1 the Role of Cask in Neuronal Morphogenesis

The Role of CASK in Neuronal Morphogenesis and Brain Size in Drosophila: A Genetic Model of Human Intellectual Disability with Microcephaly

Item Type text; Electronic Dissertation

Authors Tello Vega, Judith Arane

Publisher The University of Arizona.

Rights Copyright © is held by the author. Digital access to this material is made possible by the University Libraries, University of Arizona. Further transmission, reproduction, presentation (such as public display or performance) of protected items is prohibited except with permission of the author.

Download date 29/09/2021 20:08:46

Link to Item http://hdl.handle.net/10150/630252

THE ROLE OF CASK IN NEURONAL MORPHOGENESIS AND BRAIN SIZE IN DROSOPHILA: A GENETIC MODEL OF HUMAN INTELLECTUAL DISABILITY WITH MICROCEPHALY

Judith Arane Tello Vega

A Dissertation Submitted to the Faculty of the

GRADUATE INTERDISCIPLINARY PROGRAM IN NEUROSCIENCE

In Partial Fulfillment of the Requirements

For the Degree of

DOCTOR OF PHILOSOPHY

In the Graduate College

THE UNIVERSITY OF ARIZONA

2018

STATEMENT BY AUTHOR

This dissertation has been submitted in partial fulfillment of the requirements for an advanced degree at the University of Arizona and is deposited in the University Library to be made available to borrowers under rules of the Library.

Brief quotations from this dissertation are allowable without special permission, provided that an accurate acknowledgement of the source is made. Requests for permission for extended quotation from or reproduction of this manuscript in whole or in part may be granted by the head of the major department or the Dean of the Graduate College when in his or her judgment the proposed use of the material is in the interests of scholarship. In all other instances, however, permission must be obtained from the author.

SIGNED: Judith Tello

Acknowledgements

Firstly, I would like to express my sincere gratitude to my advisor Dr. Linda Restifo for her continuous support and patience. Her skillful guidance helped me enormously during all these years of research and writing of this dissertation. My appreciation also extends to Dr. Wulfila Gronenberg, for sharing expertise and always supporting me in numerous ways. I thank the rest of my dissertation committee: Dr. Julie Miller and Dr. Naomi Rance for insightful comments and especially for their encouragement during this process. My deepest thanks go to Dr. Lee Ryan and Dr. Konrad Zinsmaier who helped me through difficult times. Also, I wish to express my sincere thanks to Kirsten Cloutier Grabo for always taking care of me behind the scenes. I am grateful to Mark Yanagihashi, for his continuous encouragement and understanding. I am grateful to numerous Restifo Lab members who collaborated with me over the years: Monica Chaung, Dailu Chen, Rachel Bear, Lauren Pisani, and Dave Andrew. Especially I thank John Clark for the stimulating discussions. My deep gratitude goes for Sara Lewis and Cecilia Brown, my fellow lab mates and friends, for practical help and continual support, especially through hard times. I am extremely thankful and indebted to my great friends during the PhD: Romina Barria, Lilian Patron, Grace Samtani, Emilia Basantes, Claudia Barahona and Israel Lopez. Their friendship, cheering, laughs and affection were absolutely essential to survive. I take this opportunity to express gratitude to all my Chilean friends in Tucson, particularly Liss and Pablo for simply making me part of their family and keeping me sane. I am forever indebted to my surrogate parents, Mami Lupita and Mr. Heleno, for the unceasing support and love. I am also thankful to Ulises, my partner, who supported me no matter what through this venture. Above all, I would like to thank my wonderful parents, Judith and Julio, for the faith in me and for loving and supporting me in all the ways possible. To my sister Karinna and my brother Julio Cesar for the laughs and unique words of encouragement. Last but not least, I am grateful to God, for giving me the strength necessary to get through this journey.

Dedication

To my beautiful grandmother, Carmen Muñoz, my “Mami”

In loving memory of my grandfather, Ramón Tello

“I remember their love when they can no longer remember”

Table of Contents List of tables and figures ...... 8 List of abbreviations ...... 9 Abstract ...... 11

Chapter I: Introduction I.A. Intellectual Disability (ID) ...... 13 I.A.1. Definition, epidemiology, and associated features...... 13 I.A.2. Etiology and genetics of ID ...... 15 I.B. Microcephaly ...... 16 I.C. CASK mutations cause X-linked developmental brain disorders ...... 17 I.C.1. ID and microcephaly with pontine and cerebellar hypoplasia (MIM# 300749) ...... 18 I.C.2. FG syndrome 4 and X-linked ID with or without nystagmus (MIM# 300422) ...... 25 I.C.3. Genotype-phenotype relationships ...... 26 I.D. CASK a multi-domain protein, localization and hypothesized function ...... 28 I.D.1. Discovery of CASK and its structure ...... 28 I.D.2. The CaMK-like domain of CASK: a pseudo kinase ...... 29 I.D.3. CASK protein localization: guilt by association ...... 30 I.D.4. Other roles for CASK: insights from mouse models ...... 40 I.E. Drosophila CASK tools and contributions ...... 42 I.F. Problem statement and research questions addressed in this dissertation ...... 51 CHAPTER II: Neurite-arborization abnormalities of CASK-mutant neurons in dissociated cell culture II.A. Rationale ...... 54 II.B. Materials and Methods ...... 57 II.B.1. Drosophila culture and genetics ...... 57 II.B.2. Primary neuronal cell cultures ...... 60 II.B.3. Immunostaining and morphometric analysis of 2D neuron image ...... 61 II.B.4. Linear mixed modeling with nlme ...... 66 II.C. Results...... 66 II.C.1. CASKΔ18 homozygous neurons have a “bushy” phenotype ...... 66

II.C.2. Linear mixed modeling to combine results from multiple experiments ...... 74 II.C.3 CASK “bushy” phenotype is semi-dominant...... 77 II.C.4. Two chromosomal deficiencies of CASK uncover the bushy phenotype ...... 77 II.C.5. Transgenic CASK+ improves the bushy phenotype ...... 86 II.D. Discussion...... 90 II.D.1. Key findings and their significance to human CASK phenotypes ...... 90 II.D.2. Potential mechanisms underlying the CASK-mutant neurite arbor phenotype ...... 91 II.D.3. Contributions of primary culture for understanding neurite-arbor morphogenesis ....93 CHAPTER III: CNS-size abnormalities in Drosophila CASK loss-of-function mutants III.A. Rationale ...... 96 III.B. Materials and Methods ...... 101 III.B.1. Sample collection and photography ...... 101 III.B.2. Photography and estimation of body length ...... 101 III.B.3. Histology ...... 102 III.B.4. Image acquisition and analysis ...... 104 III.B.5. Statistical analysis and graphing ...... 105 III.C. Results...... 105 III.C.1. The small-brain phenotype of CASK mutant flies ...... 105 III.C.2. Relationship to head and body size: microcephaly and short stature ...... 108 III.D. Discussion ...... 113 CHAPTER IV: Discussion IV.A. Summary, synthesis and significance of findings ...... 118 IV.B. An hypothesis to explain the microcephaly of Drosophila CASK mutants ...... 122 IV.C. The functional nature of pathogenic CASK mutations ...... 124 IV.D. The role of CASK in synapse structure and function: myth or legend? ...... 127 IV.E. Future directions ...... 129 APPENDIX A: Characterization of motor behavior in Drosophila CASK LOF mutants...... 133 APPENDIX B: Neurite-arbor morphology after CNS dissociation in microfluidic device ...... 142 URL List of Web Resources Used ...... 146 Bibliography ...... 146

List of Tables and Figures

Table 1. Human mutations in CASK causing brain-development defects ...... 21

Table 2. Pairwise comparison between control Ex33 and mutant Δ18 neurite arbors across multiple neuron-culture experiments ...... 71 Table 3. Ex33 and Δ18 neurite arbors comparisons across eight replicate experiments ...... 76

Figure 1. Phenotypic spectrum of brain-structure defects caused by CASK mutations ...... 19

Figure 2. Location and binding partners of CASK ...... 32

Figure 3. Human and Drosophila CASK are highly conserved phylogenetically ...... 44

Figure 4. Drosophila CASK gene and genetic reagents ...... 46

Figure 5. Neuron-sampling and imaging output ...... 63

Figure 6. CASKΔ18 mutant neurons have a “bushy” defect in vitro ...... 68

Figure 7. Quantification of the “bushy” phenotype in CASKΔ18 mutant neurons ...... 72

Figure 8. The CASK-mutant bushy phenotype is semidominant ...... 79

Figure 9. A chromosomal deficiency, Df(3R)X307, uncovers the bushy phenotype of CASK Δ18 ...... 82 Figure 10. A chromosomal deficiency, Df(3R)X313, uncovers the bushy phenotype of CASK Δ18 ...... 84 Figure 11. CASK+ improves the bushy phenotype ...... 87

Figure 12. Estimated brain size differences in CASK mutants ...... 98

Figure 13. Relationship head and body size: microcephaly and short stature ...... 109

Figure A1. Timeline ethogram representations of spontaneous motor behavior in CASK mutants ...... 135 Figure A2. Quantitative analysis of spontaneous motor-behavior phenotypes in CASK mutants ...... 137 Figure A3. Demonstration of acclimation period in CASK-mutant flies ...... 140

Figure B. Demonstration and quantification of the “bushy” phenotype in cultured CASK-mutant neurons dissociated using a microfluidic device ...... 144

List of Abbreviations

ACC agenesis of the corpus callosum BrdU Bromodeoxyuridine CaMK Ca2+/calmodulin-dependent protein kinase CASK calcium/calmodulin-dependent serine kinase protein CBP creb binding protein CDK cyclin-dependent kinase CdLS Cornelia de Lange syndrome CINAP CASK-interacting nucleosome assembly protein CNV copy number variant(s) CNS central nervous system DIV days in vitro DS Down syndrome DYRK1A Dual-specificity tyrosine-regulated kinase 1A FBS fetal bovine serum FDA Food and Drug Administration FRMD7 FERM-domain containing 7 GRIP1 glutamate receptor interacting protein 1 HI haploinsufficiency ID intellectual disability iPSc induced pluripotent stem cell IQ intellectual quotient KO knock-out LOF loss of function MAGUK membrane-associated guanylate kinase MALS mammalian lin-seven protein MiC microcephaly MICPCH microcephaly with pontine and cerebellar hypoplasia MRI magnetic resonance image NMJ neuromuscular junction PAK p21-activated kinase

PBS phosphate-buffered saline PSD postsynaptic density PTN phosphate-triton azide WT wild-type

Abstract

CASK is a highly conserved gene with major roles in brain development and function.

CASK encodes a multi-domain synaptic protein that interacts with numerous binding partners in at least three different subcellular regions. CASK is a member of the MAGUK protein family, defined by its carboxy-terminal end, which includes a guanylate kinase, PDZ and SH3 protein- interaction domains. CASK amino-terminal end, in turn, contains the CaMKinase-like domain, known as a pseudokinase. Highly expressed in neurons, CASK is localized to both pre- and postsynaptic zones, as well as to the nucleus. Mutations in human CASK cause X-linked intellectual disability (ID). There is a phenotypic spectrum of brain-development disorders caused by CASK mutations, that has been divided in two diagnostic categories. Microcephaly with pontine- cerebellar hypoplasia is the most severe, whereas FG syndrome-4 or X-linked ID with or without nystagmus cause the milder phenotype. Both of these phenotypes are accompanied by short stature. Because the CASK-mutant phenotypes in humans, I hypothesized that CASK is essential for neuronal morphogenesis. Therefore, I studied the role of Drosophila CASK in neuronal differentiation and brain development. I used the Drosophila CASK mutation, ∆18, an imprecise- excision allele that eliminates the full-length CASK protein, and the corresponding precise- excision control, Ex33. I examined neuronal morphogenesis in vitro by using primary cultures prepared from the whole CNS of wandering third instar larvae. Morphological parameters of neurite-arbor size and shape were quantified using NeuronMetricsTM software for semi- automated image analysis. CNS neurons lacking full-length CASK grew small arbors in vitro with an altered shape. This phenotype, called “bushy” combines small size (reduced length, higher-order branches, and area) with increased branch density. In addition, I found that CASK has a semi-dominant phenotype, by introducing a transgene with a WT copy of CASK the bushy

11 phenotype was improved. To investigate whether CASK controls brain size, I studied brain morphology of Δ18 homozygous flies by histological examination of serial sections of osmium- stained, plastic-embedded pharate-adult heads. The volumes of brain and head were estimated using Olympus cellSens software. Brain and head estimated total volumes were significantly reduced in Δ18 homozygotes. Reduction in both brain and head indicates that flies have both microencephaly and microcephaly. In addition, I analyzed body size by measuring the pupal case length as a proxy for adult body size. CASK mutants have reduced body size. The small brain and reduced head found in Drosophila CASK mutants provide evidence that this is a good genetic model that parallels the CASK phenotypes in humans. In conclusion, these data suggest that the “small-brain” phenotype, associated with CASK-mutations in flies and microcephaly in children, results from decreased neuronal size and defective formation of dendritic arbors and axonal projections, rather than to a reduced number of neurons. These data strengthen the use of the Drosophila as a genetic model organism for modeling human developmental brain disorders.

It would be reasonable to adapt the insights gained to develop new strategies in the field of human medicine, and its special significance regarding human CASK mutations. For instance, by performing a screen for drugs that can normalize disrupted neurons due to CASK disease-causing mutations, potential treatment strategies could be discovered.

Chapter I: Introduction

I.A. Intellectual Disability

I.A.1. Definition, epidemiology, and associated features

Intellectual disability (ID) is a lifelong condition with onset before the age of 18 and is defined by significant limitations in both cognitive functioning and adaptive behavior

(American Psychiatric Association, 2013; see URL list) (Harris & Greenspan, 2016). ID is the relatively new term that about a decade ago replaced “mental retardation” (Schalock et al.,

2007). Cognitive functioning is most commonly measured using intelligence quotient (IQ) testing (Committee on Psychological Testing, Including Validity Testing, for Social Security

Administration Disability Determinations, 2015; see URL list). There are two IQ sub-scores, verbal and performance scales which reflect distinct aspects of cognitive function (Baron, 2005).

Adaptive behavior, in turn, is the collection of a number of personal, social and practical skills in an individual, to meet the standards of personal independence in order to copy with common life demands (National Research Council Committee on Disability Determination for Mental

Retardation, 2002; see URL list).

The estimated overall prevalence of ID in children/adolescents and adults is ~1.55%

(Maulik et al., 2011; McKenzie et al., 2016). Although there is a variation in the ratio of affected boys/girls, where epidemiological studies repeatedly showed a sex bias, with 30–50% excess of males over females, a consistent trend is observed (McKenzie et al., 2016). The fact that males have a higher prevalence of ID than females led to the assumption that much of the excess of male with ID may be due to X-linked genes. Indeed, the sex ratio difference is partially explained by the fact that males have a single X chromosome and X-linked disabilities can be attributable to them (Stevenson & Schwartz, 2009). It has been demonstrated that the human X

13 chromosome contains a disproportionately high density of ID genes influencing cognitive ability, i.e., ID genes have mainly been found on the X chromosome than on any other comparable segment of the autosomes (Inlow & Restifo, 2004). Then, if monogenic X-linked ID was to account for an excess of ⁓30% of intellectually disabled males over females, one would expect that ⁓20% of genetically based ID in males, including sporadic ID cases, are caused by X-linked

ID (Stevenson & Schwartz, 2009; McKenzie et al., 2016).

There are many ways to classify ID. Previously, DSM-IV® classified ID severity based solely on IQ (diagnostic and statistical manual of mental disorders, version IV). More recently,

DSM-5® incorporated the degree of severity of adaptive deficits with the standardized IQ test

(American Psychiatric Association, 2013). The new classification merged the adaptative deficits and IQ scores. By including now, the conceptual, social and practical domain, ID is classified as

‘mild’ (IQ 55-70); ‘moderate’ (IQ 40-55); ‘severe’ (IQ 25-40) and ‘profound’ (IQ < 25).

Clinically, ID can be subdivided into two major categories, syndromic ID, as when one or more clinical features of a syndrome are present in addition to ID, and nonsyndromic ID, as when cognitive impairment represents the only detectable manifestation (Kaufman et al., 2010;

Chiurazzi & Pirozzi, 2016).

ID is a phenotypic manifestation of complex neurodevelopmental disorders. Medical conditions commonly found in association with ID include anatomical, physiological and/or behavioral abnormalities (Leonard & Wen, 2002; Stevenson & Schwartz, 2009; American

Psychiatric Association, 2013; see URL list). Among the most common structural brain defects are corpus callosum dysplasias visible by MRI, including hypoplasia, shorter or thicker corpus callosum, partially opened septum pellucidum, and ventricular enlargement probably from accumulation of cerebrospinal fluid (Soto-Ares et al., 2003; Decobert et al., 2005). Other less

14 frequent anomalies included cortical folding abnormalities, such as pachygyria (few gyri), polymicrogyria (many gyri) and lissencephaly (absence of folds and grooves); cerebellar hypoplasia; delayed myelination and abnormal distribution of white matter (Soto-Ares et al.,

2003; Decobert et al., 2005). Additional anatomical features include facial dysmorphisms, manifest in the eyes, orbits, nose, lips, mouth, and ears (Stevenson & Schwartz, 2009). One of the most characteristic anatomical defects in ID disorders is microcephaly (Abuelo, 2007;

Passemard et al., 2013; see § I.B.). Further, reduced muscle tone and short stature, among others, have been also described in affected individuals (Abuelo, 2007; Passemard et al., 2013).

Physiological abnormalities include obesity, feeding disturbances and sleep problems, whereas autism, cerebral palsy and epilepsy are the most frequent pathophysiological manifestations of

ID (American Psychiatric Association, 2013; see URL list).

I.A.2. Etiology and genetics of ID

ID is highly heterogeneous in etiology and can be caused by environmental and/or genetic factors (Leonard & Wen, 2002; Inlow & Restifo, 2004; Chiurazzi & Pirozzi, 2016).

Because of this extraordinary heterogeneity, understanding the biology of ID is challenging. The genetic etiology of ID includes cytogenetically detectable chromosomal abnormalities (i.e., by karyotype analysis which may include fluorescence in situ hybridization), small copy-number variants (CNVs) detected by genome-wide nucleic acid hybridization, and single-gene mutations identified by DNA sequencing (Inlow & Restifo, 2004; Raymond & Tarpey, 2006; Kaufman et al., 2010; Carvill & Meffor, 2015; Chiurazzi & Pirozzi, 2016).

Alterations in chromosome structure include insertions, deletions, inversions, and translocations, as well as a change in the number of chromosomes. A clear example of a

15 chromosomal abnormality is aneuploidy, which occurs when the number of chromosomes is abnormal, as illustrated by trisomy of chromosome 21, which causes Down syndrome (Griffiths et al., 2000). CNV, in turn, is a duplication or deletion which involves relatively short segments of DNA (Zhang et al., 2009). CNVs range in size from hundreds to millions of bases. These submicroscopic CNVs are estimated to cause at least 15% of human neurodevelopmental defects

(Tang & Amon, 2013). CNVs are now routinely being detected using various high-resolution microarray platforms predominantly for the diagnosis of patients with unexplained ID. For example, a recurrent and clinically well recognizable CNV associated with ID, microcephaly, and short stature is the 16p13.1 microdeletion (Vissers et al., 2010).

As the technology advanced, the ability to detect smaller changes improved to the point that single-gene deletions, previously undetected, are being founded. To date, there are over 850 single-gene causes of ID (Inlow and Restifo, 2004; Restifo et al., in preparation, Van der Voet et al., 2014). ID genes play essential roles in signal transduction, transcription regulation, enzymatic processes involved in metabolism and the lysosomal pathway, and specific aspects of neuronal biology, e.g. molecular organization of the synapse, neuronal differentiation and maturation, and regulation of actin cytoskeleton (Inlow & Restifo, 2004). Although some of these genes have been well-studied, as exemplified by FRM1, that when mutated causes fragile

X syndrome (Macpherson & Murray, 2016), most of them in brain development and cognitive function remain unknown.

I.B. Microcephaly

Microcephaly is a condition defined as occipital-frontal head circumference (OFC) greater than two standard deviations (SD) below the mean for age and sex (Seltzer et al., 2014;

Harris, 2015). During fetal and early postnatal development (i.e., 2 years old) head growth is mainly determined by the development of the brain (Woods, 2004; Abuelo, 2007). Thus, a smaller head size is a direct indicator of abnormal brain development. This feature can be directly measured using a non-invasive and inexpensive method, by encircling the head with a measuring tape (Harris, 2015).

Microcephaly can be classified as either congenital or postnatal (Abuelo, 2007; Von der

Hagen et al., 2014). Congenital microcephaly is detected at birth and it is a manifestation of severely disrupted prenatal brain during development. Postnatal microcephaly, in turn, occurs as when the brain has a normal size at birth, but subsequently fails to grow normally (Woods,

2004).

In the United States, the estimated incidence of congenital microcephaly was 8.7 per

10,000 live births between 2009 and 2013 (Cragan et al., 2016). However, because of the recent outbreak of Zika virus, the number of congenital microcephaly cases has increased (Maklar et al., 2016; Sampathkumar & Sanchez, 2016).

Microcephaly can be caused by a variety of genetic and environmental factors (Abuelo,

2007; Gilmore & Walsh, 2013; Passemard, 2013), and is usually (but occasionally not) associated with ID (Aggarwal et al., 2013).

I.C. CASK mutations cause X-linked developmental brain disorders

Among the ⁓850 single-gene mutations causing ID is the X-linked gene CASK (MIM#

300172) (OMIM™; see URL list), located at Xp11.4 on the proximal short arm of the X chromosome. CASK-mutation-related disorders include a spectrum of phenotypes that differ in females and males, as explained below. Originally, three partially overlapping phenotypes

17 caused by presumed loss-of-functions (LOF) mutations in the CASK gene were described, but two of them have recently been merged as genotype-phenotype relationships have been clarified

(OMIM™; see URL list). Although the exact prevalence of the CASK-mutant phenotypes is unknown, the available evidence indicates these disorders are very rare, where approximately 60 cases have been molecularly characterized and reported worldwide (Piluso et al., 2009; Tarpet et al., 2009; Moog et al., 2011; Burglen et al., 2012; Hayashi et al., 2012; Moog et al., 2015;

Hayashi et al., 2017).

I.C.1. ID and microcephaly with pontine and cerebellar hypoplasia (MICPCH, MIM# 300749)

The core phenotype of the most severe CASK disorder includes congenital microcephaly

(range OFC: -2.5 to -10 SD), severe neurodevelopmental delay, and variable degrees of pontocerebellar hypoplasia (Figure 1) (Najm et al., 2008; Moog et al., 2011; Burglen et al., 2012;

Hayashi et al., 2012; Takanashi et al., 2012; Moog et al., 2015; Hayashi et al., 2017). The term

‘pontocerebellar’ refers to the pons and the cerebellum, while ‘hypoplasia’ refers to incomplete or underdeveloped organ or tissue due to a decrease in the number of cells (US National Library of medicine; see URL list).

MICPCH is typically seen in females. Based on the ~50 females cases reported in the literature (Table 1), the clinical features of MICPCH can be described as follows. Most of the affected individuals have very limited psychomotor development with lack of independent ambulation (i.e., most of them never attained the ability to walk) and absence of speech (Moog et al., 2011 and 2015, Hayashi et al., 2012 and 2017). Their neurological exams revealed axial hypotonia, muscle weakness, and increased deep tendon reflexes (Moog et al. 2011 and 2015).

External structural abnormalities include short stature and dysmorphic facial features (Burglen et

Figure 1. Phenotypic spectrum of brain-structure defects caused by CASK mutations.

Figure 1. Phenotypic spectrum of brain-structure defects caused by CASK mutations. A. Measuring head circumference. Frame from an instructional video showing measurement of the occipitofrontal head circumference using a cloth tape. (left). The tape is positioned around the widest part of the head, 1-2 finger breadths above the eyebrows and across the most prominent part of the occiput (Centers for Disease Control, 2016; see URL list). B-J. Brain MRI images in the midsagittal plane, T1-weighted. B-C. Age-matched neonates, adapted from Najm et al. (2008) Note the scale bars; image in C is somewhat magnified. B. Neurotypical brain structure and size. F: frontal cortex, cc: corpus callosum, p: pons, crb: cerebellum. C. A two- week-old male with MICPCH. White arrows, thin, poorly myelinated corpus callosum; red arrow, reduced size of the pons; yellow arrow, cerebellum hypoplasia; asterisk, hypoplastic frontal lobe. D-J. Phenotypic spectrum of MICPCH from patients with different CASK mutations. D-F. Females, 3-8 years old, adapted from Burglen et al. (2012). Images show pontine hypoplasia, especially the inferior (caudal) portion, and varying degrees of cerebellar hypoplasia. There is a selective reduction of frontal cortex volume with variable abnormalities of the gyri. G-J. Male babies, 2-16 months old, adapted from Moog et al. (2015). G-H. Images show a thinner corpus callosum and a mildly reduced number and complexity of the frontal gyri.

G-I. Pontocerebellar hypoplasia is decreasing in order of severity from G-I, and a reduction of the frontal cortex is apparent. J. This patient shows no hypoplasia, although it presents with rapidly progressive postnatal microcephaly and mild to moderate developmental delay.

Human Mutations in CASK (Xp11.4). Patients are grouped according to clinical disorder phenotype, in descending order of severity, and first Table 1. publication. Patient Affected Predicted protein DNA alteration(s) Mutation category Comments Clinical Phenotype Ref(s) No./Sex exon(s) change F ~3.2-Mb deletion of 5’ region copy-number variant de novo MICPCH (2; 6) P1 F Inversion breakpoint (Xp11.4) Found in 2 girls MICPCH (2) P2 F ~740-kb deletion copy-number variant MICPCH (2) ~15-kb 5' del and ~170-kb 3' P3 F del copy-number variant Exons 2 and 24 MICPCH (2) P4 F c.1915C>T nonsense Exon 21 p.R639X (Arg639Ter) truncated protein MICPCH (2) exon 9 skipped in ~20% P5 M c.915G>A synonymous, splice Exon 9 p.K305K (Lys305Lys) transcripts MICPCH (2) F ~4.0-Mb deletion copy-number variant all protein null Deletion of 9 genes; de novo MICPCH (1; 6) P11 F 1-bp deletion: c.68delT indel p.F23SfsX18 MICPCH (6) P15 F c.430–2A>T splice-site disruption MICPCH (6) P12 F c.831+2T>G splice-site disruption MICPCH (6) P13 F c.1668+1G>A splice-site disruption MICPCH (6) P8 F c.173-2A>C and c.174T>A splice-site disruption Exon 3 p.D58E (Asp58Glu) MICPCH (6)

P16 F c.379C>T nonsense p.E127X (Glu127Ter) MICPCH (6) p.Q547X P28 F c.1639C>T nonsense Exon 17 (Gln547Ter) MICPCH (6) P29 F c.2074C>T nonsense p.Q692X (Gln692Ter) MICPCH (6)

P19/P2 F c.316C>T nonsense Exon 4 p.R106X (Arg106Ter) MICPCH (6; 7; 8) P7 F ~2-Mb deletion copy-number variant all protein null del(X)(p11.3-p11.4) MICPCH (6) P22 F ~2.1-Mb deletion copy-number variant all protein null del(X)(p11.3-p11.4) MICPCH (6) P23 F ~4.24-Mb deletion copy-number variant all protein null del(X)(p11.3-p11.4) MICPCH (6) P15 F ~4.5-Mb deletion copy-number variant all protein null del(X)(p11.3-p11.4) MICPCH (6) P24 F ~60-kb intragenic deletion copy-number variant MICPCH (6) P18 F ~120-kb intragenic deletion copy-number variant MICPCH (6) P27 F ~479-kb deletion copy-number variant del(X)(p11.4) MICPCH (6) P20 F 11.2-kb intragenic deletion copy-number variant MICPCH (6) P26 F ~100-kb deletion copy-number variant MICPCH (6) P21 F ~215-kb intragenic duplication copy-number variant MICPCH (6) P25 F ~300-kb duplication copy-number variant MICPCH (6) P1 F c.79C>T nonsense p.R27X (Arg27Ter) MICPCH (7; 8) P3 F c.2632C>T nonsense p.Q878X (Gln878Ter) MICPCH (7; 8) P4 F c.243_244delTA indel p.Y81X (Tyr81Ter) MICPCH (7; 8) p.S119Rfs7X and Skipping of exon 5 ± P5 F c.357-1G>A splice-site disruption Exon 5 p.H120Pfs22X insertion of partial intron 5 MICPCH (7; 8) p.W680Cfs29X and P6 F c.2040-1G>C splice-site disruption Exon 22 p.W680Cfs3X Skipping of all or part of exon 22 MICPCH (7; 8) P7 F del(X)(p11.3-p11.4) copy-number variant de novo MICPCH (7; 8) P8 F del(X)(p11.3-p11.4) copy-number variant de novo MICPCH (7; 8) P9 F dup(X)(p11.4) copy-number variant MICPCH (7; 8) P10 F dup(X)(p11.4), dup(X)(p11.21) copy-number variant MICPCH (7; 8) P1 F c.2302+delT MICPCH (8) Table 1. HumanP2 F Mutationsc.173_173+1delGG in CASK (Xp11.4). Patients are grouped according to clinical disorder phenotype,MICPCH in descending(8) order of severity, andP3 F firstc.316C>G publication (continuednonsense on next page). p.R106X (Arg106Ter) MICPCH (8) p.G637D P4 F c.1910G>A missense (Gly637Asp) MICPCH (8) P5 F del(X)(p11.3-p11.4) copy-number variant MICPCH (8)

P16 M c.1061T>C missense p.L348P (Leu348Pro) MICPCH (8) 21 P1 M c.704_708del Exon 7 p.K236Efs*10 de novo MICPCH (9) P2 M Intragenic duplication Exons 10-16 de novo MICPCH (9) P3 M c.1A>G Exon 1 de novo MICPCH (9) P4 M c.79C>T Exon 2 p.R27* de novo MICPCH (9) P5 M Intragenic duplication Exons 4-20 somatic mosaicism MICPCH (9) P6 M Intragenic deletion Exon 1 somatic mosaicism MICPCH (9) P7 M Intragenic deletion Exons 3-9 somatic mosaicism MICPCH (9) missense & splice-site M c.83G>T Exon 2 p.R28L (Arg28Leu) partial skipping of exon 2 FG s. 4 (3M in 2-gen family) (3) disruption p.Ala841_Lys843del M c.2521-2A>G splice-site disruption Exon 26 + skipping of exon 26 and a 9 bp-del FG s. 4 w/ nystagmus & MiC (10) p.Ala841_Glu868del M (fam16) c.802T>C missense Exon 8 p.Y268H (Tyr268His) 4 affected in 3 generations Severe ID w/ epilepsy (4; 5) p.D710G missense & splice-site altered splice product has 27-bp in- M (fam74) c.2129A>G Exon 22 (Asp710Gly) and 9- Mild ID w/ nystagmus (4; 5) disruption frame deletion aa deletion M (fam123) c.2756T>C missense Exon 27 p.W919R (Trp919Arg)3 brothers; carrier mom has tremorMild ID w/ nystagmus (5) M c.2767C>T missense p.W914R (Trp914Arg) Mild ID ± Nystagmus (4) M & F (fam245)c.1186C>T missense Exon 13 p.P396S (Pro396Ser) 2 brothers, 1 female cousin Severe-profound ID (4; 5) 2 brothers (mod/severe ID), 1 M & F (famV) c.2183A>G missense Exon 23 p.Y728C (Tyr728Cys) sister (mild ID); mom has ID w/ nystagmus (5) nystagmus 2 aberrant transcripts: one skips p.841_868 del and M (fam683) c.2521-2A>T splice-site disruption Exons 25-27 exon 26, other deletes 9 nt Mild ID w/ nystagmus (5) p.841_843 del inframe P8 M Intragenic duplication copy-number variant Exons 1-5 maternally inherited ID w/ MiC (9)

Refs: (1) Hayashi et al., 2008; (2) Najm et al., 2008; (3) Piluso et al., 2009; (4) Tarpey et al., 2009; (5) Hackett et al., 2010; (6) Moog et al., 2011; (7) Hayashi et al., 2012; (8) Takanashi et al., 2012; (9) Moog et al., 2015; (10) Dunn et al., 2017 Human Mutations in CASK (Xp11.4). Patients are grouped according to clinical disorder phenotype, in descending order of severity, and first Table 1. publication. Patient Affected Predicted protein DNA alteration(s) Mutation category Comments Clinical Phenotype Ref(s) No./Sex exon(s) change F ~3.2-Mb deletion of 5’ region copy-number variant de novo MICPCH (2; 6) P1 F Inversion breakpoint (Xp11.4) Found in 2 girls MICPCH (2) P2 F ~740-kb deletion copy-number variant MICPCH (2) ~15-kb 5' del and ~170-kb 3' P3 F del copy-number variant Exons 2 and 24 MICPCH (2) P4 F c.1915C>T nonsense Exon 21 p.R639X (Arg639Ter) truncated protein MICPCH (2) exon 9 skipped in ~20% P5 M c.915G>A synonymous, splice Exon 9 p.K305K (Lys305Lys) transcripts MICPCH (2) F ~4.0-Mb deletion copy-number variant all protein null Deletion of 9 genes; de novo MICPCH (1; 6) P11 F 1-bp deletion: c.68delT indel p.F23SfsX18 MICPCH (6) P15 F c.430–2A>T splice-site disruption MICPCH (6) P12 F c.831+2T>G splice-site disruption MICPCH (6) P13 F c.1668+1G>A splice-site disruption MICPCH (6) P8 F c.173-2A>C and c.174T>A splice-site disruption Exon 3 p.D58E (Asp58Glu) MICPCH (6)

P16 F c.379C>T nonsense p.E127X (Glu127Ter) MICPCH (6) p.Q547X P28 F c.1639C>T nonsense Exon 17 (Gln547Ter) MICPCH (6)

P29 F c.2074C>T nonsense p.Q692X (Gln692Ter) MICPCH (6)

P3 F c.2632C>T nonsense p.Q878X (Gln878Ter) MICPCH (7; 8) P4 F c.243_244delTA indel p.Y81X (Tyr81Ter) MICPCH (7; 8) p.S119Rfs7X and Skipping of exon 5 ± P5 F c.357-1G>A splice-site disruption Exon 5 p.H120Pfs22X insertion of partial intron 5 MICPCH (7; 8) p.W680Cfs29X and P6 F c.2040-1G>C splice-site disruption Exon 22 p.W680Cfs3X Skipping of all or part of exon 22 MICPCH (7; 8) P7 F del(X)(p11.3-p11.4) copy-number variant de novo MICPCH (7; 8) P8 F del(X)(p11.3-p11.4) copy-number variant de novo MICPCH (7; 8) P9 F dup(X)(p11.4) copy-number variant MICPCH (7; 8) P10 F dup(X)(p11.4), dup(X)(p11.21) copy-number variant MICPCH (7; 8) P1 F c.2302+delT MICPCH (8) P2 F c.173_173+1delGG MICPCH (8)

P3 F c.316C>G nonsense p.R106X (Arg106Ter) MICPCH (8) p.G637D P4 F c.1910G>A missense (Gly637Asp) MICPCH (8) P5 F del(X)(p11.3-p11.4) copy-number variant MICPCH (8)

P16 M c.1061T>C missense p.L348P (Leu348Pro) MICPCH (8) P1 M c.704_708del Exon 7 p.K236Efs*10 de novo MICPCH (9) P2 M Intragenic duplication Exons 10-16 de novo MICPCH (9) P3 M c.1A>G Exon 1 de novo MICPCH (9) P4 M c.79C>T Exon 2 p.R27* de novo MICPCH (9) P5 M Intragenic duplication Exons 4-20 somatic mosaicism MICPCH (9) P6 M Intragenic deletion Exon 1 somatic mosaicism MICPCH (9) P7 M Intragenic deletion Exons 3-9 somatic mosaicism MICPCH (9) missense & splice-site M c.83G>T Exon 2 p.R28L (Arg28Leu) partial skipping of exon 2 FG s. 4 (3M in 2-gen family) (3) disruption p.Ala841_Lys843del M c.2521-2A>G splice-site disruption Exon 26 + skipping of exon 26 and a 9 bp-del FG s. 4 w/ nystagmus & MiC (10) p.Ala841_Glu868del M (fam16) c.802T>C missense Exon 8 p.Y268H (Tyr268His) 4 affected in 3 generations Severe ID w/ epilepsy (4; 5) p.D710G missense & splice-site altered splice product has 27-bp in- M (fam74) c.2129A>G Exon 22 (Asp710Gly) and 9- Mild ID w/ nystagmus (4; 5) disruption frame deletion aa deletion M (fam123) c.2756T>C missense Exon 27 p.W919R (Trp919Arg)3 brothers; carrier mom has tremorMild ID w/ nystagmus (5) M c.2767C>T missense p.W914R (Trp914Arg) Mild ID ± Nystagmus (4) M & F (fam245)c.1186C>T missense Exon 13 p.P396S (Pro396Ser) 2 brothers, 1 female cousin Severe-profound ID (4; 5) 2 brothers (mod/severe ID), 1 M & F (famV) c.2183A>G missense Exon 23 p.Y728C (Tyr728Cys) sister (mild ID); mom has ID w/ nystagmus (5) nystagmus 2 aberrant transcripts: one skips p.841_868 del and M (fam683) c.2521-2A>T splice-site disruption Exons 25-27 exon 26, other deletes 9 nt Mild ID w/ nystagmus (5) p.841_843 del inframe P8 M Intragenic duplication copy-number variant Exons 1-5 maternally inherited ID w/ MiC (9)

Table 1. Human Mutations in CASK (Xp11.4), continued.

22 al., 2012, Moog et al., 2011, Hayashi et al., 2017). The latter are a constellation of variable components, including oval face, broad nasal bridge and tip, large ears, long philtrum, mandibular hypoplasia, and hypertelorism (abnormally increased distance between the eyes)

(Burglen et al., 2012; Takanashi et al., 2012). Many of these children also have sensorineural hearing loss or visual system abnormalities (optic nerve hypoplasia, abnormal visual evoked potentials, and strabismus) (Moog et al. 2011 and 2015, Hayashi et al., 2017). In a few cases, seizures have been also reported (Hayashi et al., 2017).

A similar but even more severe disorder has been seen in ~15 males (Table 1), of which two died in the perinatal period (Najm et al., 2008; Moog et al., 2015). The vast majority of affected females and males are isolated cases, i.e., the only affected family member, resulting from de novo germline CASK mutations. These mutations are believed to be null. Because the phenotype is manifested by heterozygous females (CASK–/+), a phenotypically normal mother cannot carry and transmit the MICPCH-causing CASK-pathogenic mutation. This phenomenon is a clear indication of LOF with haploinsufficiency (i.e., when 50% of gene function is insufficient for a wild-type (WT) phenotype) (Huang et al., 2010).

Hemizygous boys (CASK–/Y) manifest the most severe phenotype of MICPCH, sometimes including early-onset and intractable seizures, such as Ohtahara syndrome (Saitsu et al., 2012), West syndrome (Takanashi et al., 2012), or myoclonic epilepsy (Nakamura et al.,

2014). The smaller-than-expected number of males reported is likely to be a consequence of reduced fetal viability or early postnatal lethality. The more severe phenotypes observed in males relative to affected females is an indication of the semi-dominant nature of these mutations.

Many different types of CASK mutations cause MICPCH. By contrast, in Fragile X syndrome, a more common disorder, about 95% of the patients have the same mutation

(Macpherson & Murray, 2016). Moog et al. (2011) reported 25 female patients with heterozygous presumed loss-of-function CASK mutations, including four previously reported by

Najm et al. (2008) and one reported by Froyen et al. (2007). Of these, 11 showed submicroscopic

CNVs, including nine deletions and two duplications, all spanning the entire region of CASK.

The rest of patients reported by Moog et al. (2011) had point mutations of both nonsense and splice-site types. These studies were expanded by Hayashi et al. (2012) who found ten girls with

MICPCH, including one reported by Moog et al. (2011), and by Takanashi et al. (2012) who confirmed and further defined those ten patients. Among those ten females, there were found five

CNVs, three nonsense mutations, and two splice-site mutations. Takanashi found additional five girls with CASK gene mutations, including two cases with a CNVs, one splice-site mutation, one nonsense and one missense mutation. During the same year, Burglen et al. (2012) reported 11 female patients, eight of whom had intragenic point mutations, and the three others had Xp11.4 submicroscopic deletions.

To date, 15 males with pathogenic, presumed loss-of-function mutations in CASK have been described. The first case was reported by Najm et al (2008), a boy who died at two weeks of age, and who had a splice-site mutation (Figure 1). An additional two patients were described by Burglen et al. (2012), another two by Saitsu et al. (2012), one case by Takanashi et al. (2012), and one case by Nakamura et al. (2014). Recently, a cohort of eight male patients were also found by Moog et al. (2015).

It is evident that when addressing the structural nature of the CASK mutations, the data obtained seems to follow general trends. Published literature affirms that CNVs and single-gene point mutations are the most common molecular alterations in the CASK gene. Therefore, the occurrence of new cases with mutations of the same nature can be expected.

I.C.2. FG syndrome-4 or X-linked ID with or without nystagmus (MIM# 300422)

The name 'FG' derives from the description of one family with a heterogenous X-linked developmental disorder, FG syndrome (Opitz & Kaveggia, 1974). Optiz named it 'FG syndrome' by using the first patient’s initials. The first locus for FG syndrome was called FGS1, whereas

FGS4 is the newest identified locus. FGS4 syndrome is characterized by marked ID, relative macrocephaly, congenital hypotonia, severe constipation, hyperactive behavior with some violent episodes, and dysmorphic facial features, including prominent forehead and frontal upsweep of the hair, broad nasal root, long philtrum, and half-open mouth (Piluso et al., 2003,

2009). Noteworthy is the fact that FGS4 syndrome-affected individuals have relative macrocephaly, in contrast to the microcephaly observed in the MICPCH phenotype. Piluso et al.

(2009) argued that this difference highlights the unclear role of CASK in embryogenesis and

CNS development. FGS4 is inherited in a X-linked recessive pattern. The observed phenotype, presumed to be hypomorphic, is only seen in males, whereas heterozygous females are phenotypically normal and obligate carriers (Piluso et al., 2003, 2009). To date, only four 4 individuals in two families have been reported. Piluso et al. (2009) identified a missense CASK mutation (c.83G>T) which was present in three affected males and 2 unaffected carrier females from the same family. This mutation (c.83G>T) lead to a partial skipping of the exon 2 of CASK, resulting in the accumulation of truncated CASK protein (Piluso et al., 2009). Recently, Dunn et al. (2017) reported one male patient with a de novo splice mutation (c.2521-2A>G) that produces two mutant transcripts. Mutant transcripts included skipping of exon 26 as well as a 9-base-pair deletion associated with a cryptic splice site, leading to two deletions, a 28- (p. Ala841-

Glu868del) and a 3-aminoacid (p. Ala841-Lys843del), respectively in the C-terminal region of

CASK. This individual was diagnosed with FGS4 and nystagmus, but his phenotype does not

25 match with the definition of FGS4, especially because microcephaly (-2 S.D.) was present.

Nowadays, it is evident that there is a spectrum of X-linked ID disorders caused by CASK mutations, and the definition of FGS is not clearly defined.

As for ID with or without nystagmus, this phenotype is also variable, and inherited in a

X-linked recessive pattern. It ranges from mild non-syndromic ID to severe ID with microcephaly (Tarpey et al., 2009; Hackett et al., 2010). In addition, ID with or without nystagmus presents with dysmorphic facial features, including flat midface, flat nasal bridge, open mouth, upslanting palpebral fissures, and short neck (Tarpey et al., 2009; Hackett et al.,

2010).). Nystagmus, a striking feature not previously reported, is defined as an involuntary, rapid and repetitive movement of the eyes (Gottlob, 2000). In addition to nystagmus, some affected individuals had reduced visual acuity and/or strabismus. In half of the cases, other neurological and physical features, such as seizures, tremor and unsteady gait, were also present (Hackett et al., 2010). Males are affected, and five unusual heterozygous symptomatic females have been reported. Tarpey et al. (2009) founded four missense mutations in the CASK gene in 4 independent families. ID was mild to moderate in these four families and accompanied by nystagmus in two of the four. In a separate study, Hackett et al. (2010) reported five missense mutations, including three reported by Tarpey et al. (2009), and an additional splice-site mutation.

I.C.3. Genotype-phenotype relationships

There is a striking difference in clinical severity among individuals with CASK mutations, even within the same sex. Molecular genetic analysis is indicative that there is no mutational hotspot in this gene leading to ID (Tarpey et al., 2009; Moog et al., 2011). Rather,

CASK mutations appear to be distributed throughout the gene. The current research literature gave rise to a working hypothesis, CASK pathogenic variants that result in MICPCH are LOF and null mutations (Moog et al., 2011, Burglen et al., 2012, Hayashi et al., 2012). However, direct laboratory evidence of LOF is lacking.

In contrast to the severely affected females with MICPCH (Najm et al., 2008; Moog et al., 2011), Hackett et al. (2010), reported that some carrier females with X-linked ID with or without nystagmus were either phenotypically normal or less severely affected compared with the males within their family. This observation would suggest that missense mutations in CASK are being less deleterious than nonfunctional protein product obtained from the already described nonsense, splice-site mutations, and small deletions abnormalities found in MICPCH.

However, it seems like there is no correlation between the clinical features or severity of

MICPCH and any particular molecular class of variants, e.g., CASK nonsense and splice-site mutations, larger rearrangements comprising CASK or intragenic deletion that inactivates CASK variants (Moog et al., 2011, Burglen et al., 2012, Hayashi et al., 2012). Unfortunately, most of the published studies regarding CASK mutants in humans, have not assessed protein accumulation, making it difficult to evaluate the functional consequences of the CASK disease- causing mutations. Thus, it is largely unknown the effect of these mutations on CASK protein expression, structure and function.

Similarly, the working hypothesis proposes that the missense CASK mutations causing

X-linked ID with or without nystagmus in males, are hypomorphic. i.e., these variants are likely less deleterious than inactivating pathogenic variants (null-mutation). Nonetheless, as occurs with MICPCH, there is no evidence for a correlation between the clinical features of X-linked ID with or without nystagmus any particular molecular type of mutation. Interestingly, all the

CASK-mutations reported that cause ID with nystagmus are located at the C-terminus of the

CASK protein (Hackett et al., 2010). Bioinformatic analyses of the four missense mutations described in Hackett et al. (2009), showed that all were predicted to be destabilizing of the

CASK protein. Further information about protein structure is missing. Three of these pathogenic variants or mutations have been shown to disrupt the interaction between CASK and FRMD7.

Mutations in FRMD7 are a major cause of X-linked familial idiopathic infantile nystagmus

(Watkins et al., 2013), supporting a correlation between nystagmus and missense mutations affecting the CASK C-terminal region.

I.D. CASK a multi-domain protein, localization and hypothesized function.

I.D.1. Discovery of CASK and its structure

CASK was originally discovered in a yeast two-hybrid system screen to identify molecules that bind the C-terminal region of mouse neurexins (Hata el al., 1996). In the same year, mutations in the Caenorhabditis elegans CASK homolog, LIN-2, were reported to affect vulval development (Hoskins et al., 1996). Also, around the same time and while studying the role of CaMK in locomotion, Martin & Ollo (1996) cloned CASK in Drosophila melanogaster.

In the animal species being studied thus far, a single gene codes for CASK (La Conte &

Mukherjee, 2013). CASK is a highly conserved member of the MAGUK (Membrane-Associated

Guanylate Kinase) protein family (Hsueh, 2006 and 2009). MAGUK proteins are group of specialized proteins that tether adhesion molecules, receptors, and intracellular signaling enzymes organizing macromolecular protein complexes (Funke et al., 2005). Hence, they are knowing as scaffolding proteins. The MAGUK family, especially the discs large subfamily, is extensively expressed in the brain and well conserved throughout evolution (Oliva et al., 2012).

In mammals, CASK is critical for neurodevelopment and essential for survival (La Conte

& Mukherjee, 2013; Srivastava et al., 2016), but whether CASK is an essential gene in invertebrates remains unknown. The varied phenotypes observed in vertebrates and invertebrates, argues for a functional divergence of CASK in different animal models (La Conte

& Mukherjee, 2013).

CASK (Figure 3) is composed of a unique combination of domains. Briefly, it consists of an N-terminal region highly homologous to Ca2+/calmodulin-dependent protein kinases (CaMK- like) and a central region comprising two L27 domains (L27A and L27B). The L27 domain was originally found in the C. elegans Lin-2 and Lin-7 proteins and is required for the establishment and maintenance of cell polarity in vulval precursor and epithelial cells (Feng et al., 2005). The

C-terminal region contains three domains: PDZ, SH3 (Src homology 3) and GUK (guanylate kinase). The arrangement and order of these three domains is a defining characteristic of all

MAGUK proteins. Thus, CASK is a MAGUK family member (Hata et al., 1996; Hsueh, 2006;

Oliva et al., 2012).

I.D.2. The CaMK-like domain of CASK: a pseudo kinase

Although CASK is an acronym that stands for calcium/calmodulin-dependent serine protein kinase (Hata et al., 1996) and it contains a CaMK-like domain, there is limited evidence that CASK possesses Ca2+/calmodulin-dependent kinase activity (Hsueh, 2006). Mukherjee et al.

(2008) determined the crystal structure of a recombinant CASK CaMK-like domain from E. coli strain BL21. They found that the CASK CaMK-like domain exhibits a typical protein kinase conformation, resembling the typical catalytically active conformation of other protein kinases.

In all known kinases Mg2+ acts as an obligate cofactor for ATP binding and phosphotransfer

(Adams, 2001). In vitro experiments from the same study demonstrated that CASK CaMK-like domain catalyzes phosphotransfer from ATP, and autophosphorylation, both processes occurring in the absence of Mg2+. Mukherjee et al. (2008) also showed that in the absence of Mg2+, by using both in vitro and in vivo assays, CASK CaMK-like domain phosphorylates neurexin-1, a synaptic cell-adhesion molecule (see § I.D.3.a.). Suggesting that equivalents and essential elements of protein kinases can also be found in the CASK CaMK-like domain. Although the

CASK CaMK-like domain has a constitutively active conformation, it lacks the canonical DFG motif (Asp-Phe-Gly) required for Mg2+-binding, which is thought to be indispensable for kinase activity. Hence, CASK is classified as a pseudokinase (Boudeau et al., 2006).

The literature addressing the kinase activity of CASK’s CaMK-like domain is contradictory and inconclusive, nothing further about it has been published, but the role of this domain likely has important functional consequences. This is suggested by the evidence that this domain has many different binding partners (see § I.D.3 below). Because of this controversial role as pseudokinase, studies of CASK function have focused mainly on its role as a MAGUK- type scaffolding protein.

I.D.3. CASK protein localization: guilt by association

CASK proteins are widely distributed in different regions of the brain, as well as in nonneural tissues (Hsueh 2006; Oliva et al., 2012). However, the experiments in this dissertation focused only on the brain functions of CASK. In neurons, CASK is present in different subcellular compartments, localizing to both pre-and post-synaptic zones, as well as to the nucleus.

I.D.3.a. Presynpatic CASK

Because CASK is located at presynaptic sites and it associates to many different binding partners, it has been proposed a role of CASK in the differentiation of the presynaptic terminal

(Figure 2) (Hata et al., 1996; Sun et al., 2009; Chen et al., 2011). Through its CaMK-like domain and L27B domain, CASK binds with Mint1 (also called X11, Lin-10, or APBA1) and Veli (also called mammalian LIN-7, MALS), respectively (Butz et al., 1998; Borg et al., 1999). Mint proteins, a group of Munc-interacting proteins, are part of a multimeric complex containing

Munc18-1. Munc18-1 in turn, is the major brain protein that binds to syntaxin 1, a synaptic vesicle fusion protein, required for synaptic vesicle exocytosis (Okamoto & Südhof, 1997). In C. elegans, several Veli proteins (LIN-2, LIN-7 and LIN-10) work together downstream of a receptor tyrosine kinase pathway (EGFR) required for induction of vulval development (Hoskins et al., 1996). The Mint1/CASK/Veli protein complex takes place on both sides of the synapse, linking trans-synaptic adhesion molecules to the cytoskeleton (Olsen et al., 2005). In addition, through its CaMK-like, and L27A domains, CASK binds cytoplasmic Liprin-α. Ultrastructural and physiological analysis revealed abnormal active zone morphology and synaptic function in

Drosophila Liprin-α (Kaufmann et al., 2002). Olsen et al., (2005) suggested a model where Veli-

CASK- Liprin-α form a tripartite complex that recruits components of the synaptic release machinery to adhesive proteins of the active zone.

Based on cellular and biochemical studies, Wei et al. (2011) demonstrated that CASK and

Liprin-α2 binding is specific to mammals. The specific residues that form the Liprin-α2/CASK

Figure 2. Location and binding partners of CASK. The role of CASK in synaptic interaction and formation. CASK acts as a scaffolding protein and localizes to both pre- and post-synaptic compartments, as well as to the nucleus. CASK has numerous binding partners in different subcellular domains. At the presynaptic structure, CASK binds neurexin, Mint1, Veli and N-type calcium channel, among others, thereby promoting the differentiation of the presynaptic terminal. CDK5 phosphorylates CASK regulating the subcellular distribution of CASK. At the postsynaptic sites, CASK maintains the morphology of dendritic spines. In addition, CASK also enters the nuclei of neurons where it regulates the expression of NMDAR2b through an interaction with Tbr-1 and CINAP. Remarkably, several CASK binding partners are encoded by other ID genes: NRXN1(NEUREXIN 1), CNTNAP2 (CONTACTIN-ASSOCIATED PROTEIN-

LIKE 2, member of the neurexin superfamily), and CDK5. (Modified from Hsueh 2009).

32 complex is conserved among mammals, but do not exist in Liprin-α orthologues from C.elegans

(syd-2) and Drosophila (syd) (Wei et al., 2011). One of the mutations that causes MICPCH is p.Y268H (Najm et al., 2008; Tarpey et al., 2009) (Table1). Consistently, the complex crystal structure including the regions LH/CaMK-like, as part of Liprin-α2 and CASK respectively, revealed an H-bond between the Asn-1058 of Liprin-α2 and the Tyr-268 of CASK. The consequent substitution of Tyr-268 by His or Ala leads to a 3- to 4-fold decrease in the affinity of

CASK CaMK-like for Liprin-α2 (Wei et al., 2011).

As mentioned above, mammalian CASK was first identified in a yeast two-hybrid screen, using β-neurexins as bait (Hata el al., 1996). Specifically, the PDZ domain of CASK binds the C- terminal tail of neurexin. Pre-synaptic β-neurexins are transmembrane cell-adhesion molecules localized on the pre-synaptic membrane (Craig & Kang, 2007; Südhof, 2008). The extracellular domain of neurexins binds its partner at the postsynaptic side, neuroligin. The intracellular cytoplasmic portion of neurexin interacts with proteins associated with exocytosis (Südhof,

2008), although this has not been directly demonstrated. By bridging the synaptic cleft, neurexin- neuroligin are postulated to play an important role in synaptic function (Südhof, 2008).

Separated initial studies evidenced that when either neurexin or neuroligin were expressed in a non-neuronal cell, they were able to induce co-cultured neurons to form respectively, presynaptic or postsynaptic specialization onto the non-neuronal cell. These presynaptic and postsynaptic differentiation, however, might be trigger by other several molecules that bind transmembrane components of such networks (Craig & Kang, 2007; Südhof, 2008). However, this in vitro evidence appears to contradict the KO analysis in vivo. Triple-knockout (KO) mice for the three

α-neurexin genes showed that neurexins are not required for synapse formation but are essential for Ca2+-triggered neurotransmitter release (Missler et al., 2003). Furthermore, although the

33 triple-KO mouse lacking neuroligin1, 2 and 3 die at birth, it exhibits relatively normal synapse numbers with an apparently normal ultrastructure (Varoqueaux et al., 2006). Electrophysiological analyses in acute brain slices showed that neuroligin triple-KO mice display a severe impairment of synaptic transmission, revealed by the dramatic decrease in spontaneous

GABAergic/glycinergic activity. However, the density of synaptic contacts was not altered in neuroligin-deficient brains and cultured neuron indicating that neuroligin is not required for the initial formation of synaptic contacts (Varoqueaux et al., 2006). Interestingly, alterations in human neurexin (NRXN1, 2, 3) and neuroligin genes (NLGN1, 2, 3, 4), have been shown to cause neurodevelopmental disorders, including autism and ID (Geschwind & Levitt, 2007). In vitro studies showed that CASK-Neurexin-Liprin-α and Mint1 form a complex at the plasma membrane (LaConte et al., 2016). Biochemical assays using HEK293 cells demonstrated that

Liprin-α and Mint1compete for direct binding to CASK, but when β-neurexin is present, this competition is eliminated and all four proteins (Liprin-α, Mint, CASK and β-neurexin) form a complex (LaConte et al., 2016).

CASK also associates with N-type calcium channels. N-type calcium channel (Cav2.2) mediate Ca2+ influx into the presynaptic terminal triggering synaptic vesicle fusion and neurotransmitter release (Maximov et al., 1999). The cytosolic C-terminus of the N-type calcium channel α1B subunit, has a proline-rich region that binds to the SH3 domain of CASK (Maximov et al., 1999). Maximov et al. (2002) showed that Mint1, CASK, and Cav2.2-L colocalized at mature synapses using rat hippocampal neuronal cultures. In addition, Khanna et al. (2006) demonstrated that immunoprecipitation of Cav2.2-L (long C-terminal splice variant) from brain lysate coprecipitated CASK. However, there was no colocalization of Cav2.2-L and CASK at the nerve terminal, confirming the idea that these two proteins can only form a complex. These

34 studies suggest that Cav2.2-L is recruited to presynaptic locations by means of interactions with the adaptor proteins Mint1 and CASK.

CASK also connects to the F-actin cytoskeleton thorough its binding to protein 4.1

(Biederer & Südhof, 2001). Biederer & Südhof (2001) showed that CASK binds a brain-enriched isoform of protein 4.1 and nucleates local assembly of actin/spectrin filaments. In addition,

CASK could be recovered with actin filaments prepared from rat brain extracts, and neurexins are recruited together with CASK and protein 4.1 into these actin filaments. Thus, these data suggest that intercellular junctions formed by neurexins, e.g., neurexin-neuroligin, are at least partially coupled to the actin cytoskeleton via an interaction with CASK and protein 4.1

(Biederer & Südhof, 2001).

In addition, cyclin-dependent kinase 5 (CDK5), which has a role in synaptic vesicle endocytosis by rephosphorylating endocytotic proteins such as dynamin (Tan et al., 2003;

Barclay et al., 2004) has been shown to coimmunoprecipitate with CASK from adult rat brain extracts (Samuels et al., 2007). CDK5 phosphorylates the CaMK-like (Ser-51) and L27 (Ser-395) domains of CASK, as demonstrated by autoradiography data from in vitro kinase assays performed with purified CDK5. To analyze phosphorylation of CASK in vivo, Samuels et al.

(2007) made phosphorylation state-specific antibodies, phospho-Ser 51 (pS51) and phospho-Ser

395 (pS395) detecting decreased phosphorylation of CASK when comparing CDK5-deficient knock-out (KO) mice and WT. Furthermore, by using subcellular fractionation of embryonic brain in CDK5 KO mice, it was found that CASK was significantly reduced in the membrane- associated fraction and increased in the cytosol (soluble pool) when compared to WT embryonic brains (Samuels et al., 2007). Elimination of CASK phosphorylation by CDK5 was conducted by mutations of phosphorylation sites in CASK. In cells expressing the double mutant S51/395,

CASK was depleted from the membrane-associated fractions and increased in the cytosol, recapitulating the localization phenotype seen in the CDK5-deficient KO mice. Thus, by abolishing CDK5-dependent phosphorylation, CASK is not recruited to presynaptic termini.

These data indicate that CDK5 activity is necessary for the appropriate membrane localization of

CASK, suggesting that the phosphorylation by CDK5 of both CaMK-like and L27 domains can directly regulate the subcellular distribution of CASK. Samuels et al. (2007) suggest that loss of

CDK5-dependent phosphorylation is necessary for removal of CASK from the membrane- associated fraction. Finally, by using a synapse-formation assay, Samuels et al. (2007) demonstrated significantly less clustering of CASK at synaptic membrane-associated fractions, which suggests that CDK5-mediated phosphorylation is important for CASK recruitment to developing synapses. Finally, a biochemical pulldown assays using GST fusion protein of the

Liprin-α2, demonstrated that CDK5-dependent phosphorylation reduces the interaction between

CASK and Liprin-α when comparing CDK5-deficient brain lysate to WT (Samuels et al., 2007).

In humans, CDK5 LOF mutations cause severe lissencephaly with cerebellar hypoplasia and death infancy (Magen et al., 2015). In addition, mutations in CDK5R encoding CDK5 regulatory subunit 1, cause nonsyndromic ID (Moncini et al., 2016), whereas mutations in

CDK5RAP2 encoding a regulatory subunit-associated protein cause congenital microcephaly

(Bond et al., 2005). Interestingly, mutations of CDK5, CDK5R1, CDK5RAP2, cause autosomal recessive primary microcephaly (MCPH), whereas mutations of CASK cause MICPCH, all essential for brain development and leading these partially overlapping but distinctive disorder phenotypes.

I.D.3.b. Postsynaptic CASK

At postsynaptic sites, the PDZ domain of CASK binds directly to the C-terminal tail of syndecan-2 (Hsueh et al., 1998; Hsueh & Sheng, 1999). Syndecan-2 was the first postsynaptic binding partner of CASK identified. Immunogold-electron microscopy revealed the post synaptic localization of CASK and syndecan-2 in neurons of adult rat brain (Hsueh et al., 1998). By using a yeast two-hybrid screen a binding specificity between CASK-PDZ and the C-terminal tail of syndecan-2 was demonstrated (Hsueh et al., 1998). This interaction between CASK and syndecan-2 was further confirmed by colocalization of these two proteins in rat brain by double- label confocal microscopy (Hsueh et al., 1998). Syndecans belong to the transmembrane heparan sulfate proteoglycan (HSPG) family (Couchman, 2003). In vitro, syndecan-2 induces dendritic filopodia formation, and later transforms filipodia into mature dendritic spines, thereby regulating dendritic spinogenesis (Ethell & Yamaguchi, 1999). Deletion of the CASK biding site

(EYFA motif) on syndecan-2 did not affect filopodia formation, suggesting that syndecan-2 does not require CASK to induce filopodia formation (Lin et al., 2007). Rather CASK maintains the morphology of dendritic spines, which is indicated by a time-course study using CASK- knockdown in cultured hippocampal neurons. At18 div CASK-knockdown induces dendritic spine retraction and the spine heads fail to enlarge, whereas at 15 div CASK knockdown does not affect spine density, size or length (Chao et al., 2008; Hu et al., 2016). The functional data described above do not necessary support the role of CASK in synaptic interaction and formation, rather it demonstrated that CASK is able to bind many different binding partners at both sides of the synapse.

Because CASK also interacts with protein 4.1 (Cohen et al., 1998; Biederer & Südhof,

2001) and because Protein 4.1 promotes association of F-actin with tetrameric postsynaptic

37 density protein spectrin, stabilizing the cytoskeleton (Correas et al., 1986), it has been proposed that CASK stabilizes the morphology of dendritic spines, by acting as a bridge between the F- actin cytoskeleton and the plasma membrane (Chao et al., 2008). The role of CASK in dendritic spine maturation is further regulated by SUMOylation, a post-translation modification.

Conjugation of small-ubiquitin-like modifier (SUMO) to CASK reduces the interaction between

CASK and protein 4.1, resulting in impaired spine formation (Chao et al., 2008).

In addition, CASK PDZ domain associates with glutamate receptor-interacting protein 1

(GRIP1) demonstrated by immunoprecipitation–immunoblotting analysis using CASK antibody

(Hong & Hsueh, 2006). CASK and GRIP1 co-precipitate from adult rat brain tissue. Through immunoprecipitation and immunoblotting experiments with adult rat brain tissue, it was demonstrated that GRIP1 specifically binds the C-terminal tails of AMPA (α-amino-3-hydroxy-

5-methyl-4-isoxazolepropionic acid) type glutamate receptors GluR2 and GluR3 (Dong et al.,

1997). Via the interaction with GRIP1, GluR2 and GluR3 were also co-immunoprecipitated by

CASK antibody from rat brain. Therefore, via interaction with GRIP1, CASK may be forming a complex with the AMPA receptor.

I.D.3.c. Nuclear CASK

In addition to the described localization and binding partners of CASK at the pre and postsynaptic sides, CASK also enters the nuclei of neurons and regulates gene expression. Tbr-1,

CINAP and Bcl11A, are three CASK-interacting proteins that bind the CASK GUK domain. All of them were originally identified by a yeast two-hybrid screen (Hsueh et al., 2000; Wang et al.,

2004a; Kuo et al., 2010b).

Tbr-1 is a neuron-specific T-box transcription factor involved in cerebral cortex development (Naiche et al., 2005). Hsueh et al. (2000) found that CASK is localized in the cytoplasm when expressed in COS-7 cells. However, simultaneous coexpression of CASK with

Tbr-1, resulted in CASK redistribution into nuclei, also co-localizing with Tbr-1. Thus, the interaction between the C-terminal region of Tbr-1 and the GUK domain of CASK results in nuclear localization of CASK. This suggests that CASK acts a coactivator of Tbr-1, consequently inducing the transcription of T box-containing genes, such as reelin (Hsueh et al., 2000) and

NMDAR subunit 2b (NR2b) (Wang et al., 2004b).

CASK-interacting nucleosome assembly protein CINAP (also called TSPY-L2, DENTT, or CDA1) binds histones, and stimulates transcription-factor binding to a specific nucleotide sequence (Wang et al., 2004b). Co-immunoprecipitation from rat brain extract demonstrated the interaction between CASK and CINAP in neurons (Wang et al., 2004b). Furthermore, confocal image analysis revealed the subcellular co-localization of CASK and CINAP, in the nuclei of cultured hippocampal neurons.

High-magnification immunofluorescent staining revealed that CASK, CINAP, and Tbr-1 co-localized with each other in the nuclei of a subset of neurons (Wang et al., 2004b). Tbr-1 and

CASK were co-transfected with CINAP into neuroblastoma cells in order to assess effects on transcription using a luciferase assay (Wang et al., 2004b). For this assay it was specifically used the NR2b promoter that contains a Tbr-1 binding site. NR2b encodes NR2B protein, critical structural and functional subunit of the NMDA glutamate receptor. By using a mutant NR2b luciferase reporter lacking the Tbr-1 binding site, it was found that Tbr-1, CASK and CINAP act on the Tbr-1 binding site of NR2b promoter. Cultured hippocampal neurons were stimulated using NMDA. Immunoblotting analysis showed that NMDA stimulation reduced the CINAP

39 protein levels, relative to the control, by ⁓60% in hippocampal neurons. Taken together, that data suggests that CASK and CINAP and Tbr-1 form a tripartite protein complex in the nucleus of the neurons, that regulates the expression of NR2b gene by the stimulation of NMDAR (Wang et al.,

2004b).

Many other CASK-binding partners have been identified, including Caskin (Tabuchi et al., 2002), carom (Ohno et al., 2003), parkin (Fallon et al., 2002), SynCam (Biederer et al, 2002).

Additional neuron-specific CASK-interacting proteins have been revealed by mass spectroscopy, such as alpha-adaptin, dynamin-associated protein 160, ninaC, superoxide dismutase 2, etc.

(Mukherjee et al., 2014). Further consideration of the proposed role of CASK as a scaffolding protein can be found in the Discussion (Chapter IV).

I.D.4. Other roles for CASK: insights from mouse models

There is an extensive literature on the proposed role of CASK as the synapse, where most of the above-mentioned studies have implicated CASK with a role in either the formation or the structure of the neuronal synapse. However, these activities or roles of CASK have not been supported by functional demonstration. Hence, it is hypothesized that other than synapsis formation, CASK may play different roles in neurons. For instance, neurons lacking CASK display overall normal electrical properties and do not alter synapse formation in mouse (Atasoy et al., 2007). In C. elegans, LIN-2, the homolog of CASK, is required for EGF receptor localization and signaling in epithelial cells, inducing vulval development (Hoskins et al., 1996).

This evidence for the role of CASK in C. elegans was obtained in mutant screen for cell-lineage specificity rather than synaptic function.

In mouse, CASK is located on the X chromosome, and it is essential for survival. Wilson et al. (1993) first reported a transgenic mouse mutant line, PyLMP.5, which harbors a transgene that codes for the Epstein-Barr virus latent membrane protein-1. Later, Laverty et al., (1998) found that PyLMP.5, was inserted into the CASK gene, thereby describing the first mouse mutation of CASK. The transgene integrated into an intron of the CaMK-like domain of CASK generating two hybrid transcripts in the mutant. Hence, this first mouse mutation of CASK is not considered null, rather hypomorphic. The unique phenotype included cleft palate, microcephaly and altered cranial morphology. Male mice die during the first day of life. No further analyses were conducted to explore neuronal function (Laverty et al., 1998). The second CASK mutation, a KO allele, was generated in a study conducted by Atasoy et al. (2007). By using the Cre-loxP system, the first coding exon of CASK was removed, abolishing the expression of CASK protein.

Similar to the first mouse mutant, CASK-KO mice exhibit cleft palate. However, all CASK-KO mice die at birth, and do not exhibit major developmental abnormalities, suggesting that cleft palate is not the cause of lethality (Atasoy et al., 2007). In addition, an increase in cell death in the thalamus of the CASK-KO was observed. However, neurons lacking CASK form normal synapses. Moreover, the electrical membrane properties displayed no significant change in cultured cortical neurons from CASK-KO brain tissue. The evoked neurotransmitter release was not affected (Atasoy et al., 2007). In the same study, a CASK knock-down mutation (where

CASK expression was suppressed by ⁓70%) was also generated. By using homologous recombination, Atasoy et al., 2007 generated a knock-in mutant in which the first coding exon of the CASK gene was flanked by loxP sites. This knock-in mouse expressed only ⁓ 35% of normal

CASK, which can be considered as CASK hypomorphic allele. CASK-knock-down mice are

41 viable but smaller than their WT littermates, defined by body weight, and have not been well characterized (Atasoy et al., 2007).

Srivastava et al. (2016) published a study about the third and last mouse CASK mutation to date. A pan-neuronal-specific CASK-KO allele was generated by using Cre-loxP-mediated gene deletion, which completely eliminated CASK expression. CASK-KO mice were generated by crossing a CASK-floxed female with a mouse line expressing Cre recombinase under the control of synapsin 1 promoter, which expresses Cre exclusively in post-mitotic neurons.

Contrary to the previous mice mutations that cause complete LOF, perinatal lethality is not seen in these animals. Granule cells are formed in the external granular layer during cerebellar neurogenesis and subsequently migrate inwards. CASK-null granule cells were capable of migration, as indicated by their proper location in the adult mouse cerebellum. Taken together, the data suggest that neuron-specific CASK deletion does not alter neuronal migration or survival. Rather, they showed that neuron-specific CASK-KO mice exhibited pronounced growth retardation and epilepsy. On the other hand, these data do not reveal the whole-animal CASK-KO phenotype.

Contrary to what has been suggested across the initial findings about CASK and its role in synapse formation, numerous experiments in the last decade have changed that view. Although it is evident that CASK has an essential brain function, it may not be required for synapsis formation or structure.

I.E. Drosophila CASK tools and contributions

One major difficulty in studying human genetic neurodevelopmental disorders is the relative lack of autopsy brain tissue, or the inability to manipulate this tissue for experimental

42 procedures. Because ID genes are highly conserved (Inlow & Restifo, 2004; Oortveld et al.,

2013), the use of an experimental animal model becomes suitable. For instance, human CASK protein shares 98% amino-acid sequence similarity with CASK in rats and mice (Hsueh et al.,

2006) and 74% with CASK Drosophila (Figure 3).

It is well-known that Drosophila has several distinct advantages as a model organism.

For example, the minimal genetic redundancy, compared with mammals, and the high degree of orthology between mammalian and fly genes, make Drosophila a powerful system (Bier, 2005;

Pandey & Nichols, 2011). Moreover, Inlow & Restifo (2004) found that 87% of the 282 genes known to cause ID in humans have homologs in Drosophila and 76% have at least one candidate functional ortholog. Homolog genes that relate through speciation from a single ancestral gene present in their most recent common ancestor are known as orthologs (Kuzniar et al., 2008).

The nervous system of a fly displays a moderate level of complexity compared to human brain. Although the Drosophila brain is simpler than mammals, it has well-organized centers for distinct functions (e.g., olfactory, visual, gustatory, motor control, and learning and memory centers) (Pereanu et al., 2010).

Drosophila CASK, the sole ortholog, has a size of 39274 kb, with 18 exons. It is located on the distal right arm of chromosome 3 at position 93F10-93F12 (Figure 4A-B) (Gramates et al., 2018), and is transcribed right to left. The CASK locus has a second transcriptional start site farther downstream that results in a transcript encoding a smaller protein of unknown function called CASK-α. CASK-α encompasses the MAGUK core (PDZ, SH3 and GUK domains), but lacks CaMK-like and L27 domains. In turn, the long isoform, the canonical CASK, is called

CASK-β (Figure 4E and 4F).

Figure 3. Human and Drosophila CASK are highly conserved phylogenetically. Color-coded schematic comparison of CASK protein domains in H. sapiens and D. melanogaster showing identical organization. Percentage of aminoacids sequence similarity for each domain between the human (NP_003679.2) and fly (NP_001163681.1) Individual domains of Drosophila CASK, based on NCBI Conserved Domain Database: CaMK-like (residues 8-308), L27 (residues 347–

462), PDZ (residues 499–572), SH3 (residues 585–645) and GuK (residues 709–882).

Designated domain regions are to scale.

The first genetic disruption of Drosophila CASK was obtained during a study of the adjacent torso-like (tsl) gene (Martin et al., 1994; Figure 4B). An X-ray mutagenesis was performed using the line tsl604, which carries a P-element insertion between tsl and CASK

(originally named ‘caki’). Two zygotic-lethal deficiency chromosomes, i.e., disruptions of one or more normal genes that are essential to the survival of the zygote, were identified, with these

Figure 4. Drosophila CASK gene and genetic reagents.

Figure 4. Drosophila CASK gene and genetic reagents. A. Chromosomal region of CASK.

Polytene chromosome map showing that CASK is located on the distal right arm of chromosome

3 at position 93F10-93F12. The numbers and letters below the chromosome indicate the cytogenetic divisions and subdivisions of the Drosophila genome, respectively. B. CASK gene map. Genomic DNA sequence interval, 21,783–21,822 kb, is indicated. The scale is shown in the upper left-hand corner. CASK has two transcriptional start sites (arrows). Location of P transposable element insertion site, EY07081, indicated within the first intron at the 5’ end of the gene. In green is tsl, a neighboring gene 3’ to CASK. C. Overlapping chromosomal deficiencies that uncover CASK, diagrammed to show the breakpoints. Df(3R)X307 deletes a large portion of

CASK, and Df(3R)X313 removes not only most of CASK, but also a large piece of the 3’-flanking

DNA (indicated by the arrow). Dashed lines indicate that the breakpoint locations are uncertain.

D. CASK alleles used in these studies. Ex33, a precise excision of the above-mentioned P- element (EY07081). The imprecise excision, CASK-mutant allele, ∆18, harbors a deletion of exon 2. CASK∆18 also has a small piece of a roo transposon inserted at the deletion site, further lengthening the gap between the promoter region and the remaining coding exons. The UAS-

CASK+ 10.20 construct expresses the wild-type cDNA of the CASK-β isoform (CASK-RB mRNA in FlyBase). E. CASK transcript maps. Gray boxes represent non-coding regions; colored boxes represent translated regions. Note that CASK-α is missing the 5’ exons. F. CASK proteins. Color code matches the transcript coding regions. Note the view of the protein is flipped 180° relative to the diagrams shown above, N’ to the left to C’ to the right. Number of amino acids (aa) indicated.

47 zygotic lethal genes located respectively in each side of tsl gene. These large and partially overlapping deletions, identified as X307 and X313 (Figure 4C) were viable in the transheterozygote, i.e., X307/X313. CASK-null flies predicted to eliminate both CASK-α and

CASK-β isoforms, were then produced by crossing together these two chromosomal deficiencies

(Martin & Ollo, 1996). The resulting transheterozygous flies were infertile and showed reduced walking speed. Moreover, due to the extension of the deletion into neighboring genes, these

CASK-mutant flies presented additional gene disruptions. Because of the nature of this CASK- synthetic deletion, i.e. a partially overlapping deletions creating a null condition, only limited conclusions can be made about the role of CASK proteins in Drosophila.

Similarities regarding fundamental neurobiology processes between Drosophila and humans, include synapse formation, neuronal communication, and membrane trafficking, among others (Bolduc & Tully, 2009; Hirth, 2010). For instance, in flies with overlapping deletions

Df(3R)X307/Df(3R)X313, the spontaneous neurotransmitter release at the neuromuscular junction (NMJ) was four-fold increased, revealed by the intracellular analysis of miniature end- plate potentials in the indirect flight muscle (Zordan et al., 2005). By electrophysiological means using larval and embryonic NMJ of body wall muscles, an independent study was carried out with the same CASK-synthetic deletion (Chen et al., 2011). Presynaptically, evoked excitatory junction current amplitudes were significantly decreased in NMJs of CASK mutants compared to controls. In addition, there were fewer spontaneous synaptic events and reduced synaptic vesicle cycling. Notice that the reduction in the spontaneous neurotransmitter release reported in this study contradicts the results obtained by Zordan et al. (2005), Chen et al. (2011) state that they cannot explain the difference between those contradictory results. Loss of CASK did not affect the total NMJ area and number of boutons, but a reduction in the number of synapses per bouton,

48 and thus a decreased synapse density, was seen, as assessed by immunoreactivity using specific neuronal markers. On the other hand, the loss of postsynaptic CASK, caused reduced spontaneous synaptic current amplitudes, smaller currents, and loss of postsynaptic glutamate receptors (Chen et al., 2011).

To create a more conventional CASK LOF allele, than the synthetic deletion produced by

Martin & Ollo (1996), Slawson et al. (2011) generated a small deletion within the CASK gene, resulting from imprecise excision of a P-element at the 5’end of the gene (Figure 4D). This allele eliminates the full-length CASK protein, i.e., it is a CASK-β null mutant, hereafter called ∆18

(see § II.B.1.).

In Drosophila, CASK is highly expressed in the developing nervous system throughout embryonic and postembryonic development, as well as in the mature brain (Martin & Ollo, 1996;

Lopes et al., 2001). In the adult brain CASK is present in all neural structures, including the central brain and the optic lobes. There is also a high level of CASK expression in the eye imaginal discs and in the adult retina. In addition, a transient expression of CASK in Malpighian tubules in the embryo, and in the gut, during the larval and adult stages is detected (Martin &

Ollo, 1996; Lopes et al., 2001).

Despite the obvious physical differences between flies and mammals, and especially their neuroanatomical differences, many of the molecular mechanism underlying learning and memory are conserved (Bolduc & Tully, 2009). For instance, established paradigms in

Drosophila (e.g., olfactory aversive conditioning, courtship memory) have revealed cognitive defects in fly mutants for orthologs of genes involved in human ID (Bolduc & Tully, 2009). ID

(cognitive and adaptative deficits) resulting from CASK mutations suggest a role for WT CASK in plasticity and learning (Malik et al., 2014). CaMKII plays a central role in synaptic plasticity

49 and memory by autophosphorylating itself upon activation by calcium/calmodulin and thereby regulating its own kinase activity (Hell, 2014). It has been demonstrated that when calcium/calmodulin is present Drosophila CASK forms a complex with CaMKII in vitro (Lu et al., 2003). In addition, deletion of either full-length CASK or its CaMK-like and L27 domains impairs middle-term memory and long-term memory in Drosophila (Malik et al., 2013).

Furthermore, human CASK expression in Drosophila rescued the dysregulation of CaMKII autophosphorylation induced by the CASK deletion, suggesting that the human CASK may also regulates CaMKII autophosphorylation (Gillespie & Hodge, 2013). Although is not clear if

CASK also binds to CaMKII in mammals, the control of CaMKII and its autophosphorylation is critical for synaptic plasticity and memory in mammals and Drosophila (Malik et al., 2014).

The fruit fly, Drosophila melanogaster is an excellent animal model system for understanding the biology of brain development and to dissect the genetic basis of human diseases (Inlow & Restifo, 2004; Bier, 2005). For instance, learning and memory formation, sensory perception, integration and motor output, are complex behaviors that share similar neurobiological bases in mammals and insects (Margulies et al., 2005; Keene & Waddell, 2007).

To date, several Drosophila neurodevelopmental disorders models, including ID, have been established (McBride et al., 2005; Restifo, 2005; Gatto & Broadie, 2011; Oortveld et al., 2013).

For instance, fly models for single-gene disorders that disrupt brain development include fragile

X syndrome (Michel et al., 2004; McBride et al., 2005) and Angelman syndrome (Wu et al.,

2008), and neurofibromatosis type 1 (The et al., 1997).

Hence, the study of molecular and genetic pathways can shed light on potential therapeutic intervention (Pandey & Nichols, 2011). In this regard, primary neuronal culture

50 methods in Drosophila arises as a powerful tool for drug discovery (Kraft et al., 2013), providing opportunities for targeting specific central nervous system disorders, including ID.

I.F. Problem statement and research questions addressed in this dissertation

ID is a manifestation of many neurodevelopmental disorders characterized by significant limitations in both intellectual functioning and adaptive behavior (American Psychiatric

Association, 2013). Estimates of the prevalence of intellectual disabilities range from 1% to 3% of the population (McKenzie et al., 2016). Single-gene mutations are common causes of ID.

CASK is one of the ~850 genes whose mutations cause ID. Unfortunately, there are very few ID disorders for which the underlying defect of brain development has been well described and understood. Very little human-brain autopsy material is available for study of rare ID disorders.

There is a need to understand the pathological changes that cause defects in brain development and function.

CASK has been found to be part of the presynaptic structure interacting with several binding partners (Hata et al., 1996; Sun et al., 2009; Chen et al., 2011), and the postsynaptic sites, where it maintains the morphology of dendritic spines (Chao et al., 2008). In addition,

CASK also enters the nuclei of neurons and regulates transcription (Hsueh et al., 2000; Wang et al., 2004a and b). Interestingly, many of the binding partners of CASK are also encoded by ID genes (e.g. CDK5, NRXN1) (Samuels et al., 2007; Südhof, 2008). Although there is a large body of molecular and cellular neuroscience literature on CASK at the synapses claiming that the primary role of CASK is as a scaffolding protein, there has not been an assessment of neuron morphology, neither brain size in animal models with CASK mutations. Human CASK semidominant LOF mutations result in intellectual disability with microcephaly and pontine and

51 cerebellar hypoplasia (Najm et al., 2008; Moog et al., 2011; Burglen et al., 2012; Moog et al.,

2015; Hayashi et al., 2017). These human phenotypes demonstrate that CASK is essential for normal brain size. The human phenotype of reduced corpus callosum also suggest that CASK may regulate neuronal differentiation, specifically axon outgrowth. Except for one study assessing dendritic spines in mice (Chao et al., 2008), a role of CASK in regulating neuronal morphogenesis has not been demonstrated. Therefore, the aim of these studies was to characterize the role of CASK in neuronal morphogenesis and brain size in a rigorous quantitative manner, which had not been previously studied.

In order to understand the cellular mechanisms of ID caused by human CASK mutations and to study the role of CASK in brain development, Drosophila CASK LOF mutants (∆18/∆18) were used. Together, the questions addressed in this dissertation allow a determination of whether Drosophila can be used to model CASK disorders.

In Chapter II, I examined neurite-arbor morphogenesis in vitro by using larval CNS neurons primary cultures. The following questions were addressed:

• Is CASK essential for normal neurite-arbor morphogenesis?

• What is the phenotype of the CASK-mutant neurite arbor?

• Is the CASK phenotype recessive, dominant, or semi-dominant?

• Do CASK chromosomal deficiencies uncover the phenotype of CASK mutants?

• Does wild-type CASK+ rescue the CASK-mutant phenotype?

In Chapter III, I analyzed brain size by using histological examination in parallel with head and overall body size. I addressed the following questions:

• Does the CASK-mutant in vitro phenotype have an in vivo correlate (small-brain)?

• Is CASK required for normal brain volume?

• Does CASK regulate overall head volume?

• Is CASK controlling body length?

Chapter II: Neurite-arborization abnormalities of CASK-mutant neurons in dissociated cell culture

II.A. Rationale

The functional architecture of the brain is based upon synaptic connectivity of neurons.

Axons and dendrites, collectively known as neurites, define neuronal connectivity. Neuronal morphogenesis is critical to create functional interconnected networks (Chklovskii, 2004; Reid,

2012). Neurite outgrowth, i.e., axonal and dendritic arborization, are key morphological events that characterize neuronal differentiation and define neuronal arbor size and shape (Wen &

Chklovskii, 2008; Van Elburg & Van Ooyen, 2010; Gillette & Ascoli, 2015). Accordingly, structural abnormalities of axonal pathways and dendritic arbors are underlying many development disorders (Kaufmann & Moser, 2000; Dierssen & Ramakers, 2006; Edwards et al.,

2014). For instance, changes in dendritic arbors have been observed in post-mortem brain tissue of individuals with ID disorders, examined by Golgi staining (Kaufmann & Moser, 2000), such as Angelman syndrome (Jay et al., 1991) and autism (Bailey et al., 1998) among others. In conditions such as Down syndrome, Rett syndrome, and fragile X syndrome, dendritic spine abnormalities, such as abnormally long and thin spines, reduced dendritic length and abnormal branching have been reported (Purpura 1974; Kaufmann & Moser, 2000; Dierssen & Ramakers,

2006). Morphological studies in postmortem brain samples from individuals with Rett syndrome have shown a significant reduction in dendritic branching of cortical pyramidal neurons

(Armstrong et al. 1995; Armstrong, 2001). In addition, cortical neurons from patients with fragile X syndrome have increased spine density, longer dendritic spines and more spines with immature morphology (Hinton et al., 1991; Irwin et al., 2001). Similarly, the fragile X knockout mouse model also reproduced the main morphological abnormalities observed in human fragile

X patients (Irwin et al., 2002). Although, the fragile X mice did not exhibit increased dendritic spine density, their dendrites showed more long spines, fewer short spines and more spines with immature-like morphology (Irwin et al., 2002). Another common cause of structural abnormalities in the human brain are defects in the corpus callosum, the largest white matter tracts in the brain which consists of about 200 million axons (Edwards et al., 2014). Single-gene mutations associated with agenesis of the corpus callosum (ACC) result in either total absence

(complete ACC), shorter (partial ACC), or a thinner than usual corpus callosum (Edwards et al.,

2014). Together, the available data suggest that many ID genes are likely to regulate neuronal morphology during brain development, with effects on dendritic arbors and axonal extension and branching.

CASK is one of the ~850 genes whose mutations cause ID. CASK interacts with many different binding partners (Mukherjee et al., 2014) in different neuronal subcellular regions to perform specific functions (Hsueh, 2006 and 2009; Oliva et al., 2011). Because laboratory studies of CASK have mostly focused on the synapse (see § I.D.3.), there is limited understanding about the role of CASK before synapses have formed. Few studies have addressed the function of CASK in neurite outgrowth and the formation of neuronal arbors. For instance,

Chao et al. (2008) showed that knockdown of CASK in cultured hippocampal neurons reduced spines density and length, whereas overexpression of CASK increased the density of dendritic spines. These data suggest that CASK stabilizes the morphology of dendritic spines but did not address dendritic arbor size or complexity. In another study, it has been shown that CASK synergizes with Bcl11A-L (B cell lymphoma 11), upregulating neurite outgrowth and branching

(Kuo et al., 2010). Kuo et al. (2010) demonstrated that CASK interacts with Bcl11A, a Krüppel- like zinc-finger transcription factor. Bcl11A, expressed specifically in hematopoietic tissue and

55 brain, is involved in the regulation of axon branching and dendritic outgrowth (Kuo et al., 2009).

Overexpression of Bcl11A-L in cultures of rat hippocampal neurons caused delayed formation of the first branch, reduced axon-branch complexity (number of branching points encountered from the axon's first branch point to the tip), shortening of axonal length, and reduction in both dendrite number and total dendrite length. CASK overexpression in the same system did not cause any change. However, co-expression of CASK enhanced the suppressive effect of the

Bcl11A on axon length and axon-branch complexity (Kuo et al., 2010), suggesting that the

Bcl11A-L and CASK work together to restrict axonogenesis. In turn, Watkins et al. (2013) demonstrated that CASK recruits FRMD7, suggested to plays a role in neurite extension, to the plasma membrane, where FRMD7 enhances the neurite outgrowth by modulating the actin cytoskeleton. Despite the hypothesized differences about the role of CASK in neuronal development, it seems that CASK has some functions in neuronal morphology and structural connectivity.

Previous studies in the Restifo laboratory have demonstrated that primary neuron cultures can reveal mutant neuronal phenotypes that are either difficult or too subtle to detect at the whole-brain level (Kraft et al., 2006; Lewis, 2015). The present study aims to answer the question of whether CASK protein is necessary for the inherent morphogenetic capabilities of

CNS neurons.

A combined approach using Drosophila genetics, primary neuronal culture, and semi- automated processing of 2D neuron images, allowed me to investigate the CASK LOF phenotype in neuron morphology. To examine neurite-arbor morphogenesis in vitro and to quantify neurite- arbor size and shape, I used primary cultures prepared from the entire CNS of wandering third instar larvae of Drosophila. My analysis was focused on neurite length, neuron area, higher-

56 order branch number (branch number - primary neurites), and branch density (branch number/neuron territory). The evaluation of these morphometric parameters in vitro provides a basis for ascertaining the size and shape of neurite arbors, which in turn influence neuronal function (Wen & Chklovskii, 2008; Van Elburg & Van Ooyen, 2010; Gillette & Ascoli, 2015).

The initial experimental design involved a direct comparison between neuron cultures prepared in parallel, with two homozygous genotypes: CASK ∆18/∆18, a loss-of-function mutation, and its genetic control, CASK Ex33/Ex33. I focused on the period immediately before the metamorphic transition because genes with roles in neuron development are more likely to show phenotypic expression during this developmental interval (Truman, 1990). I hypothesized that mutations in

CASK would disrupt the size or shape of neurite arbors in vitro.

Once I detected a phenotype, I further investigated if this phenotype was recessive, dominant or semi-dominant. This allowed me to anticipate how much restoration of function would be required to rescue the CASK phenotype. Finally, I used deficiency mapping and the expression of transgenic constructs to demonstrate that the observed neurite-arbor phenotype maps to CASK.

II.B. Materials and Methods

II.B.1. Drosophila culture and genetics

Fly stocks were maintained at room temperature (21-23°C) on a corn flour-nutritional yeast-agar medium (Elgin and Miller, 1978). Experimental cultures were reared at optimal larval density under standard conditions of 25°C, 60-80% relative humidity, as previously described

(Kraft et al., 2006). The CASK LOF P18 allele (hereafter called ∆18), was created by the Griffith lab (Slawson et al., 2011) via imprecise excision of the P-element EY07081, a P{EPgy2}

57 insertion in the first intron of CASK (Figure 3). This CASK-mutant ∆18 allele is a 2624-bp deletion that results in the absence of full-length CASK protein (CASK-β isoform). However, the

CASK-α isoform can still be produced from an internal promoter. This ∆18 allele also has a small piece of a roo transposon inserted at the deletion site (Slawson et al. 2011). Precise excision of the same P-element generated Ex(EY07081)-33, referred to here as Ex33, which serves as the genetic-background control and CASK allele chromosome. The UAS-CASK+ 10.20 construct expresses the wild-type cDNA that corresponds to the full-length CASK-β isoform,

CASK-RB mRNA (FlyBase CG6703-RB). The above Drosophila CASK stocks used in this study were kind gifts from Dr. Leslie Griffith (Brandeis University, Waltham, MA). elav-

Gal4C155/elav-Gal4C155; CyO/Sp; Sb/TM3 Ser (Lin and Goodman, 1994) with the elav neuron- specific GAL4 C155 driver, was provided by Konrad Zinsmaier (University of Arizona, Tucson,

AZ).

Stocks of chromosomal deficiencies that remove CASK, Df(3R)X307 and Df(3R)X313 (FlyBase

ID FBal0048070 and FBal0048071, respectively) (Martin and Ollo, 1996) were obtained from the Bloomington Drosophila Stock Center (Bloomington, IN; see URL list), and maintained over the balancer TM6B, Tb e. To generate each deficiency, hemizygote over balancer flies were crossed with homozygous Ex33 or ∆18 flies to yield hemizygous control, Df(3R)X307/Ex33 or

Df(3R)X313/Ex33, and hemizygous mutant, Df(3R)X307/∆18 or Df(3R)X313/∆18 progeny. To obtain the heterozygotes w/w; +/+; CASK Ex33/CASKΔ18, virgin females w/w; + / +;

CASKΔ18/CASKΔ18 were crossed with w/Y; +/+; CASK Ex33/CASK Ex33 males.

To assess transgenic rescue, I used the GAL4-UAS binary system to drive selective expression of the gene of interest (Brand & Dormand 1995). Briefly, the expression of the transcription factor, GAL4, is under the control of enhancers at the transgene insertion site; this

58 is called the “driver”. UAS (upstream activating sequence) controls the expression of a target gene, which may be a reporter. GAL4 binds to UAS, activating the transcription of the target gene in a time- and space-specific manner. To determine the neuronal expression of the driver, I

+ generate flies that contain a wild-type cDNA transgene of CASK, UAS-CASK , and the driver

C155-GAL4. Specifically, the neuron-specific driver elav-Gal4C155/elav-Gal4C155 and the UAS-

+ CASK transgene, were each crossed into the CASK ∆18 mutant background. The ELAV RNA- binding protein encoded by elav (embryonic lethal abnormal visual system) is expressed in the central nervous system as well as the peripheral nervous system of embryos, larvae, pupae, and adults, exclusively in neurons (Robinow & White, 1998). Thus, by using the neuron-specific enhancer elav and the driver GAL4, pan-neuronal expression can be achieved.

Flies of these two lines were then crossed:

♀ elav-Gal4C155/ elav-Gal4C155; + / +; CASKΔ18 / CASKΔ18

+ + ♂ w / Y; UAS-CASK / UAS-CASK ; CASKΔ18 / CASKΔ18

+ to generate elav-Gal4C155 w/+; UAS-CASK /+; ∆18/∆18 progeny larvae.

To verify the presence of the driver in the CASK ∆18 background and determine what fraction of cultured larval CNS neurons express a target gene with this GAL4 driver, UAS-

SuperGFP II homozygous flies were crossed with the elav-Gal4C155 strain.

elav-Gal4C155/w; UAS-CASK/+; CASKΔ18/CASKΔ18

w/Y; UAS-SuperGFP II/ UAS-SuperGFP II; CASKΔ18/ CASKΔ18

to generate elav-Gal4C155/+; UAS-Super-GFPII /+; CASKΔ18/ CASKΔ18. Larval CNS neurons of these progeny were tested in vitro.

II.B.2. Primary neuronal cell culture

Dissociated neuronal cultures were prepared from the whole CNS of wandering third instar larva as previously described (Kraft et al., 2006, 2013; Smrt et al., 2015). Larvae selected from different genotypes were matched for sex and culture density. The CNS was explanted by microdissection in culture media, composed of Schneider’s Insect Medium (Invitrogen, San

Diego, CA), 10% FBS and 50 μg/mL bovine or human recombinant insulin (Sigma, St. Louis,

MO). Two FBS sources were used; most of the experiments were conducted using FBS from

Hyclone (Logan, UT; lots APC20860 or AZA180873), whereas in a few of them FBS from

Optima (Atlanta, GA; lot #D0118) was used. To facilitate the mechanical dissociation into individual cells, the tissue was incubated in Liberase Blendzyme (Roche, Indianapolis, IN) at a final concentration of 0.21 Wünsch units/mL. This mix of purified collagenase I, collagenase II, and the neutral protease dispase, digests the extracellular matrix. After this enzymatic digestion, mechanical trituration by manual pipetting was performed.

Unless otherwise specified, the CNS-cell suspension was plated in custom culture dishes made in-house. To fabricate them, an 8-mm hole was drilled into the center of a 35-mm polystyrene culture dish. Then, a photoetched (“gridded”) glass coverslip (Bellco, Vineland, NJ) was glued to the bottom of the dish by using a thin layer of Sylgard (Sylgard 184, Dow-Corning,

Midland, MI). For the transgenic-rescue experiments pre-made 35-mm culture dishes with 7-mm wells (MatTek, Ashland, MA) were used. Dishes were sterilized by exposing them to ultraviolet light (185-254 nm) for two hours. After sterilization and prior to the preparation of dissociated cells, dishes were coated with Concanavalin A (Sigma) and laminin (BD Biosciences, Franklin

Lakes, NJ). CNS-cell suspension was then distributed into five or six culture dishes by pipetting

⁓ 100 µl into each well. After a 2–3 h incubation at 25°C to allow the cells to adhere to the

60 substrate, the dishes were flooded with an additional 900 µl of culture medium and maintained at

25°C.

II.B.3. Immunostaining and morphometric analysis of 2D neuron images

Cells were cultured for three days in vitro (div) (⁓70-80 h post-plating) at 25°C before fixation. Cultures were monitored every day to assess outgrowth, survival and to check for yeast contamination. Neurite outgrowth starts within hours in these cultures, and an elaborate neurite arbor can be seen at three div (Kraft et al., 2006; Smrt et al., 2015). Nonetheless, the overlap between neighboring arbors is minimal, which allows the evaluation of individual neurons.

Fixation and immunostaining were performed as described previously (Kraft et al., 1998, 2006;

Smrt et al., 2015). Culture dishes were rinsed twice with ice-cold, calcium-free Ikeda Ringer’s saline (130 mM NaCl, 4.7 mM KCl, 1.8 mM MgCl2, 0.74 mM KH2PO4, 0.35 mM Na2HPO4) and fixed using fresh 4% EM-grade formaldehyde (Ted Pella, Redding, CA) in PBS, pH 7.4, for 30 min at 4˚C. Cells were washed three times at 4˚C in PTN (0.1 M sodium phosphate buffer pH

7.4, 0.1% Triton X-100 (v/v), 0.1% sodium azide (w/v)). To visualize neuronal membranes, the

Drosophila Nervana 2 protein (Sun and Salvaterra, 1995) was labeled with a polyclonal goat anti-horseradish peroxidase (anti-HRP) antiserum (Sigma) at 1:500 in PTN. Neurons were incubated with the above primary antiserum overnight at 4ᵒC. The next day, cells were washed with PTN six times for 15 minutes each at room temperature with orbital rotation. The primary antibody was detected with either Donkey anti-Goat IgG Alexa Fluor® 488 (Molecular Probes,

Life Technologies) or Donkey anti-Goat IgG Alexa Fluor® 568 (Molecular Probes, Life

Technologies) both diluted 1:500 in PTN. Dishes were incubated with the above secondary antisera at 4ᵒC in the dark for 3h, after which six washes with PTN for 15 minutes were

61 performed. To remove any trace of detergent (Triton), a final rinse with 0.1 M Tris-HCl buffer, pH 8.0, was performed before mounting with 12% w/v PVA (polyvinyl alcohol; Sigma, St.

Louis, MO) and 1.5% w/v DABCO (1,4-diazabicyclo [2.2.2] octane; Sigma), an anti-fading reagent that delays the loss of fluorescence.

Cultured neurons were examined on a Diaphot 300 inverted microscope (Nikon, Tokyo,

Japan) with a 60X oil-immersion objective (numerical aperture, 1.4). When viewing or imaging cells, phase-contrast optics and/or epifluorescence illumination were used. Detection of fluorescent signal was achieved with filter cube Chroma #41001 (exciter 460-500 nm; dichroic

505 nm; bandpass emitter, 510-560 nm) for Alexa Fluor® 488 and with a Nikon G-2A filter cube (exciter, 510-560 nm; dichroic 580 nm; long-pass emitter, 590 nm) for Alexa Fluor® 568.

Images were collected with a Hamamatsu ORCA285 camera (Hamamatsu Photonic Systems,

Bridgewater, NJ) and HCImage® software (v.2.0.0.0; Hamamatsu).

For quantitative analysis of each genotype, systematic sample of approximately 100 neurons (i.e., cell body and complete neurite arbor) was obtained. The photoetched glass coverslip has a gridded pattern that contains alphanumeric labeled squares. Cells were selected by following the ‘staircase’ method (Figure 5A), in which each step is composed of the tread and rise, i.e., following a path of neurons along the tread and the rise of ‘steps’. At 10X magnification, I located the boundary of the well and drew it on a coverslip-grid map. My starting point was the left-most vertical grid line. From there, my imaging proceeded down and to the right, always along the tread and the rise of ‘steps’. I included all the non-overlapping neurons along each tread and rise, that fell within my field of view at 60X. In some cases, a neuron with a small degree of overlap of its neurite arbor with a neighbor was also included.

Exclusion criteria for imaging a neuron included weak anti-HRP signal (suggesting cell death),

Figure 5. Neuron-sampling and imaging output.

Figure 5. Neuron-sampling and imaging output. A. Strategy for sampling neuron images from a single culture dish. Circle illustrates an 8-mm culture well, superimposed on a photoetched grid at the same size scale. In the “staircase” sampling method, the experimenter uses the grid to find a standard starting point (blue dot), and from there, the imaging acquisition proceeds systematically always along the tread and the rise of a staircase track (red lines, note directional arrows). B. A fluorescently labeled neuron after three div, overlaid with NeuronMetrics software output. The neuron territory is bounded by the yellow polygon. A thicker yellow line is superimposed to improve visualization. The neuronal cell body in indicated with blue. All the branches of the neurite skeleton are in green.

64 damaged cell bodies, and overlapping arbors that could not be resolved to allow accurate image analysis.

Acquired digital images were converted from 16-bit to 8-bit tiff for analysis. For neurons with larger arbors extending beyond the microscopic field, multiple overlapping images were acquired and merged by “stitching”, using PanaVue ImageAssembler® software (Panavue Inc.,

Québec, Canada).

Quantitative analysis of fluorescently labeled cultured neurons was done by using

NeuronMetricsTM (Narro et al., 2007). This specialized semi-automated software converts each neurite arbor into a one-pixel-wide skeleton connected to the cell body region of interest (ROI) and bounded by the neuron's convex polygon (Figure 5B). The software computes primary neurite number, the length of the neurite arbor, branch number estimate, area and perimeter territory. Primary neurites are determined by locating the points where they emerge from the cell body. Territory is defined as the smallest convex polygon that encloses the neuron’s skeleton arbor and cell body ROI (Narro et al., 2007).

In most of the experiments, the dishes were coded, such that image acquisition was performed blind to neuron genotypes. Coding was done by a lab member not participating in the experiment, and the code was revealed after the statistical analysis. I prepared a minimum of three independent replicates for each experiment; in most of the cases, these were set up at different times.

Pair-wise comparisons, between different genotypes, of individual neurite-arbor parameters were made using the Mann-Whitney rank-sum test, a nonparametric statistical test for non-normally-distributed data (Fay & Proschan, 2010). Initially, SigmaStat (Systat Software, San

Jose, CA) was used to run the test and graph the data. Recent, R Studio (version 3.2.1; R Core

Team, 2015) was used. Fellow doctoral candidate Sara Lewis and I developed the appropriate pairwise comparison script for varying numbers of genotypes being analyzed in a single experiment. I developed an updated version of the graphing code in 2018.

II.B.4. Linear mixed modeling with nlme

Results from several experiments that compared the same pair of genotypes using a linear mixed-effects model were analyzed. The mixed effects model compares genotype-to-genotype differences within each experiment, and combines these comparisons across experiments, thus removing inter-experiment variability from the comparison. This implements a standard statistical approach for combining results across multiple replicate experiments. The mixed effects model fits a random intercept (mean) for each experiment, and genotype fixed effects within an experiment. The procedure is implemented in the R package nlme using the lme function. This package is supplied in the R system for statistical computing (version 3.2.1; R

Core Team, 2015) under the GNU General Public License. The appropriate statistical model to use for comparison, and the corresponding R script was developed by Dr. Dean Billheimer,

Professor of Statistics and Director of Statistical Consulting, along with biostatistics graduate students Nathan Grunow and Kyle Humphrey.

II.C. Results

II.C.1. CASKΔ18 homozygous neurons have a “bushy” phenotype

My motivation to examine developing neurons in primary culture, and to quantify their ability to generate neurite arbors, came from previous observations in our lab, where striking in vitro phenotypes have been revealed and characterized. For instance, fascin-deficient singed-

66 mutant neurons have a “filagree” phenotype, with excessive clockwise curvature and failure of proximal-to-distal tapering (Kraft et al., 2006). I hypothesized that Drosophila CASK LOF mutation would cause neuron-morphology defects.

To test my hypothesis, I prepared primary cultures from the entire CNS of wandering third instar larvae of Drosophila. The larval CNS harbors neuroblasts, terminally differentiated neurons and glia, functional for larval life, as well as immature neurons and glia cells of the presumptive adult nervous system (Truman, 1990; Tissot & Stocker, 2000). In our neuron culture system glial cells are absent after three div, as demonstrated by lack of Repo-positive glial cells and neuroblast proliferation is not detectable by BrdU incorporation (Luedeman, Chaung,

Levine, and Restifo, unpublished observations). To answer the question of whether neurons from

CASK∆18 mutant have altered neurite morphology, CASK∆18 and CASK Ex33 homozygous larval CNS neurons were compared. Although the CNS of CASK mutants was visibly smaller than that of control larvae (Figure 12A), cell densities after dissociation and plating were comparable by eye, i.e., based on visual comparison at the microscope.

Neurite-arbor size and shape were quantified using NeuronMetricsTM (Figure 5B) software for semi-automated image analysis (Narro et al., 2007; Smrt et al., 2015). Total neurite length, territory area (the area occupied by the cell body and neurite arbor), number of primary neurites and branch number, are the parameters used to describe the neurite arbor morphology of a cultured neuron. In addition, higher-order branches, the number of higher-order branches per 1000 µm2, is sometimes used an additional morphometric parameter (Kraft et al.,

2006).

CNS neurons lacking CASK-β (∆18/∆18) grew small arbors in vitro with an altered shape that was obvious to experienced practitioner of primary neuron culture. (Figure 6). In

Figure 6. CASKΔ18 mutant neurons have a “bushy” defect in vitro.

Figure 6. CASKΔ18 mutant neurons have a “bushy” defect in vitro. Representative images of cultured neurons derived from the entire larval CNS after three div. (60X magnification) A, B, C.

Typical CASK Ex33/Ex33 neurons with fewer higher-order branches, covering a large territory area. D, E, F. CASK-mutant neurite arbors are reduced in size, with many higher-order branches.

A-B. Representative images from approximately the 25th percentile of length, territory area and branch densities. C-D. Representative images from approximately the median of length, territory area and branch densities. E-F. Representative images from approximately the 75th percentile of length, territory area and branch densities. Scale bar A-F = 10 μm.

69 comparison to control neurons (Ex33/Ex33) these mutant neurite arbors had significant reductions in neurite length, territory area, and higher-order branch number (branch number - primary neurites), but branch density (number of branches per 1000 µm2) was increased (Figures

6, 7). The descriptive name for this phenotype is “bushy”. Because the entire CNS tissue is dissociated and cultured, a large heterogeneity of neurons is expected. Because the CASK-mutant bushy phenotype is visible across the entire primary culture, there is no indication that this is restricted to a specific subset of neurons.

To validate these results, multiple independent replicate experiments comparing CASK

∆18 homozygous mutants with Ex33 homozygous controls were performed over a period of approximately 4 years in two different labs and with different lots of biological reagents. A particular kind of replication included the assessment of neurite-arbor morphology after CNS dissociation using microfluidic devices (see Appendix B), a collaboration with colleagues in the

University of Arizona, Dept. of Aerospace & Mechanical Engineering (Jiang et al., 2015). A compilation of the data on all the morphometric parameters analyzed can be found in Table 2. In addition, to confirm that CASK LOF is responsible for the bushy phenotype a number of phenotype-mapping experiments were also carried out and are described in the next sections.

Across eight independent replicate experiments ∆18/∆18 mutant neurons always displayed a highly significant decrease in total neurite length and territory area. This decreased arbor size was consistently accompanied by increased branch density. These three parameters have been shown as the most consistently affected by the CASK mutation and define the “bushy” phenotype.

In most of the replicates (seven of eight), a reduced number of higher-order branches was observed. An additional secondary parameter, the ratio of branches to neurite length, was

Mann-Whitney Rank-Sum Test Results Length Territory Area Branches (all) Primary Processes HO Branchesa Branch Densityb HO-Branch Densityb Branches/Length Experiment and Genotypes Median Median Median Median Median Median Median Median P value P value P value P value P value P value P value P value (μm) (μm2) (#) (#) (#) (#/area) (#/area) (#/100μ) 2-way comparison Ex33/Ex33 and Δ18/Δ18 Ex33/Ex33 305 1,822 29 7 22 17.2 13.2 9.9 JT103 0.0001 4.1*10^-9 NS 0.0014 0.0236 1.4*10^-5 0.0024 0.0029 Δ18/Δ18 240 1,124 28 8 20 22.4 16.3 10.9 2-way comparison Ex33/Ex33 and Δ18/Δ18 Ex33/Ex33 520 4,110 48 6 41 10.8 9.4 8.7 JT104 1.2*10^-15 5.5*10^-17 5.5*10^-11 0.0139 4.8*10^-11 0.0017 NS 0.0063 Δ18/Δ18 274 1,846 25 5 20 12.9 9.5 8.5 2-way comparison Ex33/Ex33 and Δ18/Δ18 Ex33/Ex33 484 2,742 46 6 40 16.0 13.8 9.7 JT106 1.2*10^-15 5.5*10^-17 5.5*10^-11 0.0139 4.8*10^-11 0.0017 NS 0.0063 Δ18/Δ18 214 1,360 24 5 20 18.8 15.6 10.6 3-way comparison Ex33/Ex33 , Ex33/Δ18 and Δ18/Δ18 Ex33/Ex33 383 2,936 35 5 29 12.2 10.3 9.1 JT130 1.4*10^-5 4.8*10^-10 0.0022 0.0059 0.0005 9.6*10^-6 0.0033 0.0076 Δ18/Δ18 296 1,661 28 6 21 19.5 13.9 10.0 3-way comparison Ex33/Ex33 , Ex33/Δ18 and Δ18/Δ18 Ex33/Ex33 547 3,704 53 6 46 12.8 11.0 9.1 JT132 1.3*10^-7 1.8*10^-13 0.0005 NS 0.0002 6.5*10^-7 5.6*10^-5 0.0004 Δ18/Δ18 375 1,908 35 6 28 20.4 15.8 10.8 4-way comparison Ex33/w(z) , Ex33/Ex33 , Ex33/Δ18 and Δ18/Δ18 Ex33/Ex33 505 4,537 41 5 36 10.0 8.5 8.5 JT133 0.0312 2.0*10^-4 NS NS NS 0.0166 0.0476 NS Δ18/Δ18 397 3,047 41 4 36 12.4 11.0 9.2 2-way comparison Ex33/Ex33 and Δ18/Δ18 [microfluidic device KD020dissociation] Ex33/Ex33 487 3,061 49 5 43 17.0 14.8 9.9 2.1*10^-7 1.9*10^-9 0.0014 2.5*10^-9 5.0*10^-5 1.8*10^-7 0.0001 2.8*10^-6 PC021 Δ18/Δ18 376 1,820 39 7 31 22.2 18.1 11.0 KD020 [same as above] Ex33/Ex33 487 3,061 49 5 43 17.0 14.8 9.9 4.8*10^-6 2.5*10^-8 0.0038 0.0007 0.0012 1.2*10^-6 0.0005 5.3*10^-7 PC022 Δ18/Δ18 396 2,007 41 6 33 20.1 17.0 11.4 mutant arbors smallermutant arbors smaller high or low mutant arbors denser

a HO: Higher-Order; HO Branches = Total Branches - Primary Processes b Branch Density: # Branches/Territory Area [Branches/1000μm2] Most consistently affected parameters

Table 2. Pairwise comparison between control Ex33 and mutant Δ18 neurite arbors across multiple neuron-culture experiments

Figure 7. Quantification of the “bushy” phenotype in CASKΔ18 mutant neurons.

Figure 7. Quantification of the “bushy” phenotype in CASKΔ18 mutant neurons. Neurite- arbor morphology parameters, depicted as box-plot distributions, from three div larval CNS neuronal cultures using CASK Ex33 (n=106) homozygous control and CASKΔ18 (n=105) homozygous mutants, to assess if there is a phenotypic difference in their neuron-arbor.

Parameters shown are total neurite length (µm), territory area (µm2), higher-order branch number and branch density (branches/1000 µm2). The middle line within the box (arrowhead) represents the 50th percentile (median). Top and bottom of each box represent the 75th and 25th percentiles, respectively. The top and bottom whiskers represent the 90th and 10th percentiles, respectively.

The nonparametric Mann–Whitney rank sum test was used to compare measures of neuron size and shape between genotypes. Statistically significant differences are indicated with asterisks: *, p < 0.05; **, p < 0.005; ***, p < 0.0005; ****, p<0.00005; ns, not significant. Total neurite length, territory area and higher-order branches are dramatically decreased in CASK-mutant CNS when compared to control animals (Ex33). In addition, CASKΔ18 mutants display increased branch density.

73 also affected, where seven out of eight experiments showed a statistically significant increase in this ratio. Quantification of the number of total branches showed either significant decrease difference in six experiments or no change in the rest of the cases. There were not remarkable differences between the primary processes across the eight replicates. The lack of significant differences in the number of primary processes suggests that the initiation of neurite outgrowth by CASK mutant neuronal somata does not vary from that in control neurons. Another ratio analyzed, the higher-order branch density also called branch-point density (number of higher- order branches per 1000 µm2), was affected as well. In all the experiments conducted, CASK- mutant neurons display increased branch-point density, with six of eight reaching statistical significance. The elevation in the branch-point density and branch:length ratio in CASK-mutant neurons is consistent with the increased branch density. The observations based on the three ratios analyzed (branch density, higher order-branch density, and branch:length ratio) indicated that even though these numbers did not reach statistical significance, they were always higher for the mutants when compared to controls. Together, these data indicate the striking difference in the neurite-arbor size and shape of cultured neurons lacking homozygous CASK-β when compared to homozygous CASK Ex33 controls. My impression is that most of the CNS neurons were affected and no subpopulation was spared.

II.C.2. Linear mixed modeling to compare results across multiple experiments

From the viewpoint of the statistician, each neuron-culture experiment represents a single sample (D. Billheimer, personal communication). Therefore, I obtained a consultation to learn how to compare all my mutant versus control data across the eight replicate experiments.

On the other hand, the eight neuronal culture experiments described here differ in several aspects, including the date when the culture was set-up, the brand and lot of FBS, the brand and type of gridded dishes and even the trituration mechanism, i.e., manual pipetting vs microfluidic device for automated dissociation (Jiang et al., 2015; see Appendix B). As a result, as reported in

Kraft et al. (1998), there was a considerable variation in the absolute values for each morphometric parameter across all these experiments. For instance, in four different experiments

I found that the median neurite length for CASK Ex33 neurons was 305, 520, 484 and 383 µm.

Nevertheless, the directionally of differences and statistical significance were consistent across all the replicates.

By using the linear mixed-effects model I was able to combine results across multiple replicate experiments, removing inter-experiment variability from the comparison. By comparing

Ex33/Ex33 and ∆18/∆18 significant differences across eight different experiments were found for neurite length (p = 0.001), territory neuron area (p = 0.0004), higher-order branches (p =

0.007) and branch density (p = 0.001) (Table 3). In brief, all comparisons were found to have substantial differences between genotypes, which means that there is striking biological variation with the bushy phenotype. In addition, the 95% confidence intervals for mean differences indicate the likely magnitude of differences between Ex33/Ex33 and ∆18/∆18 (Table 3). Thus, after removing the effects of inter-experimental variation from the analysis, we see that genotype-genotype differences are repeatable with high reliability.

Neurite Length (μm) Territory Area (μm2) Higher-Order Branches (#) BranchDensity (#/area) Parameter Ex33/Ex33 Δ18/Δ18 Ex33/Ex33 Δ18/Δ18 Ex33/Ex33 Δ18/Δ18 Ex33/Ex33 Δ18/Δ18 Mean of the means across 463 317 3273 1850 37 26 13.7 18.2 replicates Difference between 146 1423 11 4.5 genotypes mean Standard Error of the mean difference 27.4 173.2 3 0.8 95% Confidence Interval (91 - 201) (1077 - 1070) (5 - 17) (2.9 - 6.1) for difference in means p-value 0.0011 0.0004 0.0066 0.001

Table 3. Ex33 and Δ18 neurite arbors comparisons across eight replicate experiments

II.C.3. CASK “bushy” phenotype is semi-dominant The bushy phenotype was revealed in CASK-mutant neurons from the CNS of homozygous larvae, a finding that does not explain if this phenotype is recessive, dominant or semidominant. Human CASK mutations causing MICPCH are semidominant, bringing up the question of whether there is a genetic parallel between fly and human CASK phenotypes.

To determine the genetic nature of the bushy neurite-arbor phenotype I did a three-way comparison among homozygous control (Ex33/Ex33), homozygous mutant (Δ18/Δ18), and heterozygous (Ex33/Δ18) larval CNS neurons. If the bushy phenotype is recessive, I would expect no differences between Ex33/Ex33 and Ex33/Δ18, because in that condition one copy of the WT gene would not be sufficient to confer normal phenotype.

Neurons cultured from heterozygous larval CNS had significant reductions in neurite length, territory area and higher-order branch number when compared to homozygous control neurons (Figure 8). In contrast, branch density displayed a significant increase in Ex33/Δ18 neurons. The heterozygous neurons had an intermediate bushy phenotype (less severe phenotype than the homozygous mutant), with significant differences from both homozygous control

(Ex33/Ex33) and homozygous mutant (Δ18/Δ18). Therefore, indicating that Drosophila CASK

Δ18 mutation is semi-dominant for this phenotype.

II.C.4. Two chromosomal deficiencies of CASK uncover the bushy phenotype of CASK Δ18

To determine whether the bushy phenotype maps to the CASK region, two chromosomal deletions, Df(3R)X307 and Df(3R)X313, were used (see Chapter I and Figure 4C). The molecular mapping of these two deletions and their respective endpoints, have been approximated by

Southern blot analysis, where both were determined to have partial deletions of CASK (Martin &

Ollo, 1996). As shown in Figure 4C, Df(3R)X307, the “small” deletion, removes most of the

CASK gene, whereas Df(3R)X313, the “large” deficiency, not only deletes most of the CASK locus but also removes some additional proximal sequences. Each individual deficiency is

Figure 8. The CASK-mutant bushy phenotype is semidominant.

Figure 8. The CASK-mutant bushy phenotype is semidominant. Neurite-arbor morphology parameters, depicted as box-plot distributions, from three div larval CNS neuronal cultures.

CASK Ex33 homozygous control (n=103), CASK heterozygous (Δ18/Ex33) (n=102), and

CASKΔ18 homozygous mutants (n=102) were used to determine whether in-vitro CASK- associated phenotype is recessive, dominant or semi-dominant. Parameters shown are total neurite length (µm), territory area (µm2), higher-order branch number and branch density

(branches/1000 µm2). The middle line within the box (arrowhead) represents the 50th percentile

(median). Top and bottom of each box represent the 75th and 25th percentiles, respectively. The top and bottom whiskers represent the 90th and 10th percentiles, respectively. The nonparametric

Mann–Whitney rank sum test was used to compare measures of neuron size and shape between genotypes. Statistically significant differences are indicated with asterisks: *, p < 0.05; **, p <

0.005; ***, p < 0.0005; ****, p<0.00005; ns, not significant. Larval CNS neurons from CASK heterozygotes (Δ18/Ex33) display an intermediate neurite-arbor size and shape between control

(Ex33/Ex33) and mutant (∆18/∆18) animals, indicating that bushy is semi-dominant.

80 embryonic lethal, but that lethality has never been mapped. However, these two mutations complement and are viable in the trans-heterozygote. Based on the molecular characteristics of these deficiencies, I predicted that each deficiency chromosome should uncover the bushy neurite-arbor phenotype displayed in CASK∆18 homozygous mutants. For deficiency mapping, three genotypes were compared for each of the two deletions:

w/w; +; Ex33/Df(3R)X# (hemizygous control)

w/w; +; ∆18/∆18 (homozygous mutant)

w/w; +; ∆18/Df(3R)X# (hemizygous mutant)

As expected, neurite arbors from homozygous CASK∆18-mutant larval CNS were smaller than both hemizygous controls, Ex33/Df(3R)X307 and Ex33/Df(3R)X313 (Figures 9D and 10D, respectively), with reduced neurite length, territory area and higher-order branch number but increased branch density.

The data also showed that neurons from each hemizygous mutant were at least as severely affected as neurons from the homozygous ∆18 mutant. Neurite length and territory area were significantly decreased when compare each hemizygous mutant to the corresponding hemizygous control (Figure 9D, 10D). In turn, comparison between each hemizygous mutant and

∆18/∆18, displayed a dramatic reduction in both length and territory area in the hemizygous mutant. Higher-order branch number decreased, while branch density increased, when comparing each hemizygous mutant to their hemizygous control. Higher-order branches were not significantly reduced when compared to ∆18/Df(3R)X307 to ∆18 homozygous. In contrast, a significant decrease in higher-order branches was observed when comparing ∆18/Df(3R)X313 to for ∆18/∆18. For branch density, no statistically significant difference was observed for each hemizygous mutant, when comparing to CASK ∆18 homozygous.

Figure 9. A chromosomal deficiency, Df(3R)X307, uncovers the bushy phenotype of CASK Δ18.

Figure 9. A chromosomal deficiency, Df(3R)X307, uncovers the bushy phenotype of CASK

Δ18. A-C. Representative images of cultured neurons derived from the entire larval CNS after three div. (60X magnification). Selected images from approximately the median of length, territory area and branch densities A. Typical Ex33/Df(3R)X307 neurons covering a large territory area with fewer higher-order branches, B. Reduced neurite-arbor size is accompanied by many higher-order branches in Δ18/Δ18 mutants. C. Δ18/ Df(3R)X307 neurite arbors are smaller in size, with an increased number of higher-order branches. Scale bar A-F = 10 μm. D. Neurite- arbor morphology parameters from three div larval CNS neuronal cultures, are depicted as box- plot distributions. Ex33/Df(3R)X307 hemizygous control (n=104), Δ18 homozygous mutant

(n=106), and Δ18/ Df(3R)X307 hemizygous mutant (n=106), were used to determine whether or not this chromosomal deletion maps to CASK. Parameters shown are total neurite length, territory area, higher-order branches and branch density. Neurite-arbor morphology parameters are depicted as box-plot distributions. The middle line within the box (arrowhead) represents the

50th percentile (median). Top and bottom lines of each box represent the 75th and 25th percentiles, respectively. The top and bottom whiskers represent the 90th and 10th percentiles, respectively.

The nonparametric Mann–Whitney rank sum test was used to compare measures of neuron size and shape between genotypes. Statistically significant differences are indicated with asterisks: *, p < 0.05; **, p < 0.005; ***, p < 0.0005; ****, p<0.00005; ns, not significant. In all the four parameters analyzed, the “bushy” neurite-arbor phenotype is displayed when comparing to control. Total neurite length, territory area and higher-order branches are decreased, whereas branch density is increased in homozygous mutants when compared to hemizygous control

Ex33/Df(3R)X307. The chromosomal deficiency of CASK, Df(3R)X307, worsen the bushy phenotype of ∆18 homozygous mutants.

Figure 10. A chromosomal deficiency, Df(3R)X313, uncovers the bushy phenotype of CASK Δ18.

Figure 10. A chromosomal deficiency, Df(3R)X313, uncovers the bushy phenotype of CASK

Δ18. A-C. Representative images of cultured neurons derived from the entire larval CNS after three div. (60X magnification). Selected images from approximately the median of length, territory area and branch densities A. Typical Ex33/Df(3R)X313 neurons covering a large territory area with fewer higher-order branches, B. Reduced neurite-arbor size is accompanied by many higher-order branches in Δ18/Δ18 mutants. C. Δ18/ Df(3R)X313 neurite arbors are smaller in size, with an increased number of higher-order branches. Scale bar A-F = 10 μm. D Neurite- arbor morphology parameters from three div larval CNS neuronal cultures, are depicted as box- plot distributions. Ex33/Df(3R)X313 hemizygous control (n=106), Δ18 homozygous mutant

(n=105), and Δ18/Df(3R)X313 hemizygous mutant (n=106), were used to determine if this chromosomal deletion maps to CASK. Parameters shown are total neurite length, territory area, higher-order branches and branch density. The middle line within the box (arrowhead) represents the 50th percentile (median). Top and bottom lines of each box represent the 75th and 25th percentiles, respectively. The top and bottom whiskers represent the 90th and 10th percentiles, respectively. The nonparametric Mann–Whitney rank sum test was used to compare measures of neuron size and shape between genotypes. Statistically significant differences are indicated with asterisks: *, p < 0.05; **, p < 0.005; ***, p < 0.0005; ****, p<0.00005; ns, not significant. Δ18 homozygous mutants shown the “bushy” neurite-arbor phenotype in all the four parameters analyzed when comparing to control animals Ex33/Df(3R)X313. The chromosomal deficiency of

CASK, Df(3R)X313, uncovers and exacerbates the bushy phenotype of CASK. This is indicated by the significant decrease of total neurite length, territory area and higher-order branches.

In summary, the distinctive features of the bushy neurite-arbor phenotype, with reduced length, territory area and higher-order branch number, accompanied by increased branch density, are uncovered by each of the two CASK chromosomal deficiencies. Therefore, this novel phenotype of cultured larval CNS neurons maps to the CASK region.

II.C.5. Transgenic CASK+ improves the bushy phenotype

To confirm that the bushy phenotype maps to the CASK gene, I determined whether this phenotype could be rescued by introducing a transgene with a WT copy of CASK. Because the bushy phenotype is semi-dominant, I hypothesized that one transgenic copy of CASK+ would partially restore CASK function.

By putting elav-GAL4C155 and UAS-CASK+ together, i.e. in the same animal, I was able to drive the expression of CASK wild-type in all neurons.

I first tested whether either of the two transgenes used, to drive wild-type CASK expression in neurons would impact the bushy phenotype of CASK ∆18 homozygotes, I performed a 3-way comparison. The following genotypes were compared:

1) w/w; +; ∆18/∆18 (homozygous mutant)

2) elav-Gal4C155 w/+; +; ∆18/∆18 (driver transgene in a mutant background)

3) w/+; UAS-CASK+/+; ∆18/∆18 (target transgene in a mutant background)

Neither of the two transgenes improved the bushy neurite-arbor phenotype of CASK ∆18 homozygous mutants (Figure 11A). In fact, both transgenes caused modest reductions in neurite- arbor size beyond that of homozygous CASK-mutant neurons. The total length was reduced by both UAS-CASK+ and elav-GAL4C155; territory area by UAS-CASK+ only. Neither transgene affected branch density (one of the most characteristic features of the bushy phenotype) or

Figure 11. CASK+ improves the bushy phenotype.

Figure 11. CASK+ improves the bushy phenotype. Neurite-arbor morphology parameters, depicted as box-plot distributions, from three div larval CNS neuronal cultures using transgenes to test the hypothesis that the bushy phenotype is caused by loss of CASK function. Parameters shown are total neurite length, territory area, higher-order branches and branch density. The middle line within the box (arrowhead) represents the 50th percentile (median). Top and bottom lines of each box represent the 75th and 25th percentiles, respectively. The top and bottom whiskers represent the 90th and 10th percentiles, respectively. The nonparametric Mann–Whitney rank sum test was used to compare measures of neuron size and shape between genotypes.

Statistically significant differences are indicated with asterisks: *, p < 0.05; **, p < 0.005; ***, p

< 0.0005; ****, p<0.00005; ns, not significant. A. Transgene controls do not impact the bushy phenotype. Results from one of three independent replicates, each with neurons from CNS with three genotypes. Mutant: w/w; +; Δ18/Δ18 (n=104) (purple); UAS-CASK+ transgene in the

CASK-mutant background: w/+; UAS-CASK+/+; ∆18/∆18 (n=104) (gray); and the pan-neuronal gene elav, in the CASK-mutant background: elav-Gal4C155 w/+; +; ∆18/∆18 (n=105) (light- blue). Neither transgene affected branch density, the most distinctive feature of the busy phenotype, or higher-order branches. Both transgenes cause modest reductions in neurite-arbor size (total length reduced by both UAS-CASK+ and elav-GAL4C155; territory area by UAS-CASK+ only). B. To assess transgenic rescue, three independent replicates from three div larval CNS neuronal culture were analyzed. The three genotypes for comparison included: Mutant: w/w; +;

Δ18/Δ18 (n=105) (magenta); CASK+ transgene in the CASK-mutant background: elav-Gal4C155, w/+; UAS-CASK+/+; Δ18/Δ18 (n=104) (blue); and control: w/w; +; Ex33/Ex33 (n=105) (cyan).

The addition of one copy of transgenic CASK+ partially rescues all the four neurite-arbor parameters analyzed.

88 higher-order branch number. If these effects are additive, it could make demonstration of phenotypic rescue more difficult, at least with respect to arbor size.

To test for transgenic rescue of the bushy phenotype, I prepared 3 independent replicates on the same day in order to minimize biological and technical variables. Each replicate had neuronal cultures from larval CNS of three genotypes:

1) w/w; +; Ex33/Ex33 (homozygous control)

2) w/w; +; ∆18/∆18 (homozygous mutant)

3) elav-Gal4C155 w/+; UAS-CASK+/+; ∆18/∆18

The latter drives the expression of CASK in all differentiating neurons in CASK-homozygous mutants.

Based on the semi-dominance of the CASK-bushy phenotype, I predicted that one copy of

CASK+ would partially rescue the defect. My prediction was that more than 50% of wild-type

CASK function would be required to achieve full rescue of the neurite-arbor phenotype.

Figure 11B shows that the addition of a single transgenic copy of the wild-type CASK+ cDNA in a ∆18 homozygous background resulted in a highly significant increase in neurite length, territory area, higher-order branch number, and a decrease in the branch density. In other words, the addition of a single copy of a wild-type CASK+ transgene significantly improved the bushy phenotype. It yielded partial rescue of the bushy phenotype (i.e., intermediate between the phenotypes of mutant and control homozygotes). This suggests that in order to fully rescue this phenotype, two copies of wild-type CASK would be required.

II. D. Discussion

II.D.1. Key findings and their significance to human CASK phenotypes

In this study, I demonstrated that CASK-mutant neurite arbors have reduced size of a particular type, characterized by a reduction in neurite length, territory area and higher-order branches. In addition, the number of branches per unit area was increased in the mutants. The highly significant increase in branch density in these small neurite arbors gives rise to the

“bushy” appearance of the phenotype. Hence, CASK regulates neurite-arbor size and shape.

I also found that the small-arbor bushy phenotype is semidominant. This implies that

50% of wild-type CASK function is not sufficient for normal neurite arbors. The dosage- sensitive requirement for CASK in the CNS is also evidenced in humans, where differences between heterozygous females and hemizygous males with CASK mutations have been reported

(Moog et al., 2011 & 2015). A hemizygous male with a mutation in the CASK gene on his sole

X-chromosome, will have a more severe phenotype than a heterozygous female, who has one

"normal" X-chromosome and one X-chromosome with the mutated CASK gene. MICPCH- associated ID phenotype in humans is a clear example where there is a dosage-sensitive requirement for CASK in the CNS. Thus, the semidominance nature of CASK is not only true in

Drosophila, but also in humans.

Haploinsufficiency, when one copy of WT gene is not enough, is a common form of dosage sensitivity represented in human developmental disorders. For example, heterozygous

LOF mutations of NIPBL cause Cornelia de Lange syndrome 1 (Pehlivan et al., 2014), which includes ID, impaired growth, characteristic facial malformations, and limb abnormalities (Boyle et al., 2015). Similarly, heterozygous LOF mutations in the gene encoding the CREB-binding protein (CREBBP, also known as “CBP”) were found to be responsible for causing Rubinstein-

Taybi syndrome, characterized by ID, growth retardation, and dysmorphic features (Roelfsema et al., 2005). In addition, heterozygous LOF mutations of CTNNB1 cause autosomal dominant ID

19 (aka MRD19) with microcephaly (Tucci et al., 2014).

Examination of brain by MRI postmortem histology revealed abnormal dendritic arborization (Kulkarni and Firestein, 2012) and axonal projection defects (Edwards et al., 2014).

It is plausible that human CASK mutations also disrupt brain neuronal morphogenesis, causing reduced dendric and/or axonal projections. Although, there is very little evidence about the causative relationship between neuronal arbor size and brain size, it is plausible that abnormalities in brain size may be a consequence of altered neuron morphology with reduced neurite extension and arborization size. As mentioned in section II.A., a significant reduction in dendritic branching in pyramidal neurons has been found in postmortem brain samples from individuals with Rett syndrome. This diminution was accompanied by global reduction in brain volume (Armstrong et al. 1995, Baj et al., 2014). Thus, cellular defect, such as the small-neurite arbor with altered neuronal shape, as seen in CASK mutants, could underlie both microcephaly and cognitive deficits due to altered neuronal circuitry in patients with MICPCH.

II.D.2. Potential mechanisms underlying the CASK-mutant neurite arbor phenotype

Lewis (2015) showed that Pak mutations cause defects in the cytoskeleton, which is essential for neurite morphology. PAK is a member of a small, highly conserved p21-activated serine/threonine kinase family, and it is a key regulator of the actin cytoskeleton (Kreis &

Barnier, 2009). Changes in the organization of F-actin were accompanied with reduced filopodia in neurons cultured for three div from Pak mutants. In addition, it was found that there was a reduction in F-actin protrusions and spread microtubules in the growth-cones. Mutations in

PAK3, the human ortholog of Pak, result in ID and microcephaly (Peippo et al., 2007).

Drosophila Pak-mutant neurons have reduced total neurite length and higher-order branches in vitro (Lewis, 2015). In addition, Drosophila Pak mutants have reduced terminal arborization of photoreceptor in the optic lobe. Pak is specifically required for cell axon guidance and targeting but not for cell differentiation and axon outgrowth in Drosophila photoreceptor growth (Hing et al., 1999). The fact that these defects occurs later in time, suggests that the initiation of neurite outgrowth proceeds normally.

In CASK, the protein 4.1-binding domain (a lysine-rich short sequence located between the GUK and SH3 domains (Figure 3) is present (Biederer et al., 2001). The function of protein

4.1 is to interact with F-actin molecules of the cytoskeleton and promote the polymerization.

Thereby, it would be interesting to examine the distribution of F-actin in CASK mutant neurons.

The potential structural or regulatory relationship of CASK to the actin cytoskeleton it is suggested by a link between CASK and FRMD7. In humans, CASK recruits FRMD7 to the plasma membrane, FRMD7, in turn, enhances the neurite outgrowth by modulating the actin cytoskeleton (Watkins et al., 2013). FRMD7 encoded FERM, where Coracle is the Drosophila ortholog. Coracle is an actin-binding protein and a member of the Protein 4.1 superfamily. In

Drosophila, Coracle is expressed in septate junctions, dorsal ectoderm (embryo) and brain

(adult). LOF mutations result in failure of dorsal closure, late embryonic lethality and salivary gland defects (Lamb et al., 1998). I suggest that CASK binds Coracle at the 4.1 domain. Hence, I hypothesize that CASK may contribute to the cytoskeletal organization of neurons.

To assess functional relationships between CASK and Coracle, it would be worth exploring the neuronal arbor in wandering third instar larvae from Coracle LOF homozygous mutants. Non-additive gene action results from a genetic interaction between genes and is widely

92 accepted as a sign of absence of a functional relationship between the mutated genes under study

(Pérez-Pérez et al., 2009). If LOF mutations in Coracle do not reveal a neuronal-arbor phenotype, and the bushy neurite-arbor phenotype gets better when a double-mutant fly for

Coracle (Coracle/+) and CASK (∆18/+) is generated, then it would indicate that there is a genetic interaction, with suppressor effects, between these genes. It also may be the case that

Coracle LOF homozygous mutants have for instance, a recessive neurite-arbor phenotype. Then, if the bushy phenotype exhibited by CASK mutants gets worse when a mutation in Coracle is also present, it would be a sign of non-additive genetic interaction with synergistic enhancement.

A synergistic enhancement occurs when the contribution of two mutations to the phenotype of a double mutant exceeds the expectations from the additive effects of the individual mutations

(Pérez-Pérez et al., 2009). On the other hand, if mutations in both CASK and coracle are present in the same animal and the neurite-arbor phenotype does not change it would be an indication of no genetic interaction.

II.D.3. Contributions of primary culture for understanding neurite-arbor morphogenesis

To understand the neurobiology of brain development and behavior, in vivo analyses have been complemented with studies of neuronal morphogenesis in vitro. For example, defects in neuron morphogenesis associated with drugs of abuse reported in vitro (Nassogne et al., 1995) complement with in vivo analyses of brain development and behavior (Robinson & Kolb, 2004).

Specifically, embryonic mouse coculture of neuronal and glial cells have been used to test for toxicity of cocaine concentrations. Cocaine was added to this neurons-glial coculture after two div. After four div shortening of the neurites was clearly detectable A dramatic reduction of both number and length of neurites, and extensive neuronal death were found in embryonic neuronal

93 cells selectively affected by cocaine in vitro (Nassogne et al., 1995). In turn, Robinson & Kolb

(2004), found that repeated in vivo exposure to drugs of abuse, such as cocaine, produces long- lasting changes in the structure of dendrites and dendritic spines, e.g. spine density was increased in the hippocampus with the use of amphetamine.

Cell lines are often used in place of primary cells to study biological processes. By using primary cultured cells, one avoids the disadvantages of immortalized cell lines, which may not adequately represent primary neurons and may alter the phenotype, yielding different results

(Gordon et al., 2013).

Primary cultures of mammalian cells are widely used to reveal cellular mechanisms in neurobiology (Kaech & Banker, 2006; Ray et al., 2009). Moreover, Drosophila primary dissociated cultures of developing CNS neurons provide a powerful experimental system with which to investigate the effects of genetic or chemical manipulations on neuronal differentiation, morphology, and function (Kraft et al., 2006; Smrt et al., 2015). A great advantage of this approach is that it allows the study of primary neural cells derived from any available genotype

(Egger et al. 2013).

Using our standard protocol, neurons in primary culture are cultured at low-density to reveal cell-autonomous characteristics of neurons under physiological and pharmacological conditions. One of the advantages of this system, for example, is that in vitro neurite outgrowth does not reflect the influence of non-neuronal cells. This is true because the protocol used yields pure culture of neurons without glial cells around them. Therefore, the bushy phenotype observed in CASK-mutants is determined primarily by intrinsic properties of the neuron and through soluble secreted (trophic) factors. Moreover, human microcephaly caused by CASK

94 mutations may exhibit a neurite-arbor bushy phenotype, which ultimately motivated me to further investigate brain size in the Drosophila system.

Chapter III:CNS-size abnormalities in Drosophila CASK loss-of-function mutants

III.A. Rationale

Neurodevelopmental disorders associated with changes in head and brain size in humans have been the focus of attention for many decades. A clinical finding that occurs in many neurodevelopmental disorders is microcephaly (Mochida & Walsh, 2001). Microcephaly is a phenotype that results from deficient brain growth either before birth or in the early years of postnatal life (Woods, 2004; Abuelo, 2007). Microcephaly results from reduced volume of cerebral cortex, cerebellar hypoplasia, brain stem and cerebellar malformations, cortical folding abnormalities and corpus callosum agenesis (Soto-Ares et al., 2003; Woods, 2004; Decobert et al., 2005). Thus, any condition that affects important processes of brain growth can severely affect head size. For instance, cell proliferation, cell differentiation, cell death, cell cycle and

DNA repair are, among others, processes that affect pathways involved in neurogenesis

(Barkovich et al., 2012). For example, humans with one of the autosomal recessive primary microcephalies (MCPH) usually have a small brain associated with mild to moderate ID. The most common cause of MCPH is homozygous mutation of abnormal spindle-like microcephaly- associated gene, ASPM. Mutations in this gene affects progenitor proliferative capacity (Bond et al., 2003; Desir et al., 2008). Abnormal spindle (asp), is the Drosophila melanogaster orthologue of the human ASPM gene (Saunders et al., 1997). Mutations in asp lead to defects in mitosis at a variety of developmental stages as well as in male meiosis. The Drosophila protein asp is essential for normal mitotic spindle function in neuroblasts, mediates interactions between microtubule minus ends and the centrosome, and is involved in microtubule organization during spindle formation and cytokinesis (Wakefield et al., 2001).

While dissecting wandering third instar larva prior to dissociation (Chapter II), I observed that CASK-mutant larvae had smaller CNS than control larvae (Figure 12A). This observation was reminiscent of the small head phenotype in humans with CASK mutations (see § I.C.). This was my motivation to determine whether CASK controls brain size in Drosophila.

Mass is not easy to assess for the tiny fruit fly CNS. For instance, the brain of an adult human weighs about 1,500 g, a mouse brain has an approximately mass of 0.416 g, while the fruit fly brain (Drosophila melanogaster) weigh less than 0.2 mg (Herculano-Houzel 2009).

Assessing the CNS mass of Drosophila is not straightforward because for weighing, the tissue should be moist but not wet, and the tiny brains have a relatively large surface area, hence desiccation is a serious problem. Instead, I determined the brain volume. My approach consisted of estimating brain and head volumes from serial sections of osmium-stained, plastic-embedded heads at late stage of adult development. I hypothesize that mutations in CASK cause a reduced brain volume.

Initially, I conducted a pilot study focused on the brain volume quantification in males and females to compare CASK-mutants and controls. In a second study, with better design I measured brain and head size, as well as body size in males only, none of these parameters has yet been investigated in CASK-mutant flies.

Developmental mechanisms that underlie brain morphogenesis and head formation differ between invertebrates and mammals. The adult fly arises during metamorphosis, a process during which most larval cells die (Tissot & Stocker, 2000). The new adult structures are derived from imaginal cells which have been dividing during larval life, but do not contribute to the functional anatomy of the larva (Tissot & Stocker, 2000). For instance, tissues that make up the brain are derived from neuroblasts (Urbach et al., 2003), while most of the head capsule of the

Figure 12. Estimated brain size differences in CASK mutants.

Figure 12. Estimated brain size differences in CASK mutants. A. CNS-size abnormalities in third-instar CASK mutant larvae. Photomicrograph of stage- and sex-matched female homozygous Ex33 and ∆18 CNS. L= optic lobe; VG= ventral ganglia. Scale bar = 500 µm.

CASK-mutant larvae have much smaller CNS than control animals. B. Dorsal view of pupa, viewed through a stereomicroscope, showing the three distinct body parts of the adult fly: head, thorax and abdomen. Body length was measured from these photographs (yellow dotted line) C.

Microscope slide showing consecutive serial sections from one head (Photo credit: Rachel Bear).

D. Photomicrograph of histological section through pharate-adult head of a control animal

(Ex33/Ex33); frontal plane of section, dorsal is to the top. Aqua and yellow outlines, obtained from Olympus cellSens software, overlay brain and head, respectively. 20X magnification. Scale bar = 100 µm. E. Side-by-side comparison of estimated brain size obtained from serial sections of osmium-stained, plastic-embedded pharate-adult heads using Ex33 homozygous control and

Δ18 homozygous mutants, to test the hypothesis that CASK mutants have a smaller brain than

CASK-control flies. Each circle represents data from a single fly: homozygous controls

(Ex33/Ex33, cyan circles) and homozygous CASK mutants (∆18/∆18, magenta circles). Ex33 control animals had significantly larger brains than CASK-mutant flies (males p < 0.0005, females p < 0.01). Significance levels, based on nonparametric Mann–Whitney rank sum test for pairwise comparison between genotypes, indicated by asterisks: *, p < 0.05; **, p < 0.005; ***, p < 0.0005; ns, not significant.

99 adult fly originates from a pair of composite discs called the eye–antennal discs (Dominguez &

Casares, 2005). Thus, brain and head in Drosophila derive from two anatomically separated structures. The Drosophila CNS is bilaterally symmetrical and includes a bi-lobed brain and a ventral nerve cord. The brain, in turn, is composed of the supraesophageal zone (SPZ) which includes the brain and optic lobes, and the subesophageal zone (SEZ) (Truman 1990; Ito et al.,

2014). In mammals, brain and head structures develop in tight physical association synchrony

(Richtsmeier & Flaherty, 2013). The brain, which emerges comparatively early in development, originates from ectodermal tissue, the neural tube. The head, a compound structure, consists of a multitude of cell types, tissues, organs and spaces that originate separately but develop in coordination (Stiles & Jernigan, 2010). Thus, the head structure that is surrounding the developing brain grows in response to mechanical pressure, indicating that in humans, brain size affects skull size directly (Richtsmeier & Flaherty, 2013). In contrast, in fruit flies, head size is not dictated by brain size.

Individuals with LOF CASK mutations also reported to exhibit short stature (Moog et al.,

2015). In fact, it is known that among the disorders causing microcephaly, a common co- occurring phenotype is short stature. For instance, LOF mutations in the Rotatin gene (RTTN), which encodes a centrosome-associated protein, were found in patients with congenital microcephaly, ID, seizures and short stature (Shamseldin et al., 2015). As in humans, body size in Drosophila, particularly pupal case length, is also among the most heritable traits reported in this species (Visccher et al., 2010; Reeves & Tautz, 2017). In Drosophila, the final size of the adult fly is tightly correlated with the pupal size, which varies between 2.8 and 3.9 mm in length among natural (i.e., wild-type) individuals (Reeves & Tautz, 2017). In Drosophila, a developing larva feeds and grows until it reaches a critical size at which the larva stops feeding, marking the

100 end of body growth. It means that at later stages, e.g., pupal case formation, maximum size has already been reached (Shingleton et al., 2008). Thus, I measured the length of the pupal case as a proxy for adult body size (Figure 2B).

III.B. Materials and Methods

III.B.1. Sample collection and photography

The Drosophila CASK alleles used in this study were originally created via excision of the P-element EY07081 (see § II.B.1). Specifically, CASK ∆18 imprecise-excision deletion, which eliminates production of full-length CASK protein, and CASK Ex33, the precise-excision control, were used. Rearing and culture conditions were as described in Chapter II (§ II.B.1.).

Pupae of both homozygous genotypes, CASK Ex33/Ex33 and CASK ∆18/∆18, were collected as late pharate adults. Technically, the pharate-adult developmental stage begins when the transparent pupal cuticle separates from the underlying adult epidermis, resulting in a pharate or covered adult (Bainbridges & Bownes, 1981). However, our lab and many others restrict the term to the latest stages of metamorphosis, shortly before adult emergence (Restifo and White,

1991). Pupae were gently removed from the walls of their culture vials, using a moistened paintbrush to dissolve the glue, being careful to avoid breaking the pupal case. They were separated by sex and maintained at room temperature in a humid chamber until photography and dissection.

III.B.2. Photography and estimation of body length

Pupae were individually positioned dorsal side up on a glass microscope slide. A 2-mm stage micrometer with 0.01 mm subdivisions (Ward's Science, Rochester, NY) was used to

101 determine the scale. Digital images were captured using a Canon Rebel T3i EOS 600D digital

SLR camera, equipped with a Canon EF 100mm f/2.8 Macro USM lens (Canon USA, Inc.,

Melville, NY), and mounted on a tripod to standardize the distance between the camera and the bench below. For each experiment, all images were acquired on the same day. A unique identifying code was assigned to each captured image and, in the follow-up experiment, stayed with the sample throughout. The longitudinal axis of the pupal case in the digital images was measured along the midline between the anterior and posterior spiracles using the ImageJ tool

‘Measuring Distance Between Points’ (Schneider et al., 2012) and the scale micrometer as a calibration tool (Figure 12B).

III.B.3. Histology

III.B.3.a. Dissection. After photography, each pupa was transferred to the well of a glass dissecting dish containing PBS, pH 7.4. Microdissection was done under a stereomicroscope using fine forceps, Dumont #5, Dumoxel with Biologie tips (Fine Science Tools, Inc.,

Vancouver, Canada). While grasping the posterior end of the pupal case with one pair of forceps,

I used a second pair to gently open the operculum, leaving most of the thorax and abdomen inside the pupal case. Then I removed the pupal cuticle surrounding the head. Next, using one tip of the forceps as a blade, I severed the exposed head from the body by slicing through the neck.

Similarly, I removed the proboscis to open the ventral head capsule. For the follow-up experiment, exceptional care was taken to avoid trauma to the head capsule so that it could be used for head-volume estimation.

102

III.B.3.b. Fixation and embedding. The protocol was modified from one previously used for estimating brain volumes in larger insects (Riveros & Gronenberg, 2010) and carried out with guidance from Dr. Gronenberg. Unless otherwise specified, all steps were carried out at room temperature. The heads were fixed overnight in freshly made 4% formaldehyde (Electron

Microscopy Science; Hatfield, PA, USA) in PBS, pH 7.4, at 4°C on an orbital shaker. In the pilot study, groups of two to three heads of the same sex underwent fixation and subsequent processing together, which prevented them from being identified as individuals. In the better- designed follow-up experiment, each head was processed individually. After fixation, head samples were transferred in groups (pilot study) or individually (follow-up experiment) to microcentrifuge tubes and washed twice in PBS, pH 7.4, for 1 hour each at room temperature, and in distilled water for an additional hour. The washes and all subsequent steps were done on a rotator.

To stain the tissue, the heads were incubated in osmium tetroxide solution (Electron

Microscopy Science; Hatfield, PA), 1% in H2O, for 1 or 2 hours, pilot and follow-up experiment respectively, at 4°C and for an additional hour at room temperature. Heads were then rinsed in water for 3 hours. A one-step dehydration method, based on Edwards et al., 1979, was performed. Samples were placed in a mix of 2,2-dimethoxypropane (Electron Microscopy

Science; Hatfield, PA) acidified with 0.03% hydrochloric acid, for 15 min. After dehydration, heads were first immersed in 100% acetone for 20 min, then transferred to 30% acetone/ 70%

Spurr's low-viscosity embedding medium (Electron Microscopy Science; Hatfield, PA) for 6 hours, and finally transferred into 100% Spurr’s overnight. The specimens were mounted in groups of two or three (pilot study) or individually (follow-up experiment) in BEEM™ (VWR)

103 embedding capsules filled with 100% Spurr’s, and oriented to allow sectioning in the frontal or horizontal plane. Polymerization was induced by overnight incubation at 70°C.

III.B.3.c. Serial sectioning. Each polymerized block was secured in a specimen holder and trimmed into a rectangle shape around the head using a disposable single-edge razor blade.

Consecutive serial sections were cut block on a sliding microtome (Reichert, Austria) using a non-disposable stainless-steel knife, at 12-µm thickness. Sections were permanently attached to the microscope slide by using a slide-warming tray at 60°C. In the pilot study, there were two to three head samples per slide, whereas in the follow-up experiment, there was one head per slide

(Figure 12C). The sections were mounted using Cytoseal (Thermo Scientific™, Waltham, MA) and coverslipped. These are permanent preparations.

III.B.4. Image acquisition and analysis

Digital images of each histological section were acquired with an Olympus DP70® camera mounted on an Olympus BX51TF® (Tokyo, Japan) compound microscope, using a 20X objective (0.50 N.A.). For any section that was not flat several images were acquired at different focal planes, merged using Adobe Photoshop (Adobe Systems Inc., San Jose, CA), and adjusted for brightness and contrast.

To estimate volumes, I used Olympus cellSens™ Entry software (v.1.14). Outlines of the

CNS tissue (“brain”) and head (Figure 12D) were traced using the measure polygon tool, which provides the total number of pixels included within the outlined area. In some sections, CNS tissue was not contiguous, in which case there were several polygons outlined and measured. The following steps were taken for calibration:

104

1. A digital image of a stage micrometer (1 division= 0.01 mm) (Ward's Science, Rochester, NY)

was acquired using the same microscopy conditions and equipment.

2. Using the stage-micrometer image, and Adobe Photoshop, I drew a one-pixel-wide line of 100

µm and found the number of pixels contained on that line: 158 pixels. That value was

used to calculate the number of pixels in a 100 x 100 µm square (10,000 µm2), namely

24,964 pixels.

3. The summed area measurements in pixels for each section were multiplied by the conversion

factor, 10000 µm2/24964 pixels, yielding measurements in units of µm2

(pixels*µm2/pixels= µm2).

4. Brain or head volumes were then estimated from the summed area measurements in µm2

multiplied by the section thickness (µm2 *12 µm= µm3).

III.B.5. Statistics and graphing

Pair-wise comparisons between genotypes of individual parameters were made using the

Mann-Whitney rank sum test, a nonparametric statistical test for non-normally-distributed data

(Fay & Proschan, 2010). I developed the appropriate pairwise comparison script and used R

Studio (version 3.2.1; R Core Team, 2015) for statistical comparisons and graphing. Microsoft

Excel 2016 (Microsoft Corporation, Redmond, WA, USA) was also using for graphing.

III.C. Results

III.C.1. The small-brain phenotype of CASK mutant flies

To assess brain-size differences, I used a histological staining technique based on osmium tetroxide, an excellent heavy metal stain for visualization of lipids and, therefore, cell

105 membranes (Figure 12D). Staining biological tissue samples with osmium tetroxide provides very good contrast for image analysis, whether by traditional qualitative methods (Wigglesworth,

1957; Adams et al., 1967) or more recent quantitative approaches (Riveros and Gronenberg,

2010). Based on my observations of freshly dissected larval CNS, I predicted that adult CASK mutants would have a smaller brain volume than CASK controls.

In the pilot study, I compared brain sizes of CASK control and mutant pharate adults.

Because male and female flies differ considerably in body size (Stillwell et al., 2010; Rideout et al., 2015), the analysis was performed separately for each sex. Estimated brain volumes of

Δ18/∆18 CASK mutants were significantly reduced in both males and females (Figure 12E), with no overlap between mutant and control values within each sex. However, samples sizes in this pilot study were all quite low. More important, the grouping of heads inadvertently resulted in non-random sampling. In particular, at the initial steps through head fixation, many more samples were prepared than were ultimately processed. For reasons of convenience, heads from pupae of similar lengths and the same sex and genotype were grouped, additional samples were discarded. However, this resulted in the samples not reflecting the body-size range within each genotype. Even worse, by chance, it happened that the selections were biased in the following way: Ex33/Ex33 females came from the middle of the range; ∆18/∆18 females came from the very top of their body-size range. For the males, the controls came from the very bottom of their range and the mutants from the middle of their range. In short, this was an unfortunate error that resulted in a grossly non-random sample.

In an effort to mitigate this limitation in the pilot study, I attempted to estimate head volume. This had not been a planned part of the experiment, and sample handling had not been done in an optimal manner to preserve the anatomical integrity of the head capsule. Nonetheless,

106 the thought was that head size might help understand the brain-volume estimates obtained. This analysis was made only in males, because sample sizes were considerable smaller in the female group. Estimated head size revealed no significant differences between CASK Ex33 control and

CASK∆18-mutant individuals (data not shown). This suggested that CASK mutants had microencephaly (small brain) but not microcephaly. The results were intriguing and raised some questions about head volume. In addition, there was a positive linear relationship between brain volume and head volume, showing that for both mutant and control, head volume increased with brain volume (data not shown).

Using a much-improved design, I conducted a follow-up experiment with random sampling, larger numbers, and focused on one sex. From each genotype (CASK Ex33/Ex33 and

CASK Δ18/Δ18) 10 males were selected. Also, unlike in the pilot study, each sample was kept separate with its unique alphanumeric code at every experimental step. Furthermore, the microdissection procedure was improved to ensure that the head volume could be estimated. To enhance the staining of the sections, a longer incubation time with osmium tetroxide was performed. In addition, a sharper blade was used for sectioning. Finally, and as mentioned before, a single head was mounted per block, which facilitated positioning the head in the desired orientation (e.g., to obtain frontal sections). These two different kinds of improvements, the random sampling method and maintenance of individual identities, together with an enhanced technique contributed to a much better experimental design, which would enable acquisition of body length, brain volume, and head volume from every sample.

Again, estimated brain volumes obtained from serial sections of osmium-stained pharate- adult heads of CASK Δ18 homozygous mutants were compared to those of CASK Ex33 homozygous controls. Now the data points of the two genotypes did overlap but the brains of

107

CASK mutant pharate adult, were significantly reduced in volume when compared to control brains (p <0.0005; Figure 13A). Compared with the pilot-study data, the brain-volume estimates were all shifted higher, suggesting that the improved histological methods allowed complete identification of brain tissue.

III.C.2. Relationship between head and body size: microcephaly and short stature

Using the same serial sections of pharate-adult heads assessed for brain volume, I estimated head volume. This analysis showed that CASK mutants have smaller heads than control adults (p <0.001; Figure 13B). In other words, CASK Δ18 homozygous flies have microcephaly. These results are surprising, considering the developmental biology of holometabolous insects. In Drosophila melanogaster, the initial phase of brain and head growth occur as two anatomically separate processes. At pupation, the head capsule, having attained its final size, surrounds the brain, which is still growing (Tissot & Stocker, 2000). In contrast, in humans, brain size is the main determinant of head size in newborns (§ II.A.).

Quantification of the fractional brain volume, the percentage of head volume occupied by the brain, showed that CASK-mutant brains occupied a somewhat greater fraction of the head volume (p < 0.05) compared to CASK EX33 controls (Figure 13D). This suggests that the reduction in the head size in mutant flies is even more severe than their reduced brain size.

The relationship between brain and head volume can also be seen in the scatter plot shown in Figure 13E. The graph displayed a positive association between brain volume and head volume, showing that for both mutant and control, as head volume increased, so did brain volume. Although these values have a minimum overlapping, CASK mutants are localized at the

108

Figure 13. Relationship head and body size: microcephaly and short stature (continued on next page).

109

Figure 13, continued.

110

Figure 13. Relationship head and body size: microcephaly and short stature. A-H.

Comparison of estimated brain and head size from serial sections of osmium-stained, plastic- embedded pharate-adult heads, and body size measured from the length of the pupal case. CASK

Ex33 homozygous control and CASKΔ18 homozygous mutants were used to determine whether or not CASK mutants have microcephaly and short stature. Each circle represents data from a single male fly: homozygous controls (Ex33/Ex33, cyan, n = 10) and homozygous CASK mutants

(∆18/∆18, magenta, n = 10). A. Brain size. The estimated brain volume is significantly reduced in CASK Δ18 animals. B. Head size. CASK mutants also display a significant reduction in head volume. C. Body size. CASK mutants also display a significant reduction in head volume. D.

Brain-Head volume ratio. The fractional brain volume (quantification of brain volume to head volume ratio), is increased in the mutants. E. Relationship between head and brain volume. Both mutant and control animals, show a monotonical increasing association between head volume and brain volume, the larger the head size the larger the brain size. F. Relationship between head and body size. Mutant and control animals show that the lengthy the body size, the bigger the head. G. Relationship between brain and body size. The scatter plot graph displays a positive linear association between brain volume and body length, showing that for both mutant and control, as body length increases does brain volume increase. H. 3D scatter plot showing the association among brain size, head volume and body length, for both CASK mutants and controls. 3D plot created in MATLAB. E-H. Through all comparison between genotypes, data points form two clusters, mutants on the bottom left, and control animals on the top right, indicating that mutants have smaller brain, smaller head and short stature.

111 low-to-medium-sized sector of the space, whereas CASK controls are localized at the medium-to- high-sized sector of the space.

Next, I evaluated the differences in body size between CASK Δ18 homozygous mutants and CASK Ex33 homozygous controls. Because the final size of the adult fly is tightly correlated with pupal size (see § III.A.), I used the length of the pupal case (Figure 12B) as a convenient indicator of body size, and roughly comparable to length or stature in humans. CASK Δ18 homozygotes have significantly smaller body length than CASK Ex33 homozygotes (p < 0.0005)

(Figure 13C). One could say that CASK-mutants have “short stature”, which is also seen with

MICPCH due to human CASK mutations.

The relationship between brain volume and body length was plotted in Figure 13G. The scatter plot displayed a positive relationship. For both mutants and controls, as body length increases so does brain volume increase.

Another association that was analyzed, considered the relationship between body length and head volume (Figure 13F). The graph showed that head sizes tended to vary according body size, the larger the body length, the larger the head.

Finally, the association among brain size, head volume and body size are displayed in a

3D scatter plot (Figure 13H). In this global overview of the different metrics analyzed, the data are clearly clustering, with mutant and control forming two distinct groups. CASK-mutants not only have smaller brain (microencephaly) and smaller head (microcephaly) when compared to

CASK controls, but they also have short stature.

112

III.D. Discussion

In a carefully controlled experiment, I demonstrated that CASK mutants have smaller brains when compared to CASK control adults. I further showed that LOF mutations in

Drosophila CASK led to small heads and short bodies. In addition, I determined the relationship among brain and head volumes, and body length. CASK-mutants flies have a much smaller proportional size than control adults.

It is well known that microcephaly is an important indicator of neurodevelopmental diseases. A smaller brain could potentially have fewer or smaller neurons, or both. For example, in Drosophila, mutations in the minibrain gene (mnb) result in a marked reduction in the volume of the adult optic lobes and central brain hemispheres. This phenotype is caused by a reduction in neuron number rather than neuronal size (Tejedor et al., 1995). Dual-specificity tyrosine- regulated kinase 1A (Dyrk1a) is one of the mammalian orthologues of the Drosophila mnb gene.

It encodes a protein kinase that catalyzes its own activation through autophosphorylation, and the phosphorylation of exogenous substrates. DYRK1A is a dosage-sensitive gene, LOF mutations in human DYRK1A cause dominant ID with microcephaly (Bronicki et al., 2015), whereas its overexpression contributes to Down syndrome cognitive dysfunction (Duchon & Herault, 2016).

In mammals Dyrk1a controls neurogenesis and neuronal maturation (Tejedor & Hämmerle,

2011). Dyrk1a pharmacological inhibition on neuronal progenitors isolated from mice overexpressing Dyrk1A prevented premature neuronal maturation (Mazur-Kolecka et al., 2012).

In addition, Fotaki et al., (2002) showed that homozygous Dyrk1A-mutant mice died at midgestation, while heterozygous mice showed decreased neonatal viability, growth retardation and body size reduction. Brains from heterozygous Dyrk1A mice were decreased in size in a region-specific manner, although in most areas there was no evidence of alterations in the

113 cytoarchitecture and neuronal components. Besides, cell counts showed a specific decrease in the number of neurons in the superior colliculus, which exhibited a significant size reduction (Fotaki et al., 2002).

Based on the bushy phenotype (see Chapter II) and my personal impression about neuronal densities, I hypothesize that the main cause of small brain or CNS size in CASK-mutant flies is a reduction of neurite arbors, not a reduction of neuronal numbers. A pilot study assessing the role of neuron number in the small-brain phenotype of Drosophila CASK mutant was conducted by the neuroscience undergraduate student Lauren Pisani. She analyzed my three div cultures (see § II.B.2., II.C.5.) and found that the whole-CNS neuron number estimates coming from were very similar through the three samples analyzed, CASK mutant, CASK control, and the transgenic rescue sample. Further experiments are necessary to test my hypothesis. For instance, an in-depth analysis of neuron number can be conducted by using flow cytometry.

The main contributors to brain size are the number and size of neurons and glial cells, among other tissues, e.g., endothelial cells, choroid plexus (Mota & Herculano-Houzel, 2014).

The adult human brain contains an estimated number of about 86 billion neurons and an equal amount of non-neuronal cells, e.g., oligodendrocytes, astrocytes, microglia (Herculano-Houzel et al., 2015, b). Thus, a reduced number of glial cell type may lead to a smaller brain. In addition to the neuroblast lineage, Drosophila studies suggest that the glial lineage governs overall brain volume in the larval brain (Pereanu et al., 2015). However, potential contribution(s) of individual brain lineage(s) (neuroblast or glia) to the defective brain growth is currently unclear. Mutations in WDR62, which encodes a centrosome-associated protein, cause autosomal recessive primary microcephaly and other cortical abnormalities in humans, e.g., lissencephaly, polymicrogyria

(Bilgüvar et al., 2010). Loss of Wdr62 results in neural progenitor cells, mitotic delay, cell death

114 and smaller brains in mice (Chen et al., 2014). More recently, Drosophila model of Wrd62 showed that control of brain growth predominantly depends upon glial lineage function, as depletion of wdr62 specifically in the glial lineage significantly reduces brain volume (Lim et al.,

2017).

Interestingly, when analyzed CASK-mutant pupae, I also found that they have short body length. Not only brain and head were reduced, but also the body size. The mean pupal length for

CASK mutants was 2.8 mm, which is at the lower limits of the smallest WT D. melanogaster analyzed by Reeves et al. (2017). This provides evidence suggesting that the CASK gene has roles beyond brain growth. Very often, short stature has been also reported in children with

CASK-related disorders, such as MICPCH (Moog et al., 2011, 2015) and FG syndrome-4 (Piluso et al., 2003) or X-linked ID with or without nystagmus (Hackett et al., 2010). There are thus parallels between the human and the fly phenotype, where the main features are faithfully replicated.

Remarkably, Brown et al. (unpublished, in preparation) found that short stature co- occurred with 66% of monogenic congenital microcephaly disorders. This evidence shows that genes with mutations causing these disorders are rarely specialized for brain growth, but rather have generalized functions in overall body growth. For example, individuals with Seckel syndrome 5, caused by CEP152 mutations, have congenital microcephaly and ID. This disorder is also associated with severe intrauterine growth retardation and short stature. Weaker CEP152 mutations also cause autosomal recessive primary microcephaly 9 (MCPH9), this phenotype being less severe than Seckel syndrome 5. MCPH9 patients present with moderate cognitive impairment but do not have short stature (Guernsey et al., 2010). Similarly, CASK mutations seem to cause widespread developmental defects, i.e., both brain and body growth are affected.

115

Available information from CASK disease-causing mutations seems to show that these are not always accompanied by a short stature phenotype. Thus, it will be important to further investigate the difference in functional consequences of human CASK mutations to understand the cases in which the phenotype is or is not accompanied by short stature.

Najm et al (2008) reveal the cellular defects underlying cerebral cortical malformation by autopsy analysis of a neonate with MICPCH. The study described an abnormal dispersion of neurons in the deeper layers of the cortex, i.e., laminae V-VI merge together, a reduced neuronal number in the basis pontis and in the thick external granular cell layer of the cerebellum. The fact that the pons and the cerebellum are the most severely affected CNS structures in this disorder suggest that specific regions are more sensitive to the disruption of CASK function. Therefore, one of the questions that arises in Drosophila is whether the volume of different brain structures is reduced proportionately, or if certain brain regions are preferentially affected by the CASK mutation. A gross assessment of the samples analyzed did not reveal any obvious disproportion among different regions of the brain. However, uncertainty arises due to lack of a quantitative analysis and the fact that other staining methods would be better to resolve regional differences.

It remains possible that subtle defects may be revealed upon assessment of a detailed regional analysis of specific brain regions, such as optic lobes or antennal lobes, two major easily identifiable neuropil areas.

Located in the middle of the Drosophila brain is the central complex, which serves as an integration center for diverse motor, sensory, learning, and memory activities (Wolff et al.,

2015). The central complex of the adult fly brain consists of four substructures: the ellipsoid body, the fanshaped body, the noduli, and the protocerebral bridge. To date, comparative anatomy, behavioral genetics and electrophysiology, all suggest that the central complex plays a

116 cerebellar-like role and has analogous functions to the cerebellum (Strausfeld, 1999). The analysis of several Drosophila mutant strains has revealed several structural defects in one or more neuropils of the central complex (Strauss & Heisenberg, 1993). For instance, in ellipsoid body open-mutant flies the central complex is disturbed to varying degrees including the dissociation of the ring into two parts, a cleft in the fan-shaped body and hypoplasia in the protocerebral bridge (Ileus et al., 1994). Therefore, it would be interesting to investigate whether there is any central complex defect in CASK-mutant flies. It may be the case that CASK-mutant flies not only have small brain but can also have additional CNS structural defects.

In summary, my data show that estimated brain volume is reduced in CASK-mutant

Drosophila pharate adults. Furthermore, I showed that the estimated total head volume and the body length are also reduced in the mutants. Most importantly, I provide evidence to support the parallel between human and fly regarding CASK-related disorders. Taken together, results obtained from CNS-size abnormalities and body-size measurements, provided novel insights into

CASK defects in brain development and establish Drosophila CASK as a good model for future investigation of human monogenic microcephaly.

117

Chapter IV: Discussion

IV.A. Summary, synthesis and significance of findings

Motivated in part by studies in the Drosophila neurexin gene conducted in the Restifo lab

(Shim & Restifo, 2013; unpublished data), I selected CASK from among a group of ID genes for the focus of my research. The appeal of CASK included its many possible functions, diverse localizations, and furthermore, because many CASK-binding partners are also encoded by ID genes. Hence, I designed and conducted a series of experiments with the ultimate goal of characterizing neuronal and CNS abnormalities in Drosophila CASK LOF mutants.

The vast majority of the hundreds of human ID genes are very highly conserved in

Drosophila. Over 85% of human ID genes have a one or more Drosophila homologs and 76% have at candidate functional ortholog (Inlow & Restifo, 2004; Oortveld et al., 2013). Among the

⁓850 single-gene mutations causing ID is the X-linked gene CASK (Piluso et al., 2009; Tarpet et al., 2009; Moog et al., 2011; Burglen et al., 2012; Hayashi et al., 2012; Moog et al., 2015;

Hayashi et al., 2017).

To understand the cellular bases of cognitive deficits, such as ID, an animal model with strong face, construct and predictive validity is desired (Nestler & Hyman, 2010). Face validity means that the model ‘looks like’ the disease, i.e., if faithfully recapitulate clinical features or symptoms of the human disease. Construct validity means the etiologic process that cause the human disease is the same in the model and/or if there is a convincing theoretical rationale for the model. Predictive validity, in turn, means whether the model can be perturbed or treated with, for instance drugs, to mitigate the symptoms. Although it seems obvious, many of the symptoms used to establish diagnoses in human developmental brain disorders, e.g. language delay, are

118 never going to have any face validity in an animal model (Nestler & Hyman, 2010; Narayanan &

Rothenfluh, 2016). Although the correspondence may only be approximate, in fruit flies some other indicators of cognitive deficits can be detected, such as reduced learning and/or memory, as well as of social behavior and executive function (e.g., integration and motor output) among others (e.g., drug addiction) (Narayanan & Rothenfluh, 2016). To meet all three types of validity criteria results difficult, however, numerous studies have demonstrated how Drosophila can be used to systematically model developmental brain disorders (Restifo, 2005; Bolduc & Tully,

2009; Gatto & Broadie, 2011; Narayanan & Rothenfluh, 2016). In fact, the Drosophila system has provided valuable disease models with strong face validity cases. Despite the diverse approaches and the challenges of validation, in Drosophila, there are some well-known examples of single-gene ID disorders that offer at least two criterion validities e.g. face/construct, face/predictivity, etc. Among these “well-validated” disorders in fruit flies are Angelman syndrome, Rett syndrome, neurofibromatosis Type 1 and fragile X syndrome (Gatto & Broadie,

2011). Indeed, a Drosophila model for fragile X syndrome was developed as the first ID model in flies. Fragile X syndrome, the leading monogenic cause of ID, results from the loss of FMR1 gene function. Drosophila fragile X mental retardation 1 (dfmr1) is the sole Drosophila ortholog of FMR1 (Wan et al. 2000). Flies with dfmr1 LOF mutations have disrupted circadian rhythms and impaired courtship behavior (Dockendorff et al. 2002), alterations in synaptic branching (Morales et al. 2002) and motor-behavior deficits revealed by increased levels of grooming (Tauber et al. 2011). In addition, dfmr1-mutant flies, have defects in cognitive function, as indicated by their reduction in long-term associative memory (LTM) (Bolduc et al.

2008) and defective mushroom-body development (Michel et al. 2004). Thus, the phenotypes observed in dfmr1-mutant flies are consistent with the human FXS phenotype. Moreover, CASK

119 mutations in Drosophila recapitulate the CASK human phenotypes, they are not only LOF, but also have an mendelian inheritance. Hence, the Drosophila model for CASK mutations causing

ID disorders has strong construct validity.

The first goal of my project was to examine the potential role of CASK in neuronal morphogenesis by using primary neuronal culture. This emerged from the fact that primary neuron culture can reveal subtle or hidden phenotypes (Kraft et al., 2006 and 2013). For instance, neurons from grossly normal brains that are deficient in fascin, a highly conserved actin- bundling protein, have a dramatic abnormality when cultured in vitro. Fascin-deficient singed- mutant neurite arbors have a “filagree” phenotype, characterized by the excessive clockwise curvature and failure of proximal-to-distal tapering, a striking phenotype with high penetrance that was revealed in neuronal primary culture (Kraft et al., 2006). Restifo group have identified an ID-gene pathway that connect Fascin1, PAK3, CREB Binding Protein (CBP) and CASK

(Lewis, Tello & Restifo; unpublished data), all with characteristic neuronal-arbor phenotypes revealed by primary cultures prepared from the entire CNS of wandering third instar larvae of

Drosophila. PAK function stabilizes the actin cytoskeleton (Kreis & Barnier, 2009). Cultured mutant neurons from Drosophila PAK, the fly ortholog of PAK3, show a marked reduction in neurite outgrowth, especially of higher-order branching in vitro (Lewis, 2007). CBP is a transcriptional coactivator of CREB (McManus & Hendzel, 2001). In Drosophila, the single ortholog of CBP is the gene dCBP (or nejire). When cultured in vitro, dCBP-mutant neurons have a “naked” neurite-arbor phenotype, with few primary neurites and even fewer higher-order branches (Olson & Restifo, unpublished).

By using primary cultured neurons from the whole CNS of wandering third instar larvae of Drosophila, I found that homozygous CASK ∆18-mutant neurons have small neurite arbors

120 with an altered shape, that because of its qualitative appearance was named “bushy”. Neurite length, territory area, and higher-order branch number (branch number - primary neurites), were significant reduced, while branch density (number of branches per 1000 µm2) was increased. By testing CASK heterozygous (Ex33/∆18), I also demonstrated that the Drosophila CASK-mutant bushy phenotype is semi-dominant. Furthermore, by using two chromosomal deletions,

Df(3R)X307 and Df(3R)X313, I verified that the bushy phenotype of cultured larval CNS neurons maps to the CASK region (see § II.C.4.). Finally, I demonstrated that by introducing one transgenic copy of CASK+ the bushy phenotype was improved (see § II.C.5.). Together, all data indicate that there is a dosage-sensitive requirement for CASK in the CNS of both Drosophila and humans.

Three different but connected facts motivated the second goal of this project, which was to study the role of Drosophila CASK in brain size. First, the strong construct validity for

Drosophila CASK recapitulating the disease-associated genetic abnormality in humans. To date,

CASK mutations reported in humans suggest a model of LOF causality of CASK-mutant phenotypes (Piluso et al., 2009; Moog et al., 2011 and 2015; Hayashi et al., 2017). In addition, the semi-dominant nature of these mutations can be inferred based on the phenotypic data in males and females (see § II.C.3.). Drosophila CASK mutations are LOF with a semi-dominant phenotype, consistent with the human phenotypes. Second, I observed reduced brain size in

CASK mutants at wandering third instar larvae stage. Third, the human microcephaly phenotype caused by CASK mutations. To confirm and evaluate brain-size abnormalities, I analyzed Δ18 homozygous flies by histology of plastic-embedded serial sections. I found that the estimated total brain volume was reduced in CASK mutants. In addition, I showed that mutations in

Drosophila CASK lead to a small head and short body length.

121

There are several examples that connect ID, microcephaly and short stature. Indeed, it was found that short stature co-occurred with 66% of monogenic congenital microcephaly disorders (Brown et al., ms. in preparation). For example, mutations in polyglutamine binding protein 1 (PQBP1) cause an X-linked disorder called Renpenning syndrome (Stevenson et al.,

2005). This disorder is characterized by ID, microcephaly, short stature and variable malformations (Kalscheur et al., 2003). LOF PQBP1 mutations affects the alternative splicing of

RNA, influencing the mRNAs of multiple targets (Wang et al., 2013). Knockdown of PQBPQ1 impairs neurite outgrowth and branching, i.e., primary embryonic cortical neurons from PQBP1 knockdown mice, showed that the number of projections and length of neurites were decreased

(Wang et al., 2013).

Taken together, the bushy neurite-arbor defect and the small-brain phenotype accompanied by “short stature” correspond well to the human phenotypes caused by CASK mutations. Thus, my study demonstrated the main features of CASK-related disorders can be faithfully modeled in flies.

IV.B. An hypothesis to explain the microcephaly of Drosophila CASK mutants

Woods (2004) suggested that congenital microcephaly is likely to be caused by a decrease in the number of cells, while postnatal microcephaly may be due to the reduced numbers of dendritic or axonal processes of a normal number of neurons. In this context, the significantly smaller brain volume of CASK mutants could be caused by reductions in either neuronal number or neuronal-arbor size or both. My impression regarding the density of neurons after CNS dissociation, was that the number of neurons in the various genotypes analyzed across all of my neuronal cell culture experiments was not different. Preliminary data obtained from

122 these cultures verified this impression (see § III.D.). Thus, I hypothesize that the reduced brain size may is not due to the number of neurons, but rather occurs because of decreased neuronal size, i.e., neurite arbor.

An example that indicates a connection between reduced brain size and decreased neuronal size is evident in the case of PAK3, a gene on the X chromosome encoding a p21- activated kinase (see § II.D.2.) (Kreis & Barnier, 2009). Mutations in PAK3 cause ID, postnatal microcephaly, craniofacial and neuropsychiatric features (Peippo et al., 2007). Consistent with the neuroanatomical data from humans, Lewis & Restifo (unpublished) discovered that

Drosophila Pak-mutant brains are reduced in size. In addition, cultured mutant neurons have reduced neurite arbor size, especially of higher-order branching in vitro (Lewis, 2015).

Furthermore, mutations in Pak cause defects in the length, thickness, and direction of the mushroom bodies α and β lobes (Lewis, 2015). Mushroom bodies are paired neuropil structures in the Drosophila central brain that are essential for certain types of learning and memory

(Heisenberg, 2003).

In studying the G protein-coupled receptor kinase-interacting protein-1, GIT1, Hong &

Mah (2015) set another example where a small brain and reduced neuronal size are both part of the same phenotype. GIT1 is a multifunctional adaptor protein that upregulates the PIX-RAC-

PAK signaling pathway. Brain tissue of GIT1−/− mice showed reduced RAC1 and PAK3 activation, as assessed by immunoblotting analysis of whole-brain homogenates (Won et al.,

2011). It has been suggested that GIT1 has an essential role in the development of dendritic spines and neuronal synapses (Zhang et al., 2003 and 2005). GIT1−/− mice showed reduced brain size. In addition, a significant decrease in neuronal cell body size was observed in these mutants.

There were no changes in the number of neurons per area in various brain regions, including the

123 cortex and hippocampus. Hong & Mah, (2015) suggest that small brain appears to be caused by reduced neuronal size rather than number. Moreover, in a Drosophila model, dGit mutants also exhibited decreased brain size (Hong & Mah, 2015). Drosophila dGit mutants also exhibited impaired mushroom body development, e.g., missing one α or β lobe. Because GIT1 upregulates

PAK, it results quite interesting the similarities observed between the dGit- and PAK-mutant phenotypes in Drosophila.

In incorporating these findings to the evidence presented in this body of work about the

“bushy” neurite arbors of CASK-deficient fly neurons, I suggest that the small-brain phenotype results from an inability of mutant neurons to fully extend dendritic and/or axonal arbors.

Although some CNS abnormalities using CASK∆18 mutant have been noted previously, as demonstrated by impaired middle-term and long-term memory in Drosophila (Malik et al 2013), this is the first time when defects in neuronal arbor and brain size have been reported.

IV.C. The functional nature of pathogenic CASK mutations

Since the first case reported of MICPCH caused by a CASK mutation, more than 60 cases with many different molecular lesions (Table 1) have been reported (Najm et al., 2008; Moog et al., 2011; Burglen et al., 2012; Hayashi et al., 2012; Takanashi et al., 2012; Moog et al., 2015;

Hayashi et al., 2017). Moreover, it seems that this number continue to rise. As for FG syndrome-

4 or ID with or without nystagmus, ten cases have been reported. CASK alterations in all the

MICPCH and FG syndrome-4 or ID with or without nystagmus cases reported come roughly equally from point mutations, as well as small deletions of the CASK gene.

The vast majority of human data about CASK mutations comes from X chromosome array-CGH and whole-exome sequencing which reveals specific mutations types. However, there

124 is a lack of understanding of the nature of the functional change of CASK. To understand the effect of the mutations is necessary to conduct functional studies to assess protein accumulation.

However, in a vast number of cases reported, information about mRNA transcript and protein abundance levels is limited. In a few cases in silico analyses of deleterious mutations are used specifically to predict the possible consequences of such alterations, e.g., transcript level expression, splicing efficiency, etc. (Najm et al., 2008; Dunn et al., 2017).

The current working hypothesis is that all CASK mutations causing MICPCH or FG syndrome-4 or ID with or without nystagmus are LOF (Najm et al., 2008; Moog et al, 2011;

Moog et al, 2015; Piluso et al., 2009). Moog et al. (2015) identified four boys with MICPCH phenotype, each with different novel CASK mutations. They presented a severe phenotype, as seen in affected heterozygous females. RNA analysis in these four males showed that all CASK transcripts harbored a premature termination codon. In addition, by using fibroblasts from the same patients, no CASK protein was detected. (Moog et al., 2015). Additionally, RNA analysis was performed in four females affected with MICPCH, all of them carrying different CASK mutations. In each case aberrant transcripts leading to a premature top codon were identified

(Burglen et al., 2012; Hayashi et al., 2012). Unfortunately, for the rest of these cases with

MICPCH there is no additional information about functional analysis. These data give rise to an additional hypothesis, that MICPCH represents the null CASK phenotype. In milder phenotypes

(FG syndrome-4 or ID with or without nystagmus), only one out of the 10 cases reported have shown some functional analysis of CASK (Dunn et al., 2017). Piluso et al., (2009) reported a patient carrying a mutation (R28L) that affected specifically the CaMK-like domain of CASK.

By using the yeast two-hybrid system, it was shown that the binding between CaMKinase-like domain of CASK, the interacting domains of Mint-1 and Caskin1 were not modified by the

125

R28L mutation identified in this patient. Similarly, by performing in vitro GST pull-down assays, the same result was also obtained, there were not differences in binding properties of

CaMKinase-like domain of CASK as a result of R28L substitution. It suggests that the observed

FGS4-syndrome phenotype was not determined by alterations of the binding property of CaMK- like domain (Piluso et al. (2009). Three additional cases were reported to have RNA analysis in this milder phenotype (Hackett et al., 2010; Dunn et al., 2017). RNA isolated from lymphocytes from two affected males, each with a different CASK mutation, showed different aberrant transcripts that result in an in-frame deletion of nine amino acids at the C-terminal end of CASK

(the protein 4.1 binding motif) and in an in-frame deletion of 28 amino acids at GUK-kinase domain in each case (Hackett el al., 2010). Hackett et al., (2010) also reported that an obligated carrier for the latter defect (28 amino acids in-frame deletion) has the WT and both mutant transcripts. The possibility that CASK may have dominant-negative effects cannot be ruled out.

For instance, three different transcripts, a WT transcript and two-mutant transcripts, were found in an affected boy with FG syndrome-4 due to a splice-site mutation (Dunn et al., 2017). It may be the case that these two mutant transcripts result in a small mutant CASK protein stable enough to accumulate, interfering with the function of the normal CASK. Taken together, these data suggest that FG syndrome-4 or ID with or without nystagmus is due to varying degrees of partial loss of function, i.e., a hypomorphic phenotype. That is to say, less normal CASK protein is made, and/or some function of CASK has been reduced. Human CASK mutations are considered to be LOF as in CASK Δ18 homozygous flies. However, homozygous CASK-mutants flies do not correspond to the human null phenotype, i.e., MICPCH. Instead, Δ18 homozygous would fall into the category of hypomorphic phenotype, i.e., less severe.

126

IV.D. The role of CASK in synapse structure and function: myth or legend?

The available data about CASK disease-causing mutations and the human phenotypes provide direct evidence that CASK has an essential role in brain development. For years, published literature, including protein CASK-domain organization, expression and distribution of CASK in different locations and CASK protein interaction, have focused on a role for CASK in synaptic maturation development and function. Indeed, because CASK is mostly known for its alleged role as scaffolding protein, one might think that defects in synaptic structure and function are the main pathologic feature in animal models of CASK mutations. For instance, contrary to the prevailing hypothesis, CASK mutations have no effect on synapse formation in mouse, where neurons lacking CASK display overall normal electrical properties and form ultrastructurally normal synapses (Atasoy et al., 2007). Therefore, the current state of knowledge about the pre- and post-synaptic roles of CASK have not confirmed these expectations.

Furthermore, the congenital microcephaly and the pons and cerebellum hypoplasia reported in CASK-mutant patients, support a role of CASK early in brain development, probably during the main phase of neurogenesis. Therefore, based on the body of published research on

CASK function, CASK is required before synaptogenesis. Indeed, neurite outgrowth and the formation of the appropriate neuronal arbor have to occur before synapse formation. This is for instance illustrated during optic lobe formation in flies, where synaptogenesis is the last cellular event. In this case, limited growth of presumptive dendrites from lamina neurons towards presynaptic sites distributed over the photoreceptor axon terminals occurs first, later synaptogenesis is reached (Frohlich & Meinertzhagen, 1983). Interestingly, severe defects in brain size and in the optic lobe were found in Drosophila mutants for the gene asp. ASPM, the human ortholog of asp, is a well-known gene involved in neural development. Mutation in

127

ASPM are the most frequent case of autosomal recessive primary microcephaly (Rujano et al.,

2013).

It seems obvious that CASK must have a non-synaptic function. For instance, Srivastava et al. (2016) demonstrated that CASK regulates postnatal brain growth in mouse, and this is not due to the cell-autonomous loss of CASK-null brain cells. Specifically, the Cre-LoxP mediated gene excision was used to specifically delete CASK from mouse cerebellar neurons (see § I.D.4.).

These cerebellar neuron-specific deletion of CASK does not produce cerebellar hypoplasia and does not cause perinatal lethality, as the previous reported mouse models (Laverty et al., 1998;

Atasoy et al., 2007). In addition, in heterozygous female mice (CASK–/+) pronounced microcephaly was observed beginning one week after birth (Srivastava et al. 2016). A marked decrease in cerebellar size and number of cells in the cerebellum was also observed in these animals. According to Srivastava et al., (2016) the observed microcephaly in heterozygous animals (CASK–/+), was not associated with a specific loss of CASK null neurons indicating that

CASK regulates postnatal brain growth in a non-cell autonomous manner. In addition, reduction of CASK expression in human-cell or whole-mouse models caused impaired metabolic activity, with decreased oxidation rates and increased lactate production (Srivastava et al. 2016).

Furthermore, the role of CASK in a variety of molecular functions has been demonstrated.

For instance, a proteomic analysis of insulin signaling pathway revealed that CASK forms a complex with two unidentified phosphotyrosine proteins in response to insulin-stimulation.

Insulin decreases CASK nuclear localization and down-regulates the expression of CASK and its target genes (Wang et al., 2006). In addition, it has been found that CASK is over-expressed in various cancers and it has been related to cell migration and invasion in studies (Freissler et al.,

2000; Stafford et al., 2011). Indeed, high expression of CASK is detected in colorectal cancer

128

(Wei et al., 2014), which might prove valuable as a prognostic marker. Interestingly, CASK dysregulated levels not only alter neurogenesis, but also can be tumorigenic.

IV.E. Future directions

As discussed in the two previous sections (IV.B. and IV.C.), pathogenic CASK mutations are considered to be LOF as in CASK ∆18 in Drosophila. In addition, it is also apparent that these mutations failed to reveal either a synaptic structural or function phenotype. For instance,

NMJ defects found in CASK mutants seems to be contradictory and inconclusive (see § I.E.)

(Zordan et al., 2005; Chen et al., 2011). A growing body of work argues that CASK may not be essential for synaptic function. Nonetheless, it does not necessary conveys that CASK plays no role. It is possible that the role of CASK in synapse formation could be partially redundant. For instance, compensation and molecular redundancy have been reported among postsynaptic density-membrane associated guanylate kinase proteins (PSD-MAGUKs) in the task of AMPAR trafficking (Elias et al., 2006; Howard et al., 2010). Thus, in order to see the potential role for

CASK in this scenario, one or another source of functional redundancy can be knocked down.

For example, genes that regulate the same biological process than CASK can be identified using a genetic screen for suppressors and enhancers. Then, by decreasing the dosage of those genes, the

“missing” function of CASK can be revealed.

The CASK bushy phenotype can provide new insights to the distinctive function of CASK in neuronal morphogenesis. In addition, it can be used for drug discovery and development. The sensitivity of the CASK-mutant neurons can be tested with FDA-approved drugs or novel small molecules, that can either restore or worsen the phenotype. For instance, Kraft et al. (2013) conducted a drug screen using fascin-deficient singed mutant Drosophila neurons cultured in

129 vitro. This was the first drug screening analysis to use Drosophila primary neuronal culture with therapeutic goals for developmental brain disorders. An ongoing study at Restifo lab is focusing on the potential therapeutic benefit of fascin-pathway modifiers for ID-associated postnatal microcephaly. These modifiers are a small collection of fascin-pathway enhancing and fascin- pathway blocking drugs identified in a phenotypic screen based on the filagree phenotype (Kraft et al., 2013). By exposing Drosophila CASK-mutant neurons to fascin-pathway modifiers we may be able to rescue (or worsen) the bushy neurite-arbor defects observed in these mutants.

This analysis may shed light on designing future drug trials in mouse models and, subsequently, in human patients with genetic ID.

Neural cells derived from human induced pluripotent stem cell (iPSC) lines have been used to investigate neurological disorders. iPSCs can be generated from human adult somatic cells by retroviral transduction of the four specific transcription factors (Xie & Tang, 2016). By reprogramming somatic cells from MICPCH-patients to iPSC lines, the effect of these mutations on the differentiation of various cell types and on early developmental processes can be investigated. In addition, by providing special trophic factors, human iPSCs sequentially differentiate into functional neurons and glia in vitro (Bao-Yang et al., 2010). Thus, it would be interesting to study differentiated neurons from MICPCH patient-derived iPSC lines to determine whether or not they resemble the abnormal neurite arbors observed in CASK-mutant

Drosophila neurons.

Because the number of genes causing ID in humans is rapidly growing, it is necessary to develop and validate animal models in order to facilitate the study of ID pathogenesis. Hence, my data contribute to strengthening the value of Drosophila genetics, and the remarkable role of

CASK-mutant flies as models of human ID with microcephaly disorders. This Drosophila model

130 expands our current knowledge for neuronal morphogenesis and brain size, and provides novel insights into the pathogenesis of brain development disorders.

131

APPENDIX A

Characterization of motor behavior in Drosophila CASK loss-of-function mutants

In addition to cognitive deficits and behavioral disturbances, motor-system development may also be compromised in children with ID (Hartman et al., 2010). Patients with X-linked

CASK-disorders provide a clear example of this. For instance, moderate-to-severe psychomotor delay, including hypotonia and lack of ambulation is a consistent phenotype in most individuals with MICPCH, the most severe phenotype caused by CASK mutations (Najm et al., 2008; Moog et al., 2011; Takanashi et al., 2012; Moog et al., 2015). The vast majority of affected children never attained the ability to walk; mild ataxia was described in a few cases as well. The case reports of the less severe disorders (FG syndrome-4/X-linked ID with or without nystagmus), are less informative regarding motor phenotypes, although nystagmus is a specific motor-control defect. Tremor and unsteady gait are the only other motor manifestations reported (Hackett et al.,

2010; Dunn et al., 2017).

Drosophila melanogaster demonstrate a rich repertoire of complex behaviors that are relevant to human behaviors, which make it an excellent model organism (Hirth, 2010). During the last several decades, Drosophila have been used to study locomotion (Whitworth et al.,

2005), sleep and other circadian rhythms (Konopka & Benzer, 1971; Ho & Seghal, 2005; Rosato

& Kiriakou, 2006; Chiu et al., 2010) among others. In addition, learning-and-memory studies of

Drosophila have provided strong evidence that validates fruit flies as ID models (Restifo, 2005;

Bolduc & Tully, 2009). For instance, courtship-conditioning assays and olfactory shock- avoidance conditioning helped to demonstrate that deletion of either full-length CASK or its N’- terminal domains (CaMK-like and L27) impairs middle-term and long-term memory in

Drosophila (Malik et al., 2014).

132

There is strong evidence that Drosophila CASK mutants exhibit locomotor deficits, specifically markedly reduced ambulation (Martin & Ollo, 1996; Slawson et al., 2011).

Locomotion is the animal's ability to move itself from place to place and represents one of the most basic behaviors found in nature. It requires coordinated motor movements of the body or specialized parts of the body (Dickinson et al., 2000). To identify specific parameters of locomotion that are defective in Drosophila CASK mutants, Slawson et al. (2011) used a high- resolution video tracking assay. In this assay flies were placed in an open arena and, after a brief air pulse that stimulates walking, their ambulation was video recorded for 30 sec. To characterize the nature of these behavioral differences, the contribution of specific walking parameters, such as pause length, speed and acceleration was determined. CASK-mutant flies displayed a reduction in both initial pause length, i.e., time to resume movement following the air pulse, and average pause length, which can be used as measures of motor initiation. CASK mutants displayed motor-initiation defects (increases in motor initiation time) and are unable to maintain locomotion once initiated (decreases in ability to maintain active movement). This walking deficit in locomotion also includes decreases in speed and acceleration (Slawson et al. (2011). In addition, Slawson et al., (2011) showed that the performance behavior of CASK homozygous mutants was significantly different from that of CASK homozygous controls, with CASK heterozygote flies performing at an intermediate level, which demonstrates a dose-dependent increase of inactivity with loss of CASK and the semi-dominance of the locomotor phenotype.

In order to understand aspects of motor-system function that may accompany cognitive defects in fly models of ID, I examined the spontaneous motor behavior of CASK-mutants. These studies were done in partnership with Dailu Chen, a former Physiology undergraduate student, who had been trained to perform and score the assay by former postdoctoral fellow, Dr. David

133

Andrew. The assay involves recording flies in a small arena for 15 min and scoring behavior frame-by-frame. This allows for the detection of movements during the full period of recording.

In addition to walking activity, we are able to collect other information, such as grooming and falling, among others.

Individual flies (isolated at late pupal stages) were recorded for 15 min at 1 day (24-28 hr) post-eclosion in a standard 96-well-plate high-humidity arenas (Figure A1A). Each fly was gently aspirated from its isolation tube and placed into a well (or arena). The well was then quickly covered with a standard glass microscope slide through which two flies in adjacent wells were recorded at one time. Videos were manually scored using V-code (customized by Dr. David

Andrew) for walking, grooming, falling, standing, twitching and tumbling behaviors, which can then be displayed as an ethogram (Figure A1B-C). We defined “twitch” as a rapid and sudden jerking movement of the entire body, and “tumble”, as essentially the righting reflex, a very quick acrobatic roll from the wall or ceiling of the arena, with the fly landing on its feet. The V- code software allows us to quantify the percentage of time each fly spends performing a specific behavior (the “index” of that behavior), as well as bout number and duration. Bouts were defined as a period of time spent in a particular behavior.

CASK homozygous ∆18 mutants displayed significant reduced walking behavior indices

(Figure A2A; WI, p < 0.0001) compared to Ex33 controls. Consequently, CASK mutants showed a dramatic increase in standing indices (Figure A2B; SI, p < 0.0001). In addition, CASK homozygous ∆18 mutants displayed significant increases in grooming indices (Figure A2C; GI, p < 0.001) compared to Ex33 controls. Interestingly, despite their relative immobility, CASK- mutant flies groomed excessively, primarily due to longer bout duration, i.e., flies spent more time grooming, whereas grooming-bout number was not different between the genotypes. Falls,

134

Figure A1. Timeline ethogram representations of spontaneous motor behavior in CASK mutants.

135

Figure A1. Timeline ethogram representations of spontaneous motor behavior in CASK mutants. A. Photo of a fly in the observation arena (photo credit: Rachel Doser). B-C.

Ethograms showing various mutually exclusive behaviors (walk, groom, fall, stand, twitch and tumble) during a 15-minute observational period, where each row represents a single fly.

Control and mutant flies were selected from the middle quartiles for walking, grooming and standing indices. Broken vertical lines represent the separation between the Acclimation period

(0-4th min) and Test Period (5th-15th min). B. Ethograms of six typical Ex33 homozygous flies.

Genetic controls spend most of their time walking, with frequent grooming bouts, usually brief, and rarely standing. C. Ethograms of six typical ∆18 homozygous mutant flies. Mutants spend most of their time standing, with longer grooming bouts than performed by control flies.

136

Figure A2. Quantitative analysis of spontaneous motor-behavior phenotypes in CASK mutants.

137

Figure A2. Quantitative analysis of spontaneous motor-behavior phenotypes in CASK mutants. Behavior parameters are depicted as box-plot distributions of data obtained from 15- minute video recordings of CASK Ex33 homozygous controls and CASK ∆18 homozygous mutants. The number of flies analyzed is indicated below the genotypes. The middle line within the box (arrowhead) represents the 50th percentile (median). Top and bottom lines of each box represent the 75th and 25th percentiles, respectively. The top and bottom whiskers represent the

90th and 10th percentiles, respectively. The nonparametric Mann–Whitney rank sum test was used to compare the different indices between genotypes. A. Walking Index. CASK-mutant homozygotes spend significantly less time walking when compared to control animals. B.

Standing Index. CASK-mutant homozygotes spend much more time standing than do control flies which rarely stand still in this assay. C. Grooming Index. Despite their relative immobility,

CASK-mutant flies groom excessively compared to controls.

138 twitches, and tumbles were significantly reduced in CASK mutants compared to controls (data not shown).

Unlike control flies, mutant flies showed behavioral activation during the 4 min immediately after being placed in the arena, as revealed by looking at the ethograms (Figure A1).

In other words, CASK-mutant flies were far more active (i.e., more normal) during the early period of observation. We thus separated the 15-minute assay into an acclamation period (0-4 min) and a test period (5-15 min) and analyzed the difference between these two time intervals

(Figure A3). Remarkably, standing and walking metrics demonstrated significant changes between the acclimation and test periods. CASK ∆18 mutant flies showed a highly significant dramatic increase in standing during the test period compared to CASK Ex33 control animals during the same period (Figure A3A; p < 0.0005). This increase in standing by the mutants was accompanied by a concurrent reduction in walking during the test period (Figure A3B; p <

0.0001). Indeed, standing and walking quantification suggests that CASK ∆18 mutants flies respond to handling by exploring the small arena, and then spending most of their time during the test period standing. However, for CASK mutants, the grooming index during these two observation periods showed no significant difference (Figure A3C), while the mean duration of grooming bouts was significantly increased during the test period (Figure A3D; p < 0.0005).

Whereas control flies show little difference between acclimation and test periods, walking is activated in CASK-mutant flies when they are placed in this new environment, indicating that the deficit is not due to motor paralysis. These data demonstrate that, even though CASK-mutant animals engage in higher levels of sedentary behavior, they are able of initiating and sustaining ambulation in the arena.

139

Figure A3. Demonstration of acclimation period in CASK-mutant flies. Change in motor- behavior metrics between early and late phases of the 15-min observation period. Each circle represents data from a single fly: homozygous controls (Ex33/Ex33, cyan circles) and homozygous CASK mutants (∆18/∆18, magenta circles). The ∆ symbol refers to the difference between the 11-min test and 4-min acclimation periods (∆ = Test-period metric – Acclimation- period metric). Changes in behavioral metrics between the two periods were only seen in the mutants. A. Standing index. CASK-mutant flies showed a dramatic increase during the test period, whereas control animals (Ex33) displayed no change. B. In contrast, walking index of mutant flies was substantially decreased, while controls animals did not change their inclination to walk. C. Grooming index. There was no difference between the two time periods for either genotype. D. Grooming-bout duration increased markedly in CASK mutants, but not in control flies. Significance levels, based on Mann-Whitney rank sum test for pairwise comparison between genotypes, indicated by asterisks: ***, p < 0.0005; ****, p<0.00005.

140

To allow a more specific assessment of locomotion behavior in mutants than the provided by the spontaneous behavior assay, further examination assessing gross locomotor coordination and balance could be conducted. Thus, these behavioral data contribute to strengthen the value of

CASK-mutant flies as models of human ID disorders. This is valuable to deep understand the motor-behavior abnormalities in genetic models of ID, providing for instance, a starting point for drug discovery.

141

APPENDIX B

Neurite-arbor morphology after CNS dissociation in microfluidic devices

Primary culture of cells dissociated from animal tissues is a powerful technique that offers the advantage that specific cell types can be studied in vitro (Kaech & Banker, 2006; Ray et al., 2009). Cultured neurons manifest morphological and physiological phenotypes due to normal developmental transitions, mutations, and exposures to drugs and toxins (Chen & Herrup,

2008; Kraft et al., 2013). For almost five decades, the method has relied on mechanical dissociation of developing brain tissue by manual trituration. This method consists of repeated flushing of microdissected, enzyme-treated tissue pieces through a narrow pipette tip, which breaks the cells apart from each other and, if they are already differentiated, from their neurites

(axons and dendrites).

Drs. Linan Jiang and Yitshak Zohar in the Dept. of Aerospace & Mechanical Engineering designed and fabricated a microfluidic device for soft-tissue dissociation. As part of their collaboration with the Restifo lab, I assessed the neurite-arbor morphology of Drosophila CASK larval CNS tissue after dissociation using the microfluidic device. In using this microfluidic device, the dissociation of individual cells, for instance, from brain tissue can be achieved.

During operation enzyme-treated (see § II.B.2) brain tissue is loaded into the device. The microfluidic device components generate then a flow-induced shear stress on the tissue allowing cell dissociation. Because the dissociation can be directly observed, and simultaneously video recording, flow and force parameters can be adjusted (Jiang et al., 2015). The key features of this microfluidic device include one or more narrow orifices along the dissociation channel to exert sufficient stress on the tissue sample to induce cell separation.

142

Briefly, I microdissected the whole CNS of wandering third instar larva and loaded into the microfluidic device. CNS-cell suspension collected from the microdevice was then distributed into five or six culture dishes and proceeded as described in § II.B.2.

Both genetically normal neurons and CASK-mutant neurons extended larger neurite arbors when they were dissociated in the microfluidic device, compared with a parallel manual dissociation (data no shown). The neurons dissociated in the device are viable and capable of extending an arbor of neurites typical of neurons cultured after manual trituration as seen by using Drosophila CASK neurons. Moreover, when mutant and control CNS are dissociated in the device, the CASK-mutant neurons extend smaller, denser neurite arbors, compared with control neurons (Figure B). This replicates the CASK-mutant phenotype previously shown in cultures prepared after manual dissociation. In addition, duplicate preparations of mutant neurons dissociated in the device showed, in a very consistent way, that CASK-mutant neurons cultured after dissociation in the device are larger than those from a parallel manual culture (Figure B).

143

Figure B. Demonstration and quantification of the “bushy” phenotype in cultured CASK- mutant neurons dissociated using a microfluidic device.

144

Figure B. Demonstration and quantification of the “bushy” phenotype in cultured CASK- mutant neurons dissociated using a microfluidic device. Neurite-arbor morphology parameters, depicted as box-plot distributions, from three div larval CNS neuronal cultures prepared after using a microfluidic device for tissue dissociation. CASK Ex33 control neurons

(n=104) (magenta) and two separate set of mutant neurons CASKΔ18 (n=106) (light orange), and

CASKΔ18 (n=104) (dark orange) were compared. Parameters shown are total neurite length

(µm), territory area (µm2), higher-order branch number, and branch density (branches/1000

µm2). The middle line within the box (arrowhead) represents the 50th percentile (median). Top and bottom of each box represent the 75th and 25th percentiles, respectively. The top and bottom whiskers represent the 90th and 10th percentiles, respectively. When mutant and control CNS are dissociated in the device, the CASK-mutant neurons extend smaller, denser neurite arbors, compared with control neurons. This replicates the CASK-mutant phenotype previously shown in cultures prepared after manual dissociation. Duplicate preparations of mutant neurons dissociated in the device were very consistent. The nonparametric Mann-Whitney rank sum test was used to compare measures of neuron size and shape between genotypes. Statistically significant differences are indicated with asterisks: ***, p < 0.0005; ****, p<0.00005.

145

URL List of Web Resources Used

American Psychiatric Association, 2013: https://www.psychiatry.org/psychiatrists/practice/DSM/APA_DSM-5-Intellectual-Disability.pdf, last accessed on August 13th, 2018.

Centers for Disease Control: https://www.cdc.gov/ncbddd/birthdefects/microcephaly.html, last accessed on August 13th, 2018.

National Research Council (US) Committee on Disability Determination for Mental Retardation,

2002: https://www.ncbi.nlm.nih.gov/books/NBK207536/, last accessed on August 13th, 2018.

NCBI Protein (2015): https://www.ncbi.nlm.nih.gov/protein, last accessed on August 13th, 2018.

R Core Team (2015): R: A language and environment for statistical computing. Vienna, Austria:

R Foundation for Statistical Computing; 2015, last accessed on August 13th, 2018.

Online Mendelian Inheritance in Man (OMIM) database: www.omim.org, last accessed on

August 13th, 2018.

US National Library of Medicine: https://ghr.nlm.nih.gov/condition/pontocerebellar-hypoplasia, last accessed on August 13th, 2018.

Bibliography

Abuelo D (2007) Microcephaly syndromes. Semin Pediatr Neurol 14:118-127.

Adams JA (2001) Kinetic and catalytic mechanisms of protein kinases. Chem Rev 101:2271-

2290.

146

Adams CWM, Abdulla YH, Bayliss OB (1967) Osmium tetroxide as a histochemical and

histological reagent. Histochemie 9:68-77.

Aggarwal A, Mittal H, Patil R, Debnath S, Rai A (2013) Clinical profile of children with

developmental delay and microcephaly. J Neurosci Rural Pract 4:288-291.

Armstrong DD (2001) Rett syndrome neuropathology review 2000. Brain Dev 23 Suppl 1:S72-

76.

Asim A, Kumar A, Muthuswamy S, Jain S, Agarwal S (2015) Down syndrome: an insight of the

disease. J Biomed Sci 22:41.

Atasoy D, Schoch S, Ho A, Nadasy KA, Liu X, Zhang W, Mukherjee K, Nosyreva ED,

Fernandez-Chacon R, Missler M, Kavalali ET, Südhof TC (2007) Deletion of CASK in

mice is lethal and impairs synaptic function. Proc Natl Acad Sci USA 104:2525-2530.

Azevedo FA, Carvalho LR, Grinberg LT, Farfel JM, Ferretti RE, Leite RE, Jacob Filho W, Lent

R, Herculano-Houzel S (2009) Equal numbers of neuronal and nonneuronal cells make

the human brain an isometrically scaled-up primate brain. J Comp Neurol 513:532-541.

Bailey A, Luthert P, Dean A, Harding B, Janota I, Montgomery M, Rutter M, Lantos P (1998) A

clinicopathological study of autism. Brain 121:889-905.

Bainbridge SP, Bownes M (1981) Staging the metamorphosis of Drosophila melanogaster. J

Embryol Exp Morphol 66:57-80.

Baj G, Patrizio A, Montalbano A, Sciancalepore M, Tongiorgi E (2014) Developmental and

maintenance defects in Rett syndrome neurons identified by a new mouse staging system

in vitro. Front Cell Neurosci 8:18.

147

Barclay JW, Aldea M, Craig TJ, Morgan A, Burgoyne RD (2004) Regulation of the fusion pore

conductance during exocytosis by cyclin-dependent kinase 5. J Biol Chem 279:41495-

41503.

Bardoni B, Mandel JL, Fisch GS (2000) FMR1 gene and fragile X syndrome. Am J Hum Genet

97:153-163.

Barkovich AJ, Guerrini R, Kuzniecky RI, Jackson GD, Dobyns WB (2012) A developmental and

genetic classification for malformations of cortical development. Brain 135:1348-1369.

Baron IS (2005) Test review: Wechsler Intelligence Scale for Children-Fourth Edition (WISC-

IV). Child Neuropsychol 11:471-475.

Biederer T, Südhof TC (2001) CASK and protein 4.1 support F-actin nucleation on neurexins. J

Biol Chem 276:47869-47876.

Biederer T, Sara Y, Mozhayeva M, Atasoy D, Liu X, Kavalali ET, Südhof TC (2002) SynCAM,

a synaptic adhesion molecule that drives synapse assembly. Science 297:1525-1531.

Bier E (2005) Drosophila, the golden bug, emerges as a tool for human genetics. Nat Rev Genet

6:9-23.

Bilguvar K et al. (2010) Whole-exome sequencing identifies recessive WDR62 mutations in

severe brain malformations. Nature 467:207-210.

Bolduc FV, Tully T (2009) Fruit flies and intellectual disability. Fly (Austin) 3:91-104.

Bolduc FV, Bell K, Cox H, Broadie KS, Tully T (2008) Excess protein synthesis in Drosophila

fragile X mutants impairs long-term memory. Nat Neurosci 11:1143-1145.

Bond J, Scott S, Hampshire DJ, Springell K, Corry P, Abramowicz MJ, Mochida GH, Hennekam

RC, Maher ER, Fryns JP, Alswaid A, Jafri H, Rashid Y, Mubaidin A, Walsh CA, Roberts

148

E, Woods CG (2003) Protein-truncating mutations in ASPM cause variable reduction in

brain size. Am J Hum Genet 73:1170-1177.

Bond J, Roberts E, Springell K, Lizarraga SB, Scott S, Higgins J, Hampshire DJ, Morrison EE,

Leal GF, Silva EO, Costa SM, Baralle D, Raponi M, Karbani G, Rashid Y, Jafri H,

Bennett C, Corry P, Walsh CA, Woods CG (2005) A centrosomal mechanism involving

CDK5RAP2 and CENPJ controls brain size. Genet 37:353-355.

Borg JP, Lopez-Figueroa MO, de Taddeo-Borg M, Kroon DE, Turner RS, Watson SJ, Margolis

B (1999) Molecular analysis of the X11-mLin-2/CASK complex in brain. J Neurosci

19:1307-1316.

Boudeau J, Miranda-Saavedra D, Barton GJ, Alessi DR (2006) Emerging roles of pseudokinases.

Trends Cell Biol 16:443-452.

Brand AH, Dormand EL (1995) The GAL4 system as a tool for unravelling the mysteries of the

Drosophila nervous system. Curr Opin Neurobiol 5:572-578.

Bronicki LM et al. (2015) Ten new cases further delineate the syndromic intellectual disability

phenotype caused by mutations in DYRK1A. Eur J Hum Genet 23:1482-1487.

Brown C (2017) Genetic requirements for building a brain of sufficient size: insights from

Mendelian congenital microcephaly disorders. Master Thesis. University of Arizona,

Tucson, Arizona.

Burglen L et al. (2012) Spectrum of pontocerebellar hypoplasia in 13 girls and boys with CASK

mutations: confirmation of a recognizable phenotype and first description of a male

mosaic patient. Orphanet J Rare Dis 7:18.

Butz S, Okamoto M, Südhof TC (1998) A tripartite protein complex with the potential to couple

synaptic vesicle exocytosis to cell adhesion in brain. Cell 94:773-782.

149

Byrne PG, Rice WR (2006) Evidence for adaptive male mate choice in the fruit fly Drosophila

melanogaster. Proc Biol Sci 273:917-922.

Carvill GL, Mefford HC (2015) Next-generation sequencing in intellectual disability. J Pediatr

Genet 4:128-135.

Chao HW, Hong CJ, Huang TN, Lin YL, Hsueh YP (2008) SUMOylation of the MAGUK

protein CASK regulates dendritic spinogenesis. J Cell Biol 182:141-155.

Chelly J, Khelfaoui M, Francis F, Cherif B, Bienvenu T (2006) Genetics and pathophysiology of

mental retardation. Eur J Hum Genet 14:701-713.

Chen K, Featherstone DE (2011) Pre and postsynaptic roles for Drosophila CASK. Mol Cell

Neurosci 48:171-182.

Chen JF, Zhang Y, Wilde J, Hansen KC, Lai F, Niswander L (2014) Microcephaly disease gene

Wdr62 regulates mitotic progression of embryonic neural stem cells and brain size. Nat

Commun 5:3885.

Chiu JC, Low KH, Pike DH, Yildirim E, Edery I (2010) Assaying locomotor activity to study

circadian rhythms and sleep parameters in Drosophila. J Vis Exp:2157.

Chiurazzi P, Pirozzi F (2016) Advances in understanding-genetic basis of intellectual disability.

F1000Res: 5.

Chklovskii DB (2004) Synaptic connectivity and neuronal morphology: two sides of the same

coin. Neuron 43:609-617.

Cohen AR, Woods DF, Marfatia SM, Walther Z, Chishti AH, Anderson JM (1998) Human

CASK/LIN-2 binds syndecan-2 and protein 4.1 and localizes to the basolateral membrane

of epithelial cells. J Cell Biol 142:129-138.

150

Correas I, Leto TL, Speicher DW, Marchesi VT (1986) Identification of the functional site of

erythrocyte protein 4.1 involved in spectrin-actin associations. J Biol Chem 261:3310-

3315.

Couchman JR (2003) Syndecans: proteoglycan regulators of cell-surface microdomains? Nat

Rev Mol Cell Biol 4:926-937.

Cragan JD et al. (2016) Population-based microcephaly surveillance in the United States, 2009 to

2013: An analysis of potential sources of variation. Birth Defects Res A 106:972-982.

Craig AM, Kang Y (2007) Neurexin-neuroligin signaling in synapse development. Curr Opin

Neurobiol 17:43-52.

Cukier HN, Perez AM, Collins AL, Zhou Z, Zoghbi HY, Botas J (2008) Genetic modifiers of

MeCP2 function in Drosophila. PLoS Genet 4:e1000179.

Deaner RO, Isler K, Burkart J, van Schaik C (2007) Overall brain size, and not encephalization

quotient, best predicts cognitive ability across non-human primates. Brain Behav Evol

70:115-124.

Decobert F, Grabar S, Merzoug V, Kalifa G, Ponsot G, Adamsbaum C, des Portes V (2005)

Unexplained mental retardation: is brain MRI useful? Pediatr Radiol 35:587-596.

Desir J, Cassart M, David P, Van Bogaert P, Abramowicz M (2008) Primary microcephaly with

ASPM mutation shows simplified cortical gyration with antero-posterior gradient pre- and

post-natally. Am J Med Genet 146a:1439-1443.

Dickinson MH, Farley CT, Full RJ, Koehl MA, Kram R, Lehman S (2000) How animals move:

an integrative view. Science 288:100-106.

Dierssen M, Ramakers GJ (2006) Dendritic pathology in mental retardation: from molecular

genetics to neurobiology. Genes Brain Behav 5 Suppl 2:48-60.

151

Dockendorff TC, Su HS, McBride SM, Yang Z, Choi CH, Siwicki KK, Sehgal A, Jongens TA

(2002) Drosophila lacking dfmr1 activity show defects in circadian output and fail to

maintain courtship interest. Neuron 34:973-984.

Dominguez M, Casares F (2005) Organ specification-growth control connection: new insights

from the Drosophila eye-antennal disc. Dev Dyn 232:673-684.

Dong H, O'Brien RJ, Fung ET, Lanahan AA, Worley PF, Huganir RL (1997) GRIP: a synaptic

PDZ domain-containing protein that interacts with AMPA receptors. Nature 386:279-

284.

Duchon A, Herault Y (2016) DYRK1A, a dosage-sensitive gene involved in neurodevelopmental

disorders, is a target for drug development in Down syndrome. Front Behav Neurosci

10:104.

Dunn P, Prigatano G, Szelinger S, Roth J, Siniard A, Claasen A, Richholt R, De Both M,

Corneveaux J, Moskowitz A, Balak C, Piras I, Russell M, Courtright A, Belnap N,

Rangasamy S, Ramsey K, Opitz J, Craig D, Narayanan V, Huentelman M, I. Schrauwen I

(2017) A de novo splice site mutation in CASK causes FG syndrome-4 and congenital

nystagmus. Am J Med Genet Part A 173:611-617.

Edwards CA, Walker GK, Booth HD, Fredenburgh J (1979) The use of 2,2 dimethoxypropane in

rapid dehydration of tissues for light microscopy. J Histotechnol 2:114-116.

Edwards TJ, Sherr EH, Barkovich AJ, Richards LJ (2014) Clinical, genetic and imaging findings

identify new causes for corpus callosum development syndromes. Brain 137:1579-1613.

Egger B, Van Giesen L, Moraru M, Sprecher SG (2013) In vitro imaging of primary neural cell

culture from Drosophila. Nat Protoc 8:958-965.

152

Elgin SCR, Miller DW (1980) Mass rearing of flies and mass production and harvesting of

embryos. In The Genetics and Biology of Drosophila, Vol 2a. Academic Press; New

York:112–121.

Elias GM, Funke L, Stein V, Grant SG, Bredt DS, Nicoll RA (2006) Synapse-specific and

developmentally regulated targeting of AMPA receptors by a family of MAGUK

scaffolding proteins. Neuron 52:307–320.

Ethell IM, Yamaguchi Y (1999) Cell surface heparan sulfate proteoglycan syndecan-2 induces

the maturation of dendritic spines in rat hippocampal neurons. J Cell Biol 144:575-586.

Fallon L, Moreau F, Croft BG, Labib N, Gu WJ, Fon EA (2002) Parkin and CASK/LIN-2

associate via a PDZ-mediated interaction and are co-localized in lipid rafts and

postsynaptic densities in brain. J Biol Chem 277:486-491.

Fay MP, Proschan MA (2010) Wilcoxon-Mann-Whitney or t-test? On assumptions for

hypothesis tests and multiple interpretations of decision rules. Stat Surv 4:1-39.

Feng W, Long J-f, Zhang M (2005) A unified assembly mode revealed by the structures of

tetrameric L27 domain complexes formed by mLin-2/mLin-7 and Patj/Pals1 scaffold

proteins. Proc Natl Acad Sci USA 102:6861-6866.

Fiala JC, Feinberg M, Popov V, Harris KM (1998) Synaptogenesis via dendritic filopodia in

developing hippocampal area CA1. J Neurosci 18:8900-8911.

Fotaki V, Dierssen M, Alcantara S, Martinez S, Marti E, Casas C, Visa J, Soriano E, Estivill X,

Arbones ML (2002) Dyrk1A haploinsufficiency affects viability and causes

developmental delay and abnormal brain morphology in mice. Mol Cell Biol 22:6636-

6647.

153

Fox MH, Bonardi A, Krahn GL (2015) Expanding public health surveillance for people with

intellectual and developmental disabilities in the United States. Int Rev Res Dev Disabil

48:73-114.

Freissler E, Meyer auf der Heyde A, David G, Meyer TF, Dehio C (2000) Syndecan-1 and

syndecan-4 can mediate the invasion of OpaHSPG-expressing Neisseria gonorrhoeae

into epithelial cells. Cell Microbiol 2:69-82.

Frohlich A, Meinertzhagen IA (1983) Quantitative features of synapse formation in the fly's

visual system. I. The presynaptic photoreceptor terminal. J Neurosci 3:2336-2349.

Froyen G, Van Esch H, Bauters M, Hollanders K, Frints SG, Vermeesch JR, Devriendt K, Fryns

JP, Marynen P (2007) Detection of genomic copy number changes in patients with

idiopathic mental retardation by high-resolution X-array-CGH: important role for

increased gene dosage of XLMR genes. Hum Mutat 28:1034-1042.

Funke L, Dakoji S, Bredt DS (2005) Membrane-associated guanylate kinases regulate adhesion

and plasticity at cell junctions. Annu Rev Biochem 74:219-245.

Gatto CL, Broadie K (2011) Drosophila modeling of heritable neurodevelopmental disorders.

Curr Opin Neurobiol 21:834-841.

Geschwind DH, Levitt P (2007) Autism spectrum disorders: developmental disconnection

syndromes. Curr Opin Neurobiol 17:103-111.

Gillespie JM, Hodge JJ (2013) CASK regulates CaMKII autophosphorylation in neuronal

growth, calcium signaling, and learning. Front Mol Neurosci 6:27.

Gillette TA, Ascoli GA (2015) Topological characterization of neuronal arbor morphology via

sequence representation: I-motif analysis. BMC Bioinformatics 16:216.

154

Gilmore EC, Walsh CA (2013) Genetic causes of microcephaly and lessons for neuronal

development. Dev Biol 2:461-478.

Gordon J, Amini S, White MK (2013) General overview of neuronal cell culture. Methods Mol

Biol 1078:1-8.

Gottlob I (2000) Nystagmus. Curr Opin Ophthalmol 11:330-335.

Gramates LS, Marygold SJ, Santos Gd, Urbano J-M, Antonazzo G, Matthews BB, Rey AJ,

Tabone CJ, Crosby MA, Emmert DB, Falls K, Goodman JL, Hu Y, Ponting L, Schroeder

AJ, Strelets VB, Thurmond J, Zhou P, the FlyBase Consortium (2017) FlyBase at 25:

looking to the future. Nucleic Acids Res 45:D663-D671.

Griffith GM, Hastings RP, Nash S, Hill C (2010) Using matched groups to explore child

behavior problems and maternal well-being in children with Down syndrome and autism.

J Autism Dev Disord 40:610-619.

Guedj F, Pereira PL, Najas S, Barallobre MJ, Chabert C, Souchet B, Sebrie C, Verney C, Herault

Y, Arbones M, Delabar JM (2012) DYRK1A: a master regulatory protein controlling

brain growth. Neurobiol Dis 46:190-203.

Guernsey DL et al. (2010) Mutations in centrosomal protein CEP152 in primary microcephaly

families linked to MCPH4. Am J Med Genet 87:40-51.

Hackett A et al. (2010) CASK mutations are frequent in males and cause X-linked nystagmus and

variable XLMR phenotypes. Eur J Hum Genet 18:544-552.

Harris SR (2015) Measuring head circumference: Update on infant microcephaly. Can Fam

Physician 61:680-684.

155

Harris JC, Greenspan S (2016) Definition and nature of intellectual disability. In Handbook of

evidence-based practices in intellectual and developmental disabilities. Springer; New

York:11–39.

Hartman E, Houwen S, Scherder E, Visscher C (2010) On the relationship between motor

performance and executive functioning in children with intellectual disabilities. J Intellect

Disabil Res 54:468-477.

Hata Y, Butz S, Südhof TC (1996) CASK: a novel dlg/PSD95 homolog with an N-terminal

calmodulin-dependent protein kinase domain identified by interaction with neurexins. J

Neurosci 16:2488-2494.

Hayashi S, Okamoto N, Chinen Y, Takanashi J, Makita Y, Hata A, Imoto I, Inazawa J (2012)

Novel intragenic duplications and mutations of CASK in patients with mental retardation

and microcephaly with pontine and cerebellar hypoplasia (MICPCH). Hum Genet

131:99-110.

Hayashi S, Uehara DT, Tanimoto K, Mizuno S, Chinen Y, Fukumura S, Takanashi JI, Osaka H,

Okamoto N, Inazawa J (2017) Comprehensive investigation of CASK mutations and other

genetic etiologies in 41 patients with intellectual disability and microcephaly with

pontine and cerebellar hypoplasia (MICPCH). PloS One 12:e0181791.

Healy SD, Rowe C (2007) A critique of comparative studies of brain size. Proc R Soc Lond B

274:453-464.

Heisenberg M (2003) Mushroom body memoir: from maps to models. Nature Rev Neurosci

4:266-275.

Hell JW (2014) CaMKII: claiming center stage in postsynaptic function and organization.

Neuron 81:249-265.

156

Herculano-Houzel S (2009) The human brain in numbers: a linearly scaled-up primate brain.

Front Hum Neurosci 3:31.

Herculano-Houzel S, Messeder DJ, Fonseca-Azevedo K, Pantoja NA (2015) When larger brains

do not have more neurons: increased numbers of cells are compensated by decreased

average cell size across mouse individuals. Front Neuroanat 9:64.

Hing H, Xiao J, Harden N, Lim L, Zipursky SL (1999) Pak functions downstream of Dock to

regulate photoreceptor axon guidance in Drosophila. Cell 97:853-863.

Hinton VJ, Brown WT, Wisniewski K, Rudelli RD (1991) Analysis of neocortex in three males

with the fragile X syndrome. Am J Med Genet 41:289-294.

Hirth F (2010) Drosophila melanogaster in the study of human neurodegeneration. CNS Neurol

Disord Drug Targets 9:504-523.

Ho KS, Sehgal A (2005) Drosophila melanogaster: an insect model for fundamental studies of

sleep. Methods Enzymol 393:772-793.

Hong CJ, Hsueh YP (2006) CASK associates with glutamate receptor interacting protein and

signaling molecules. Biochem Biophys Res Commun 351:771-776.

Hong CJ, Hsueh YP (2007) Cytoplasmic distribution of T-box transcription factor Tbr-1 in adult

rodent brain. J Chem Neuroanat 33:124-130.

Hong ST, Mah W (2015) A critical role of GIT1 in vertebrate and invertebrate brain

development. Exp Neurobiol 24:8-16.

Hoskins R, Hajnal AF, Harp SA, Kim SK (1996) The C. elegans vulval induction gene lin-2

encodes a member of the MAGUK family of cell junction proteins. Development 122:97-

111.

157

Howard MA, Elias GM, Elias LA, Swat W, Nicoll RA (2010) The role of SAP97 in synaptic

glutamate receptor dynamics. Proc Natl Acad Sci USA 107:3805–3810.

Hsueh YP (2006) The role of the MAGUK protein CASK in neural development and synaptic

function. Curr Med Chem 13:1915-1927.

Hsueh YP (2009) Calcium/calmodulin-dependent serine protein kinase and mental retardation.

Ann Neurol 66:438-443.

Hsueh YP, Sheng M (1999) Regulated expression and subcellular localization of syndecan

heparan sulfate proteoglycans and the syndecan-binding protein CASK/LIN-2 during rat

brain development. J Neurosci19:7415-7425.

Hsueh YP, Wang TF, Yang FC, Sheng M (2000) Nuclear translocation and transcription

regulation by the membrane-associated guanylate kinase CASK/LIN-2. Nature 404:298-

302.

Hsueh YP, Yang FC, Kharazia V, Naisbitt S, Cohen AR, Weinberg RJ, Sheng M (1998) Direct

interaction of CASK/LIN-2 and syndecan heparan sulfate proteoglycan and their

overlapping distribution in neuronal synapses. J Cell Biol 142:139-151.

Hu HT, Shih PY, Shih YT, Hsueh YP (2016) The involvement of neuron-specific factors in

dendritic spinogenesis: molecular regulation and association with neurological disorders.

Neural Plast 2016:5136286.

Hu B-Y, Weick JP, Yu J, Ma L-X, Zhang X-Q, Thomson JA, Zhang S-C (2010) Neural

differentiation of human induced pluripotent stem cells follows developmental principles

but with variable potency. Proc Natl Acad Sci USA 107:4335-4340.

158

Huang TN, Chang HP, Hsueh YP (2010) CASK phosphorylation by PKA regulates the protein-

protein interactions of CASK and expression of the NMDAR2b gene. J Neurochem

112:1562-1573.

Huang N, Lee I, Marcotte EM, Hurles ME (2010) Characterising and predicting

haploinsufficiency in the human genome. PLoS Genet 6:e1001154.

Hutsler JJ, Zhang H (2010) Increased dendritic spine densities on cortical projection neurons in

autism spectrum disorders. Brain Res 1309:83-94.

Ilius M, Wolf R, Heisenberg M (1994) The central complex of Drosophila melanogaster is

involved in flight control: studies on mutants and mosaics of the gene ellipsoid body

open. J Neurogenet 9:189-206.

Inlow JK, Restifo LL (2004) Molecular and comparative genetics of mental retardation. Genetics

166:835-881.

Irwin SA, Patel B, Idupulapati M, Harris JB, Crisostomo RA, Larsen BP, Kooy F, Willems PJ,

Cras P, Kozlowski PB, Swain RA, Weiler IJ, Greenough WT (2001) Abnormal dendritic

spine characteristics in the temporal and visual cortices of patients with fragile-X

syndrome: a quantitative examination. Am J Med Genet 98:161-167.

Irwin SA, Idupulapati M, Gilbert ME, Harris JB, Chakravarti AB, Rogers EJ, Crisostomo RA,

Larsen BP, Mehta A, Alcantara CJ, Patel B, Swain RA, Weiler IJ, Oostra BA, Greenough

WT (2002) Dendritic spine and dendritic field characteristics of layer V pyramidal

neurons in the visual cortex of fragile-X knockout mice. Am J Med Genet 111:140-146.

Jay V, Becker LE, Chan FW, Perry TL, Sr. (1991) Puppet-like syndrome of Angelman: a

pathologic and neurochemical study. Neurology 41:416-422.

159

Jiang L, Kraft R, Restifo LL, Zohar Y (2015) Dissociation of brain tissue into viable single

neurons in a microfluidic device. IEEE Nano/Molecular Medicine & Engineering 9: 29-

32.

Jinnou H, Okanishi T, Enoki H, Ohki S (2012) Pontocerebellar hypoplasia type 3 with tetralogy

of Fallot. Brain Dev 34:392-395.

Kaas JH (2000) Why is brain size so important: design problems and solutions as neocortex gets

bigger or smaller. Brain Mind 1:7-23.

Kaech S, Banker G (2006) Culturing hippocampal neurons. Nat Protoc 1:2406-2415.

Kalscheuer VM et al. (2003) Mutations in the polyglutamine binding protein 1 gene cause X-

linked mental retardation. Nat Genet 35:313-315.

Kaufman L, Ayub M, Vincent JB (2010) The genetic basis of non-syndromic intellectual

disability: a review. J Neurodev Disord 2:182-209.

Kaufmann WE, Moser HW (2000) Dendritic anomalies in disorders associated with mental

retardation. Cereb Cortex 10:981-991.

Kaufmann N, DeProto J, Ranjan R, Wan H, Van Vactor D (2002) Drosophila liprin-alpha and

the receptor phosphatase Dlar control synapse morphogenesis. Neuron 34:27-38.

Keene AC, Waddell S (2007) Drosophila olfactory memory: single genes to complex neural

circuits. Nat Rev Neurosci 8:341-354.

Khanna R, Sun L, Li Q, Guo L, Stanley EF (2006) Long splice-variant N-type calcium channels

are clustered at presynaptic transmitter release sites without modular adaptor proteins.

Neurosci 138:1115-1125.

Konopka RJ, Benzer S (1971) Clock mutants of Drosophila melanogaster. Proc Natl Acad Sci

USA 68:2112-2116.

160

Koushika SP, Lisbin MJ, White K (1996) ELAV, a Drosophila neuron-specific protein, mediates

the generation of an alternatively spliced neural protein isoform. Curr Biol 6:1634-1641.

Kraft R, Levine RB, Restifo LL (1998) The steroid hormone 20-hydroxyecdysone enhances

neurite growth of Drosophila mushroom body neurons isolated during metamorphosis. J

Neurosci 18:8886-8899.

Kraft R, Escobar MM, Narro ML, Kurtis JL, Efrat A, Barnard K, Restifo LL (2006) Phenotypes

of Drosophila brain neurons in primary culture reveal a role for fascin in neurite shape

and trajectory. J Neurosci 26:8734-8747.

Kraft R, Kahn A, Medina-Franco JL, Orlowski ML, Baynes C, Lopez-Vallejo F, Barnard K,

Maggiora GM, Restifo LL (2013) A cell-based fascin bioassay identifies compounds with

potential anti-metastasis or cognition-enhancing functions. Dis Model Mech 6:217-235.

Kreis P, Barnier JV (2009) PAK signaling in neuronal physiology. Cell Signal 21:384-393.

Kulkarni VA, Firestein BL (2012) The dendritic tree and brain disorders. Mol Cell Neurosci

50:10-20.

Kuo TY, Hong CJ, Hsueh YP (2009) Bcl11A/CTIP1 regulates expression of DCC and MAP1b in

control of axon branching and dendrite outgrowth. Mol Cell Neurosci 42:195-207.

Kuo TY, Chen CY, Hsueh YP (2010a) Bcl11A/CTIP1 mediates the effect of the glutamate

receptor on axon branching and dendrite outgrowth. J Neurochem 114:1381-1392.

Kuo TY, Hong CJ, Chien HL, Hsueh YP (2010b) X-linked mental retardation gene CASK

interacts with Bcl11A/CTIP1 and regulates axon branching and outgrowth. J Neurosci

Res 88:2364-2373.

Kuzniar A, van Ham RC, Pongor S, Leunissen JA (2008) The quest for orthologs: finding the

corresponding gene across genomes. Trends in genetics 24:539-551.

161

LaConte L, Mukherjee K (2013) Structural constraints and functional divergences in CASK

evolution. Biochem Soc Trans 41:1017-1022.

LaConte LE, Chavan V, Liang C, Willis J, Schonhense EM, Schoch S, Mukherjee K (2016)

CASK stabilizes neurexin and links it to liprin-alpha in a neuronal activity-dependent

manner. Cell Mol Life Sci 73:3599-3621.

Lamb RS, Ward RE, Schweizer L, Fehon RG (1998) Drosophila coracle, a member of the

protein 4.1 superfamily, has essential structural functions in the septate junctions and

developmental functions in embryonic and adult epithelial cells. Mol Biol Cell 9:3505-

3519.

Laverty HG, Wilson JB (1998) Murine CASK is disrupted in a sex-linked cleft palate mouse

mutant. Genomics 53:29-41.

Leonard H, Wen X (2002) The epidemiology of mental retardation: challenges and opportunities

in the new millennium. Ment Retard Dev Disabil Res Rev 8:117-134.

Lewis S, (2015) Functions of Drosophila PAK (p21-activated kinase) in morphogenesis: a

mechanistic model based on cellular, molecular, and genetic studies. Doctoral

dissertation. University of Arizona, Tucson, Arizona.

Lim NR, Yeap YY, Zhao TT, Yip YY, Wong SC, Xu D, Ang CS, Williamson NA, Xu Z,

Bogoyevitch MA, Ng DC (2015) Opposing roles for JNK and Aurora A in regulating the

association of WDR62 with spindle microtubules. J Cell Sci 128:527-540.

Lin DM, Goodman CS (1994) Ectopic and increased expression of Fasciclin II alters motoneuron

growth cone guidance. Neuron 13:507-523.

Lin YL, Lei YT, Hong CJ, Hsueh YP (2007) Syndecan-2 induces filopodia and dendritic spine

formation via the neurofibromin-PKA-Ena/VASP pathway. J Cell Biol 177:829-841.

162

Lopes C, Gassanova S, Delabar JM, Rachidi M (2001) The CASK/Lin-2 Drosophila homologue,

Camguk, could play a role in epithelial patterning and in neuronal targeting. Biochem

Biophys Res Commun 284:1004-1010.

Lu CS, Hodge JJ, Mehren J, Sun XX, Griffith LC (2003) Regulation of the Ca2+/CaM-responsive

pool of CaMKII by scaffold-dependent autophosphorylation. Neuron 40:1185-1197.

Macpherson JN, Murray A (2016) Development of genetic testing for fragile x syndrome and

associated disorders, and estimates of the prevalence of FMR1 expansion mutations.

Genes 7:110.

Magen D, Ofir A, Berger L, Goldsher D, Eran A, Katib N, Nijem Y, Vlodavsky E, Tzur S, Behar

DM, Fellig Y, Mandel H (2015) Autosomal recessive lissencephaly with cerebellar

hypoplasia is associated with a loss-of-function mutation in CDK5. Hum Genet 134:305-

314.

Malik BR, Hodge JJ (2014) CASK and CaMKII function in Drosophila memory. Front Neurosci

8:178.

Malik BR, Gillespie JM, Hodge JJ (2013) CASK and CaMKII function in the mushroom body

alpha'/beta' neurons during Drosophila memory formation. Front Neural Circuits 7:52.

Margulies C, Tully T, Dubnau J (2005) Deconstructing memory in Drosophila. Curr Biol

15:R700-713.

Martin JR, Ollo R (1996) A new Drosophila Ca2+/calmodulin-dependent protein kinase (Caki)

is localized in the central nervous system and implicated in walking speed. Embo J

15:1865-1876.

Martin JR, Raibaud A, Ollo R (1994) Terminal pattern elements in Drosophila embryo induced

by the torso-like protein. Nature 367:741-745.

163

Maulik PK, Mascarenhas MN, Mathers CD, Dua T, Saxena S (2011) Prevalence of intellectual

disability: a meta-analysis of population-based studies. Res Dev Disabil 32:419-436.

Maximov A, Bezprozvanny I (2002) Synaptic targeting of N-type calcium channels in

hippocampal neurons. J Neurosci 22:6939-6952.

Maximov A, Südhof TC, Bezprozvanny I (1999) Association of neuronal calcium channels with

modular adaptor proteins. J Biol Chem 274:24453-24456.

Mazur-Kolecka B, Golabek A, Kida E, Rabe A, Hwang YW, Adayev T, Wegiel J, Flory M,

Kaczmarski W, Marchi E, Frackowiak J (2012) Effect of DYRK1A activity inhibition on

development of neuronal progenitors isolated from Ts65Dn mice. J Neurosci Res 90:999-

1010.

McBride SM, Choi CH, Wang Y, Liebelt D, Braunstein E, Ferreiro D, Sehgal A, Siwicki KK,

Dockendorff TC, Nguyen HT, McDonald TV, Jongens TA (2005) Pharmacological

rescue of synaptic plasticity, courtship behavior, and mushroom body defects in a

Drosophila model of fragile X syndrome. Neuron 45:753-764.

McCarroll SA, Altshuler DM (2007) Copy-number variation and association studies of human

disease. Genet 39:S37-42.

McKenzie K, Milton M, Smith G, Ouellette-Kuntz H (2016) Systematic review of the prevalence

and incidence of intellectual disabilities: current trends and issues. Curr Dev Disord Rep

3:104-115.

McManus KJ, Hendzel MJ (2001) CBP, a transcriptional coactivator and acetyltransferase. Int J

Biochem Cell Biol 79:253-266.

Michel CI, Kraft R, Restifo LL (2004) Defective neuronal development in the mushroom bodies

of Drosophila fragile X mental retardation 1 mutants. J Neurosci 24:5798-5809.

164

Missler M, Zhang W, Rohlmann A, Kattenstroth G, Hammer RE, Gottmann K, Südhof TC

(2003) Alpha-neurexins couple Ca2+ channels to synaptic vesicle exocytosis. Nature

423:939-948.

Mlakar J, Korva M, Tul N, Popovic M, Poljsak-Prijatelj M, Mraz J, Kolenc M, Resman Rus K,

Vesnaver Vipotnik T, Fabjan Vodusek V, Vizjak A, Pizem J, Petrovec M, Avsic Zupanc

T (2016) Zika Virus associated with microcephaly. N Engl J Med 374:951-958.

Mochida GH, Walsh CA (2001) Molecular genetics of human microcephaly. Curr Opin Neurol

14:151-156.

Moncini S, Castronovo P, Murgia A, Russo S, Bedeschi MF, Lunghi M, Selicorni A, Bonati MT,

Riva P, Venturin M (2016) Functional characterization of CDK5 and CDK5R1 mutations

identified in patients with non-syndromic intellectual disability. J Hum Genet 61:283-

293.

Moog U, Bierhals T, Brand K, Bautsch J, Biskup S, Brune T, Denecke J, de Die-Smulders CE,

Evers C, Hempel M, Henneke M, Yntema H, Menten B, Pietz J, Pfundt R, Schmidtke J,

Steinemann D, Stumpel CT, Van Maldergem L, Kutsche K (2015) Phenotypic and

molecular insights into CASK-related disorders in males. Orphanet J Rare Dis 10:44.

Moog U, Kutsche K, Kortüm F, Chilian B, Bierhals T, Apeshiotis N, Balg S, Chassaing N,

Coubes C, Das S, Engels H, Van Esch H, Grasshoff U, Heise M, Isidor B, Jarvis J,

Koehler U, Martin T, Oehl-Jaschkowitz B, Ortibus E, Pilz D, Prabhakar P, Rappold G,

Rau I, Rettenberger G, Schlüter G, Scott R, Shoukier M, Wohlleber E, Zirn B, Dobyns

W, Uyanik G (2011) Phenotypic spectrum associated with CASK loss-of-function

mutations. Journal of medical genetics 48:741-751.

165

Mukherjee K, Slawson JB, Christmann BL, Griffith LC (2014) Neuron-specific protein

interactions of Drosophila CASK-beta are revealed by mass spectrometry. Front Mol

Neurosci 7:58.

Mukherjee K, Sharma M, Urlaub H, Bourenkov GP, Jahn R, Südhof TC, Wahl MC (2008)

CASK Functions as a Mg2+-independent neurexin kinase. Cell 133:328-339.

Naiche LA, Harrelson Z, Kelly RG, Papaioannou VE (2005) T-box genes in vertebrate

development. Annu Rev Genet 39:219-239.

Najm J, Horn D, Wimplinger I, Golden JA, Chizhikov VV, Sudi J, Christian SL, Ullmann R,

Kuechler A, Haas CA, Flubacher A, Charnas LR, Uyanik G, Frank U, Klopocki E,

Dobyns WB, Kutsche K (2008) Mutations of CASK cause an X-linked brain

malformation phenotype with microcephaly and hypoplasia of the brainstem and

cerebellum. Genet 40:1065-1067.

Nakamura K, Nishiyama K, Kodera H, Nakashima M, Tsurusaki Y, Miyake N, Matsumoto N,

Saitsu H, Jinnou H, Ohki S, Yokochi K, Okanishi T, Enoki H (2014) A de novo CASK

mutation in pontocerebellar hypoplasia type 3 with early myoclonic epilepsy and

tetralogy of Fallot. Brain Dev 36:272-273.

Narayanan AS, Rothenfluh A (2016) I believe I can fly!: use of Drosophila as a model organism

in neuropsychopharmacology research. Neuropsychopharmacol 41:1439-1446.

Narro ML, Yang F, Kraft R, Wenk C, Efrat A, Restifo LL (2007) NeuronMetrics: software for

semi-automated processing of cultured neuron images. Brain Res 1138:57-75.

Nassogne MC, Evrard P, Courtoy PJ (1995) Selective neuronal toxicity of cocaine in embryonic

mouse brain cocultures. Proc Natl Acad Sci USA 92:11029-11033.

166

Nestler EJ, Hyman SE (2010) Animal models of neuropsychiatric disorders. Nat Neurosci

13:1161-1169.

Ohno H, Hirabayashi S, Kansaku A, Yao I, Tajima M, Nishimura W, Ohnishi H, Mashima H,

Fujita T, Omata M, Hata Y (2003) Carom: a novel membrane-associated guanylate

kinase-interacting protein with two SH3 domains. Oncogene 22:8422-8431.

Okamoto M, Südhof TC (1997) Mints, Munc18-interacting proteins in synaptic vesicle

exocytosis. J Biol Chem 272:31459-31464.

Oliva C, Escobedo P, Astorga C, Molina C, Sierralta J (2012) Role of the MAGUK protein

family in synapse formation and function. Dev Neurobiol 72:57-72.

Olsen O, Moore KA, Fukata M, Kazuta T, Trinidad JC, Kauer FW, Streuli M, Misawa H,

Burlingame AL, Nicoll RA, Bredt DS (2005) Neurotransmitter release regulated by a

MALS-liprin-alpha presynaptic complex. J Biol Chem 170:1127-1134.

Oortveld MA, Keerthikumar S, Oti M, Nijhof B, Fernandes AC, Kochinke K, Castells-Nobau A,

van Engelen E, Ellenkamp T, Eshuis L, Galy A, van Bokhoven H, Habermann B,

Brunner HG, Zweier C, Verstreken P, Huynen MA, Schenck A (2013) Human

intellectual disability genes form conserved functional modules in Drosophila. PLoS

Genet 9:e1003911.

Opitz JM, Kaveggia EG (1974) The FG syndrome. An X-linked recessive syndrome of multiple

congenital anomalies and mental retardation. Z Kinderheilkd 117:1-18.

Pandey UB, Nichols CD (2011) Human disease models in Drosophila melanogaster and the role

of the fly in therapeutic drug discovery. Pharmacol Rev 63:411-436.

Parrish JZ, Emoto K, Kim MD, Jan YN (2007) Mechanisms that regulate establishment,

maintenance, and remodeling of dendritic fields. Annu Rev Neurosci 30:399-423.

167

Passemard S, Kaindl AM, Verloes A (2013) Microcephaly. Handb Clin Neurol 111:129-141.

Pehlivan D, Hullings M, Carvalho CM, Gonzaga-Jauregui CG, Loy E, Jackson LG, Krantz ID,

Deardorff MA, Lupski JR (2012) NIPBL rearrangements in Cornelia de Lange syndrome:

evidence for replicative mechanism and genotype-phenotype correlation. Genet Med

14:313-322.

Peippo M, Koivisto AM, Sarkamo T, Sipponen M, von Koskull H, Ylisaukko-oja T, Rehnstrom

K, Froyen G, Ignatius J, Jarvela I (2007) PAK3-related mental disability: further

characterization of the phenotype. Am J Med Genet Part A 143A:2406-2416.

Pereanu W, Shy D, Hartenstein V (2005) Morphogenesis and proliferation of the larval brain glia

in Drosophila. Dev Biol 283:191-203.

Pereanu W, Kumar A, Jennett A, Reichert H, Hartenstein V (2010) Development-based

compartmentalization of the Drosophila central brain. J Comp Neurol 518:2996-3023.

Perez-Perez JM, Candela H, Micol JL (2009) Understanding synergy in genetic interactions.

Trends in genetics 25:368-376.

Picker JD, Walsh CA (2013) New innovations: therapeutic opportunities for intellectual

disabilities. Ann Neurol 74:382-390.

Piluso G, Carella M, D'Avanzo M, Santinelli R, Carrano EM, D'Avanzo A, D'Adamo AP,

Gasparini P, Nigro V (2003) Genetic heterogeneity of FG syndrome: a fourth locus

(FGS4) maps to Xp11.4-p11.3 in an Italian family. Hum Genet 112:124-130.

Piluso G, D'Amico F, Saccone V, Bismuto E, Rotundo IL, Di Domenico M, Aurino S, Schwartz

CE, Neri G, Nigro V (2009) A missense mutation in CASK causes FG syndrome in an

Italian family. Am J Med Genet 84:162-177.

168

Puram SV, Bonni A (2013) Cell-intrinsic drivers of dendrite morphogenesis. Development

140:4657-4671.

Purpura DP (1974) Dendritic spine "dysgenesis" and mental retardation. Science 186:1126-1128.

Ray B, Bailey JA, Sarkar S, Lahiri DK (2009) Molecular and immunocytochemical

characterization of primary neuronal cultures from adult rat brain: Differential expression

of neuronal and glial protein markers. J Neurosci Methods 184:294-302.

Raymond FL, Tarpey P (2006) The genetics of mental retardation. Hum Mol Genet 2: R110-116.

Reeves RG, Tautz D (2017) Automated phenotyping indicates pupal size in Drosophila is a

highly heritable trait with an apparent polygenic basis. G3 7:1277-1286.

Reid R (2012) From functional architecture to functional connectomics. Neuron 75: 209-217.

Restifo LL (2005) Mental retardation genes in Drosophila: new approaches to understanding and

treating developmental brain disorders. Ment Retard Dev Disabil Res Rev 11:286-294.

Restifo LL, White K (1991) Mutations in a steroid hormone-regulated gene disrupt the

metamorphosis of the central nervous system in Drosophila. Dev Biol 148:174-194.

Rice AM, McLysaght A (2017) Dosage-sensitive genes in evolution and disease. BMC Biol

15:78.

Richtsmeier JT, Flaherty K (2013) Hand in glove: brain and skull in development and

dysmorphogenesis. Acta Neuropathol 125:469-489.

Rideout EJ, Narsaiya MS, Grewal SS (2015) The sex determination gene transformer regulates

male-female differences in Drosophila body size. PLoS Genet 11: e1005683.

Riveros AJ, Gronenberg W (2010) Brain allometry and neural plasticity in the bumblebee

Bombus occidentalis. Brain Behav Evol 75:138-148.

169

Robinow S, White K (1988) The locus elav of Drosophila melanogaster is expressed in neurons

at all developmental stages. Dev Biol 126:294-303.

Robinson TE, Kolb B (2004) Structural plasticity associated with exposure to drugs of abuse.

Neuropharmacol 47 Suppl 1:33-46.

Roelfsema JH, White SJ, Ariyürek Y, Bartholdi D, Niedrist D, Papadia F, Bacino CA,

den Dunnen JT, van Ommen G-JB, Breuning MH, Hennekam RC, Peters DJM (2005)

Genetic Heterogeneity in Rubinstein-Taybi Syndrome: Mutations in Both the CBP and

EP300 Genes Cause Disease. Am J Hum Genet 76:572-580.

Rosato E, Kyriacou CP (2006) Analysis of locomotor activity rhythms in Drosophila. Nat Protoc

1:559-568.

Roth G, Dicke U (2005) Evolution of the brain and intelligence. Trends Cogn Sci 9:250-257.

Rujano MA, Sanchez-Pulido L, Pennetier C, le Dez G, Basto R (2013) The microcephaly protein

Asp regulates neuroepithelium morphogenesis by controlling the spatial distribution of

myosin II. Nat Cell Biol 15:1294-1306.

Rushton JP, Ankney CD (2009) Whole brain size and general mental ability: a review. The

International Journal of Neuroscience 119:692-732.

Saitsu H, Kato M, Osaka H, Moriyama N, Horita H, Nishiyama K, Yoneda Y, Kondo Y,

Tsurusaki Y, Doi H, Miyake N, Hayasaka K, Matsumoto N (2012) CASK aberrations in

male patients with Ohtahara syndrome and cerebellar hypoplasia. Epilepsia 53:1441-

1449.

Sampathkumar P, Sanchez JL (2016) Zika virus in the Americas: a review for clinicians. Mayo

Clin Proc 91:514-521.

170

Samuels BA, Hsueh YP, Shu T, Liang H, Tseng HC, Hong CJ, Su SC, Volker J, Neve RL, Yue

DT, Tsai LH (2007) Cdk5 promotes synaptogenesis by regulating the subcellular

distribution of the MAGUK family member CASK. Neuron 56:823-837.

Saunders RD, Avides MC, Howard T, Gonzalez C, Glover DM (1997) The Drosophila gene

abnormal spindle encodes a novel microtubule-associated protein that associates with the

polar regions of the mitotic spindle. J Cell Biol 137:881-890.

Schalock RL, Luckasson RA, Shogren KA, Borthwick-Duffy S, Bradley V, Buntinx WH,

Coulter DL, Craig EM, Gomez SC, Lachapelle Y, Reeve A, Snell ME, Spreat S, Tasse

MJ, Thompson JR, Verdugo MA, Wehmeyer ML, Yeager MH (2007) The renaming of

mental retardation: understanding the change to the term intellectual disability. Intellect

Dev Disabil 45:116-124.

Schneider CA, Rasband WS, Eliceiri KW (2012) NIH Image to ImageJ: 25 years of image

analysis. Nature Methods 9:671-675.

Seltzer LE, Paciorkowski AR (2014) Genetic disorders associated with postnatal microcephaly.

Am J Med Genet C Semin Med Genet 166C:140-155.

Setou M, Nakagawa T, Seog DH, Hirokawa N (2000) Kinesin superfamily motor protein KIF17

and mLin-10 in NMDA receptor-containing vesicle transport. Science 288:1796-1802.

Shamseldin H, Alazami AM, Manning M, Hashem A, Caluseiu O, Tabarki B, Esplin E, Schelley

S, Innes AM, Parboosingh JS, Lamont R, Majewski J, Bernier FP, Alkuraya FS (2015)

RTTN mutations cause primary microcephaly and primordial dwarfism in humans. Am

J Hum Genet 97:862-868.

Sherwood CC, Subiaul F, Zawidzki TW (2008) A natural history of the human mind: tracing

evolutionary changes in brain and cognition. J Anat 212:426-454.

171

Shingleton AW, Mirth CK, Bates PW (2008) Developmental model of static allometry in

holometabolous insects. Proc Biol Sci 275:1875-1885.

Slawson JB, Kuklin EA, Ejima A, Mukherjee K, Ostrovsky L, Griffith LC (2011) Central

regulation of locomotor behavior of Drosophila melanogaster depends on a CASK

isoform containing CaMK-like and L27 domains. Genet 187:171-184.

Smrt RD, Lewis S A, Kraft R, Restifo L L (2015) Primary culture of Drosophila larval neurons

with morphological analysis using NeuronMetrics. Drosophila Inf Serv 98:125-140.

Soto-Ares G, Joyes B, Lemaitre MP, Vallee L, Pruvo JP (2003) MRI in children with mental

retardation. Pediatr Radiol 33:334-345.

Srivastava S, McMillan R, Willis J, Clark H, Chavan V, Liang C, Zhang H, Hulver M,

Mukherjee K (2016) X-linked intellectual disability gene CASK regulates postnatal brain

growth in a non-cell autonomous manner. Acta Neuropathol Commun 4:30.

Stafford RL, Ear J, Knight MJ, Bowie JU (2011) The molecular basis of the Caskin1 and Mint1

interaction with CASK. J Mol Biol 412:3-13.

Stevenson RE, Schwartz CE (2009) X-linked intellectual disability: unique vulnerability of the

male genome. Dev Disabil Res Rev 15:361-368.

Stevenson RE, Bennett CW, Abidi F, Kleefstra T, Porteous M, Simensen RJ, Lubs HA, Hamel

BC, Schwartz CE (2005) Renpenning syndrome comes into focus. Am J Med Genet A

134:415-421.

Stiles J, Jernigan TL (2010) The basics of brain development. Neuropsychol Rev 20:327-348.

Stillwell RC, Blanckenhorn WU, Teder T, Davidowitz G, Fox CW (2010) Sex differences in

phenotypic plasticity affect variation in sexual size dimorphism in insects: from

physiology to evolution. Annu Rev Entomol 55:227-245.

172

Strausfeld NJ (1999) A brain region in insects that supervises walking. Prog Brain Res 123:273-

284.

Strauss R, Heisenberg M (1993) A higher control center of locomotor behavior in the Drosophila

brain. J Neurosci 13:1852-1861.

Südhof TC (2008) Neuroligins and neurexins link synaptic function to cognitive disease. Nature

455:903-911.

Sun B, Salvaterra PM (1995) Characterization of nervana, a Drosophila melanogaster neuron-

specific glycoprotein antigen recognized by anti-horseradish peroxidase antibodies. J

Neurochem 65:434-443.

Sun M, Liu L, Zeng X, Xu M, Liu L, Fang M, Xie W (2009) Genetic interaction between

Neurexin and CAKI/CMG is important for synaptic function in Drosophila

neuromuscular junction. Neurosci Res 64:362-371.

Tabuchi K, Biederer T, Butz S, Südhof TC (2002) CASK participates in alternative tripartite

complexes in which Mint 1 competes for binding with caskin 1, a novel CASK-binding

protein. J Neurosci 22:4264-4273.

Takanashi J, Okamoto N, Yamamoto Y, Hayashi S, Arai H, Takahashi Y, Maruyama K, Mizuno

S, Shimakawa S, Ono H, Oyanagi R, Kubo S, Barkovich AJ, Inazawa J (2012) Clinical

and radiological features of Japanese patients with a severe phenotype due to CASK

mutations. Am J Med Genet A 158a:3112-3118.

Tan TC, Valova VA, Malladi CS, Graham ME, Berven LA, Jupp OJ, Hansra G, McClure SJ,

Sarcevic B, Boadle RA, Larsen MR, Cousin MA, Robinson PJ (2003) Cdk5 is essential

for synaptic vesicle endocytosis. Nat Cell Biol 5:701-710.

173

Tang YC, Amon A (2013) Gene copy-number alterations: a cost-benefit analysis. Cell 152:394-

405.

Tarpey PS et al. (2009) A systematic, large-scale resequencing screen of X-chromosome coding

exons in mental retardation. Genet 41:535-543.

Tauber JM, Vanlandingham PA, Zhang B (2011) Elevated levels of the vesicular monoamine

transporter and a novel repetitive behavior in the Drosophila model of fragile X

syndrome. PloS One 6: e27100.

Tejedor FJ, Hammerle B (2011) MNB/DYRK1A as a multiple regulator of neuronal development.

FEBS J 278:223-235.

Tejedor F, Zhu XR, Kaltenbach E, Ackermann A, Baumann A, Canal I, Heisenberg M,

Fischbach KF, Pongs O (1995) Minibrain: a new protein kinase family involved in

postembryonic neurogenesis in Drosophila. Neuron 14:287-301.

The I, Hannigan GE, Cowley GS, Reginald S, Zhong Y, Gusella JF, Hariharan IK, Bernards A

(1997) Rescue of a Drosophila NF1 mutant phenotype by protein kinase A. Science

276:791-794.

Thorpe JR, Harvey DM (1979) Optimization and investigation of the use of 2,2-

dimethoxypropane as a dehydration agent for plant tissues in transmission electron

microscopy. J Ultrastruct Res 68:186-194.

Tissot M, Stocker RF (2000) Metamorphosis in Drosophila and other insects: the fate of neurons

throughout the stages. Prog Neurobiol 62:89-111.

Tkacs NC, Thompson HJ (2006) From bedside to bench and back again: research issues in

animal models of human disease. Biol Res Nurs 8:78-88.

174

Truman JW (1990) Metamorphosis of the central nervous system of Drosophila. J Neurobiol

21:1072-1084.

Tucci V et al. (2014) Dominant β-catenin mutations cause intellectual disability with

recognizable syndromic features. J Clin Invest 124:1468-1482.

Urbach R (2003) The pattern of neuroblast formation, mitotic domains and proneural gene

expression during early brain development in Drosophila. Development 130:3589-3606.

Vaillend C, Poirier R, Laroche S (2008) Genes, plasticity and mental retardation. Behav Brain

Res 192:88-105.

Van der Voet M, Nijhof B, Oortveld MA, Schenck A (2014) Drosophila models of early onset

cognitive disorders and their clinical applications. Neurosci Biobehav Rev 2:326-342.

Van Elburg RA, Van Ooyen A (2010) Impact of dendritic size and dendritic topology on burst

firing in pyramidal cells. PLoS Comput Biol 6:e1000781.

Van Maldergem L, Hou Q, Kalscheuer VM, Rio M, Doco-Fenzy M, Medeira A, de Brouwer AP,

Cabrol C, Haas SA, Cacciagli P, Moutton S, Landais E, Motte J, Colleaux L, Bonnet C,

Villard L, Dupont J, Man HY (2013) Loss of function of KIAA2022 causes mild to severe

intellectual disability with an autism spectrum disorder and impairs neurite outgrowth.

Hum Mol Genet 22:3306-3314.

Varoqueaux F, Aramuni G, Rawson RL, Mohrmann R, Missler M, Gottmann K, Zhang W,

Südhof TC, Brose N (2006) Neuroligins determine synapse maturation and function.

Neuron 51:741-754.

Venken KJ, Simpson JH, Bellen HJ (2011) Genetic manipulation of genes and cells in the

nervous system of the fruit fly. Neuron 72:202-230.

175

Visscher PM, McEvoy B, Yang J (2010) From Galton to GWAS: quantitative genetics of human

height. Genet Res 92:371-379.

Vissers LE, de Vries BB, Veltman JA (2010) Genomic microarrays in mental retardation: from

copy number variation to gene, from research to diagnosis. Am J Med Genet 47:289-297.

Von der Hagen M, Pivarcsi M, Liebe J, von Bernuth H, Didonato N, Hennermann JB, Buhrer C,

Wieczorek D, Kaindl AM (2014) Diagnostic approach to microcephaly in childhood: a

two-center study and review of the literature. Dev Med Child Neurol 56:732-741.

Wakefield JG, Bonaccorsi S, Gatti M (2001) The Drosophila protein ASP is involved in

microtubule organization during spindle formation and cytokinesis. J Cell Biol 153:637-

648.

Wan L, Dockendorff TC, Jongens TA, Dreyfuss G (2000) Characterization of dFMR1, a

Drosophila melanogaster homolog of the fragile X mental retardation protein. Mol Cell

Biol 20:8536-8547.

Wang H, Doering LC (2012) Induced pluripotent stem cells to model and treat neurogenetic

disorders. Neural Plast 2012:346053.

Wang Q, Moore MJ, Adelmant G, Marto JA, Silver PA (2013) PQBP1, a factor linked to

intellectual disability, affects alternative splicing associated with neurite outgrowth.

Genes Dev 27:615-626.

Wang Y, Li R, Du D, Zhang C, Yuan H, Zeng R, Chen Z (2006) Proteomic analysis reveals

novel molecules involved in insulin signaling pathway. J Proteome Res 5:846-855.

Wang TF, Ding CN, Wang GS, Luo SC, Lin YL, Ruan Y, Hevner R, Rubenstein JL, Hsueh YP

(2004a) Identification of Tbr-1/CASK complex target genes in neurons. J Neurochem

91:1483-1492.

176

Wang GS, Hong CJ, Yen TY, Huang HY, Ou Y, Huang TN, Jung WG, Kuo TY, Sheng M,

Wang TF, Hsueh YP (2004b) Transcriptional modification by a CASK-interacting

nucleosome assembly protein. Neuron 42:113-128.

Ward REt, Schweizer L, Lamb RS, Fehon RG (2001) The protein 4.1, ezrin, radixin, moesin

(FERM) domain of Drosophila Coracle, a cytoplasmic component of the septate junction,

provides functions essential for embryonic development and imaginal cell proliferation.

Genet 159:219-228.

Watkins RJ, Patil R, Goult BT, Thomas MG, Gottlob I, Shackleton S (2013) A novel interaction

between FRMD7 and CASK: evidence for a causal role in idiopathic infantile nystagmus.

Hum Mol Genet 22:2105-2118.

Wei Z, Zheng S, Spangler SA, Yu C, Hoogenraad CC, Zhang M (2011) Liprin-mediated large

signaling complex organization revealed by the liprin-alpha/CASK and liprin-

alpha/liprin-beta complex structures. Mol Cell 43:586-598.

Wei JL, Fu ZX, Fang M, Zhou QY, Zhao QN, Guo JB, Lu WD, Wang H (2014) High expression

of CASK correlates with progression and poor prognosis of colorectal cancer. Tumour

Biol 35:9185-9194.

Wen Q, Chklovskii DB (2008) A cost-benefit analysis of neuronal morphology. J Neurophysiol

99:2320-2328.

Whitworth AJ, Theodore DA, Greene JC, Benes H, Wes PD, Pallanck LJ (2005) Increased

glutathione S-transferase activity rescues dopaminergic neuron loss in a Drosophila

model of Parkinson's disease. Proc Natl Acad Sci USA 102:8024-8029.

Wigglesworth VB (1957) The use of osmium in the fixation and staining of tissues. Proc R Soc

Lond B 147:185-199.

177

Wilson JB, Ferguson MW, Jenkins NA, Lock LF, Copeland NG, Levine AJ (1993) Transgenic

mouse model of X-linked cleft palate. Cell Growth Differ 4:67-76.

Wolff T, Iyer NA, Rubin GM (2015) Neuroarchitecture and neuroanatomy of the Drosophila

central complex: A GAL4-based dissection of protocerebral bridge neurons and circuits.

J Comp Neurol 523:997-1037.

Won H, Mah W, Kim E, Kim JW, Hahm EK, Kim MH, Cho S, Kim J, Jang H, Cho SC, Kim

BN, Shin MS, Seo J, Jeong J, Choi SY, Kim D, Kang C, Kim E (2011) GIT1 is

associated with ADHD in humans and ADHD-like behaviors in mice. Nat Med 17:566-

572.

Woods CG (2004) Human microcephaly. Curr Opin Neurobiol 14:112-117.

Wu Y, Bolduc FV, Bell K, Tully T, Fang Y, Sehgal A, Fischer JA (2008) A Drosophila model

for Angelman syndrome. Proc Natl Acad Sci USA 105:12399-12404.

Xie N, Tang B (2016) The application of human IPSCs in neurological diseases: from bench to

bedside. Stem Cells Int 2016:6484713.

Yang Y, Baboolal TG, Siththanandan V, Chen M, Walker ML, Knight PJ, Peckham M, Sellers

JR (2009) A FERM domain autoregulates Drosophila myosin 7a activity. Proc Natl Acad

Sci USA 106:4189-4194.

Zhang H, Webb DJ, Asmussen H, Horwitz AF (2003) Synapse formation is regulated by the

signaling adaptor GIT1. J Cell Biol 161:131-142.

Zhang F, Gu W, Hurles ME, Lupski JR (2009) Copy number variation in human health, disease,

and evolution. Annu Rev Genomics Hum Genet 10:451-481.

178

Zhang H, Webb DJ, Asmussen H, Niu S, Horwitz AF (2005) A GIT1/PIX/Rac/PAK signaling

module regulates spine morphogenesis and synapse formation through MLC. J Neurosci

25:3379-3388.

Zordan MA, Massironi M, Ducato MG, Te Kronnie G, Costa R, Reggiani C, Chagneau C,

Martin JR, Megighian A (2005) Drosophila CAKI/CMG protein, a homolog of human

CASK, is essential for regulation of neurotransmitter vesicle release. J Neurophysiol

94:1074-1083.

179