V., Willems, P. J., 1996. The clinical spectrum of mutations in L1, a neuronal cell adhesion molecule. American Journal of Medical Genetics. 64, 73-77. Freese, J. L., Pino, D., Pleasure, S. J., Wnt signaling in development and disease. Neurobiol Dis. 38, 148-53. Fuerer, C., Nusse, R., ten Berge, D., 2008. Wnt signalling in development and disease. EMBO Rep. 9, 134-138. Fujimori, T., Miyatani, S., Takeichi, M., 1990. Ectopic expression of N-cadherin perturbs histogenesis in Xenopus embryos. Development. 110, 97-104. Fukushima, N., Yokouchi, K., Kawagishi, K., Ren, G., Higashiyama, F., Moriizumi, T., 2003. Proliferating cell populations in experimentally-induced hydrocephalus in developing rats. Journal of Clinical Neuroscience. 10, 334-337. Funato, H., Saito-Nakazato, Y., Takahashi, H., 2000. Axonal Growth from the Habenular Nucleus along the Neuromere Boundary Region of the Diencephalon Is Regulated by Semaphorin 3F and Netrin-1. Molecular and Cellular Neuroscience. 16, 206-220. Furukawa, T., Morrow, E. M., Li, T., Davis, F. C., Cepko, C. L., 1999. Retinopathy and attenuated circadian entrainment in Crx-deficient mice. Nat Genet. 23, 466-470. Furuta, Y., Piston, D. W., Hogan, B. L., 1997. Bone morphogenetic proteins (BMPs) as regulators of dorsal forebrain development. Development. 124, 2203-2212. Gabriel, J. M., Merchant, M., Ohta, T., Ji, Y., Caldwell, R. G., Ramsey, M. J., Tucker, J. D., Longnecker, R., Nicholls, R. D., 1999. A transgene insertion creating a heritable chromosome deletion mouse model of Prader-Willi and Angelman syndromes. Proceedings of the National Academy of Sciences of the United States of America. 96, 9258-9263. Gadbury, G. L., Page, G. P., Edwards, J., Kayo, T., Prolla, T. A., Weindruch, R., Permana, P. A., Mountz, J. D., Allison, D. B., 2004. Power and sample size estimation in high dimensional biology. Statistical Methods in Medical Research. 13, 325-338. Galarza, M., 2002. Evidence of the subcommissural organ in humans and its association with hydrocephalus. Neurosurgical Review. 25, 205-215. Gobron, S., Creveaux, I., Meiniel, R., Didier, R., Dastugue, B., Meiniel, A., 1999. SCO- spondin is evolutionarily conserved in the central nervous system of the chordate phylum. Neuroscience. 88, 655-664. Gobron, S., Monnerie, H., Meiniel, R., Creveaux, I., Lehmann, W., Lamalle, D., Dastugue, B., Meiniel, A., 1996. SCO-spondin: a new member of the thrombospondin family secreted by the subcommissural organ is a candidate in the modulation of neuronal aggregation. J Cell Sci. 109, 1053-1061. Goncalves-Mendes, N., Simon-Chazottes, D., Creveaux, I., Meiniel, A., Guenet, J.-L., Meiniel, R., 2003. Mouse SCO-spondin, a gene of the thrombospondin type 1 repeat (TSR) superfamily expressed in the brain. Gene. 312, 263-270. Gordon, M. D., Nusse, R., 2006. Wnt signaling: multiple pathways, multiple receptors, and multiple transcription factors. J Biol Chem. 281, 22429-33. Graham, V., Khudyakov, J., Ellis, P., Pevny, L., 2003. SOX2 functions to maintain neural progenitor identity. Neuron. 39, 749-65. Graves, J. A., 1998. Interactions between SRY and SOX genes in mammalian sex determination. Bioessays. 20, 264-9. IX
Graves, J. A., 2001. From brain determination to testis determination: evolution of the mammalian sex-determining gene. Reprod Fertil Dev. 13, 665-72. Greene, N. D. E., Copp, A. J., 2009. Development of the vertebrate central nervous system: formation of the neural tube. Prenatal Diagnosis. 29, 303-311. Grigoryan, T., Wend, P., Klaus, A., Birchmeier, W., 2008. Deciphering the function of canonical Wnt signals in development and disease: conditional loss- and gain-of- function mutations of beta-catenin in mice. Genes Dev. 22, 2308-41. Grondona, J. M., Fernández-Llebrez, P., Pérez, J., Cifuentes, M., Pérez-Fígares, J. M., Rodríguez, E. M., 1994a. Class-specific epitopes detected by polyclonal antibodies against the secretory products of the subcommissural organ of the dogfish <i>Scyliorhinus canicula</i>. Cell and Tissue Research. 276, 515-522. Grondona, J. M., Perez, J., Cifuentes, M., Lopez-Avalos, M. D., Nualart, F. J., Peruzzo, B., Fernandez, L. P., Rodriguez, E. M., 1994b. Analysis of the secretory glycoproteins of the subcommissural organ of the dogfish (Scyliorhinus canicula). Brain Res Mol Brain Res. 26, 299-308. Gross, C. G., 2000. Neurogenesis in the adult brain: death of a dogma. Nat Rev Neurosci. 1, 67-73. Guenther, C., Pantalena-Filho, L., Kingsley, D. M., 2008. Shaping Skeletal Growth by Modular Regulatory Elements in the Bmp5 Gene. PLoS Genet. 4, e1000308. Guillemot, F., 2005. Cellular and molecular control of neurogenesis in the mammalian telencephalon. Curr Opin Cell Biol. 17, 639-47. Halfter, W., Schurer, B., 1998. Disruption of the pial basal lamina during early avian embryonic development inhibits histogenesis and axonal pathfinding in the optic tectum. The Journal of Comparative Neurology. 397, 105-117. Hamel, B. C., Smits, A. P., Otten, B. J., van den Helm, B., Ropers, H. H., Mariman, E. C., 1996. Familial X-linked mental retardation and isolated growth hormone deficiency: clinical and molecular findings. Am J Med Genet. 64, 35-41. Hargrave, M., Karunaratne, A., Cox, L., Wood, S., Koopman, P., Yamada, T., 2000. The HMG Box Transcription Factor Gene Sox14 Marks a Novel Subset of Ventral Interneurons and Is Regulated by Sonic Hedgehog. Developmental Biology. 219, 142- 153. Harrington, M. J., Hong, E., Brewster, R., 2009. Comparative analysis of neurulation: First impressions do not count. Molecular Reproduction and Development. 76, 954-965. Harris, M. J., Juriloff, D. M., 2007. Mouse mutants with neural tube closure defects and their role in understanding human neural tube defects. Birth Defects Research Part A: Clinical and Molecular Teratology. 79, 187-210. Harris, M. J., Juriloff, D. M., 2010. An update to the list of mouse mutants with neural tube closure defects and advances toward a complete genetic perspective of neural tube closure. Birth Defects Research Part A: Clinical and Molecular Teratology. 88, 653- 669. Hashimoto-Torii, K., Motoyama, J., Hui, C.-C., Kuroiwa, A., Nakafuku, M., Shimamura, K., 2003. Differential activities of Sonic hedgehog mediated by Gli transcription factors define distinct neuronal subtypes in the dorsal thalamus. Mechanisms of Development. 120, 1097-1111. Hébert, J. M., Mishina, Y., McConnell, S. K., 2002. BMP Signaling Is Required Locally to Pattern the Dorsal Telencephalic Midline. Neuron. 35, 1029-1041. Hegardt, F. G., 1999. Mitochondrial 3-hydroxy-3-methylglutaryl-CoA synthase: a control enzyme in ketogenesis. Biochem J. 338 ( Pt 3), 569-82. X
Henke, R. M., Meredith, D. M., Borromeo, M. D., Savage, T. K., Johnson, J. E., 2009. Ascl1 and Neurog2 form novel complexes and regulate Delta-like3 (Dll3) expression in the neural tube. Developmental Biology. 328, 529-540. Hernandez, D., Janmohamed, A., Chandan, P., Omar, B. A., Phillips, I. R., Shephard, E. A., 2009. Deletion of the mouse Fmo1 gene results in enhanced pharmacological behavioural responses to imipramine. Pharmacogenet Genomics. 19, 289-99. Hikosaka, O., Sesack, S. R., Lecourtier, L., Shepard, P. D., 2008. Habenula: Crossroad between the Basal Ganglia and the Limbic System. The Journal of Neuroscience. 28, 11825-11829. Hill, R. E., Favor, J., Hogan, B. L. M., Ton, C. C. T., Saunders, G. F., Hanson, I. M., Prosser, J., Jordan, T., Hastie, N. D., Heyningen, V. v., 1991. Mouse Small eye results from mutations in a paired-like homeobox-containing gene. Nature. 354, 522-525. Hirano, S., Kimoto, N., Shimoyama, Y., Hirohashi, S., Takeichi, M., 1992. Identification of a neural [alpha]-catenin as a key regulator of cadherin function and multicellular organization. Cell. 70, 293-301. Hirata, T., Nakazawa, M., Muraoka, O., Nakayama, R., Suda, Y., Hibi, M., 2006a. Zinc- finger genes Fez and Fez-like function in the establishment of diencephalon subdivisions. Development. 133, 3993-4004. Hirata, T., Nakazawa, M., Yoshihara, S.-i., Miyachi, H., Kitamura, K., Yoshihara, Y., Hibi, M., 2006b. Zinc-finger gene Fez in the olfactory sensory neurons regulates development of the olfactory bulb non-cell-autonomously. Development. 133, 1433- 1443. Ho, I. C., Pai, S. Y., 2007. GATA-3 - not just for Th2 cells anymore. Cell Mol Immunol. 4, 15-29. Hol, F. A., Schepens, M. T., van Beersum, S. E. C., Redolfi, E., Affer, M., Vezzoni, P., Hamel, B. C. J., Karnes, P. S., Mariman, E. C. M., Zucchi, I., 2000. Identification and Characterization of an Xq26-q27 Duplication in a Family with Spina Bifida and Panhypopituitarism Suggests the Involvement of Two Distinct Genes. Genomics. 69, 174-181. Hollyday, M., McMahon, J. A., McMahon, A. P., 1995. Wnt expression patterns in chick embryo nervous system. Mechanisms of Development. 52, 9-25. Holmberg, J., Clarke, D. L., Frisen, J., 2000. Regulation of repulsion versus adhesion by different splice forms of an Eph receptor. Nature. 408, 203-206. Holmberg, J., Hansson, E., Malewicz, M., Sandberg, M., Perlmann, T., Lendahl, U., Muhr, J., 2008. SoxB1 transcription factors and Notch signaling use distinct mechanisms to regulate proneural gene function and neural progenitor differentiation. Development. 135, 1843-1851. Hong, H.-K., Chakravarti, A., Takahashi, J. S., 2004. The gene for soluble N-ethylmaleimide sensitive factor attachment protein α is mutated in hydrocephaly with hop gait (hyh) mice. Proceedings of the National Academy of Sciences of the United States of America. 101, 1748-1753. Hoser, M., Potzner, M. R., Koch, J. M. C., Bosl, M. R., Wegner, M., Sock, E., 2008. Sox12 Deletion in the Mouse Reveals Nonreciprocal Redundancy with the Related Sox4 and Sox11 Transcription Factors. Mol. Cell. Biol. 28, 4675-4687. Hoyo-Becerra, C., López-Ávalos, M., Cifuentes, M., Visser, R., Fernández-Llebrez, P., Grondona, J., 2010. The subcommissural organ and the development of the posterior commissure in chick embryos. Cell and Tissue Research. 339, 383-395. Hoyo-Becerra, C., López-Ávalos, M., Pérez, J., Miranda, E., Rojas-Ríos, P., Fernández- Llebrez, P., Grondona, J., 2006. Continuous delivery of a monoclonal antibody against Reissner’s fiber into CSF reveals CSF-soluble material immunorelated to the XI
subcommissural organ in early chick embryos. Cell and Tissue Research. 326, 771- 786. Hubbard, T. J. P., Aken, B. L., Beal, K., Ballester, B., Caccamo, M., Chen, Y., Clarke, L., Coates, G., Cunningham, F., Cutts, T., Down, T., Dyer, S. C., Fitzgerald, S., Fernandez-Banet, J., Graf, S., Haider, S., Hammond, M., Herrero, J., Holland, R., Howe, K., Howe, K., Johnson, N., Kahari, A., Keefe, D., Kokocinski, F., Kulesha, E., Lawson, D., Longden, I., Melsopp, C., Megy, K., Meidl, P., Ouverdin, B., Parker, A., Prlic, A., Rice, S., Rios, D., Schuster, M., Sealy, I., Severin, J., Slater, G., Smedley, D., Spudich, G., Trevanion, S., Vilella, A., Vogel, J., White, S., Wood, M., Cox, T., Curwen, V., Durbin, R., Fernandez-Suarez, X. M., Flicek, P., Kasprzyk, A., Proctor, G., Searle, S., Smith, J., Ureta-Vidal, A., Birney, E., 2006. Ensembl 2007. Nucl. Acids Res., gkl996. Huelsken, J., Vogel, R., Brinkmann, V., Erdmann, B., Birchmeier, C., Birchmeier, W., 2000. Requirement for β-Catenin in Anterior-Posterior Axis Formation in Mice. The Journal of Cell Biology. 148, 567-578. Huh, M. S., Todd, M. A., Picketts, D. J., 2009. SCO-ping out the mechanisms underlying the etiology of hydrocephalus. Physiology (Bethesda). 24, 117-26. Hunter, S., Apweiler, R., Attwood, T. K., Bairoch, A., Bateman, A., Binns, D., Bork, P., Das, U., Daugherty, L., Duquenne, L., Finn, R. D., Gough, J., Haft, D., Hulo, N., Kahn, D., Kelly, E., Laugraud, A. l., Letunic, I., Lonsdale, D., Lopez, R., Madera, M., Maslen, J., McAnulla, C., McDowall, J., Mistry, J., Mitchell, A., Mulder, N., Natale, D., Orengo, C., Quinn, A. F., Selengut, J. D., Sigrist, C. J. A., Thimma, M., Thomas, P. D., Valentin, F., Wilson, D., Wu, C. H., Yeats, C., 2009. InterPro: the integrative protein signature database. Nucleic Acids Research. 37, D211-D215. Ibanez-Tallon, I., Gorokhova, S., Heintz, N., 2002. Loss of function of axonemal dynein Mdnah5 causes primary ciliary dyskinesia and hydrocephalus. Hum Mol Genet. 11, 715-21. Ibanez-Tallon, I., Pagenstecher, A., Fliegauf, M., Olbrich, H., Kispert, A., Ketelsen, U.-P., North, A., Heintz, N., Omran, H., 2004. Dysfunction of axonemal dynein heavy chain Mdnah5 inhibits ependymal flow and reveals a novel mechanism for hydrocephalus formation. Hum. Mol. Genet. 13, 2133-2141. Ihaka, R., Gentleman, R., 1996. R: A Language for Data Analysis and Graphics. Journal of Computational and Graphical Statistics. 5, 299-314. Ishibashi, M., Ang, S. L., Shiota, K., Nakanishi, S., Kageyama, R., Guillemot, F., 1995. Targeted disruption of mammalian hairy and Enhancer of split homolog-1 (HES-1) leads to up-regulation of neural helix-loop-helix factors, premature neurogenesis, and severe neural tube defects. Genes & Development. 9, 3136-3148. Ishibashi, M., Moriyoshi, K., Sasai, Y., Shiota, K., Nakanishi, S., Kageyama, R., 1994. Persistent expression of helix-loop-helix factor HES-1 prevents mammalian neural differentiation in the central nervous system. Embo J. 13, 1799-805. Jackson, S. R., Guner, Y. S., Woo, R., Randolph, L. M., Ford, H., Shin, C. E., 2009. L1CAM mutation in association with X-linked hydrocephalus and Hirschsprung's disease. Pediatr Surg Int. 25, 823-5. Janmohamed, A., Hernandez, D., Phillips, I. R., Shephard, E. A., 2004. Cell-, tissue-, sex- and developmental stage-specific expression of mouse flavin-containing monooxygenases (Fmos). Biochemical Pharmacology. 68, 73-83. Jeanmougin, M., de Reynies, A., Marisa, L., Paccard, C., Nuel, G., Guedj, M., 2010. Should We Abandon the t-Test in the Analysis of Gene Expression Microarray Data: A Comparison of Variance Modeling Strategies. PLoS ONE. 5, e12336. XII
Johansson, P., Dziegielewska, K., Ek, C., Habgood, M., Møllgård, K., Potter, A., Schuliga, M., Saunders, N., 2005. Aquaporin-1 in the choroid plexuses of developing mammalian brain. Cell and Tissue Research. 322, 353-364. Jones, H. C., Bucknall, R. M., 1988. Inherited prenatal hydrocephalus in the H-Tx rat: a morphological study. Neuropathol Appl Neurobiol. 14, 263-74. Jones, H. C., Dack, S., Ellis, C., 1987. Morphological aspects of the development of hydrocephalus in a mouse mutant (SUMS/NP). Acta Neuropathologica. 72, 268-276. Jouet, M., Rosenthal, A., MacFarlane, J., Kenwrick, S., Donnai, D., 1993. A missense mutation confirms the L1 defect in X-linked hydrocephalus (HSAS). Nat Genet. 4, 331. Kadokawa, Y., Marunouchi, T., 2002. Chimeric analysis of Notch2 function: a role for Notch2 in the development of the roof plate of the mouse brain. Dev Dyn. 225, 126- 34. Kahane, N., Kalcheim, C., 1998. Identification of early postmitotic cells in distinct embryonic sites and their possible roles in morphogenesis. Cell and Tissue Research. 294, 297-307. Kamachi, Y., Sockanathan, S., Liu, Q., Breitman, M., Lovell-Badge, R., Kondoh, H., 1995. Involvement of SOX proteins in lens-specific activation of crystallin genes. Embo J. 14, 3510-9. Kamachi, Y., Uchikawa, M., Collignon, J., Lovell-Badge, R., Kondoh, H., 1998. Involvement of Sox1, 2 and 3 in the early and subsequent molecular events of lens induction. Development. 125, 2521-32. Kamachi, Y., Uchikawa, M., Tanouchi, A., Sekido, R., Kondoh, H., 2001. Pax6 and SOX2 form a co-DNA-binding partner complex that regulates initiation of lens development. Genes Dev. 15, 1272-86. Kamata, T., Katsube, K.-i., Michikawa, M., Yamada, M., Takada, S., Mizusawa, H., 2004. R- spondin, a novel gene with thrombospondin type 1 domain, was expressed in the dorsal neural tube and affected in Wnts mutants. Biochimica et Biophysica Acta (BBA) - Gene Structure and Expression. 1676, 51-62. Kan, L., Israsena, N., Zhang, Z., Hu, M., Zhao, L.-R., Jalali, A., Sahni, V., Kessler, J. A., 2004. Sox1 acts through multiple independent pathways to promote neurogenesis. Developmental Biology. 269, 580-594. Kan, L., Jalali, A., Zhao, L.-R., Zhou, X., McGuire, T., Kazanis, I., Episkopou, V., Bassuk, A. G., Kessler, J. A., 2007. Dual function of Sox1 in telencephalic progenitor cells. Developmental Biology. 310, 85-98. Kanai, Y., Kanai-Azuma, M., Noce, T., Saido, T. C., Shiroishi, T., Hayashi, Y., Yazaki, K., 1996. Identification of two Sox17 messenger RNA isoforms, with and without the high mobility group box region, and their differential expression in mouse spermatogenesis. J. Cell Biol. 133, 667-681. Kaprielian, Z., Imondi, R., Runko, E., 2000. Axon guidance at the midline of the developing CNS. The Anatomical Record. 261, 176-197. Kaprielian, Z., Runko, E., Imondi, R., 2001. Axon guidance at the midline choice point. Developmental Dynamics. 221, 154-181. Kazanskaya, O., Ohkawara, B., Heroult, M., Wu, W., Maltry, N., Augustin, H. G., Niehrs, C., 2008. The Wnt signaling regulator R-spondin 3 promotes angioblast and vascular development. Development. 135, 3655-3664. Kelberman, D., de Castro, S. C. P., Huang, S., Crolla, J. A., Palmer, R., Gregory, J. W., Taylor, D., Cavallo, L., Faienza, M. F., Fischetto, R., Achermann, J. C., Martinez- Barbera, J. P., Rizzoti, K., Lovell-Badge, R., Robinson, I. C. A. F., Gerrelli, D., XIII
Dattani, M. T., 2008. SOX2 Plays a Critical Role in the Pituitary, Forebrain, and Eye during Human Embryonic Development. J Clin Endocrinol Metab. 93, 1865-1873. Kelly, R. G., Buckingham, M. E., 2000. Modular regulation of the MLC1F/3F gene and striated muscle diversity. Microsc Res Tech. 50, 510-21. Kiefer, J. C., 2007. Back to basics: Sox genes. Developmental Dynamics. 236, 2356-2366. Kim, H.-J., McMillan, E., Han, F., Svendsen, C. N., 2009. Regionally Specified Human Neural Progenitor Cells Derived from the Mesencephalon and Forebrain Undergo Increased Neurogenesis Following Overexpression of ASCL1. STEM CELLS. 27, 390-398. Kim, K.-A., Wagle, M., Tran, K., Zhan, X., Dixon, M. A., Liu, S., Gros, D., Korver, W., Yonkovich, S., Tomasevic, N., Binnerts, M., Abo, A., 2008. R-Spondin Family Members Regulate the Wnt Pathway by a Common Mechanism. Mol. Biol. Cell. 19, 2588-2596. Kim, T.-H., Goodman, J., Anderson, K. V., Niswander, L., 2007. Phactr4 Regulates Neural Tube and Optic Fissure Closure by Controlling PP1-, Rb-, and E2F1-Regulated Cell- Cycle Progression. Developmental Cell. 13, 87-102. Kimura, C., Yoshinaga, K., Tian, E., Suzuki, M., Aizawa, S., Matsuo, I., 2000. Visceral Endoderm Mediates Forebrain Development by Suppressing Posteriorizing Signals. Developmental Biology. 225, 304-321. Kirchhamer, C. V., Yuh, C. H., Davidson, E. H., 1996. Modular cis-regulatory organization of developmentally expressed genes: two genes transcribed territorially in the sea urchin embryo, and additional examples. Proc Natl Acad Sci U S A. 93, 9322-8. Kishi, M., Mizuseki, K., Sasai, N., Yamazaki, H., Shiota, K., Nakanishi, S., Sasai, Y., 2000. Requirement of Sox2-mediated signaling for differentiation of early Xenopus neuroectoderm. Development. 127, 791-800. Klingensmith, J., Ang, S.-L., Bachiller, D., Rossant, J., 1999. Neural Induction and Patterning in the Mouse in the Absence of the Node and Its Derivatives. Developmental Biology. 216, 535-549. Ko, T.-L., Chien, C.-L., Lu, K.-S., 2005. The expression of α-internexin and peripherin in the developing mouse pineal gland. Journal of Biomedical Science. 12, 777-789. Kobayashi, Y., Watanabe, M., Okada, Y., Sawa, H., Takai, H., Nakanishi, M., Kawase, Y., Suzuki, H., Nagashima, K., Ikeda, K., Motoyama, N., 2002. Hydrocephalus, Situs Inversus, Chronic Sinusitis, and Male Infertility in DNA Polymerase λ-Deficient Mice: Possible Implication for the Pathogenesis of Immotile Cilia Syndrome. Mol. Cell. Biol. 22, 2769-2776. Koster, R. W., Kuhnlein, R. P., Wittbrodt, J., 2000. Ectopic Sox3 activity elicits sensory placode formation. Mech Dev. 95, 175-87. Kostiuk, M. A., Keller, B. O., Berthiaume, L. G., 2010. Palmitoylation of ketogenic enzyme HMGCS2 enhances its interaction with PPARα and transcription at the Hmgcs2 PPRE. The FASEB Journal. 24, 1914-1924. Koukouritaki, S. B., Simpson, P., Yeung, C. K., Rettie, A. E., Hines, R. N., 2002. Human hepatic flavin-containing monooxygenases 1 (FMO1) and 3 (FMO3) developmental expression. Pediatr Res. 51, 236-43. Kovacevic-Grujicic, N., Mojsin, M., Djurovic, J., Petrovic, I., Stevanovic, M., 2007. Comparison of promoter regions of SOX3, SOX14 and SOX18 orthologs in mammals. DNA Seq. 1. Krebs, D. L., Metcalf, D., Merson, T. D., Voss, A. K., Thomas, T., Zhang, J. G., Rakar, S., O'Bryan M, K., Willson, T. A., Viney, E. M., Mielke, L. A., Nicola, N. A., Hilton, D. J., Alexander, W. S., 2004. Development of hydrocephalus in mice lacking SOCS7. Proc Natl Acad Sci U S A. 101, 15446-51. XIV
Krispin, S., Nitzan, E., Kalcheim, C., 2010a. The dorsal neural tube: A dynamic setting for cell fate decisions. Developmental Neurobiology. 70, 796-812. Krispin, S., Nitzan, E., Kassem, Y., Kalcheim, C., 2010b. Evidence for a dynamic spatiotemporal fate map and early fate restrictions of premigratory avian neural crest. Development. 137, 585-595. Krstic, A., Mojsin, M., Stevanovic, M., 2007. Regulation of SOX3 gene expression is driven by multiple NF-Y binding elements. Archives of Biochemistry and Biophysics. 467, 163-173. Krueger, S. K., Williams, D. E., 2005. Mammalian flavin-containing monooxygenases: structure/function, genetic polymorphisms and role in drug metabolism. Pharmacol Ther. 106, 357-87. Kubo, F., Takeichi, M., Nakagawa, S., 2005. Wnt2b inhibits differentiation of retinal progenitor cells in the absence of Notch activity by downregulating the expression of proneural genes. Development. 132, 2759-2770. Kuroda, T., Tada, M., Kubota, H., Kimura, H., Hatano, S. Y., Suemori, H., Nakatsuji, N., Tada, T., 2005. Octamer and Sox elements are required for transcriptional cis regulation of Nanog gene expression. Mol Cell Biol. 25, 2475-85. Lagerstrom-Fermer, M., Sundvall, M., Johnsen, E., Warne, G. L., Forrest, S. M., Zajac, J. D., Rickards, A., Ravine, D., Landegren, U., Pettersson, U., 1997. X-linked recessive panhypopituitarism associated with a regional duplication in Xq25-q26. Am J Hum Genet. 60, 910-6. Lang, B., Song, B., Davidson, W., MacKenzie, A., Smith, N., McCaig, C. D., Harmar, A. J., Shen, S., 2006. Expression of the human PAC1 receptor leads to dose-dependent hydrocephalus-related abnormalities in mice. J Clin Invest. 116, 1924-34. Laronda, M. M., Jameson, J. L., 2011. Sox3 Functions in a Cell-Autonomous Manner to Regulate Spermatogonial Differentiation in Mice. Endocrinology. 152, 1606-1615. Laumonnier, F., Ronce, N., Hamel, B. C., Thomas, P., Lespinasse, J., Raynaud, M., Paringaux, C., Van Bokhoven, H., Kalscheuer, V., Fryns, J. P., Chelly, J., Moraine, C., Briault, S., 2002. Transcription factor SOX3 is involved in X-linked mental retardation with growth hormone deficiency. Am J Hum Genet. 71, 1450-5. Le Goff, C., Somerville, R. P. T., Kesteloot, F., Powell, K., Birk, D. E., Colige, A. C., Apte, S. S., 2006. Regulation of procollagen amino-propeptide processing during mouse embryogenesis by specialization of homologous ADAMTS proteases: insights on collagen biosynthesis and dermatosparaxis. Development. 133, 1587-1596. Lechtreck, K. F., Delmotte, P., Robinson, M. L., Sanderson, M. J., Witman, G. B., 2008. Mutations in Hydin impair ciliary motility in mice. J Cell Biol. 180, 633-43. Lechtreck, K. F., Witman, G. B., 2007. Chlamydomonas reinhardtii hydin is a central pair protein required for flagellar motility. J Cell Biol. 176, 473-82. Lee, J. E., Wu, S. F., Goering, L. M., Dorsky, R. I., 2006. Canonical Wnt signaling through Lef1 is required for hypothalamic neurogenesis. Development. 133, 4451-61. Lee, K. J., Dietrich, P., Jessell, T. M., 2000a. Genetic ablation reveals that the roof plate is essential for dorsal interneuron specification. Nature. 403, 734-740. Lee, K. J., Mendelsohn, M., Jessell, T. M., 1998. Neuronal patterning by BMPs: a requirement for GDF7 in the generation of a discrete class of commissural interneurons in the mouse spinal cord. Genes Dev. 12, 3394-407. Lee, L., Campagna, D. R., Pinkus, J. L., Mulhern, H., Wyatt, T. A., Sisson, J. H., Pavlik, J. A., Pinkus, G. S., Fleming, M. D., 2008. Primary Ciliary Dyskinesia in Mice Lacking the Novel Ciliary Protein Pcdp1. Mol. Cell. Biol. 28, 949-957. XV
Lee, S. K., Kang, M. J., Jin, C., In, M. K., Kim, D. H., Yoo, H. H., 2009. Flavin-containing monooxygenase 1-catalysed N, N-dimethylamphetamine N-oxidation. Xenobiotica. 39, 680-686. Lee, S. M., Tole, S., Grove, E., McMahon, A. P., 2000b. A local Wnt-3a signal is required for development of the mammalian hippocampus. Development. 127, 457-467. Lefebvre, V., Dumitriu, B., Penzo-Mendez, A., Han, Y., Pallavi, B., 2007. Control of cell fate and differentiation by Sry-related high-mobility-group box (Sox) transcription factors. Int J Biochem Cell Biol. Lein, E. S., Hawrylycz, M. J., Ao, N., Ayres, M., Bensinger, A., Bernard, A., Boe, A. F., Boguski, M. S., Brockway, K. S., Byrnes, E. J., Chen, L., Chen, L., Chen, T.-M., Chi Chin, M., Chong, J., Crook, B. E., Czaplinska, A., Dang, C. N., Datta, S., Dee, N. R., Desaki, A. L., Desta, T., Diep, E., Dolbeare, T. A., Donelan, M. J., Dong, H.-W., Dougherty, J. G., Duncan, B. J., Ebbert, A. J., Eichele, G., Estin, L. K., Faber, C., Facer, B. A., Fields, R., Fischer, S. R., Fliss, T. P., Frensley, C., Gates, S. N., Glattfelder, K. J., Halverson, K. R., Hart, M. R., Hohmann, J. G., Howell, M. P., Jeung, D. P., Johnson, R. A., Karr, P. T., Kawal, R., Kidney, J. M., Knapik, R. H., Kuan, C. L., Lake, J. H., Laramee, A. R., Larsen, K. D., Lau, C., Lemon, T. A., Liang, A. J., Liu, Y., Luong, L. T., Michaels, J., Morgan, J. J., Morgan, R. J., Mortrud, M. T., Mosqueda, N. F., Ng, L. L., Ng, R., Orta, G. J., Overly, C. C., Pak, T. H., Parry, S. E., Pathak, S. D., Pearson, O. C., Puchalski, R. B., Riley, Z. L., Rockett, H. R., Rowland, S. A., Royall, J. J., Ruiz, M. J., Sarno, N. R., Schaffnit, K., Shapovalova, N. V., Sivisay, T., Slaughterbeck, C. R., Smith, S. C., Smith, K. A., Smith, B. I., Sodt, A. J., Stewart, N. N., Stumpf, K.-R., Sunkin, S. M., Sutram, M., Tam, A., Teemer, C. D., Thaller, C., Thompson, C. L., Varnam, L. R., Visel, A., Whitlock, R. M., Wohnoutka, P. E., Wolkey, C. K., Wong, V. Y., Wood, M., Yaylaoglu, M. B., Young, R. C., Youngstrom, B. L., Feng Yuan, X., Zhang, B., Zwingman, T. A., Jones, A. R., 2007. Genome-wide atlas of gene expression in the adult mouse brain. Nature. 445, 168-176. Levine, A. J., Brivanlou, A. H., 2007. Proposal of a model of mammalian neural induction. Dev Biol. 308, 247-56. Levine, E., Lee, C. H., Kintner, C., Gumbiner, B. M., 1994. Selective disruption of E- cadherin function in early Xenopus embryos by a dominant negative mutant. Development. 120, 901-909. Liem, K. F., Tremml, G., Jessell, T. M., 1997. A Role for the Roof Plate and Its Resident TGF[beta]-Related Proteins in Neuronal Patterning in the Dorsal Spinal Cord. Cell. 91, 127-138. Liem, K. F., Tremml, G., Roelink, H., Jessell, T. M., 1995. Dorsal differentiation of neural plate cells induced by BMP-mediated signals from epidermal ectoderm. Cell. 82, 969- 979. Lien, W.-H., Klezovitch, O., Vasioukhin, V., 2006. Cadherin-catenin proteins in vertebrate development. Current Opinion in Cell Biology. 18, 499-506. Lim, Y., Golden, J. A., 2007. Patterning the developing diencephalon. Brain Research Reviews. 53, 17-26. Lindeman, G. J., Dagnino, L., Gaubatz, S., Xu, Y., Bronson, R. T., Warren, H. B., Livingston, D. M., 1998. A specific, nonproliferative role for E2F-5 in choroid plexus function revealed by gene targeting. Genes & Development. 12, 1092-1098. Lindwall, C., Fothergill, T., Richards, L. J., 2007. Commissure formation in the mammalian forebrain. Current Opinion in Neurobiology. 17, 3-14. Liu, P., Wakamiya, M., Shea, M. J., Albrecht, U., Behringer, R. R., Bradley, A., 1999. Requirement for Wnt3 in vertebrate axis formation. Nat Genet. 22, 361-5. XVI
Liu, Y., Helms, A. W., Johnson, J. E., 2004. Distinct activities of Msx1 and Msx3 in dorsal neural tube development. Development. 131, 1017-1028. Liu, Y., Shi, J., Lu, C.-C., Wang, Z.-B., Lyuksyutova, A. I., Song, X.-J., Zou, Y., 2005. Ryk- mediated Wnt repulsion regulates posterior-directed growth of corticospinal tract. Nat Neurosci. 8, 1151-1159. López-Avalos, M. D., Pérez, J., Pérez-Fígares, J. M., Peruzzo, B., Grondona, J. M., Rodríguez, E. M., 1995. Secretory glycoproteins of the subcommissural organ of the dogfish (Scyliorhinus canicula): evidence for the existence of precursor and processed forms. Cell and Tissue Research. 283, 75-84. Lösecke, W., Naumann, W., Sterba, G., 1986. Immuno-electron-microscopic analysis of the basal route of secretion in the subcommissural organ of the rabbit. Cell and Tissue Research. 244, 449-456. Louvi, A., Wassef, M., 2000. Ectopic engrailed 1 expression in the dorsal midline causes cell death, abnormal differentiation of circumventricular organs and errors in axonal pathfinding. Development. 127, 4061-4071. Lowery, L. N., Sive, H., 2009. Totally tubular: the mystery behind function and origin of the brain ventricular system. BioEssays. 31, 446-458. Ma, D. K., Bonaguidi, M. A., Ming, G.-l., Song, H., 2009. Adult neural stem cells in the mammalian central nervous system. Cell Res. 19, 672-682. Ma, X., Bao, J., Adelstein, R. S., 2007. Loss of Cell Adhesion Causes Hydrocephalus in Nonmuscle Myosin II-B-ablated and Mutated Mice. Mol. Biol. Cell. 18, 2305-2312. MacFarlane, J. R., Du, J.-S., Pepys, M. E., Ramsden, S., Donnai, D., Charlton, R., Garrett, C., Tolmie, J., Yates, J. R. W., Berry, C., Goudie, D., Moncla, A., Lunt, P., Hodgson, S., Jouet, M., Kenwrick, S., 1997. Nine novel L1 CAM mutations in families with X- linked hydrocephalus. Human Mutation. 9, 512-518. Mailhot, G., Yang, M., Mason-Savas, A., Mackay, C. A., Leav, I., Odgren, P. R., 2008. BMP-5 expression increases during chondrocyte differentiation in vivo and in vitro and promotes proliferation and cartilage matrix synthesis in primary chondrocyte cultures. J Cell Physiol. 214, 56-64. Malas, S., Postlethwaite, M., Ekonomou, A., Whalley, B., Nishiguchi, S., Wood, H., Meldrum, B., Constanti, A., Episkopou, V., 2003. Sox1-deficient mice suffer from epilepsy associated with abnormal ventral forebrain development and olfactory cortex hyperexcitability. Neuroscience. 119, 421-432. Mao, X., Enno, T. L., Del Bigio, M. R., 2006. Aquaporin4 changes in rat brain with severe hydrocephalus. European Journal of Neuroscience. 23, 2929-2936. Mashayekhi, F., Draper, C. E., Bannister, C. M., Pourghasem, M., Owen-Lynch, P. J., Miyan, J. A., 2002. Deficient cortical development in the hydrocephalic Texas (H-Tx) rat: a role for CSF. Brain. 125, 1859-1874. Mastick, G. S., Andrews, G. L., 2001. Pax6 Regulates the Identity of Embryonic Diencephalic Neurons. Molecular and Cellular Neuroscience. 17, 190-207. Matsuo, I., Kuratani, S., Kimura, C., Takeda, N., Aizawa, S., 1995. Mouse Otx2 functions in the formation and patterning of rostral head. Genes & Development. 9, 2646-2658. Megason, S. G., McMahon, A. P., 2002. A mitogen gradient of dorsal midline Wnts organizes growth in the CNS. Development. 129, 2087-2098. Meiniel, A., 2007. The secretory ependymal cells of the subcommissural organ: which role in hydrocephalus? Int J Biochem Cell Biol. 39, 463-8. Meiniel, O., Meiniel, A., 2007. The complex multidomain organization of SCO-spondin protein is highly conserved in mammals. Brain Research Reviews. 53, 321-327. Meiniel, R., Meiniel, A., 1985. Analysis of the secretions of the subcommissural organs of several vertebrate species by use of fluorescent lectins. Cell Tissue Res. 239, 359-64. XVII
Merrill, R. A., Ahrens, J. M., Kaiser, M. E., Federhart, K. S., Poon, V. Y., Clagett-Dame, M., 2004. All-trans retinoic acid-responsive genes identified in the human SH-SY5Y neuroblastoma cell line and their regulated expression in the nervous system of early embryos. Biological Chemistry. 385, 605-614. Mertin, S., McDowall, S. G., Harley, V. R., 1999. The DNA-binding specificity of SOX9 and otherSOX proteins. Nucleic Acids Research. 27, 1359-1364. Milhorat, T. H., 1975. The third circulation revisited. J Neurosurg. 42, 628-45. Millevoi, S., Thion, L., Joseph, G., Vossen, C., Ghisolfi-Nieto, L., Erard, M., 2001. Atypical binding of the neuronal POU protein N-Oct3 to noncanonical DNA targets. European Journal of Biochemistry. 268, 781-791. Millonig, J. H., Millen, K. J., Hatten, M. E., 2000. The mouse Dreher gene Lmx1a controls formation of the roof plate in the vertebrate CNS. Nature. 403, 764-9. Minoux, M., Rijli, F. M., 2010. Molecular mechanisms of cranial neural crest cell migration and patterning in craniofacial development. Development. 137, 2605-2621. Miquelajauregui, A., Varela-Echavarraa, A., Ceci, M. L., Garcia-Moreno, F., Ricano, I., Hoang, K., Frade-Perez, D., Portera-Cailliau, C., Tamariz, E., De Carlos, J. A., Westphal, H., Zhao, Y., 2010. LIM-Homeobox Gene Lhx5 Is Required for Normal Development of Cajal-Retzius Cells. The Journal of Neuroscience. 30, 10551-10562. Miyan, J. A., Nabiyouni, M., Zendah, M., 2003. Development of the brain: a vital role for cerebrospinal fluid. Can J Physiol Pharmacol. 81, 317-28. Miyan, J. A., Zendah, M., Mashayekhi, F., Owen-Lynch, P. J., 2006. Cerebrospinal fluid supports viability and proliferation of cortical cells in vitro, mirroring in vivo development. Cerebrospinal Fluid Res. 3, 2. Mizuseki, K., Kishi, M., Matsui, M., Nakanishi, S., Sasai, Y., 1998. Xenopus Zic-related-1 and Sox-2, two factors induced by chordin, have distinct activities in the initiation of neural induction. Development. 125, 579-587. Mojsin, M., Grujicic, N. K., Nikcevic, G., Krstic, A., Savic, T., Stevanovic, M., 2006. Mapping of the RXRα binding elements involved in retinoic acid induced transcriptional activation of the human SOX3 gene. Neuroscience Research. 56, 409- 418. Møller, M., Baeres, F., 2002. The anatomy and innervation of the mammalian pineal gland. Cell and Tissue Research. 309, 139-150. Monnerie, H., Boespflug-Tanguy, O., Dastugue, B., Meiniel, A., 1995. Reissner’s fibre supports the survival of chick cortical neurons in primary mixed cultures. Cell and Tissue Research. 282, 81-91. Monnerie, H., Boespflug-Tanguy, O., Dastugue, B., Meiniel, A., 1996. Soluble material from Reissner's fiber displays anti-aggregative activity in primary cultures of chick cortical neurons. Developmental Brain Research. 96, 120-129. Monnerie, H., Dastugue, B., Meiniel, A., 1997. Reissner's fibre promotes neuronal aggregation and influences neuritic outgrowth in vitro. Cell and Tissue Research. 287, 285-295. Monnerie, H., Dastugue, B., Meiniel, A., 1998. Effect of synthetic peptides derived from SCO-spondin conserved domains on chick cortical and spinal-cord neurons in cell cultures. Cell and Tissue Research. 293, 407-418. Monuki, E. S., Porter, F. D., Walsh, C. A., 2001. Patterning of the Dorsal Telencephalon and Cerebral Cortex by a Roof Plate-Lhx2 Pathway. Neuron. 32, 591-604. Mortlock, D. P., Guenther, C., Kingsley, D. M., 2003. A General Approach for Identifying Distant Regulatory Elements Applied to the Gdf6 Gene. Genome Research. 13, 2069- 2081. XVIII
Muroyama, Y., Fujihara, M., Ikeya, M., Kondoh, H., Takada, S., 2002. Wnt signaling plays an essential role in neuronal specification of the dorsal spinal cord. Genes Dev. 16, 548-53. Naar, A. M., Taatjes, D. J., Zhai, W., Nogales, E., Tjian, R., 2002. Human CRSP interacts with RNA polymerase II CTD and adopts a specific CTD-bound conformation. Genes Dev. 16, 1339-44. Nabi, U., Nagi, A. H., Sami, W., 2008. Ki-67 proliferating index and histological grade, type and stage of colorectal carcinoma. J Ayub Med Coll Abbottabad. 20, 44-8. Nakagawa, Y., O'Leary, D. D. M., 2001. Combinatorial Expression Patterns of LIM- Homeodomain and Other Regulatory Genes Parcellate Developing Thalamus. The Journal of Neuroscience. 21, 2711-2725. Nakamura, Y., Sakakibara, S.-i., Miyata, T., Ogawa, M., Shimazaki, T., Weiss, S., Kageyama, R., Okano, H., 2000. The bHLH Gene Hes1 as a Repressor of the Neuronal Commitment of CNS Stem Cells. J. Neurosci. 20, 283-293. Nam, J.-S., Turcotte, T. J., Smith, P. F., Choi, S., Yoon, J. K., 2006. Mouse Cristin/R-spondin Family Proteins Are Novel Ligands for the Frizzled 8 and LRP6 Receptors and Activate Catenin-dependent Gene Expression. Journal of Biological Chemistry. 281, 13247-13257. Nardelli, J., Thiesson, D., Fujiwara, Y., Tsai, F.-Y., Orkin, S. H., 1999. Expression and Genetic Interaction of Transcription Factors GATA-2 and GATA-3 during Development of the Mouse Central Nervous System. Developmental Biology. 210, 305-321. Navratilova, P., Fredman, D., Hawkins, T. A., Turner, K., Lenhard, B., Becker, T. S., 2009. Systematic human/zebrafish comparative identification of cis-regulatory activity around vertebrate developmental transcription factor genes. Developmental Biology. 327, 526-540. Nechiporuk, T., Fernandez, T. E., Vasioukhin, V., 2007. Failure of epithelial tube maintenance causes hydrocephalus and renal cysts in dlg5(-/-) mice. Dev Cell. 13, 338-50. Neves, J., Kamaid, A., Alsina, B., Giraldez, F., 2007. Differential expression of Sox2 and Sox3 in neuronal and sensory progenitors of the developing inner ear of the chick. J Comp Neurol. 503, 487-500. Nicholson, D. W., Ali, A., Thornberry, N. A., Vaillancourt, J. P., Ding, C. K., Gallant, M., Gareau, Y., Griffin, P. R., Labelle, M., Lazebnik, Y. A., Munday, N. A., Raju, S. M., Smulson, M. E., Yamin, T.-T., Yu, V. L., Miller, D. K., 1995. Identification and inhibition of the ICE/CED-3 protease necessary for mammalian apoptosis. Nature. 376, 37-43. Nikcevic, G., Savic, T., Kovacevic-Grujicic, N., Stevanovic, M., 2008. Up-regulation of the SOX3 gene expression by retinoic acid: characterization of the novel promoter- response element and the retinoid receptors involved. J Neurochem. 107, 1206-15. Nishida, A., Furukawa, A., Koike, C., Tano, Y., Aizawa, S., Matsuo, I., Furukawa, T., 2003. Otx2 homeobox gene controls retinal photoreceptor cell fate and pineal gland development. Nat Neurosci. 6, 1255-1263. Nishiguchi, S., Wood, H., Kondoh, H., Lovell-Badge, R., Episkopou, V., 1998. Sox1 directly regulates the gamma-crystallin genes and is essential for lens development in mice. Genes Dev. 12, 776-81. Nitta, K. R., Takahashi, S., Haramoto, Y., Fukuda, M., Onuma, Y., Asashima, M., 2006. Expression of Sox1 during Xenopus early embryogenesis. Biochemical and Biophysical Research Communications. 351, 287-293. XIX
Nualart, F., Hein, S., 2001. Biosynthesis and molecular biology of the secretory proteins of the subcommissural organ. Microsc Res Tech. 52, 468-83. Nualart, F., Hein, S., Rodriguez, E. M., Oksche, A., 1991. Identification and partial characterization of the secretory glycoproteins of the bovine subcommissural organ- Reissner's fiber complex. Evidence for the existence of two precursor forms. Brain Res Mol Brain Res. 11, 227-38. Nualart, F., Rodriguez, E. M., 1996. Immunochemical analysis of the subcommissural organ- Reissner's fiber complex using antibodies against alkylated and deglycosylated glycoproteins of the bovine Reissner's fiber. Cell Tissue Res. 286, 23-31. Ohkawara, B., Glinka, A., Niehrs, C., 2011. Rspo3 Binds Syndecan 4 and Induces Wnt/PCP Signaling via Clathrin-Mediated Endocytosis to Promote Morphogenesis. Developmental Cell. 20, 303-314. Ohtsuka, T., Sakamoto, M., Guillemot, F., Kageyama, R., 2001. Roles of the Basic Helix- Loop-Helix Genes Hes1 and Hes5 in Expansion of Neural Stem Cells of the Developing Brain. J. Biol. Chem. 276, 30467-30474. Okamoto, N., Del Maestro, R., Valero, R., Monros, E., Poo, P., Kanemura, Y., Yamasaki, M., 2004. Hydrocephalus and Hirschsprung’s disease with a mutation of L1CAM. Journal of Human Genetics. 49, 334-337. Okamoto, N., Wada, Y., Goto, M., 1997. Hydrocephalus and Hirschsprung's disease in a patient with a mutation of L1CAM. J Med Genet. 34, 670-1. Okuda, Y., Yoda, H., Uchikawa, M., Furutani-Seiki, M., Takeda, H., Kondoh, H., Kamachi, Y., 2006. Comparative genomic and expression analysis of group B1 sox genes in zebrafish indicates their diversification during vertebrate evolution. Dev Dyn. 235, 811-25. Overholser, M. D., Whitley, J. R., O'Dell, B. L., Hogan, A. G., 1954. The ventricular system in hydrocephalic rat brains produced by a deficiency of vitamin B12 or of folic acid in the maternal diet. Anat Rec. 120, 917-33. Overton, P. M., Meadows, L. A., Urban, J., Russell, S., 2002. Evidence for differential and redundant function of the Sox genes Dichaete and SoxN during CNS development in Drosophila. Development. 129, 4219-28. Page, G., Edwards, J., Gadbury, G., Yelisetti, P., Wang, J., Trivedi, P., Allison, D., 2006. The PowerAtlas: a power and sample size atlas for microarray experimental design and research. BMC Bioinformatics. 7, 84. Pattyn, A., Guillemot, F., Brunet, J.-F., 2006. Delays in neuronal differentiation in Mash1/Ascl1 mutants. Developmental Biology. 295, 67-75. Paul, L. K., Brown, W. S., Adolphs, R., Tyszka, J. M., Richards, L. J., Mukherjee, P., Sherr, E. H., 2007. Agenesis of the corpus callosum: genetic, developmental and functional aspects of connectivity. Nat Rev Neurosci. 8, 287-299. Peng, G., Westerfield, M., 2006. Lhx5 promotes forebrain development and activates transcription of secreted Wnt antagonists. Development. 133, 3191-3200. Pennacchio, L. A., Ahituv, N., Moses, A. M., Prabhakar, S., Nobrega, M. A., Shoukry, M., Minovitsky, S., Dubchak, I., Holt, A., Lewis, K. D., Plajzer-Frick, I., Akiyama, J., De Val, S., Afzal, V., Black, B. L., Couronne, O., Eisen, M. B., Visel, A., Rubin, E. M., 2006. In vivo enhancer analysis of human conserved non-coding sequences. Nature. 444, 499-502. Penzel, R., Oschwald, R., Chen, Y., Tacke, L., Grunz, H., 1997. Characterization and early embryonic expression of a neural specific transcription factor xSOX3 in Xenopus laevis. Int J Dev Biol. 41, 667-77. Perez-Figares, J. M., Jimenez, A. J., Perez-Martin, M., Fernandez-Llebrez, P., Cifuentes, M., Riera, P., Rodriguez, S., Rodriguez, E. M., 1998. Spontaneous congenital XX
hydrocephalus in the mutant mouse hyh. Changes in the ventricular system and the subcommissural organ. J Neuropathol Exp Neurol. 57, 188-202. Pérez-Fígares, J. M., Jimenez, A. J., Rodríguez, E. M., 2001. Subcommissural organ, cerebrospinal fluid circulation, and hydrocephalus. Microscopy Research and Technique. 52, 591-607. Pérez, J., Peruzzo, B., Estivill-Torrús, G., Cifuentes, M., Schoebitz, K., Rodriguez, E., Fernández-Llebrez, P., 1995. Light- and electron-microscopic immunocytochemical investigation of the subcommissural organ using a set of monoclonal antibodies against the bovine Reissner's fiber. Histochemistry and Cell Biology. 104, 221-232. Peruzzo, B., Perez, J., Fernandez-Llebrez, P., Perez-Figares, J. M., Rodriguez, E. M., Oksche, A., 1990. Ultrastructural immunocytochemistry and lectin histochemistry of the subcommissural organ in the snake Natrix maura with particular emphasis on its vascular and leptomeningeal projections. Histochemistry. 93, 269-77. Pevny, L. H., Sockanathan, S., Placzek, M., Lovell-Badge, R., 1998. A role for SOX1 in neural determination. Development. 125, 1967-1978. Picketts, D. J., 2006. Neuropeptide signaling and hydrocephalus: SCO with the flow. J Clin Invest. 116, 1828-32. Pillai, A., Mansouri, A., Behringer, R., Westphal, H., Goulding, M., 2007. Lhx1 and Lhx5 maintain the inhibitory-neurotransmitter status of interneurons in the dorsal spinal cord. Development. 134, 357-366. Popovic, J., Klajn, A., Petrovic, I., Stevanovic, M., 2010. Tissue-specific Forkhead protein FOXA2 up-regulates SOX14 gene expression. Biochimica et Biophysica Acta (BBA) - Gene Regulatory Mechanisms. 1799, 411-418. Puelles, E., Acampora, D., Gogoi, R., Tuorto, F., Papalia, A., Guillemot, F., Ang, S. L., Simeone, A., 2006. Otx2 controls identity and fate of glutamatergic progenitors of the thalamus by repressing GABAergic differentiation. J Neurosci. 26, 5955-64. Puelles, L., Rubenstein, J. L. R., 2003. Forebrain gene expression domains and the evolving prosomeric model. Trends in Neurosciences. 26, 469-476. Pyrgaki, C., Trainor, P., Hadjantonakis, A.-K., Niswander, L., 2010. Dynamic imaging of mammalian neural tube closure. Developmental Biology. 344, 941-947. Radice, G. L., Rayburn, H., Matsunami, H., Knudsen, K. A., Takeichi, M., Hynes, R. O., 1997. Developmental Defects in Mouse Embryos Lacking N-Cadherin. Developmental Biology. 181, 64-78. Rahmanzadeh, R., Hüttmann, G., Gerdes, J., Scholzen, T., 2007. Chromophore-assisted light inactivation of pKi-67 leads to inhibition of ribosomal RNA synthesis. Cell Proliferation. 40, 422-430. Rakic, P., Sidman, R. L., 1968. Subcommissural organ and adjacent ependyma: autoradiographic study of their origin in the mouse brain. Am J Anat. 122, 317-35. Raverot, G., Weiss, J., Park, S. Y., Hurley, L., Jameson, J. L., 2005. Sox3 expression in undifferentiated spermatogonia is required for the progression of spermatogenesis. Dev Biol. 283, 215-25. Raynaud, M., Ronce, N., Ayrault, A. D., Francannet, C., Malpuech, G., Moraine, C., 1998. X-linked mental retardation with isolated growth hormone deficiency is mapped to Xq22-Xq27.2 in one family. Am J Med Genet. 76, 255-61. Reese, T. A., Liang, H. E., Tager, A. M., Luster, A. D., Van Rooijen, N., Voehringer, D., Locksley, R. M., 2007. Chitin induces accumulation in tissue of innate immune cells associated with allergy. Nature. 447, 92-6. Remenyi, A., Lins, K., Nissen, L. J., Reinbold, R., Scholer, H. R., Wilmanns, M., 2003. Crystal structure of a POU/HMG/DNA ternary complex suggests differential assembly of Oct4 and Sox2 on two enhancers. Genes Dev. 17, 2048-2059. XXI
Rex, M., Orme, A., Uwanogho, D., Tointon, K., Wigmore, P. M., Sharpe, P. T., Scotting, P. J., 1997a. Dynamic expression of chicken Sox2 and Sox3 genes in ectoderm induced to form neural tissue. Dev Dyn. 209, 323-32. Rex, M., Uwanogho, D. A., Orme, A., Scotting, P. J., Sharpe, P. T., 1997b. cSox21 exhibits a complex and dynamic pattern of transcription during embryonic development of the chick central nervous system. Mech Dev. 66, 39-53. Richards, L. J., Plachez, C., Ren, T., 2004. Mechanisms regulating the development of the corpus callosum and its agenesis in mouse and human. Clin Genet. 66, 276-89. Rizzoti, K., Brunelli, S., Carmignac, D., Thomas, P. Q., Robinson, I. C., Lovell-Badge, R., 2004. SOX3 is required during the formation of the hypothalamo-pituitary axis. Nat Genet. 36, 247-55. Rizzoti, K., Lovell-Badge, R., 2007. SOX3 activity during pharyngeal segmentation is required for craniofacial morphogenesis. Development. 134, 3437-3448. Robb, L., Tam, P. P. L., 2004. Gastrula organiser and embryonic patterning in the mouse. Seminars in Cell & Developmental Biology. 15, 543-554. Rodda, D. J., Chew, J.-L., Lim, L.-H., Loh, Y.-H., Wang, B., Ng, H.-H., Robson, P., 2005. Transcriptional Regulation of Nanog by OCT4 and SOX2. Journal of Biological Chemistry. 280, 24731-24737. Rodríguez, E. M., Garrido, O., Oksche, A., 1990. Lectin histochemistry of the human fetal subcommissural organ. Cell and Tissue Research. 262, 105-113. Rodríguez, E. M., Herrera, H., Peruzzo, B., Rodríguez, S., Hein, S., Oksche, A., 1986. Light and electron-microscopic immunocytochemistry and lectin histochemistry of the subcommissural organ: Evidence for processing of the secretory material. Cell and Tissue Research. 243, 545-559. Rodríguez, E. M., Oksche, A., Hein, S., Rodríguez, S., Yulis, R., 1984a. Comparative immunocytochemical study of the subcommissural organ. Cell and Tissue Research. 237, 427-441. Rodríguez, E. M., Oksche, A., Hein, S., Rodríguez, S., Yulis, R., 1984b. Spatial and structural interrelationships between secretory cells of the subcommissural organ and blood vessels. Cell and Tissue Research. 237, 443-449. Rodriguez, E. M., Oksche, A., Montecinos, H., 2001. Human subcommissural organ, with particular emphasis on its secretory activity during the fetal life. Microsc Res Tech. 52, 573-90. Rodriguez, E. M., Rodriguez, S., Hein, S., 1998. The subcommissural organ. Microsc Res Tech. 41, 98-123. Rodriguez, S., Batiz, L., Ortloff, A., Vio, K., Munoz, R., DeGraff, L., Graves, J., Stumpo, D., Blackshear, P., Zeldin, D., Goto, J., Tezuka, T., Yamamoto, T., Rodriguez, E., 2007. Lack of formation of Reissner fiber leads to hydrocephalus. Cerebrospinal Fluid Research. 4, S25. Rodríguez, S., Caprile, T., 2001. Functional aspects of the subcommissural organ-Reissner's fiber complex with emphasis in the clearance of brain monoamines. Microscopy Research and Technique. 52, 564-572. Rodríguez, S., Vio, K., Wagner, C., Barría, M., Navarrete, E. H., Ramírez, V. D., Pérez- Fígares, J. M., Rodríguez, E. M., 1999. Changes in the cerebrospinal-fluid monoamines in rats with an immunoneutralization of the subcommissural organ- Reissner’s fiber complex by maternal delivery of antibodies. Experimental Brain Research. 128, 278-290. Rogers, C. D., Harafuji, N., Archer, T., Cunningham, D. D., Casey, E. S., 2009. Xenopus Sox3 activates sox2 and geminin and indirectly represses Xvent2 expression to induce XXII
neural progenitor formation at the expense of non-neural ectodermal derivatives. Mech Dev. 126, 42-55. Rolf, B., Kutsche, M., Bartsch, U., 2001. Severe hydrocephalus in L1-deficient mice. Brain Research. 891, 247-252. Rosenquist, T. A., Martin, G. R., 1995. Visceral endoderm-1 (VE-1): an antigen marker that distinguishes anterior from posterior embryonic visceral endoderm in the early post- implantation mouse embryo. Mechanisms of Development. 49, 117-121. Sah, V. P., Attardi, L. D., Mulligan, G. J., Williams, B. O., Bronson, R. T., Jacks, T., 1995. A subset of p53-deficient embryos exhibit exencephaly. Nat Genet. 10, 175-80. Sandberg, M., Kallstrom, M., Muhr, J., 2005. Sox21 promotes the progression of vertebrate neurogenesis. Nat Neurosci. 8, 995-1001. Santagati, F., Rijli, F. M., 2003. Cranial neural crest and the building of the vertebrate head. Nat Rev Neurosci. 4, 806-818. Sapiro, R., Kostetskii, I., Olds-Clarke, P., Gerton, G. L., Radice, G. L., Strauss, I. J., 2002. Male infertility, impaired sperm motility, and hydrocephalus in mice deficient in sperm-associated antigen 6. Mol Cell Biol. 22, 6298-305. Saunders, N. R., Knott, G. W., Dziegielewska, K. M., 2000. Barriers in the immature brain. Cell Mol Neurobiol. 20, 29-40. Savare, J., Bonneaud, N., Girard, F., 2005. SUMO represses transcriptional activity of the Drosophila SoxNeuro and human Sox3 central nervous system-specific transcription factors. Mol Biol Cell. 16, 2660-9. Scaffidi, P., Bianchi, M. E., 2001. Spatially Precise DNA Bending Is an Essential Activity of the Sox2 Transcription Factor. Journal of Biological Chemistry. 276, 47296-47302. Schambra, U., 2008. Prenatal Mouse Brain Atlas. Springer Science, New York. Schepers, G. E., Teasdale, R. D., Koopman, P., 2002. Twenty pairs of sox: extent, homology, and nomenclature of the mouse and human sox transcription factor gene families. Dev Cell. 3, 167-70. Schlegel, J. r., Peters, I., Orrenius, S., Miller, D. K., Thornberry, N. A., Yamin, T.-T., Nicholson, D. W., 1996. CPP32/Apopain Is a Key Interleukin 1 Converting Enzyme- like Protease Involved in Fas-mediated Apoptosis. Journal of Biological Chemistry. 271, 1841-1844. Schlosser, A., Thomsen, T., Moeller, J. B., Nielsen, O., Tornoe, I., Mollenhauer, J., Moestrup, S. K., Holmskov, U., 2009. Characterization of FIBCD1 as an acetyl group-binding receptor that binds chitin. J Immunol. 183, 3800-9. Serbedzija, G. N., Bronner-Fraser, M., Fraser, S. E., 1992. Vital dye analysis of cranial neural crest cell migration in the mouse embryo. Development. 116, 297-307. Shawlot, W., Behringer, R. R., 1995. Requirement for LIml in head-organizer function. Nature. 374, 425-430. Shih, Y.-H., Kuo, C.-L., Hirst, C. S., Dee, C. T., Liu, Y.-R., Laghari, Z. A., Scotting, P. J., 2010. SoxB1 transcription factors restrict organizer gene expression by repressing multiple events downstream of Wnt signalling. Development. Shimamura, K., Hirano, S., McMahon, A., Takeichi, M., 1994. Wnt-1-dependent regulation of local E-cadherin and alpha N-catenin expression in the embryonic mouse brain. Development. 120, 2225-2234. Shimogori, T., Lee, D. A., Miranda-Angulo, A., Yang, Y., Wang, H., Jiang, L., Yoshida, A. C., Kataoka, A., Mashiko, H., Avetisyan, M., Qi, L., Qian, J., Blackshaw, S., 2010. A genomic atlas of mouse hypothalamic development. Nat Neurosci. 13, 767-75. Shu, T., Richards, L. J., 2001. Cortical Axon Guidance by the Glial Wedge during the Development of the Corpus Callosum. J. Neurosci. 21, 2749-2758. XXIII
Smukler, S. R., Runciman, S. B., Xu, S., van der Kooy, D., 2006. Embryonic stem cells assume a primitive neural stem cell fate in the absence of extrinsic influences. The Journal of Cell Biology. 172, 79-90. Solomon, N. M., Nouri, S., Warne, G. L., Lagerstrom-Fermer, M., Forrest, S. M., Thomas, P. Q., 2002. Increased Gene Dosage at Xq26-q27 Is Associated with X-Linked Hypopituitarism. Genomics. 79, 553-559. Solomon, N. M., Ross, S. A., Morgan, T., Belsky, J. L., Hol, F. A., Karnes, P. S., Hopwood, N. J., Myers, S. E., Tan, A. S., Warne, G. L., Forrest, S. M., Thomas, P. Q., 2004. Array comparative genomic hybridisation analysis of boys with X linked hypopituitarism identifies a 3.9 Mb duplicated critical region at Xq27 containing SOX3. J Med Genet. 41, 669-78. Somera, K. C., Jones, H. C., 2004. Reduced subcommissural organ glycoprotein immunoreactivity precedes aqueduct closure and ventricular dilatation in H-Tx rat hydrocephalus. Cell and Tissue Research. 315, 361-373. Spassky, N., Merkle, F. T., Flames, N., Tramontin, A. D., Garcia-Verdugo, J. M., Alvarez- Buylla, A., 2005. Adult Ependymal Cells Are Postmitotic and Are Derived from Radial Glial Cells during Embryogenesis. J. Neurosci. 25, 10-18. Specht, C. G., Schoepfer, R., 2001. Deletion of the alpha-synuclein locus in a subpopulation of C57BL/6J inbred mice. BMC Neurosci. 2, 11. Stanic, K., Montecinos, H., Caprile, T., 2010. Subdivisions of chick diencephalic roof plate: Implication in the formation of the posterior commissure. Developmental Dynamics. 239, 2584-2593. Stankiewicz, P., Thiele, H., Schlicker, M., Cseke-Friedrich, A., Bartel-Friedrich, S., Yatsenko, S. A., Lupski, J. R., Hansmann, I., 2005. Duplication of Xq26.2-q27.1, including SOX3, in a mother and daughter with short stature and dyslalia. Am J Med Genet A. 138, 11-7. Stepniak, E., Radice, G. L., Vasioukhin, V., 2009. Adhesive and Signaling Functions of Cadherins and Catenins in Vertebrate Development. Cold Spring Harbor Perspectives in Biology. 1. Sterba, G., Kießig, C., Naumann, W., Petter, H., Kleim, I., 1982. The secretion of the subcommissural organ. Cell and Tissue Research. 226, 427-439. Sterba, G., Kleim, I., Naumann, W., Petter, H., 1981. Immunocytochemical investigation of the subcommissural organ in the rat. Cell and Tissue Research. 218, 659-662. Stevanovic, M., 2003. Modulation of SOX2 and SOX3 gene expression during differentiation of human neuronal precursor cell line NTERA2. Mol Biol Rep. 30, 127-32. Stevanovic, M., Lovell-Badge, R., Collignon, J., Goodfellow, P. N., 1993. SOX3 is an X- linked gene related to SRY. Hum Mol Genet. 2, 2013-8. Strain, L., Gosden, C. M., Brock, D. J., Bonthron, D. T., 1994. Genetic heterogeneity in X- linked hydrocephalus: linkage to markers within Xq27.3. Am J Hum Genet. 54, 236- 43. Sturrock, R. R., 1979. A morphological study of the development of the mouse choroid plexus. J Anat. 129, 777-93. Sutton, E., Hughes, J., White, S., Sekido, R., Tan, J., Arboleda, V., Rogers, N., Knower, K., Rowley, L., Eyre, H., Rizzoti, K., McAninch, D., Goncalves, J., Slee, J., Turbitt, E., Bruno, D., Bengtsson, H., Harley, V., Vilain, E., Sinclair, A., Lovell-Badge, R., Thomas, P., 2011. Identification of SOX3 as an XX male sex reversal gene in mice and humans. J Clin Invest. 121, 328-41. Taatjes, D. J., Naar, A. M., Andel, F., 3rd, Nogales, E., Tjian, R., 2002. Structure, function, and activator-induced conformations of the CRSP coactivator. Science. 295, 1058-62. XXIV
Takeuchi, I. K., Kimura, R., Matsuda, M., Shoji, R., 1987. Absence of subcommissural organ in the cerebral aqueduct of congenital hydrocephalus spontaneously occurring in MT/HokIdr mice. Acta Neuropathologica. 73, 320-322. Takeuchi, I. K., Kimura, R., Shoji, R., 1988. Dysplasia of subcommissural organ in congenital hydrocephalus spontaneously occurring in CWS/Idr rats. Cellular and Molecular Life Sciences (CMLS). 44, 338-340. Takeuchi, I. K., Takeuchi, Y. K., 1986. Congenital hydrocephalus following X-irradiation of pregnant rats on an early gestational day. Neurobehav Toxicol Teratol. 8, 143-50. Tanaka, S., Kamachi, Y., Tanouchi, A., Hamada, H., Jing, N., Kondoh, H., 2004. Interplay of SOX and POU factors in regulation of the Nestin gene in neural primordial cells. Mol Cell Biol. 24, 8834-46. Tapanes-Castillo, A., Weaver, E., Smith, R., Kamei, Y., Caspary, T., Hamilton-Nelson, K., Slifer, S., Martin, E., Bixby, J., Lemmon, V., 2010. A modifier locus on chromosome 5 contributes to <i>L1 cell adhesion molecule</i> X-linked hydrocephalus in mice. neurogenetics. 11, 53-71. Thomas, P., Beddington, R., 1996. Anterior primitive endoderm may be responsible for patterning the anterior neural plate in the mouse embryo. Curr Biol. 6, 1487-96. Thomas, P. Q., Brown, A., Beddington, R. S., 1998. Hex: a homeobox gene revealing peri- implantation asymmetry in the mouse embryo and an early transient marker of endothelial cell precursors. Development. 125, 85-94. Thomsen, T., Moeller, J. B., Schlosser, A., Sorensen, G. L., Moestrup, S. K., Palaniyar, N., Wallis, R., Mollenhauer, J., Holmskov, U., 2009. The recognition unit of FIBCD1 organizes into a noncovalently linked tetrameric structure and uses a hydrophobic funnel (S1) for acetyl group recognition. J Biol Chem. 285, 1229-38. Threadgill, D. W., Yee, D., Matin, A., Nadeau, J. H., Magnuson, T., 1997. Genealogy of the 129 inbred strains: 129/SvJ is a contaminated inbred strain. Mamm Genome. 8, 390-3. Tilleman, H., Hakim, V., Novikov, O., Liser, K., Nashelsky, L., Di Salvio, M., Krauthammer, M., Scheffner, O., Maor, I., Mayseless, O., Meir, I., Kayam, G., Sela-Donenfeld, D., Simeone, A., Brodski, C., 2010. Bmp5/7 in concert with the mid-hindbrain organizer control development of noradrenergic locus coeruleus neurons. Molecular and Cellular Neuroscience. 45, 1-11. Timmer, J. R., Wang, C., Niswander, L., 2002. BMP signaling patterns the dorsal and intermediate neural tube via regulation of homeobox and helix-loop-helix transcription factors. Development. 129, 2459-2472. Tullio, A. N., Bridgman, P. C., Tresser, N. J., Chan, C.-C., Conti, M. A., Adelstein, R. S., Hara, Y., 2001. Structural abnormalities develop in the brain after ablation of the gene encoding nonmuscle myosin II-B heavy chain. The Journal of Comparative Neurology. 433, 62-74. Urase, K., Fujita, E., Miho, Y., Kouroku, Y., Mukasa, T., Yagi, Y., Momoi, M. Y., Momoi, T., 1998. Detection of activated Caspase-3 (CPP32) in the vertebrate nervous system during development by a cleavage site-directed antiserum. Developmental Brain Research. 111, 77-87. Uwanogho, D., Rex, M., Cartwright, E. J., Pearl, G., Healy, C., Scotting, P. J., Sharpe, P. T., 1995. Embryonic expression of the chicken Sox2, Sox3 and Sox11 genes suggests an interactive role in neuronal development. Mech Dev. 49, 23-36. Valera, E., Isaacs, M. J., Kawakami, Y., Izpisua Belmonte, J. C., Choe, S., 2010. BMP-2/6 heterodimer is more effective than BMP-2 or BMP-6 homodimers as inductor of differentiation of human embryonic stem cells. PLoS ONE. 5, e11167. Vieira, C., Garda, A.-L., Shimamura, K., Martinez, S., 2005. Thalamic development induced by Shh in the chick embryo. Developmental Biology. 284, 351-363. XXV
Vieira, C., Pombero, A., Garcia-Lopez, R., Gimeno, L., Echevarria, D., Martinez, S., 2010. Molecular mechanisms controlling brain development: an overview of neuroepithelial secondary organizers. Int J Dev Biol. 54, 7-20. Vio, K., Rodríguez, S., Navarrete, E. H., Pérez-Fígares, J. M., Jiménez, A. J., Rodríguez, E. M., 2000. Hydrocephalus induced by immunological blockage of the subcommissural organ-Reissner's fiber (RF) complex by maternal transfer of anti-RF antibodies. Experimental Brain Research. 135, 41-52. Vio, K., Rodriguez, S., Yulis, C. R., Oliver, C., Rodriguez, E. M., 2008. The subcommissural organ of the rat secretes Reissner's fiber glycoproteins and CSF-soluble proteins reaching the internal and external CSF compartments. Cerebrospinal Fluid Res. 5, 3. Visel, A., Thaller, C., Eichele, G., 2004. GenePaint.org: an atlas of gene expression patterns in the mouse embryo. Nucleic Acids Res. 32, D552-6. Vue, T. Y., Bluske, K., Alishahi, A., Yang, L. L., Koyano-Nakagawa, N., Novitch, B., Nakagawa, Y., 2009. Sonic hedgehog signaling controls thalamic progenitor identity and nuclei specification in mice. J Neurosci. 29, 4484-97. Wagner, C., Batiz, L. F., Rodriguez, S., Jimenez, A. J., Paez, P., Tome, M., Perez-Figares, J. M., Rodriguez, E. M., 2003. Cellular mechanisms involved in the stenosis and obliteration of the cerebral aqueduct of hyh mutant mice developing congenital hydrocephalus. J Neuropathol Exp Neurol. 62, 1019-40. Wahlsten, D., 1981. Prenatal schedule of appearance of mouse brain commissures. Developmental Brain Research. 1, 461-473. Wang, T. W., Stromberg, G. P., Whitney, J. T., Brower, N. W., Klymkowsky, M. W., Parent, J. M., 2006. Sox3 expression identifies neural progenitors in persistent neonatal and adult mouse forebrain germinative zones. J Comp Neurol. 497, 88-100. Wegner, M., 2009. ALL PURPOSE SOX: The many roles of Sox proteins in gene expression. Int J Biochem Cell Biol. Weiss, J., Meeks, J. J., Hurley, L., Raverot, G., Frassetto, A., Jameson, J. L., 2003. Sox3 is required for gonadal function, but not sex determination, in males and females. Mol Cell Biol. 23, 8084-91. Weller, S., Gartner, J., 2001. Genetic and clinical aspects of X-linked hydrocephalus (L1 disease): Mutations in the L1CAM gene. Hum Mutat. 18, 1-12. Wilson, G. R., Tan, J. T., Brody, K. M., Taylor, J. M., Delatycki, M. B., Lockhart, P. J., 2009. Expression and localization of the Parkin Co-Regulated Gene in mouse CNS suggests a role in ependymal cilia function. Neuroscience Letters. 460, 97-101. Wilson, L. D., Ross, S. A., Lepore, D. A., Wada, T., Penninger, J. M., Thomas, P. Q., 2005. Developmentally regulated expression of the regulator of G-protein signaling gene 2 (Rgs2) in the embryonic mouse pituitary. Gene Expr Patterns. 5, 305-11. Wissmuller, S., Kosian, T., Wolf, M., Finzsch, M., Wegner, M., 2006. The high-mobility- group domain of Sox proteins interacts with DNA-binding domains of many transcription factors. Nucleic Acids Res. 34, 1735-44. Wodarczyk, C., Rowe, I., Chiaravalli, M., Pema, M., Qian, F., Boletta, A., 2009. A Novel Mouse Model Reveals that Polycystin-1 Deficiency in Ependyma and Choroid Plexus Results in Dysfunctional Cilia and Hydrocephalus. PLoS ONE. 4, e7137. Wolff, D. J., Gustashaw, K. M., Zurcher, V., Ko, L., White, W., Weiss, L., Dyke, D. L. V., Schwartz, S., Willard, H. F., 1997. Deletions in Xq26.3–q27.3 including FMR1 result in a severe phenotype in a male and variable phenotypes in females depending upon the X inactivation pattern. Human Genetics. 100, 256-262. Wong, J., Farlie, P., Holbert, S., Lockhart, P., Thomas, P. Q., 2007. Polyalanine expansion mutations in the X-linked hypopituitarism gene SOX3 result in aggresome formation and impaired transactivation. Front Biosci. 12, 2085-95. XXVI
Wood, H. B., Episkopou, V., 1999. Comparative expression of the mouse Sox1, Sox2 and Sox3 genes from pre-gastrulation to early somite stages. Mech Dev. 86, 197-201. Woods, K. S., Cundall, M., Turton, J., Rizotti, K., Mehta, A., Palmer, R., Wong, J., Chong, W. K., Al-Zyoud, M., El-Ali, M., Otonkoski, T., Martinez-Barbera, J. P., Thomas, P. Q., Robinson, I. C., Lovell-Badge, R., Woodward, K. J., Dattani, M. T., 2005. Over- and underdosage of SOX3 is associated with infundibular hypoplasia and hypopituitarism. Am J Hum Genet. 76, 833-49. Yamasaki, M., Thompson, P., Lemmon, V., 1997. CRASH syndrome: mutations in L1CAM correlate with severity of the disease. Neuropediatrics. 28, 175-8. Yuan, H., Corbi, N., Basilico, C., Dailey, L., 1995. Developmental-specific activity of the FGF-4 enhancer requires the synergistic action of Sox2 and Oct-3. Genes Dev. 9, 2635-2645. Yuen, T., Wurmbach, E., Pfeffer, R. L., Ebersole, B. J., Sealfon, S. C., 2002. Accuracy and calibration of commercial oligonucleotide and custom cDNA microarrays. Nucleic Acids Res. 30, e48. Zaki, P. A., Quinn, J. C., Price, D. J., 2003. Mouse models of telencephalic development. Curr Opin Genet Dev. 13, 423-37. Zakin, L. D. J., Mazan, S., Maury, M., Martin, N., Guénet, J.-L., Brûlet, P., 1998. Structure and expression of Wnt13, a novel mouse Wnt2 related gene. Mechanisms of Development. 73, 107-116. Zechner, D., Muller, T., Wende, H., Walther, I., Taketo, M. M., Crenshaw, E. B., 3rd, Treier, M., Birchmeier, W., Birchmeier, C., 2007. Bmp and Wnt/beta-catenin signals control expression of the transcription factor Olig3 and the specification of spinal cord neurons. Dev Biol. 303, 181-90. Zeltser, L. M., 2005. Shh-dependent formation of the ZLI is opposed by signals from the dorsal diencephalon. Development. 132, 2023-33. Zhang, C., Basta, T., Hernandez-Lagunas, L., Simpson, P., Stemple, D. L., Artinger, K. B., Klymkowsky, M. W., 2004. Repression of nodal expression by maternal B1-type SOXs regulates germ layer formation in Xenopus and zebrafish. Dev Biol. 273, 23-37. Zhang, C., Basta, T., Jensen, E. D., Klymkowsky, M. W., 2003. The beta-catenin/VegT- regulated early zygotic gene Xnr5 is a direct target of SOX3 regulation. Development. 130, 5609-24. Zhang, J., Cerny, M. A., Lawson, M., Mosadeghi, R., Cashman, J. R., 2007. Functional activity of the mouse flavin-containing monooxygenase forms 1, 3, and 5. Journal of Biochemical and Molecular Toxicology. 21, 206-215. Zhang, J., Williams, M. A., Rigamonti, D., 2006. Genetics of human hydrocephalus. J Neurol. 253, 1255-66. Zhao, S., Nichols, J., Smith, A. G., Li, M., 2004. SoxB transcription factors specify neuroectodermal lineage choice in ES cells. Molecular and Cellular Neuroscience. 27, 332-342. Zhao, Y., Kwan, K. M., Mailloux, C. M., Lee, W. K., Grinberg, A., Wurst, W., Behringer, R. R., Westphal, H., 2007. LIM-homeodomain proteins Lhx1 and Lhx5, and their cofactor Ldb1, control Purkinje cell differentiation in the developing cerebellum. Proc Natl Acad Sci U S A. 104, 13182-6. Zhao, Y., Sheng, H. Z., Amini, R., Grinberg, A., Lee, E., Huang, S., Taira, M., Westphal, H., 1999. Control of hippocampal morphogenesis and neuronal differentiation by the LIM homeobox gene Lhx5. Science. 284, 1155-8. Zhou, K. W., Zheng, X. M., Yang, Z. W., Zhang, L., Chen, H. D., 2009. Overexpression of CIRP may reduce testicular damage induced by cryptorchidism. Clinical & Investigative Medicine. 32, E103-11. XXVII
Zorn, A. M., Barish, G. D., Williams, B. O., Lavender, P., Klymkowsky, M. W., Varmus, H. E., 1999. Regulation of Wnt signaling by Sox proteins: XSox17 alpha/beta and XSox3 physically interact with beta-catenin. Mol Cell. 4, 487-98. Zucchi, I., Jones, J., Affer, M., Montagna, C., Redolfi, E., Susani, L., Vezzoni, P., Parvari, R., Schlessinger, D., Whyte, M. P., Mumm, S., 1999. Transcription Map of Xq27: Candidates for Several X-Linked Diseases. Genomics. 57, 209-218.