GEMIN FUNCTION IN SMALL NUCLEAR RNP BIOGENESIS
By
KARL BRYAN SHPARGEL
Submitted in partial fulfillment of the requirements
for the degree of Doctor of Philosophy
Thesis Advisor: Dr. A. Gregory Matera
Department of Genetics
CASE WESTERN RESERVE UNIVERSITY
August, 2006 CASE WESTERN RESERVE UNIVERSITY
SCHOOL OF GRADUATE STUDIES
We hereby approve the dissertation of
______
candidate for the Ph.D. degree *.
(signed)______(chair of the committee)
______
______
______
______
______
(date) ______
*We also certify that written approval has been obtained for any proprietary material contained therein.
Table of Contents
List of figures………………………………………………………………………….....6
Acknowledgements………………………………………………………………...... …..8
Abbreviations………………………………………………………………………..…...9
Abstract………………………………………………………………………………….10
Chapter I: Introduction and Research Objectives………………………...…………12
Introduction………………………………………………………………………13
Spinal Muscular Atrophy…………………………………………...……13
The SMN complex……………………………………………………….16
snRNP biogenesis……………………………………………………..…18
The Cajal body…………………………………………………………...22
Dilemmas in SMA pathogenesis and snRNP biogenesis…………...……26
Research Objectives……………………………………………………...………29
Chapter II: Gemin proteins are required for efficient assembly of Sm-class RNPs………………………………………………………….30
Abstract………………………………………………………………………..…31
Introduction………………………………………………………………………32
Materials and Methods…………………………………………………………...35
Results and Discussion…………………………………………………………..38
3 Chapter III: Drosophila Gemin3 is essential for larval motor function, pupation, and viability………………………………….……56
Abstract…………………………………………………………………………..57
Introduction…………………………………………………………………...….58
Materials and Methods………………………………………………………...…62
Results………………………………………………………………………...….65
Discussion……………………………………………………………..…………85
Chapter IV: Discussion and future directions………………………………..………88
What factors regulate Cajal body homeostasis?...... 89
Gemin proteins function in snRNP assembly, but by what mechanism?...... 91
Are Gemins required for additional snRNP biogenesis steps?...... 94
Can Gemin snRNP biogenesis defects promote Spinal Muscular Atrophy?...... 95
Which SMN functions are required for appropriate neuromuscular development?...... 97
Does the primary SMA defect originate in neurons or muscle (or both)? ...... 98
Why does snRNP biogenesis have such a profound effect on motor neurons?...... 103
Appendix Chapter I: Control of Cajal body number is mediated by the coilin C-terminus……………………………………...108
Abstract…………………………………………………………………...….…109
Introduction……………………………………………………………….....….110
Materials and Methods……………………………………………………...…..114
Results…………………………………………………………………………..115
Discussion……………………………………………………………………....134
4 Appendix Chapter II: Exogenous Gemin4 expression enhances Gemin3 and SMN nuclear localization…………….……….138
Abstract…………………………………………………………...…………….139
Materials and Methods…………………………………….……………………140
Results and Discussion…………………………………………………………141
Bibliography…………………………………………………..……………………….147
5 List of Figures
Chapter I
Figure 1-1: Mutations in SMN1 result in SMA…………………………………..14
Figure 1-2: SMN forms a large macromolecular complex that interacts with Sm proteins………………………………………17
Figure 1-3: SMN protein domain structure………………………………………17
Figure 1-4: snRNP biogenesis overview…………………………………………21
Figure 1-5: The Cajal body functions in snRNP, snoRNP, and teRNP maturation and assembly………………………………..25
Figure 1-6: SMN is involved in several cellular functions………………………28
Chapter II
Figure 2-1: SMN and Gemin protein levels are interdependent…………..……..39
Figure 2-2: SMN, Gemin2, Gemin3, and Gemin4 are required for efficient Sm core assembly…………………………………..….41
Figure 2-3: Loss of SMN and snurportin results in breakdown of nuclear Cajal bodies………………………………………………44
Figure 2-4: SMN and Gemins regulate Sm core assembly in vivo………………48
Figure 2-5: SMA patient-derived SMN mutations are defective in Sm core assembly……………………………………………..…52
Figure 2-6: Supplemental figure…………………………………………………54
Chapter III
Figure 3-1: Drosophila Gemin3 is conserved throughout helicase motifs..…..…66
Figure 3-2: dGemin3 interacts with dSMN in vitro and in vivo…………………68
Figure 3-3: Drosophila SMN and Gemin3 are required for efficient snRNA Sm core assembly…………………………...…….70
Figure 3-4: Smn and Gemin3 mutant larvae exhibit viability and pupation defects………………………………………...………74
6 Figure 3-5: Smn and Gemin3 mutants exhibit growth defects……………..…….77
Figure 3-6: Smn and Gemin3 mutant larvae exhibit defects in motor function…………………………………………………….79
Figure 3-7: Smn and Gemin3 mutants exhibit neuronal pathfinding defects…………………………………………..………80
Figure 3-8: Smnex33, SmD2, and Gemin3 interact genetically……………...…….83
Chapter IV
Figure 4-1: Models for SMN complex function in snRNP Sm core assembly……………………………………..……...……..94
Figure 4-2: Experimental analysis of tissue specificity in SMA………….……103
Figure 4-3: Models for tissue specific SMA pathogenesis………………..……107
Chapter AI
Figure A1-1: Deletion of the coilin self-association domain affects epitope recognition………………………………….……116
Figure A1-2: Coilin mutants display unregulated nuclear body formation…………………………………………………...119
Figure A1-3: Coilins from different species display variation in nuclear body formation…………………………………….….122
Figure A1-4: Schematic of human and mouse coilin………………...…………123
Figure A1-5: Mutations in mouse coilin affect nuclear body formation……….127
Figure A1-6: Ectopically expressed SMN produces numerous SMN foci……..132
Chapter AII
Figure A2-1: Gemin3 overexpression localizes SMN to the Golgi complex…..142
Figure A2-2: Golgi dissociation alters Gemin3 localization………………...…143
Figure A2-3: Gemin4 overexpression imports Gemin3 and SMN into the nucleus……………………………...………..……145
7 Acknowledgements
I would like to take this opportunity to thank my research advisor, Greg Matera.
Greg’s enthusiastic approach to research has left a very positive impression on me. There were times when I doubted whether I would ever be able to finish my project, but Greg was always convinced that the end was somewhere within reach. He has been a great mentor and I will always cherish our discussions about research, life, or football. I also would like to thank all members of the Matera Lab who have been instrumental in my training as well as helping me keep my sanity. Thank you to my thesis committee for their advise in the direction of my project.
I would like to thank my parents, Jerold and Shirley, and sisters, Sara and
Rebecca, for their unending support and inspiration to follow my dreams in life.
Finally and perhaps most importantly, I would like to thank my wife, Tarah. She
has put up with all the late worknights, weekend trips to the lab, and life with a graduate
student stipend. Throughout the tumultuous graduate school life, she has been my smile
at the end of the day and continued to give me confidence that this graduation day will
come. And a special thanks to her for giving me my son Jack, a fantastic graduation gift!
8 List of abbreviations
ATP Adenosine triphosphate CB Cajal body CBC Cap binding complex DAPI 4′,6-diamidino-2-phenylindole DTT Dithiothreitol EDTA Ethylenediaminetetraacetic acid Gem3 Gemin3 GFP Green fluorescent protein GST Glutathione S-transferase GTP Guanosine triphosphate His Histidine IP Immunoprecipitation IPTG Isopropyl-beta-D-Thiogalactopyranoside m7G 7-methylguanosine mRIPA Modified radioimmunoprecipitaion assay myc c-Myc epitope NLS Nuclear localization signal NP-40 Nonidet-40 NPC Nuclear pore complex PBS Phosphate buffered saline PCR Polymerase chain reaction PHAX Phosphorylated adaptor for snRNA export PRMT5 Protein arginine methyltransferase rRNA ribosomal RNA RNA Ribonucleic acid RNP Ribonucleoprotein SMA Spinal Muscular Atrophy SMN Survival of Motor Neurons scaRNP small Cajal body specific RNP snRNA small nuclear RNA snRNP small nuclear RNP snoRNP small nucleolar RNP SPN snurportin teRNP telomerase RNP Tgs1 trimethylguanosine synthase TMG 2,2,7-trimethylguanosine Unrip unr interacting protein WT Wildtype
9 Gemin function in small nuclear RNP biogenesis
and Spinal Muscular Atrophy
Abstract
By
KARL BRYAN SHPARGEL
The heart of genetics research lies in gaining knowledge and understanding of the effects that gene alterations have on human health and development. Ultimately, a thorough analysis of genetic diseases spanning from mutation identification to protein function will hopefully lead to the development of therapeutic treatments. Spinal Muscular Atrophy
(SMA) is one of the most prevalent and dire genetic diseases. Hallmarks of the disease include degeneration of spinal motor neurons and skeletal muscular atrophy. The disease is particularly crippling due to early onset in infants and children.
Research in the field of genetics has made great strides toward understanding this grave disease. The causative gene, the Survival of Motor Neurons 1 gene (SMN1), has been identified by positional cloning. Remarkably, research has unraveled a critical function for SMN in the assembly of the small nuclear ribonucleoproteins (snRNPs) essential for splicing of pre-messenger RNAs (pre-mRNAs). As an interesting exception to this global function, SMN has neuronal and muscle specific localization patterns suggesting that an ulterior tissue specific defect may be causing SMA. Altogether, the underlying cause of SMA is still shrouded in mystery, warranting further investigation.
10 In an attempt to better comprehend the involvement of snRNP assembly in SMA pathogenesis, my studies focused on SMN and its interacting partners. The goal of my thesis is to examine members of the SMN protein complex to determine first if they function in snRNP assembly, and next whether their loss of function leads to an SMA- like phenotype when mutated in a model organism. Notably by making use of biochemical and genetic tools, I have discovered that several Gemin proteins of the SMN complex are also required for efficient assembly of snRNPs. Further, SMN point mutations corresponding to the most severe phenotypic class of SMA are also defective in snRNP assembly. Finally, mutations in the Drosophila ortholog of Gemin3, a DEAD box RNA helicase, produced motor defects and larval lethality, phenotypes similar to mutations in fly Smn. Therefore, it appears that snRNP biogenesis is a real contributing factor to SMA pathology. My research adds substantial contributions in understanding the relationship between snRNP assembly and SMA.
11 Chapter I
Introduction and Research Objectives
12 Introduction
Spinal Muscular Atrophy
As the second leading cause of autosomal recessive lethality, Spinal Muscular Atrophy
(SMA) is a devastating neuromuscular disease. This prominent disease afflicts nearly 1 in
8000 live births and is carried by a 50th of the Caucasian population (Pearn, 1978). The
underlying etiology of SMA is characterized by degeneration of motor neurons in the anterior horn of the lower spinal cord and accompanying atrophy of proximal skeletal muscle. The phenotypic manifestations of SMA are incredibly variable. While SMA exists as a continuum of severity, patients are classified into subtypes to facilitate patient care. Type I SMA, otherwise known as Werdnig-Hoffman disease or infantile SMA, is the most common and severe form. These infants display onset before 6 months of age as general weakness and inability to sit up (Pearn, 1980). Infant survival does not usually exceed 2 years due to respiratory failure. Type II or intermediate SMA is characterized by onset at approximately the first year of life and survival into adolescence. Type III
SMA patients, also known as Kugelberg-Welander syndrome or juvenile SMA, have a relatively mild phenotype with onset after 18 months of age and a normal life expectancy.
While Type II SMA patients exhibit severe muscle wasting, large portions of Type III
SMA patients are able to walk early on, but progressively lose this ability. More recently, extreme cases of SMA have been classified as Type 0 (embryonic lethal) or Type IV
(adult onset) (Brahe and Bertini, 1996).
Most classes of SMA are allelic (96%), resulting from mutations in the Survival of
Motor Neurons 1 (SMN1) gene (Lefebvre et al., 1995). Interestingly, a duplication event
specific to the human lineage has created a unique situation whereby two copies of SMN
13 exist in the genome (SMN1 and SMN2). However, SMA causative mutations have only been identified in the telomeric copy, SMN1. The centromeric copy, SMN2, differs by a single substitution in exon 7, a C-T transition resulting in high levels of alternative splicing skipping this exon (Figure 1-1) (Lorson et al., 1999; Monani et al., 1999). This truncated SMN protein is unstable, cannot oligomerize, and cannot compensate for the loss of full-length SMN1 transcripts (Lorson and Androphy, 1998). However, in patients homozygous for SMN1 mutations, SMN2 produces a small percentage of full-length transcripts (Frugier et al., 2002). As the vast majority of SMA cases result from SMN1 deletion, a hypomorphic phenotype exists because of the diminished production of functional SMN from the SMN2 locus. Furthermore, there is a good correlation between the disease severity, SMN2 copy number, and overall levels of SMN protein (Coovert et al., 1997; Lefebvre et al., 1997).
14 Figure 1-1: Mutations in SMN1 result in SMA. SMN1 and SMN2 differ by a single
nucleotide in exon7. This synonymous substitution promotes skipping of exon7 in the
majority of SMN2 transcripts producing a truncated unstable protein. When SMN1 is
mutated in cases of SMA, the minor, full-length transcript from SMN2 allows for
postnatal survival and can modify SMA severity.
Several mouse models of SMA have been generated to aid in the study of disease
pathogenesis and progression. While two copies of the SMN gene are encoded within the human genome, the mouse genome has only one copy, the Smn locus (DiDonato et al.,
1997). When this single copy Smn gene was knocked out, homozygous mutants exhibited extensive cell death early in development (Schrank et al., 1997). These embryos failed to
form a blastocyst and died before implantation. Heterozygotes exhibited phenotypically
normal motor functions, but displayed an adult-onset degeneration of spinal motor
neurons at six months of age (Jablonka et al., 2000). In an attempt to re-create the human condition, the human SMN2 transgene was introduced into mice homozygous for the Smn
knockout (Monani et al., 2000). One or two copies of the human SMN2 transgene rescued
the embryonic lethality in Smn-/- knockout mice, but these animals displayed severe
motor, neuronal, and muscular defects culminating in death prior to postnatal day 6. In
stark contrast, Smn-/- mice carrying eight copies of the human SMN2 transgene appeared
phenotypically normal. Just as levels of the SMN protein are correlated with SMA
severity in humans, mice exhibited a similar dosage sensitive effect.
15 The Smn complex
SMN research has focused on the identification of interacting proteins in an attempt to
gain insight into function of this critical protein. Purification and mass spectrometry of
the macromolecular SMN complex from both cytoplasmic and nuclear human cell culture
extracts has identified several interacting partners (Charroux et al., 1999; Meister et al.,
2000). SMN directly interacts with other complex members including Sip1 (SMN
interacting protein 1, Gemin2), Gemin3, Gemin5, and Gemin7 (Figure 1-2) (Gubitz et al.,
2004) (See map of SMN functional domains, Fig. 1-3). SMN also indirectly interacts
with Gemin4 through Gemin3 and Gemin6, Unrip, and Gemin8 through Gemin7
(Carissimi et al., 2005; Carissimi et al., 2006; Gubitz et al., 2004). Amazingly, the
majority of these novel proteins contain no recognizable functional domains. Gemin3, a
DEAD box RNA helicase, and Gemin5, a WD motif protein, are the exceptions
(Charroux et al., 1999; Gubitz et al., 2002). Notably, most members of the SMN complex interact with several Sm proteins that form a stable complex with the Uridine rich small nuclear Ribonucleoproteins (U snRNPs) that catalyze the splicing of pre-messenger
RNAs (pre-mRNAs) (Gubitz et al., 2004; Liu et al., 1997). Upstream of splicing, the assembly of these Sm proteins on small nuclear RNAs (snRNAs) is a critical step in the formation of a functional snRNP (the snRNP biogenesis pathway) (Figure 1-2).
16
Figure 1-2: SMN forms a large macromolecular complex that interacts with Sm proteins.
SMN self-interacts to initiate incorporation of Gemin proteins and Sm proteins into a
common complex. This illustrated complex assembles Sm proteins on snRNAs.
Figure 1-3: Domain structure of the SMN protein illustrating the Gemin2 interaction domain, the Tudor domain (Sm binding), YG box (oligomerization), and exon 7 (deleted from the major SMN2 transcript).
17 snRNP biogenesis
The snRNP biogenesis pathway constitutes all events leading to the proper assembly and
localization of splicing snRNPs (Fig. 1-4) (Kiss, 2004; Matera and Shpargel, 2006; Will
and Luhrmann, 2001). This pathway is highly regulated and spans both the nucleus and
cytoplasm. Biogenesis of a snRNP is initiated by the nuclear transcription of snRNA
genes. Major spliceosomal U snRNAs (U1, U2, U4, and U5) are largely transcribed by
RNA polymerase II (with the exception of U6 snRNA transcribed by RNA polymerase
III) with co-transcriptional 7-methyl guanosine cap methylation (m7G) (Cougot et al.,
2004; Lobo, 1994). This cap modification initiates assembly of the snRNA nuclear export
complex. The cap binding complex recognizes the snRNA which in turn is bound by the
phosphorylated adaptor for snRNA export (PHAX) (Izaurralde et al., 1995; Izaurralde et
al., 1994; Ohno et al., 2000). As the name implies, PHAX phosphorylation is required for
formation of the export complex. Finally, binding of Ran-GTP and the CRM1 nuclear
export receptor signify that the snRNA is ready for export (Fornerod et al., 1997;
Izaurralde et al., 1997). This export complex guides the snRNA to the cytoplasm where the first stage of protein assembly occurs.
Upon arrival in the cytoplasm, the snRNA is first received by the SMN complex.
The SMN complex binds directly to all major spliceosomal U snRNAs at a unique
docking site on each snRNA (Golembe et al., 2005; Yong et al., 2004; Yong et al., 2002).
All members of the SMN complex interact directly with a distinct subset of Sm protein
heterocomplexes, SmD1/D2, SmB/D3, or SmF/E/G (Gubitz et al., 2004). These Sm sub- complexes are assembled into a heptameric protein core on the snRNA (Kambach et al.,
1999). Depletion of the SMN complex from cellular extracts eliminates Sm core
18 formation (Meister et al., 2001; Pellizzoni et al., 2002) . Remarkably, add-back of recombinant SMN alone is not sufficient to recover this activity. Appropriate rescue by addition of the entire purified SMN complex further illustrates that contribution of other factors within this complex are essential for snRNP assembly. Notably, several Sm proteins are symmetrically dimethylated by the protein arginine methyl transferase 5 complex (PRMT5) (Brahms et al., 2001; Brahms et al., 2000). This post-translational modification substantially increases the affinity of Sm proteins for the Tudor domain of
SMN and therefore may facilitate assembly with the snRNA (Fig. 1-4).
Subsequent to Sm core assembly, the snRNA undergoes 3’ exonucleolytic trimming (Seipelt et al., 1999). Additionally, the tri-methyl guanosine synthase enzyme
(Tgs1) hypermethylates the RNA cap of the snRNP from m7G to yield a 2,2,7 tri-methyl- guanosine (TMG) cap (Mouaikel et al., 2002; Plessel et al., 1994). The Tgs1 enzyme may be recruited to the snRNP through a direct interaction with SMN (Mouaikel et al., 2003).
It is the TMG cap modification that promotes binding of the import adaptor, snurportin
(Huber et al., 1998). The import receptor, importin-B, can bind to both snurportin at the
TMG cap as well as the SMN complex, which remains bound to the snRNP through the
Sm core, as a mechanism to drive snRNP entry into the nucleus (Narayanan et al., 2004b;
Narayanan et al., 2002). Thus the SMN complex dependent assembly of the Sm core is required for further snRNP modifications and snRNP nuclear import in addition to snRNP stability and function in splicing.
Just as the cytoplasmic phase of snRNP biogenesis follows a rigid step-by-step blueprint, subsequent nuclear snRNP biogenesis steps are also highly scripted. Following import, nascent snRNPs first accumulate in the Cajal body (CB) (Sleeman and Lamond,
19 1999a). The fact that the snRNP import machinery (snurportin, SMN complex) also concentrates in CBs suggests that this complex may be retained on the snRNP for CB targeting. Direction to this specific nuclear locale promotes post-transcriptional modifications of the snRNA. Small Cajal body specific RNAs (scaRNAs) guide psuedouridylation and 2’-O-methylation of crucial residues within the snRNAs (Darzacq et al., 2002; Jady et al., 2003). Finally, higher order snRNP protein assembly may signify completion of a mature snRNP product and allow release of the snRNP from the CB to be destined for sites of splicing (Nesic et al., 2004; Schaffert et al., 2004).
20
21 Figure 1-4: snRNP biogenesis overview. snRNAs are complexed with the Cap Binding
Complex (CBC), PHAX (phosphorylated), RAN-GTP, and CRM1 export complex
following nuclear transcription. This complex dissociates after traversing the nuclear pore. The SMN complex binds to snRNAs, recruits Sm proteins, and regulates their assembly into a core on the snRNA. SmB, SmD1, and SmD3 are symmetrically dimethylated on arginine residues to increase the affinity for SMN in this process.
Following core formation, Trimethyl Guanosine Synthase (Tgs1) hypermethylates the
snRNA cap and an exonuclease (Exo) trims the 3’ end. This cap modification promotes
binding of the snurportin (SPN) import adaptor. Importin-β (Impβ) can bind to SPN and
the SMN complex to import the snRNP into the nucleus. The snRNP is targeted to CBs
for post-transcriptional modifications and protein assembly. The snRNP is now ready for
transport to active sites of splicing.
The Cajal body
The CB was originally identified by Santiago Ramon y Cajal in the early 1900s by silver
staining and light microscopy (Ramon y Cajal, 1903). These nuclear organelle are
visualized by silver staining due to the high concentrations of RNAs that reside there.
Only recently have various CB components become defined with the aid of antibodies
and localization studies. One autoimmune serum specifically and exclusively recognized
CBs (Andrade et al., 1991). This prompted the identification of the CB marker protein,
coilin. While the SMN complex (SMN and Gemins) localizes diffusely throughout the
cytoplasm, in the nucleus these proteins are housed within CBs (Carvalho et al., 1999). A
minority of cell lines and embryonic tissue exhibit a separation of the SMN complex
22 from CBs in structures termed “Gemini of Cajal bodies” (Gems) (Liu and Dreyfuss,
1996). Additional studies uncovered CBs as potential sites for maturation and processing of several families of RNAs. The snRNAs (splicing), small nucleolar RNAs (snoRNAs, ribosomal RNA maturation), and telomerase RNAs (telomere length maintenance) all reside within CBs along with various assembly and maturation factors (Figure 1-5) (Jady et al., 2003; Jady et al., 2006; Matera and Shpargel, 2006; Tomlinson et al., 2006;
Verheggen et al., 2002).
The CB brings together snRNAs with corresponding scaRNAs, guiding post- transcriptional modifications within the CB (Jady et al., 2003). These guide scaRNAs direct 2’-o-methylation and pseudouridylation of specific snRNA bases. In addition to post-transcriptional maturation, snRNPs also undergo additional protein assembly and aggregation into multi-snRNP complexes that signify completion of snRNP biogenesis and release from the CB. For example, Splicing factor 3a (SF3a) is comprised of three subunits that assemble to form the mature 17S U2 snRNP. RNAi or mutation of SF3a subunits results in an increased accumulation of these factors and immature U2 snRNPs in the Cajal body suggesting that assembly of the mature U2 snRNP occurs in Cajal bodies and is a pre-requisite for CB release (Nesic et al., 2004). Moreover, U4/U6•U5 tri- snRNP assembly is essential for spliceosome formation and also is thought to occur in
CBs (Schaffert et al., 2004). Therefore, CBs are the nuclear sites of snRNP post- transcriptional maturation and higher order assembly that mark completed snRNPs to be destined for active sites of splicing in the nucleus.
Some nuclear snoRNAs require both cap hypermethylation by Tgs1 as well as assembly with protein factors (Fibrillarin, Nopp140, etc.), all of which concentrate in
23 CBs (Verheggen et al., 2002). Telomerase also requires cap hypermethylation and
assembly with reverse transcriptase (hTERT) and other proteins (Fibrillarin) prior to
telomere function (Matera and Shpargel, 2006). Fascinatingly, not only do all these
factors concentrate in the CB, but CBs also localize to telomeres, possibly as a delivery
messenger (Jady et al., 2006; Tomlinson et al., 2006). In a similar fashion, CBs can also
be observed in a close association with the nucleolus, the destination site for snoRNPs
(Young et al., 2001). The Cajal body marker protein, coilin, is conserved throughout
mammals, arachnids, and plants. While other members of the Animal and Fungi
Kingdoms (such as flies and yeast) lack a recognizable coilin ortholog, they do retain a
homologous CB structure implying a conserved essential function (Liu et al., 2006;
Verheggen et al., 2002). Remarkably, mouse embryonic fibroblast cell lines (MEFs)
derived from coilin knockout mice exhibit a separation of CB components into “residual
CBs” (Tucker et al., 2001). The SMN complex, snRNPs/scaRNAs, and snoRNPs separate into 3 classes of residual bodies (Jady et al., 2003; Tucker et al., 2001). These coilin
knockout mice have viability and fertility defects. Therefore the CB is essential in
bringing these pathways together into a common locale, possibly in the interests of efficiency.
24
25 Figure 1-5: The Cajal body functions in snRNP, snoRNP, and teRNP maturation and assembly. Several classes of small RNPs, including snRNPs, scaRNP, snoRNPs and telomerase (teRNP), all traffic through the Cajal body (CB). Sm-class snRNPs target to
CBs, possibly through the SMN/snurportin (SPN) import complex. The trimethylguanosine (TMG)-capped snoRNPs (i.e. U3, U8 and U13) require CRM1 and
PHAX for CB localization, whereas scaRNPs and teRNPs use a CAB box element. The scaRNPs are known to guide pseudouridylation and ribose 2́-O-methylation of snRNAs
(red stars) and perhaps other RNAs as well. Tgs1 (TMG synthase) localizes to CBs and probably directs hypermethylation of both telomerase RNA and the major snoRNAs.
Secondary snRNP and snoRNP maturation events involve binding of additional snRNP and snoRNP proteins. These higher order interactions presumably signal their release from the CB. Telomerase reverse transcriptase (TERT) and teRNP proteins are assembled with teRNA. The SMN protein interacts with all three distinct classes of RNPs and might be a common RNP assembler, consistent with its function in the cytoplasm.
Dilemmas in SMA pathogenesis and snRNP biogenesis
The field of Spinal Muscular Atrophy is extremely advanced. Not only has the disease causing gene (SMN) been identified by positional cloning, but a cellular function (snRNP biogenesis) has also been ascribed to the causative protein. Identification of an underlying cause of the disease is vital to the development of therapeutic strategies for treatment. For all that is known about the disease, the true cause of SMA is an enigma.
The major function ascribed to SMN, snRNP biogenesis, is a ubiquitous essential function while the disease is incredibly tissue specific in nature. Due to this discrepancy,
26 several alternative models of pathogenesis have arisen pertaining to SMN function. These
hypotheses center on the identification of novel tissue specific SMN functions that hope
to explain this disparity (Fig. 1-6).
In support of a tissue specific SMN function, the protein exhibits novel
localization patterns in neuronal and muscular cells. SMN is found in growth cones of
neuronal like cell lines (Fan and Simard, 2002). These growth cones are essential in
creating local concentrations of mRNAs and proteins required for neurite extension in
axon pathfinding and muscle innervation. One of these mRNAs, β-actin, has a high
affinity for hnRNP R which in turn directly interacts with SMN (Rossoll et al., 2002).
Intriguingly, primary motor neurons derived from Smn knockout mice develop shorter neurites with reduced B-actin localization to growth cones (Rossoll et al., 2003). In vivo, in mouse tissues, Smn localizes to neuromuscular junctions, demonstrating a potential function in synapse formation between neurons and muscles (Fan and Simard, 2002).
Additionally, RNAi knockdown of SMN in muscle cell precursors (myoblasts) inhibits
muscle cell fusion and myotube formation (Shafey et al., 2005). Consistent with all of
these ideas, neuronal or muscular tissue specific Smn knockout mice displayed motor
defects and cell death, however it is still unclear if these phenotypes are due to general
cellular lethality or tissue specific loss of function (Cifuentes-Diaz et al., 2001; Frugier et
al., 2000). Therefore generating an organismal model separating tissue specific loss of
function from general snRNP biogenesis loss of function will be critical in deciphering
the true nature of SMA pathogenesis.
27
Figure 1-6: SMN is involved in several cellular functions. In addition to ubiquitous function in snRNP biogenesis, SMN is also hypothesized to function in transport and concentration of specific mRNAs in neuronal growth cones. Cell culture studies also point to a function for SMN in myoblast cell fusion. Finally due to localization at neuromuscular junctions, SMN may have further neuronal and muscular functions.
28 Research Objectives
SMN forms a large macromolecular cytoplasmic complex with Gemins that as a whole is
required for snRNP Sm core assembly in vitro. SMN depleted extracts exhibit Sm
assembly defects, and rescue is only achieved by addition of the entire purified SMN
complex. By utilizing RNAi to knock down individual Gemin proteins, I propose to
assess the requirements of individual proteins within the SMN complex in the process of
snRNP biogenesis via in vitro, in vivo, and cell biological assays. Furthermore, rescue of
SMN knockdown with exogenous constructs containing SMA causative point mutations
in SMN will determine whether SMA pathology might result from snRNP biogenesis
defects.
SMN function in snRNP biogenesis is required ubiquitously throughout all
organismal tissues. Conversely, SMA pathogenesis is restricted to motor neurons and
corresponding skeletal muscle. While the defect may stem from cell specific SMN function, it is instead possible that the neuromusculature is sensitive to perturbations in
the efficiency of snRNP biogenesis. Neurons in particular have tremendous amounts of
transcriptional activity and alternative splicing. If snRNP biogenesis is a causative factor
in SMA pathology, then mutations in other snRNP biogenesis genes should also display
similar neuromuscular defects. I will focus my examination on Gemin3, a DEAD box
RNA helicase within the SMN complex. This helicase may be critical for mediating
protein and RNA rearrangements necessary for Sm core formation. I will analyze a
putative Gemin3 homolog in Drosophila to determine if it is required for snRNP biogenesis and if when mutated, the flies develop SMA-like phenotypes.
29 Chapter II
Gemin proteins are required for efficient assembly of Sm-class RNPs
Karl B. Shpargel and A. Gregory Matera*
Department of Genetics Case Western Reserve University School of Medicine Cleveland, OH 44106-4955 USA
* Corresponding Author Tel: 216-368-4922 Fax: 216-368-0491 Email: [email protected]
Note: This chapter is a manuscript published in the Proceedings of the National Academy of Sciences 102(48): 17372-17377.
30 Abstract
Spinal Muscular Atrophy (SMA) is a neurodegenerative disease characterized by loss of spinal motor neurons. The gene encoding the survival of motor neurons (SMN) protein is mutated in over 95% of SMA cases. SMN is the central component of a large oligomeric complex, including Gemins2-7, which is necessary and sufficient for the in vivo assembly of Sm proteins onto the small nuclear (sn)RNAs that mediate pre-mRNA splicing. Following cytoplasmic assembly of the Sm core, both SMN and splicing snRNPs are imported into the nucleus, accumulating in Cajal bodies for additional snRNA maturation steps before targeting splicing factor compartments known as
“speckles.” In this study, we analyzed the function of individual SMN complex members by RNA interference (RNAi). RNAi-mediated knockdown of SMN, Gemin2, Gemin3 and Gemin4 each disrupted Sm core assembly, whereas knockdown of Gemin5 and
Snurportin1 had no effect. Assembly activity was rescued by expression of a GFP-SMN construct that is refractive to RNAi, but not by similar constructs that contain SMA patient-derived mutations. Our results also demonstrate that Cajal body homeostasis requires SMN and ongoing snRNP biogenesis. Perturbation of SMN complex function results in disassembly of Cajal bodies and relocalization of the marker protein, coilin, to nucleoli. Moreover, in SMN-deficient cells, newly-synthesized SmB proteins fail to associate with U2 snRNA or to accumulate in Cajal bodies. Collectively, our results identify a novel function for Gemin3 and Gemin4 in Sm core assembly and correlate the activity of this pathway with SMA.
31 Introduction
Spinal Muscular Atrophy (SMA) is a severe autosomal recessive disease characterized by
degeneration of motor neurons in the anterior horn of the spinal cord, resulting in
subsequent atrophy of skeletal muscle (Frugier et al., 2002; Talbot and Davies, 2001;
Wirth, 2000). The disease has an incidence rate of 1 in ~8,000 live births and a carrier
frequency of about 1 in 50 (Pearn, 1980). SMA patients can be divided into 3 classes,
based on phenotypic severity. Type I, Werdnig-Hoffmann, or infantile SMA is
characterized by onset within 6 months of birth and death before 2 years of age. Type II,
or intermediate, SMA patients exhibit onset at 6 months of age and survival into
adolescence. Type III, Kugelberg-Welander, or juvenile SMA patients typically display a
late onset (after 18 months of age) and can survive into adulthood (Pearn, 1980).
The Survival of Motor Neurons 1 gene (SMN1) was identified as the SMA
disease-causing gene by Melki and colleagues (Lefebvre et al., 1995). This region of the genome has undergone a duplication to create a second copy of the gene, SMN2. The key difference between SMN1 and SMN2 is a C to T transition within exon 7 (Lorson et al.,
1999). This mutation causes skipping of exon7 in a majority of SMN2 transcripts, resulting in a dearth of functional protein (Cartegni and Krainer, 2002; Lorson and
Androphy, 2000). Notably, gene conversion events can increase SMN2 copy number and reduce SMA severity (Burghes, 1997; Campbell et al., 1997). Although a majority of
SMA cases (92%) result from homozygous deletions of SMN1, a growing list of point mutations identified in SMN1 account for 3% of SMA patients (Sun et al., 2005; Wirth,
2000). Overall, total levels of functional SMN protein correlate with a reduction in SMA
32 severity establishing the basis for dramatic phenotypic variation in affected individuals
(Coovert et al., 1997; Lefebvre et al., 1997).
Whereas the SMN protein shows strong, diffuse cytoplasmic localization, the
protein also accumulates in discrete nuclear foci known as Cajal bodies (CBs) (Carvalho
et al., 1999; Liu and Dreyfuss, 1996; Matera, 1998). In fetal tissues and a small subset of
cell lines, SMN localizes to distinct nuclear structures known as Gemini bodies (gems),
so named because of their typical close proximity to CBs (Liu and Dreyfuss, 1996). SMN
is the central member of a large macromolecular complex (Gubitz et al., 2004; Meister et
al., 2002). Members of this, so-called, SMN complex are termed “Gemins” as they co-
localize with SMN in gems and CBs. Some of the most notable members of this complex
are Gemin2 (alias SMN interacting protein 1, SIP1), Gemin3 (DP103, Ddx20), and
Gemin4 (GIP1) (Gubitz et al., 2004; Meister et al., 2002). Gemin2 forms a very stable
direct interaction with SMN, whereas Gemin3 is a putative DEAD box RNA
helicase/unwindase that directly interacts with both SMN and Gemin4.
Critical insight into SMN function came from the identification that the protein
interacts with Sm proteins, core components of small nuclear ribonucleoproteins
(snRNPs) (Fischer et al., 1997; Liu et al., 1997). In metazoans, pre-snRNA transcripts are exported to the cytoplasm for assembly into stable Sm-core particles. In vivo, this assembly is mediated by the SMN complex (Meister et al., 2001; Pellizzoni et al., 2002).
Following additional cytoplasmic remodeling steps, the RNPs are imported back into the nucleus, where they undergo further maturation in CBs before ultimately functioning in the spliceosome (Kiss, 2004; Will and Luhrmann, 2001). SMN and Gemins2-7 also
localize to CBs due to a direct interaction between SMN and coilin, the CB marker
33 protein (Hebert et al., 2002b; Hebert et al., 2001). Thus, the biogenesis of Sm snRNPs is a multi-step process that takes place in distinct subcellular compartments.
To identify the roles of individual SMN complex proteins in the process of
snRNP biogenesis, we have used RNA interference (RNAi) to ablate the expression of
SMN complex proteins in HeLa cells. Our results demonstrate that SMN, Gemin2,
Gemin3 and Gemin4 are required for efficient Sm core assembly. In addition, we show that loss of SMN protein leads to disassembly of nuclear CBs and a redistribution of coilin to the nucleolus. Furthermore, we found that various SMA-causing point mutations
failed to rescue Sm core assembly in vitro, consistent with the idea that snRNP
biogenesis defects underlie the pathogenesis of the disease.
34 Materials and Methods
RNAi
HeLa-ATCC cells were transfected with siRNAs targeting SMN, Gemins2-7, Snurportin,
or a Control sequence (Ambion). Cells were transfected using the DharmaFECT1
lipofection reagent (Dharmacon), as directed. The mRNA sequences targeted by siRNA
duplexes were as follows: SMN, GGAGCAAAAUCUGUCCGAU; Gemin2,
GGUUUCGAUCCCUCGGUAC; Gemin3, GGAAAUAAGUCAUACUUGG; Gemin4,
GGCACUGGCAGAAUUAACA; Gemin5, GGGUCUCUGGCUUCACAUU; Gemin6,
GGAUGGGUUUUAACUACAG; Gemin7, GGCCAGAGGUUCCUGAAAU;
Snurportin, GGAATGGATTGTGGTCGTG. Control siRNAs that do not target human
mRNAs include Silencer Negative Control #1 siRNA (Ambion) or mouse Gemin3
siRNAs: GGATTAGAATGTCATGTCT. The specificity of SMN and Gemin3
knockdown was confirmed by western blotting and Sm core assembly assays utilizing a
second set of siRNAs targeting each message (GAAGAAUACUGCAGCUUCC for
SMN (Feng et al., 2005) or GGCUUAGAGUGUCAUGUCU for Gemin3). The siRNA
transfections were allowed to proceed for 48-60 hours before analysis. GFP-SmB DNA
was electroporated 48 hours after the initial siRNA transfection (BIO RAD GenePulser
XCell) and cell lysates were collected 24 hours later. SMN siRNAs and GFP-SMN rescue constructs were co-transfected using Lipofectamine 2000 (Invitrogen) and allowed to proceed for 48 hours.
35 Western blotting and Immunofluorescence
Cytoplasmic cellular lysates were collected using Ne-Per Nuclear/Cytoplasmic extraction
kit (Pierce). 10μg cytoplasmic lysate was loaded per lane and blotted with α-SMN (BD
Transduction Labs), α-Gemin2 (BD Transduction Labs), α-Gemin3 (BD Transduction
Labs), α-Gemin4 (Santa Cruz), or α-Karyopherin (BD Transduction Labs). Western blots
were exposed to film and quantified using the Quantity One (BIO RAD) software
package. Cells were fixed with 4% paraformaldehyde and extracted with 0.5% Triton X-
100. Immunofluorescence was performed with α-SMN (BD Transduction Labs), α-
Gemin3 (BD Transduction Labs), α-coilin (pAb R124), or α-Fibrillarin (mAb 72B9).
Sm Core assembly assays
U1 snRNA was transcribed in vitro by standard procedures in the presence of m7G cap analogue (Promega) and 32P UTP. 100,000 counts of U1 snRNA were incubated with
40μg cytoplasmic lysate for 20 minutes at 30ºC. Sm core assembly reactions were pre-
cleared with protein-G beads (Pierce) followed by immunoprecipitation with αY12
(Labvision) in RSB 100 buffer (Pellizzoni et al., 2002). Immunoprecipitates were run on
a 6% acrylamide TBE-Urea denaturing gel and exposed to a phosphoimager. Sm core
assembly assays and western blots were quantified with Quantity One (BIO RAD). U2
snRNA IP northerns were carried out following GFP-SmB transfection in NET buffer
(150mM NaCl, 5mM EDTA, 50mM Tris pH 7.5, 0.5% NP40) with α-GFP (Roche).
Products were run on 10% TBE-Urea gels. Northern probes were generated by random-
primed labeling of a U2 snRNA PCR product with 32P dCTP.
36 DNA Constructs
pT7U1, GFP-Spn, and GFP-SMN were cloned as previously described (McConnell et al.,
2003; Narayanan et al., 2002; Sleeman et al., 2001). GFP-SMN* (siRNA target mutant) was generated using the Quickchange PCR mutagenesis kit (Stratagene), along with 5'-
AGAACAGAACTTAAGTGACCTACTTTCCCCAATCTGTGAAGTAGC-3' and 5'-
GTCACTTAAGTTCTGTTCTTCTCTATTTCCATATCCAGTGTAAAC-3' primers.
SMA point mutations were developed by mutagenesis of GFP-SMN* with primers
spanning 15-nt on each side of the amino acid codon change.
37 Results and Discussion
SMN and Gemin protein levels are interdependent
We used RNAi to systematically knock down expression of SMN complex proteins in human cells by transfection of small interfering (si)RNA triggers (Elbashir et al., 2001).
Western blotting demonstrated efficient reduction of SMN, Gemin2, Gemin3, and
Gemin4 protein levels, as compared to the importin-α loading control (Fig. 2-1A).
Control siRNAs that do not recognize a cellular target (Ctl) or those targeting
Snurportin1 (SPN), a protein that is not involved in Sm core assembly had no effect on the levels of SMN complex proteins. Interestingly, SMN knockdown also resulted in a
concomitant decrease in the levels of Gemin2 protein and, to a lesser extent, Gemin3.
Moreover, knockdown of Gemin4 significantly reduced Gemin3 protein levels.
Immunofluorescence following RNAi of SMN and Gemin3 displayed effective knockdown in 97-98% of transfected cells (data not shown). Quantification of these results confirmed the fact that we achieved efficient knockdown of SMN, Gemin2,
Gemin3 and Gemin4 (Fig. 2-1B).
It is pertinent to note that the interdependence of protein levels within the SMN
complex has previously been illustrated using Smn knockout mice. Heterozygous mutant
mice exhibit a corresponding reduction of Gemin2 protein levels (Jablonka et al., 2002).
Furthermore, SMA patients often display reduced Gemin2 and Gemin3 levels in addition
to those of SMN (Helmken et al., 2003). One likely explanation is that formation of the
SMN complex is required for protein stability. Reduction in the amount of SMN might
disrupt formation of the entire complex, leading to degradation of several other proteins
(e.g. Gemin2 and Gemin3). Loss of Gemin4 may not be as critical for SMN complex
38 formation, however, Gemin4 forms an independent, stable complex with Gemin3
(Charroux et al., 2000; Mourelatos et al., 2002). Thus an absence of Gemin4 might lead
to specific degradation of Gemin3 due to breakdown of this independent complex.
Figure 2-1: SMN and Gemin protein levels are interdependent. (A) siRNAs were
transfected into HeLa-ATCC cells and cytoplasmic cell lysates were collected 60 hours
later for analysis by western blotting. Columns denote siRNA transfections with mock
(no siRNA), SMN, G2 (Gemin2), G3 (Gemin3), G4 (Gemin4), Spn (Snurportin), or Ctl
(control) siRNAs. Rows are blotted with the corresponding antibodies; importin-α was used as a loading control. (B) Quantification of protein levels from 3 separate siRNA
39 transfections. Protein levels were normalized to importin-α, then graphed as a fraction of the mock transfection. The siRNA transfections effectively reduced SMN (13%), Gemin2
(11%), Gemin3 (12%), and Gemin4 (6%) protein levels.
SMN and Gemins2-4 are required for efficient Sm core assembly
To determine the relative contribution of each member of the SMN complex to the process of Sm core assembly, we performed in vitro assays using siRNA-treated cytosolic extracts. Radiolabeled U1 snRNA was incubated with cytoplasmic extracts to allow for Sm core assembly. The reaction was then immunoprecipitated with monoclonal antibody Y12, which is specific for a subset of methylated Sm proteins (Brahms et al.,
2000). Whereas mock and SPN siRNA transfections had little effect on Sm core assembly, siRNAs targeting SMN, Gemin2, Gemin3, and Gemin4 each displayed significant defects (Fig. 2-2A). SMN knockdown showed the most pronounced effect, whereas siRNAs targeting Gemin2, Gemin3, and Gemin4 had moderate Sm core assembly defects (Fig. 2-2B). We also tested siRNAs targeting Gemins5-7 in this assay, although we were unable to confirm knockdown of these proteins due to a lack of the appropriate antibodies. Targeting of Gemin6 and Gemin7 had modest effects on Sm core assembly, whereas siRNAs against Gemin5 had no effect (see supplemental data, Fig 2-
6).
Using similar procedures (but different siRNAs and transfection reagents), Feng et al. (2005) recently showed that RNAi-mediated knockdown of SMN, Gemin2 and
Gemin6 inhibited Sm core assembly, however, siRNAs targeting Gemin3, Gemin4 and
Gemin5 had little effect (Gemin7 was not tested) (Feng et al., 2005). Thus the major
40 difference between the two sets of experiments is that we observed a significant defect in
Sm core assembly upon loss of Gemin3 or Gemin4 (Fig. 2-2B); the results for other
members of the SMN complex are in good agreement. In reconciling these differences, it
is important to note that Feng et al. (Feng et al., 2005) were unable achieve efficient
knockdown of Gemin3 (~70%) or Gemin4 (~55%). In contrast, we were able to reduce
Gemin3 and Gemin4 levels by ~90% (Fig. 1B). We therefore conclude that Gemin3 and
Gemin4 are required for efficient assembly of Sm core particles.
Figure 2-2: SMN, Gemin2, Gemin3, and Gemin4 are required for efficient Sm core assembly. (A) Cytoplasmic lysates were collected 60 hours after mock, SMN, Gemin2,
Gemin3, Gemin4, or Snurportin siRNA transfections. The lysates were incubated with either wildtype U1 snRNA (+) or U1∆Sm snRNA (∆, containing an Sm site deletion), at
41 either non-permissive (4°C) or permissive (30°C) Sm core assembly temperatures for 20
min. Sm core assembly reactions were immunoprecipitated with monoclonal antibody
Y12, run on a denaturing gel, and exposed to a phosphoimager. (B) Quantification of
three separate Sm core assembly assays was graphed as a fraction of the mock
transfection. Sm core assembly activity was significantly inhibited (p-values < 0.04)
following RNAi for SMN (28%), Gemin2 (48%), Gemin3 (59%), Gemin4 (55%), but not
by Snurportin (122%) or control (112%) siRNA transfections (p-values > 0.7).
Loss of SMN and snurportin results in breakdown of nuclear Cajal bodies
Cajal bodies are thought to be sites of post-transcriptional snRNP modification (Darzacq
et al., 2002; Jady et al., 2003; Kiss, 2004; Matera, 1999a)}. To determine the
consequence of reduction in the levels of SMN complex proteins on CB homeostasis, we
performed RNAi followed by immunofluorescence microscopy using antibodies against the CB marker protein, coilin. As shown in Fig. 2-3A, RNAi for SMN resulted in the complete loss of coilin foci, whereas mock treatment had little effect on CB number. Of all the SMN complex proteins tested, only knockdown of SMN had a significant effect on the number of cells displaying CBs. The results are quantified in Fig. 2-3C (grey bars), showing that control cells or those treated with siRNAs targeting Gemins2-4 displayed roughly the same number of coilin foci (i.e. 80-85% of cells showed at least one CB).
SMN knockdown reduced CB numbers significantly from 2.5 CBs/cell to 0.8 CB/cell (p- value < 10-20) and from 80-85% of cells containing a CB to only 35-40% (Fig. 3C).
In addition to CB disassembly, we found that SMN knockdown often resulted in
relocalization of coilin to nucleoli and/or nucleolar caps. As shown in Fig. 2-3B, several
42 different cellular phenotypes were observed. In some cells, coilin was fragmented into
smaller foci, whereas in others the protein was localized throughout the nucleolus (Fig. 2-
3B, middle panels). Cells were also scored for nucleolar coilin accumulation following
RNAi of other SMN-associated proteins, including Gemins2-4 and SPN. SMN and SPN
knockdown had the greatest effect, with 35-45% of the cells relocalizing coilin to the
nucleolus. Gemin2, Gemin3, and Gemin4 had no effect (Fig. 2-3C, black bars).
Additionally, an increased proportion of cells transfected with SMN siRNAs localized
coilin to nucleolar caps (Fig. 2-3B bottom panel, Fig. 2-3C striped bars). Collectively, the
results suggest that SMN and SPN, but not Gemins2-4, are critical for Cajal body
homeostasis. Because SMN is important for both Sm core assembly (Meister et al., 2001;
Pellizzoni et al., 2002) and UsnRNP import (Narayanan et al., 2004a), it is difficult to
separate the two functions. However, the relocalization/disassembly of CBs upon SPN
knockdown suggests that ongoing UsnRNP import is the key factor.
Our findings are consistent with the previous observations of Lamond and
colleagues, who showed that Sm protein expression enhances the formation of CBs in cells that typically lack these nuclear suborganelles (Sleeman et al., 2001). Coilin
contains a cryptic nucleolar localization signal, and in some tissues, the protein normally
forms associations with nucleoli in the form of peri-nucleolar caps (Hebert and Matera,
2000; Young et al., 2001). Leptomycin B, a drug that disrupts the first step of snRNP
biogenesis, namely snRNA export to the cytoplasm, also causes coilin localization to the
nucleolus (Carvalho et al., 1999). Alternatively, SMN reduction might disrupt other
cellular pathways (e.g. cell cycle, transcription, snoRNP biogenesis, splicing) in addition
to snRNP biogenesis, resulting in fewer Cajal bodies per cell. Thus assembly and nuclear
43 import of new snRNPs are critical for CB formation and coilin subnuclear localization
(Sleeman and Lamond, 1999a). In the absence of ongoing UsnRNP import, CBs disassemble and coilin relocalizes to the nucleolus either by default or because of a continued association and function in some other pathway, such as snoRNP biogenesis
(Kiss, 2004; Verheggen et al., 2002).
44 Figure 2-3: Loss of SMN and snurportin results in breakdown of nuclear Cajal bodies.
(A) CB localization was detected using anti-coilin immunofluorescence (left column) in mock (top) or SMN siRNA (bottom) transfections. In the top panels, note the prominent
CBs, whereas in the bottom panel no CBs were detected and coilin was localized in a granular pattern. All scale bars represent 2.5μM. (B) We observed three different types of coilin nucleolar phenotypes after SMN siRNA treatment. CBs were visualized by coilin immunofluorescence (left column), while anti-fibrillarin stains both CBs and nucleoli
(right column). Mock transfections exhibited typical coilin localization in CBs (top panel). Following SMN RNAi, CBs appear to break into smaller fragments (second panel), and coilin is partially relocalized to the nucleolus. The third panel shows a cell similar to the one in Fig. 2-3A, where the coilin signal granular in appearance, with prominent nucleolar accumulations. In the bottom panel, SMN siRNA treatment results in coilin redistribution to nucleolar caps. (C) Quantification of cellular phenotypes. Cells transfected with mock, SMN, Gemin2, Gemin3, Gemin4, Snurportin, or control siRNAs were scored (n = 160 cells each) for the fraction of cells lacking CBs (grey bars), localizing coilin within the nucleolus (coilin NoL, black bars), or containing coilin in nucleolar caps (striped bars).
SMN is required for targeting Sm proteins to Cajal bodies
Plausibly, defects in Sm core assembly might result in either a cytoplasmic accumulation or a nuclear reduction in UsnRNP levels. RNAi experiments targeting SMN, followed by immunofluorescense against Sm proteins, failed to detect significant differences in cytoplasmic versus nuclear distributions of Sm proteins (data not shown). In order to
45 visualize only newly-assembled UsnRNPs , we therefore analyzed the localization of
GFP-tagged Sm proteins following RNAi. Cells were treated with siRNAs targeting
either SMN, Gemin2 or Gemin3/Gemin4 for 48 hrs., and then transfected and incubated
overnight with constructs that express GFP-SmB (Fig. 2-4A). Mock RNAi treatments
showed the typical GFP-SmB localization patterns in both CBs and nucleoplasmic
speckles. SMN RNAi followed by GFP-SmB expression did not result in a significant
buildup of cytoplasmic fluorescence, however, snRNP localization in the nucleus was
much more diffuse and GFP-SmB failed to concentrate in the CBs (i.e. nucleoplasmic
coilin foci) that remain after SMN knockdown. However, GFP-SmB did localize to large
SC35-positive speckles in the absence of SMN (Fig. 2-4A). In contrast, RNAi of Gemin3
and Gemin4 together did not have a significant effect on GFP-SmB localization (Fig. 2-
4A).
Quantification of the number of GFP-SmB foci (corresponding to CBs and large speckles) in these experiments revealed that cells treated with control or anti-Gemin siRNAs displayed an average of 2.4 foci/cell, whereas only 0.2 foci/cell were observed following RNAi for SMN (Fig. 2-4B, p-value < 10-16). To verify that the defects observed
upon RNAi were correlated with a loss of Sm core assembly, we repeated the GFP-SmB
reporter experiment and performed immunoprecipitation with anti-GFP antibodies,
followed by northern analysis of U2 snRNA (Fig. 2-4C). As shown, SMN knockdown
had the greatest effect on recovery of U2 snRNA with GFP-SmB, while knockdown of
Gemin2 or Gemin3/Gemin4 brought down intermediate amounts of U2. Note that the
results of this in vivo pulldown (Fig. 2-4C) are in good agreement with those of the in
vitro assembly assay (Fig. 2-2).
46 Interestingly, even in the absence of proper snRNP assembly, GFP-SmB did not significantly accumulate in the cytoplasm. Moreover, we routinely recovered reduced amounts of GFP-SmB from cytoplasmic extracts of cells treated siRNAs targeting SMN.
To normalize the input levels of GFP-SmB in Fig. 2-4C, we transfected 1.5 fold greater amounts of GFP-SmB plasmid. The fact that we do not see a clear increase in cytoplasmic accumulation of Sm proteins suggests that the unassembled Sm subunits are either imported into the nucleus without complexing with snRNAs or they are degraded.
These two options are not mutually exclusive; overexpression of GFP-SmB might swamp the machinery that normally degrades unassembled Sm proteins, allowing their import.
Consistent with this scenario, recent experiments have demonstrated a nuclear localization activity associated with the basic C-terminal tails of Sm proteins (Girard et al., 2004). Our results suggest that this import pathway is independent of snRNP assembly and interaction with SMN.
Thus a consequence of SMN reduction is that we not only detect fewer CBs, but the remaining CBs do not accumulate detectable amounts of GFP-SmB. Therefore the decline in snRNP import precedes CB disassembly. As GFP-SmB can still localize to large speckles after SMN RNAi, it is possible that the unassembled GFP-SmB proteins can still interact with other splicing factors leading to accumulation in this nuclear locale.
Although Sm proteins localize to the nucleus in the absence of ongoing snRNP assembly, our findings illustrate that proper targeting of newly-synthesized Sm proteins to CBs is dependent on SMN.
47
Figure 2-4: SMN and Gemins regulate Sm core assembly in vivo. (A) Cells were transfected with mock, SMN, or a cocktail containing Gemin3 and Gemin4 siRNAs.
After 48 hrs. of siRNA treatment, a GFP-SmB reporter construct was transfected; the cells were fixed after 66 hrs total incubation time. Immunofluorescence was performed to visualize CBs (coilin, left panels) or speckles (SC35, right panel). The GFP-SmB reporter construct exhibited a normal distribution in CBs (arrows) and nuclear speckles following mock or Gemin3/4 siRNA treatments (first and third columns). In contrast, RNAi for
SMN failed to localize GFP-SmB to CBs (second column) but did localize the reporter to large speckles (last column). All scale bars represent 2.5μm. (B) Quantification of GFP-
SmB nuclear foci following mock, SMN, Gemin2, Gemin3/Gemin4, or control siRNA
48 treatment (n = 90 cells; SMN RNAi p-value < 10-17, Gemin3/Gemin4, or control RNAi p- values > 0.7). (C) The experiment in panel (A) was repeated and cytoplasmic cell extracts collected after 66 hours. IP-northerns were performed using anti-GFP antibodies and blotted for U2 snRNA. SMN RNAi greatly reduces GFP-SmB association with U2 snRNA compared to the controls, whereas Gemin2 or Gemin3/Gemin4 siRNA transfections have an intermediate effect. A GFP-only reporter was used as a negative control for the immunoprecipitation.
SMA type I, but not type III, SMN point mutations are defective in Sm assembly
The severity of SMA is inversely proportional to SMN protein levels (Coovert et al.,
1997; Lefebvre et al., 1997). However, in the absence of an SMN1 gene deletion, a number of disease-causing mis-sense mutations have also been described (Clermont et al., 2004; Cusco et al., 2004; Hahnen et al., 1997; Sun et al., 2005; Wirth, 2000). As shown in Fig. 2-5A, these mutations are scattered throughout the length of the coding region. Because SMA is a recessive disease and because the SMN complex is oligomeric, it has been particularly difficult to assay the effects of these patient-derived mutants by overexpressing them a wildtype background. In order to assay the activities of individual
SMN mutations, in the absence of the wildtype protein, we generated a ‘wildtype’ SMN construct (called GFP-SMN*) containing synonymous point-substitutions in the region targeted by the siRNAs, making it refractive to RNAi (data not shown). A number of
SMA-causing point mutations were next introduced into this background and then assayed for their abilities to rescue Sm core formation in vitro.
49 As shown in Fig. 2-5B, co-transfection of anti-SMN siRNAs with GFP-
SMN*(Wt) substantially rescued Sm core assembly; co-transfection with the empty GFP
vector had a nominal effect. Note that the SMN knockdown was slightly less effective,
and that cell death was slightly increased when cells were co-transfected with plasmids
and siRNAs (data not shown). Thus the partial rescue of Sm-core assembly activity
detected in the GFP-only transfections (Fig. 2-5B, lane 4) is likely due to the incomplete
inactivation of the SMN complex in those cells. Conversely, the incomplete rescue of
Sm-core activity by the GFP-SMN*(Wt) construct (Fig. 2-5B, lane 5) may be due to a
difference in the expression of the exogenous construct and/or the presence of the GFP
tag. In any event, the assembly activity of the wildtype rescue construct was reproducibly
higher and significantly different (p value < 0.02) from that of GFP alone (Fig. 2-5C).
Next, we analyzed the abilities of eight separate SMN mutations (six SMA type I
alleles and two SMA type III alleles) to rescue Sm core assembly. As shown in Figs. 2-
5B and 2-5C, five out of the six SMA type I alleles tested (I116F, E134K, Q136E,
Y272C, and ΔEx7) showed only basal levels of Sm-core assembly activity, whereas both
of the SMA type III alleles (D30N and T274I) functioned as well as the wildtype construct (Fig 2-5B,C). Three of the SMA type I mutations (E134K, Y272C, and ΔEx7) have been studied previously in vitro. The SMN(E134K) is thought to reduce binding of
SMN to Sm proteins and importin beta (Bühler et al., 1999; Narayanan et al., 2004a; Sun
et al., 2005). Moreover, the Y272C and ΔEx7 mutations have been shown to disrupt
SMN oligomerization, with concomitant or downstream defects in Sm protein binding
(Lorson et al., 1998; Pellizzoni et al., 1999). Consistent with these previous findings, we
now demonstrate that E134K, Y272C, and ΔEx7, as well as the uncharacterized I116F
50 and Q136E SMN mutations are defective in Sm core assembly. Interestingly, I116F,
E134K, and Q136E cluster together within the Tudor domain of SMN, which is required for high-affinity Sm protein interaction (Friesen et al., 2001; Selenko et al., 2001).
Recently, two other laboratories have shown that the SMN2 gene product (i.e. primarily
SMN∆Ex7) is partially defective in Sm core assembly by performing in vitro assays in extracts derived from SMA patient fibroblasts (Wan et al., 2005; Winkler et al., 2005).
Furthermore, we found that SMA Type I point mutations (E134K and Y272C) were unable to rescue concomitant loss of Gemin2 upon SMN RNAi (supplemental data, Fig.
2-6). In contrast, GFP-SMN*(Wt), GFP-SMN*(D30N) and GFP-SMN*(T274I) restored expression of Gemin2 to control levels (see supplemental data, Fig. 2-6). Therefore it is possible that the Type I SMA point mutations prevent proper formation of the SMN complex and are thus defective in Sm core assembly. Notably, we found that both of the
SMA type III alleles (D30N and T274I) and one SMA type I allele (A111G) were functional in our in vitro assembly assay. It is possible that these regions of SMN may be required for some other aspect of SMN function, such as cap hypermethylation
(Mouaikel et al., 2003) or nuclear import (Narayanan et al., 2004a), thus establishing a potential basis for SMA type I-type III phenotypic variation. Alternatively, the D30N,
T274I and A111G mutations might be required for some novel SMN function, such as in trafficking of messenger RNAs to neuronal growth cones (Rossoll et al., 2003). In summary, we have shown that SMN-deficient cells display defects in Sm core assembly activity that can be rescued by addition of ectopically expressed GFP-SMN.
51
Figure 2-5: SMA patient-derived SMN mutations are defective in Sm core assembly. (A)
An alignment of SMN orthologues, showing the evolutionary conservation of SMN protein sequences from worms to humans. In the top row, the locations of SMA-causing point mutations (red boxes for SMA type I and green boxes for type II-III) are shown for comparison. The Tudor domain, responsible for Sm protein interaction, is boxed in blue
52 and the sequences encoded by human exon 7, skipped in the SMN2 major transcript, is boxed in red. (B) Cells were either mock- or co-transfected with SMN siRNAs and a GFP or GFP-SMN* DNA construct (containing synonymous substitutions in the siRNA target site, see text for details). Cells were collected after 48 hours and Sm core assembly assays
were performed using the various GFP-SMN* mutant cytoplasmic extracts. (C)
Quantification of the Sm assembly activity of each GFP-SMN* construct in the absence
of endogenous SMN protein. Notably, the activities of SMA type I SMN mutations
I116F, E134K, Q136E, Y272C, and ΔEx7 were each significantly lower (p-values <
0.03), than the GFP-SMN*(Wt) rescue construct. Conversely, the activities of the SMA type III mutations D30N and T274I were not significantly different from the GFP-
SMN*(Wt) control (p-values > 0.6).
Gemin3 and Gemin4: Active participants in assembly of the Sm core?
Our results highlight the importance of Gemin3 and Gemin4 in the Sm core assembly
process; previous efforts to investigate this question were inconclusive (Feng et al.,
2005). However, additional experiments will clearly be required in order to elucidate the precise functions of each of the members of the SMN complex. For example, it will be interesting to determine whether Gemin3, the putative RNA helicase/unwindase, is
actively involved in adding Sm proteins to the snRNA or if it merely aids by providing a
framework for assembly. Alternatively, the participation of micro (mi)RNAs in the
assembly of snRNPs has not been strictly ruled out. In addition to interacting with SMN,
Gemin3 and Gemin4 have been shown to form independent complexes with miRNAs
(Mourelatos et al., 2002). It is unclear whether or not these complexes are functionally
53 related. Future experiments, including development of in vivo model systems, will be
essential in sorting out the different pathways that contribute to SMA pathology.
Figure 2-6: Supplemental figure. (A) Snurportin expression was reduced by treatment with corresponding short interfering RNAs (siRNAs). A GFP-Snurportin reporter construct was transfected into HeLa-ATCC cells, followed by Gemin3 or Snurportin siRNA treatment the next day. A Western blot for GFP-Snurportin confirms specific
54 targeting by Snurportin siRNAs. (B) SMN, Gemin2, Gemin3, Gemin4, Gemin6, and
Gemin7 are required for efficient Sm core assembly. Cytoplasmic lysates were collected
60 h after mock, survival of motor neuron (SMN), Gemin2–7, or Snurportin siRNA transfections. The lysates were incubated with either wild-type U1 small nuclear RNA
(snRNA) (+) or U1ΔSm snRNA (Δ, containing an Sm site deletion), at either nonpermissive (4°C) or permissive (30°C) Sm core assembly temperatures for 20 min.
Sm core assembly reactions were immunoprecipitated with monoclonal antibody Y12, run on a denaturing gel, and exposed to a PhosphoImager. (C) Quantification from three separate Sm core assembly assays graphed as a fraction of the mock transfection. Sm core assembly was effectively reduced after RNA interference (RNAi) for SMN (28%),
Gemin2 (48%), Gemin3 (59%), Gemin4 (55%), Gemin6 (75%), or Gemin7 (74%) but not after Gemin5 (105%), Snurportin (122%), or control (112%) siRNA transfections. (D)
GFP-SMN*(Wt) rescues Gemin2 levels reduced by SMN RNAi. (Note: the * denotes a mutated siRNA target site). GFP, GFP-SMN*(Wt), or the indicated GFP-SMN* point mutation constructs were cotransfected into HeLa along with anti-SMN siRNAs. After 48 h of RNAi treatment, cell lysates were analyzed by Western blotting for Gemin2, SMN, or Importin-α. SMN Type III point mutations of D30N and T247I were found to restore
Gemin2 to control levels, whereas Type I SMA point mutations of E134K and Y272C did not.
55 Chapter III
Drosophila Gemin3 is essential for larval motor function, pupation, and viability
Karl B. Shpargel, T. K. Rajendra, and A. Gregory Matera*
Department of Genetics Case Western Reserve University School of Medicine Cleveland, OH 44106-4955 USA
* Corresponding Author Tel: 216-368-4922 Fax: 216-368-0491 Email: [email protected]
Note: This chapter is a manuscript in preparation
56 Abstract
Small nuclear ribonucleoproteins (snRNPs) that catalyze pre-mRNA splicing are
assembled and matured in an elaborate and highly regulated cellular pathway. A major
feature of this snRNP biogenesis pathway is the assembly of SmB/D3, SmD1/D2, and
SmF/E/G into a heptomeric protein core on small nuclear RNAs (snRNAs). Sm protein assembly is dependent on the macromolecular, cytoplasmic Survival of Motor Neurons protein (SMN) complex. This complex is comprised of SMN and several Gemin proteins which are individually required for appropriate Sm core formation. Intriguingly, the
SMN1 gene is mutated in Spinal Muscular Atrophy (SMA), a neuromuscular disease characterized by motor neuron degeneration and atrophy of skeletal muscle. However the relationship between this disease and snRNP biogenesis is not well understood. We now identify the Drosophila homolog of the Gemin3 DEAD box RNA helicase. Drosophila
Gemin3 (dGemin3) interacts with dSMN directly in vitro and in vivo. RNAi knockdown
of dGemin3 and dSMN inhibits snRNA Sm core assembly similar to their human
counterparts. P-element and piggyBac insertions in Drosophila Smn and Gemin3 are
lethal during the second and third instar larval stages and exhibit motor defects similar to
previously characterized Smn alleles. Remarkably, some Smn and Gemin3 mutants escape
early lethality, and live as larvae for several weeks without undergoing pupation. Finally,
Smn and Gemin3 alleles genetically interact with mutations in SmD2, tying their
phenotypes into snRNP biogenesis function. Our results demonstrate a conservation of
SMN complex function in Drosophila snRNP assembly and exemplify that loss of SMN
complex function contributes to SMA pathogenesis and possibly additional aspects of
organismal development.
57 Introduction
Splicing is an essential housekeeping function required for all intron containing pre- mRNAs. Splicing, catalyzed by small nuclear ribonucleoproteins (snRNPs), is conserved throughout eukaryotic organisms. snRNP biogenesis refers to the cellular pathway whereby the small nuclear RNAs (snRNAs) are post-transcriptionally modified, complexed with proteins, and destined for nuclear sites of splicing (Kiss, 2004; Will and
Luhrmann, 2001). This pathway initiates with nuclear snRNA transcription followed by cytoplasmic export. In a major cytoplasmic snRNP biogenesis step, heteromeric Sm protein complexes (SmB/D3, SmD1/D2, and SmF/E/G) are assembled into a heptameric core on the snRNA (Yong et al., 2004). This process is carried out by a large multimeric protein complex, the Survival of Motor Neurons protein (SMN) complex (Meister et al.,
2001; Pellizzoni et al., 2002). Formation of the snRNP Sm core is a prerequisite for further snRNA modifications and subsequent nuclear import (Mattaj, 1986; Narayanan et al., 2004a; Plessel et al., 1994; Seipelt et al., 1999). In nuclear translocation, the snRNP binds the Importin-β import receptor through snurportin bound to the snRNA cap and the
SMN complex that remains associated with the snRNP (Narayanan et al., 2002). Upon nuclear re-entry, the snRNP is first targeted to the Cajal body for additional post- transcriptional modifications and protein assembly before the final snRNP product is delivered to sites of splicing (Jady et al., 2003; Matera and Shpargel, 2006; Sleeman and
Lamond, 1999b).
The Survival of Motor Neurons 1 gene (SMN1) was identified by positional cloning to be the gene responsible for Spinal Muscular Atrophy (SMA) (Lefebvre et al.,
1995). SMA is one of the most prevalent autosomal recessive genetic diseases
58 distinguished by a degeneration of motor neurons in the anterior horn of the lumbar
regions of the spinal cord and atrophy of skeletal muscle (Frugier et al., 2002). The
disease has a carrier frequency of 1 in 50 with an incidence rate of 1 in 8000 live births
(Pearn, 1978). Three phenotypic classes of SMA have been established due to the
extreme variability in patient severity (Frugier et al., 2002). SMA type I is most severe
with infant onset before 6 months of age. SMA type II patients experience an
intermediate onset by 18 months of age, while SMA type III is characterized by late onset after 18 months. SMA type I patients die before 2 years of age, however SMA type II patients can survive into adolescence. The leading cause of death amongst SMA patients
is respiratory failure that accompanies atrophy of the proximal skeletal muscles. Even
though SMA type III patients develop muscular atrophy disability later in life, they tend
to exhibit a normal life span. All three classes of SMA are allelic, caused by mutations in
SMN1. Amazingly, the human genome contains a second locus, SMN2, which cannot
compensate for the loss of SMN1. SMN2 contains a mutation in exon7 that promotes
skipping of the exon in the majority of transcripts producing a truncated nonfunctional
protein product (Lorson et al., 1999). A minority of SMN2 transcripts include the exon7
full length SMN product. Thus SMA is defined as a hypomorphic disease with a
reduction, but not absence of SMN protein. This aspect of the disease is critical because
SMN function in snRNP biogenesis is an essential cellular function and complete loss of
this pathway would cause very early embryonic lethality. Consistently, knock-out of the
single copy Smn gene in mice is early embryonic lethal (Schrank et al., 1997).
The SMN complex is a large 20S protein complex that localizes diffusely
throughout the cytoplasm with intense punctuate nuclear signal in the Cajal body (Gubitz
59 et al., 2004). The nuclear and cytoplasmic human SMN complexes have been purified for the identification of their components by mass spectrometry (Charroux et al., 1999;
Meister et al., 2000). The identified contents have been termed Gemin proteins due to altered localization in some cell lines to Gemini bodies, focal nuclear structures that often are associated with Cajal bodies (Gubitz et al., 2004; Liu and Dreyfuss, 1996).
Organization of the complex based on known protein-protein interactions centers around
SMN which directly interacts with itself, Gemin2, Gemin3, Gemin5, and Gemin7 (Gubitz et al., 2004). Gemin7 is thought to recruit Gemin6, Gemin8, and unr-interacting protein
(unrip) while Gemin3 brings Gemin4 into the complex (Baccon et al., 2002; Carissimi et al., 2005; Carissimi et al., 2006; Charroux et al., 2000). The SMN complex binds directly to the snRNA and to Sm proteins to coordinate snRNP protein assembly (Gubitz et al.,
2004). We have previously demonstrated by RNAi knockdown that SMN, Gemin2,
Gemin3, and Gemin4 are all required for efficient snRNP assembly (Shpargel and
Matera, 2005). Therefore it is likely that several Gemins function together with an unidentified mechanism to mediate snRNP Sm core formation.
Genetic analysis in model organisms provides a unique opportunity to study
contributing factors to disease pathogenesis. To date, the SMN complex has not been isolated in Drosophila. Drosophila SMN (dSMN) has been identified based on sequence conservation, and point mutations created within the gene are larval lethal in the second- third instar stage (Chan et al., 2003). These larvae exhibit motor and neuromuscular defects. We have subsequently generated several adult models for Drosophila SMA. A hypomorphic mutation in Smn, Smnex33, was created by imprecise excision of a p-element residing in an Smn enhancer (Rajendra et al., manuscript in preparation). Smnex33
60 homozygotes exhibit reduced dSMN protein levels in the thorax of the adult fly. This
deficiency leads to severe motor neuron, muscular, and flightless defects, all of which can
be rescued by a GFP-Smn transgene (Rajendra et al., manuscript in preparation). Notably,
dSMN localizes to structural components of flight muscles in wild-type flies (Rajendra et
al., manuscript in preparation). The muscle is a multinucleate cell that is divided into
repeating segments known as sarcomeres. Within each sarcomere, actin and myosin
filaments overlap to provide the contractile force of the muscle. Myosin clusters together
at the M-line of the sarcomeres, whereas actin clusters within the Z-disc. dSMN co-
localizes with actin at the Z-disc and biochemically interacts with α-actinin, another
structural component of actin filaments. Therefore, the severe muscular phenotypes
observed in Smnex33 may be due to a novel function for dSMN in developing appropriate
muscle structure. We have also obtained a hypomorphic SmD2 p-element insertion
(unpublished observations). This insertion lies in the only intron of the SmD2 gene, reducing the ability to splice properly into the second exon. The SmD2 homozygotes are also flightless, albeit at incomplete penetrance (unpublished observations).
We now characterize Drosophila Gemin3 (dGemin3), a DEAD box RNA helicase
within the SMN complex. Just like its human counterpart, dGemin3 interacts directly
with dSMN in vitro and in vivo. Furthermore, they co-localize in the Drosophila Cajal
body and are required for efficient snRNP Sm protein assembly. P-element insertions in
Smn and Gemin3 exhibit larval lethality, motor function and pupation defects, and genetic
interactions with the P-element insertion in SmD2. Our results demonstrate conservation
of the SMN complex in Drosophila and define its function in various aspects of
Drosophila development.
61 Materials and Methods
DNA constructs
Smn and Gemin3 full-length cDNAs were PCR amplified with primers flanked by
Gateway recombination sequences (Invitrogen). These products were recombined
initially into pDONR221 before entry into GST-tagged pDEST15, His-tagged pDEST17,
GFP-tagged pAGW, or myc-tagged pAMW vectors (Invitrogen and DGRC-T. Murphy collection).
Recombinant protein expression and S2 cell transfections
GST-dSMN and His-dGemin3 were expressed in BL21-star bacteria (Invitrogen) by
1mM IPTG induction for 3 hours. Lysate was extracted by sonication and passed over
glutathione or Ni-agarose beads. S2 cells were transfected using Cellfectin as directed
(Invitrogen).
Antibodies
GST (Santa Cruz, 1:1000), His (Lab Vision, 1:1000), GFP (Roche, 1:1000), myc (Santa
Cruz, 1:1000), SMN (Transduction Labs, 1:5000), and SNF (U2B’’, 1:1000) antibodies
were used for western blotting. Myc (Santa Cruz, 1:40) and 22C10 (1:250) were used for
immunofluorescence or staining. Myc antibody (5µl) was used for immunoprecipitation
in mRIPA buffer.
62 Sm Assembly Assay
Smn and LacZ dsRNAs were transcribed in vitro from PCR products flanked with T7
promoters. Drosophila S2 cells were placed in SF-900 media containing 14ug/mL of dsRNA. Extracts were generated 3 days after transfection using the Ne-Per nuclear/cytoplasmic extraction kit as directed (Pierce) and dialyzed in reconstitution buffer (20mM Hepes-KOH pH 7.9 / 50mM KCl / 5mM MgCl2 / 0.2mM EDTA;
Pellizzoni et al., 2002). 40ug of cytoplasmic extract was loaded on a gel for Western
blotting analysis to confirm knockdown. For the assembly assay, wild type U1 snRNA
and U1 snRNA containing a deletion of the Sm assembly site were in vitro transcribed
from PCR products in the presence of P32-rUTP and m7G cap analogue (Promega).
100,000 counts of radiolabeled U1 snRNA were incubated in 100ug of cytoplasmic
extract at 22ºC for 40 min in reconstitution buffer. Assembled snRNPs were pre-cleared
with protein-G beads before immunoprecipitation with 4ul Y12 antibody in RSB-100
buffer (600mM NaCl / 20mM Tris-HCl pH 7.4 / 2.5mM MgCl2 / 0.01% NP40: Pellizzoni
et al., 2002). Immunoprecipitation products were denatured in Formamide loading buffer,
run on a 6% acrylamide TBE-Urea gel, and exposed to a phosphorimager.
Fly stocks
SmnA (G202S: Chan et al., 2003), SmnB (S201F: Chan et al., 2003), SmnC (f05960:
Thibault et al., 2004), SmnD (f01109: Thibault et al., 2004), and SmnF (PL00733:
Spradling et al., 1999) were maintained over TM3, GFP balancer chromosomes. Smnex33
(excision of EY14384: Spradling et al., 1999) Gem3A (e03688: Thibault et al., 2004),
Gem3B (rL562: Spradling et al., 1999), and SmD2 (EP3399: Spradling et al., 1999) were
63 maintained on TM6, BTb balancer chromosomes. Alleles were recombined to create
multiple insertions on a single chromosome. Gem3B rev was created by precise excision of
Gem3B. Timed matings were allowed to proceed for 24 hours and larvae were collected for phenotypic analysis 3, 5, or 8 days later.
Tissue histology
Mutant larvae (4 days after egg laying) were prepared for immunofluorescence. An
incision was made in the posterior of the larvae. The head was then pushed entirely
through the body to expose interior epitopes. Tissue was fixed in 3.7% formaldehyde and
processed for immunofluorescence.
64 Results
Drosophila Dhh1 encodes the Gemin3 homolog
The genome wide Drosophila yeast-2-hybrid analysis has been published with an accompanying searchable online database (Giot et al., 2003). Analysis of dSMN identified a high confidence interaction with a DEAD box RNA helicase, Dhh1
(CG6539). Comparison of this protein sequence to Gemin3 homologues from various species demonstrated strong conservation of amino acid sequence within the helicase motifs (Fig. 3-1, red boxes) and weak conservation within the SMN interaction domain
(Fig. 3-1, blue box). The Carboxy-terminus of the protein is very extensive and not conserved between Drosophila and vertebrates.
65
66 Figure 3-1: Drosophila Gemin3 is conserved throughout its helicase motifs. Aligned are
Gemin3 sequences from human, mouse, chicken, zebrafish, and fly. The seven helicase motifs are boxed in red and the SMN interaction domain is boxed in blue.
Human Gemin3 is defined by a direct interaction with SMN, co-localization with
SMN throughout the cytoplasm and in nuclear Cajal bodies, and function in efficient
snRNP Sm core formation. The putative Gemin3 and Smn cDNAs were cloned into
expression vectors for binding and localization assays. Bacterially expressed recombinant
GST-dSMN specifically and efficiently pulled down His-dGemin3 (Fig. 3-2A). This
interaction was verified in vivo by performing immunoprecipitations. S2 cells were co-
transfected with myc-dGemin3 and GFP-dSMN. Immunoprecipitation of GFP-dSMN
pulled down myc-dGemin3 out of S2 lysate (Fig. 3-2B). Furthermore, fixation and
immunofluorescence of these S2 cells revealed complete co-localization of dGemin3 and
dSMN throughout the cytoplasm and in nuclear Cajal bodies (Fig. 3-2C). Interestingly,
dSMN overexpression formed cytoplasmic aggregates, similar to its human homolog
(Shpargel et al., 2003).
67
Figure 3-2: dGemin3 interacts with dSMN in vitro and in vivo. (A) Bacterially expressed
His-dGemin3 (His-dGEM3) was passed over GST beads, glutathione beads (Glut), or
GST-dSMN beads in a pull down reaction. Western blotting demonstrated a specific, direct interaction with GST-dSMN. (B) Drosophila S2 cells were transfected with GFP- dSMN alone or co-transfected with myc-dGemin3. Immunoprecipitation with myc antibody only pulled down GFP-dSMN when myc-Gemin3 was present. Inputs represent
1% of immunoprecipitation reaction. (C) S2 co-transfected with GFP-dSMN and myc- dGemin3 exhibit cytoplasmic and nuclear Cajal body (arrow) co-localization.
68 The dSMN complex is essential for efficient snRNP assembly
SMN and Gemin3 are essential for snRNA Sm protein core formation in human cell lines
(Feng et al., 2005; Shpargel and Matera, 2005). We utilized an Sm core assembly assay
to investigate whether dSMN and dGemin3 play similar conserved functions. This assay
takes advantage of cytoplasmic extracts prepared from deficient lysates. A radiolabeled
U1 snRNA is incubated in these extracts to promote assembly of Sm proteins. Next, an
immunoprecipitation using Y12 (αSm) antibody quantifies how much assembly has occurred. We performed RNAi on S2 cells to knock down levels of dSMN and dGemin3
proteins (Fig. 3-3A). Western blotting of cytoplasmic extracts derived from untransfected
S2 cells (Mock) or S2 cells transfected with LacZ dsRNA, Smn dsRNA, or Gemin3 dsRNA demonstrated efficient and specific knockdown of dSMN and dGemin3
compared to the Tubulin loading control. The myc-dGemin3 construct was utilized to
analyze levels of dGemin3 knockdown due to the unavailability of an antibody.
Interestingly, RNAi of dGemin3 also resulted in a moderate reduction of dSMN, possibly
due to decreased protein stability when this interaction is lost. These extracts were
incubated with either radiolabeled + U1 snRNA (wild-type) or Δ U1 snRNA (Sm
assembly site deleted) at either non-permissive (4ºC) or permissive (22ºC) temperatures
for the assembly assay (Fig. 3-3B). RNAi of dSMN and dGemin3 significantly reduced
Sm core assembly (p-values < 0.005) relative to Mock or LacZ controls. Quantification from three separate experiments verified a 50% reduction in Sm core assembly when dSMN and dGemin3 were knocked down (Fig. 3-3C). Thus SMN function in snRNP assembly is conserved in invertebrates.
69
Figure 3-3: Drosophila SMN and Gemin3 are required for efficient snRNA Sm core assembly. A) Drosophila S2 cells were left untreated (Mock) or transfected with either
Smn, Gemin3 (dGEM3), or LacZ dsRNAs. These S2 cells were initially transfected with a myc-dGemin3 reporter construct. Cytoplasmic extracts were collected 3 days following transfection and Western blotting confirmed efficient knockdown of dSMN and dGEM3 relative to the Tubulin loading control. B) A radiolabeled U1 snRNA transcript was
70 incubated in cytoplasmic extract and immunoprecipitated with Y12 (αSm) antibody to assay for Sm core assembly. The U1 snRNA containing a deletion of the Sm protein assembly site (Δ) or incubating a wild type U1 snRNA (+) at a non-permissive temperature (4ºC) serve as negative controls in the experiment. RNAi of dSMN and dGEM3 significantly disrupted Sm core assembly compared to Mock and LacZ dsRNA transfections. C) Quantification of Sm core assembly assays from three separate experiments. The results, normalized relative to the Mock control, revealed significant (p- values < .005), approximately 50% reductions in Sm core assembly for dSMN and dGEM3 knockdown. LacZ RNAi had no significant effect (p-value > 0.2).
Mutations in Smn and Gemin3 are larval lethal
Previously, Smn alleles containing point mutations, SmnA (G202S) and SmnB (S201F), were characterized with larval lethality, motor and neuromuscular defects (Fig. 3-4A)
(Chan et al., 2003). We have obtained three novel alleles, SmnC (f05960: piggyBac
insertion downstream of the Tudor domain), SmnD (f01109: piggyBac insertion upstream
of the Tudor domain), and SmnF (PL00733: P-element insertion in the 3’ untranslated
region). Lysate collected from all homozygous second-third instar larvae exhibited absent
or very low dSMN levels (Fig. 3-4B). At first glance, all alleles appeared to be larval
lethal since no homozygous pupae were observed. Temporal analysis of heterozygous
intercrosses revealed that some lethality of SmnD, SmnC, and SmnB homozygotes has
already occurred by three days after egg laying (second instar larvae, Chi-Square p-
values < 0.001) (Fig. 3-4C). SmnD and SmnC appeared to be the most severely affected
with very few larvae surviving past day five (third instar) (Fig. 3-4D). Notably, while
71 SmnB and SmnF exhibited moderate viability defects at the third instar larval timepoint
(Chi-Square p-values < 0.0002), approximately 30% of homozygous larvae survived past
day eight, a period where wild-type (WT) and heterozygous larvae have already pupated
(Fig. 3-4E). Incredibly, some SmnF larvae survived for weeks without pupating. SmnB seemed to be the weakest allele because 30% of homozygous larvae survived to pupation timepoints and some even appeared to initiate the formation of a thin pupal case (data not shown). All these viability defects were rescued by expression of a GFP-dSMN transgene.
Two Gemin3 alleles were obtained for contrast to Smn phenotypes (Fig. 3-4A).
Similar to all Smn alleles, Gem3A (e03688: piggyBac insertion in the Gemin3 promoter)
and Gem3B (rL562: P-element insertion immediately following the translation start site) appeared to be lethal before pupation. Furthermore, the two alleles fail to complement each other. Lysate collected from second-third instar homozygous Gemin3 mutant larvae exhibited a consistent reduction in dSMN levels by western blotting (Fig. 3-4B). These
results are reminiscent of those seen by RNAi of dGemin3 in cell culture (Fig. 3-3A).
Temporal analysis of Gemin3 heterozygous intercrosses indicated larval lethality occurs around second-third instar timepoints (Fig. 3-4C,D) (second instar Chi-Square p-values <
0.004). Similar to SmnF, 30% of Gem3A and 15% of Gem3B homozygous larvae survive
to periods of pupation, but fail to pupate (Fig. 3-4E). Double homozygosity for Gem3B and SmnF (illustrated as Gem3B,SmnF in Fig. 3-4) displayed a phenotype similar to the
Gem3B insertion, demonstrating a lack of genetic interaction between these two alleles.
Therefore, while dSMN and dGemin3 work together in snRNP assembly, complete loss
of function of both genes is equivalent to loss of any one of them. Notably, a Gemin3
72 revertant allele, Gem3B rev, recovers the ability to pupate. Thus, Smn and Gemin3 are essential for larval viability and pupation.
73
74 Figure 3-4: Smn and Gemin3 mutant larvae exhibit viability and pupation defects. (A)
Schematic illustrating locations of Smn, Gemin3, and SmD2 alleles. (B) Western blotting of larval lysates four days after egg laying. All Smn alleles were protein nulls, while
Gemin3 alleles displayed reduced dSMN levels compared to WT and the revertant,
Gem3B rev. A SNF antibody (that recognizes Drosophila Sans fille, a homolog of U2
snRNP B'' protein) was used as the loading control. (C) Percentage of collected
homozygous larvae during second instar stage. Heterozygous intercrosses over balancer
chromosomes expect that 33% of progeny should be homozygous (N=at least 100 larvae
scored for each genotype, SmnD, SmnC, and SmnA Chi–Square p-values < 0.001, p-values
for remaining genotypes > 0.3). (D) Percentage of collected homozygous larvae five days
after egg laying (third instar stage) (Chi–Square p-values for all genotypes < 0.004). (E)
Percentage of collected homozygous larvae eight days after egg laying (pupation
timepoint) (Chi-Square p-values for all genotypes < 3x10-5). Note that several Smn and
Gemin3 mutant larvae exhibited extended survival late in larval development, but failed to pupate.
Smn and Gemin3 mutant larvae exhibit growth defects
Four days after egg laying, Smn and Gemin3 mutant larvae appeared runted in size (Fig.
3-5A). WT or Gem3B rev control larvae measured over 3mm in length and averaged
approximately 0.8mm in width. Conversely, Smn and Gemin3 mutant alleles generally
averaged only 1.5-2mm in length and 0.4mm in width (Fig. 3-5B,C) (except for SmnB, all p-values < 0.0001) SmnB homozygotes were intermediate in size averaging a significant
reduction in width (0.6mm, p-value < 0.002), but not length (2.9mm, p-value > 0.3).
75 Therefore, aside from reduced viability at the second-third instar timepoint, Smn and
Gemin3 homozygotes also showed overall growth defects.
76
77 Figure 3-5 Smn and Gemin3 mutants exhibit growth defects. (A) Images of Smn and
Gemin3 second-third instar larvae compared to WT and revertant (Gem3B rev) controls.
Tick marks on the ruler are 1mm apart. (B) Graph of average length of 2nd-3rd instar larvae (N=20 larvae scored for each genotype) (mutant length p-values < 0.0001 except for SmnB with a p-value > 0.3). (C) Graph of average width of 2nd-3rd instar larvae
(mutant width p-values < 0.002).
Smn and Gemin3 are required for proper motor function
SmnA and SmnB are previously characterized as having defects in motor function (Chan et
al., 2003). To test for a lack of movement, we placed larvae on plates and stimulated
movement with a needle. The distance they traveled over 20 seconds was traced and
measured (Fig. 3-6A,B). Second-third instar larvae were selected that were still alive, but
had an observable lack of movement. While Wt or Gem3B rev larvae traveled
approximately 14mm over 20 seconds, Smn and Gemin3 mutant larvae traveled at most
5mm over 20 seconds (mutant p-values < 5x10-6). Again, SmnB on average displayed a relatively less severe phenotype. Thus, the motor defects observed in SmnA and SmnB are also seen in other Smn and Gemin3 mutant larvae.
78
Figure 3-6: Smn and Gemin3 mutant larvae exhibit defects in motor function. (A) Control
(WT or Gem3B rev), Smn, or Gemin3 homozygous larvae were prodded with a needle to stimulate movement and traced over 20 seconds (yellow line). White scale bar represents
5mm. (B) Graph of average distance traveled over 20 seconds (N=10 larvae scored for
each genotype). Smn and Gemin3 mutant larvae movement was impaired relative to WT
controls (p-values < 5x10-6).
79
Smn and Gemin3 mutants develop neuronal abnormalities
The motor defects observed in Smn and Gemin3 mutant larvae may stem from an SMA- like neuromuscular defect. To test this possibility, we performed immunofluorescence for a neuronal specific antibody on larval preparations. Immunofluorescence with 22C10, an antibody that recognizes neurons, illustrates an organized network of motor neurons in wild-type larvae (Fig. 3-7). In contrast, SmnF mutant larvae instead exhibited excessive terminal arborization of motor neurons (arrows) Likewise, Gem3A and Gem3B P-element insertions also elicit similar neuronal branching phenotypes. Therefore, both members of the SMN complex appear to have a similar requirement in neuronal development.
80
Figure 3-7: Smn and Gemin3 mutants exhibit neuronal pathfinding defects. Larval tissue immunofluorescence with 22C10 marks neurons (scale bars represent 2mm). Compared to WT, mutant alleles of SmnF, Gem3A, and Gem3B larvae display excessive motor neuron
arborization (arrows).
Smn, Gemin3, and SmD2 mutants interact genetically
Initially, we failed to observe a genetic interaction between SmnF and Gem3B (Fig. 3-4),
so we tested genetic interactions between less severe alleles of snRNP biogenesis
components. Smnex33 was previously created by imprecise excision of SmnE (EY14384: P-
element insertion in the enhancer of Smn, Fig. 3-4A) (Rajendra et al., manuscript in
preparation). Smnex33 is a hypomorphic allele that exhibits reduced dSMN levels in the
thorax of adult flies. These flies exhibit severe muscular phenotypes (Rajendra et al.,
2006). While homozygous Smnex33 flies are viable and 100% flightless, heterozygotes
over a balancer are 88% flightless (Fig. 3-8A,B). The SmD2 P-element insertion
(EP3399) reduces appropriate splicing of the gene resulting again in a hypomorphic allele
(Fig. 3-4A, unpublished observations). SmD2 homozygotes are also viable (Fig. 3-8A), but 40% are flightless (Fig. 3-8B). Compounding the flightless problem are defects in
wing spreading and formation seen in 23% of SmD2 homozygotes (Fig. 3-8C). Doubly
homozygous Smnex33; SmD2 flies have reduced viability (Fig. 3-8A, Chi-Square p-value
= 0.001), and 100% are flightless with fully penetrant wing defects (Fig. 3-8B,C).
Interestingly, there is also a genetic interaction in heterozygous Smnex33; SmD2 flies
81 which appear normal, whereas 88% of heterozygous Smnex33 flies were flightless (Fig. 3-
8B).
Gemin3 mutant flies also interacted genetically with Smnex33 and SmD2 alleles.
Heterozygous Gem3B; homozygous SmD2 flies significantly exacerbate the SmD2
flightless phenotype from 40% to 100% and the wing phenotype from 23% to 90% (Fig.
3-8B,C, p-values < 1x10-10). Heterozygous Gem3B; heterozygous Smnex33 flies significantly increase the heterozygous Smnex33 flightless phenotype from 88% to 100%
(Fig. 3-8B, p-value = 0.03). Therefore, the SMN complex and snRNP proteins (Smn,
Gemin3, and SmD2) interact genetically in the development of musculature and wings
for flight.
82
83 Figure 3-8: Smnex33, SmD2, and Gemin3 interact genetically. (A) Observed genotypes
from Smnex33+/-, SmD2+/-, or Smnex33+/-; SmD2+/- intercrosses. Additionally,
Smnex33+/- flies were crossed with Gem3B+/-; Smnex33+/- flies and SmD2+/- flies were
crossed with Gem3B+/-; SmD2+/- flies. All alleles were maintained over a TM6 balancer
chromosome. Deviations from expected genotypic ratio were determined by calculating the χ2 p-value. Notably, fewer Smnex33-/-; SmD2-/- flies were observed than expected (B)
Flies with the given genotypes from the crosses in (A) were scored for flightlessness
(N=at least 50 flies for each genotype). Compared to 40% of SmD2-/- flies being
flightless, 100% of Gem3B-/-; SmD2-/- flies are flightless (p-value < 1x10-10).
Furthermore, compared to Smnex33+/- flies, Gem3B+/-, Smnex33+/- flies demonstrated a significantly increase in flightlessness from 88% to 100% (p-value = 0.03) (C) Flies with the given genotypes from the crosses in (A) were scored for the illustrated wing defects.
While 23% of SmD2-/- flies have wing defects, 100% of Smnex33-/-; SmD2-/- and 90% of
Gem3B+/-; SmD2-/- flies have the wing phenotypes (p-values < 1x10-10).
84 Discussion
In this manuscript we have characterized the Drosophila Gemin3 homolog and identified
key roles that this gene plays in organismal development. Sequence based alignments
demonstrated only a 56% similarity of dGemin3 compared to human Gemin3 across the
helicase motifs. While this conservation is not overwhelming, similar conservations are
also observed when comparing dSMN to human SMN as well as putative dGemin2
(CG10419) and dGemin5 (CG30149) homologues. To definitively demonstrate
homologous Gemin3 activity, we have verified a direct interaction between dGemin3 and dSMN. Additionally, dGemin3 and dSMN both localize to Drosophila Cajal bodies and ultimately function in snRNP Sm core assembly. Interestingly, the requirement of dSMN in Drosophila Sm core formation is not nearly as critical as that observed in human HeLa lysate (Shpargel and Matera, 2005). Thus the lack of conservation observed for members of the Drosophila SMN complex may reflect a diminished requirement or slightly different function in snRNP biogenesis.
Smn and Gemin3 are both essential for Drosophila life. Several mutant larvae die
at approximately the same second-third instar timepoints. Maternal deposition of Smn
transcripts was previously demonstrated and the observed lethality late in larval
development occurs when maternal stores become depleted (Chan et al., 2003). We
expect that this is also the case for Gemin3 due to the extreme similarity between the
mutant larvae. Amazingly, some Smn and Gemin3 mutant larvae survived for several
weeks. These larvae appeared to escape the initial wave of lethality at six-eight days after
egg laying, but failed to initiate pupation. Interestingly, Ensembl predicts that the
Drosophila homolog of Gemin5 is a gene known as rigor mortis (CG30149). Rigor
85 mortis mutations also exhibit delayed larval development with defects in larval molting
and pupation (Gates et al., 2004). However the protein directly interacts with several
members of the ecdysone signaling pathway required for pupal initiation. It will be
interesting to determine if the Drosophila SMN complex has developed a novel function
in this pathway.
The connection between snRNP biogenesis and SMA is certainly complicated and
not well understood. To date, several Smn and Gemin2 loss of function model organisms
have been characterized. Smn knockout mice expressing the human SMN2 transgene display severe motor neuron degeneration and muscular atrophy (Monani et al., 2000).
Smn heterozygous knockout mice have mild motor neuron degeneration that is enhanced on a Gemin2 heterozygous knockout background (Jablonka et al., 2002; Jablonka et al.,
2000). RNAi or morpholino mediated Smn or Gemin2 depletion in frogs or zebrafish
elicit motor neuron degeneration and uncontrolled pathfinding (McWhorter et al., 2003;
Winkler et al., 2005). We now demonstrate that mutation of two members of the
Drosophila SMN complex, Smn and Gemin3, both develop defects in motor function and neuronal development. The architecture of motor neurons in these mutants is comparable
to Smn and Gemin2 zebrafish loss of function models whereby uncontrolled, excessive
neuronal pathfinding is evident. Interestingly, mutant models lacking acetylcholine
receptors provoke the destruction of the neuromuscular junction, retraction of motor
neurons, and subsequent axonal overgrowth and branching (Hesser et al., 2006). This
phenotype may represent a futile attempt by motor neurons to re-establish the
neuromuscular junction. In addition to larval dSMN mutants, our laboratory has
previously observed SMA-like phenotypes in adult flies with hypomorphic Smnex33 and
86 SmD2 mutations (Rajendra et al., 2006). Interestingly, the SmD2 phenotypes are
enhanced on the homozygous Smnex33 or heterozygous Gemin3 mutant backgrounds.
Thus the SMA-like pathology observed in Smn and Gemin3 mutants may at some part be
due to snRNP Sm core assembly defects. Consistently, motor neuron defects generated through RNAi knockdown of SMN and Gemin2 in zebrafish and frog embryos can be rescued by injection of purified, assembled snRNPs (Winkler et al., 2005). Therefore, snRNP biogenesis may really be a major causative factor in SMA pathology. Additional tissue specific functions for SMN in neuromuscular development may compound the problems initiated by snRNP biogenesis defects (Fan and Simard, 2002; Rossoll et al.,
2002; Shafey et al., 2005). Understanding the localization and role of Gemin3 in SMN related tissue specific functions such as axon pathfinding will require further future analysis.
87 Chapter IV
Discussion and future directions
88 Over the course of my graduate studies, I have addressed several topics that aid in understanding SMN complex function in snRNP biogenesis and Spinal Muscular
Atrophy. Notably, by using RNAi, I have discovered that SMN, Gemin2, Gemin3, and
Gemin4 are all required for proficient snRNP Sm core assembly in vitro and in vivo.
Cellular deprivation of the snRNP biogenesis pathway incites Cajal body (CB) breakdown. Several Type I SMA point mutations in SMN are defective in snRNP assembly. And finally mutation of an additional snRNP biogenesis pathway member,
Gemin3, elicits Spinal Muscular Atrophy (SMA) like phenotypes (similar to SMN mutants). These studies have shed further insight into the fields of snRNP biogenesis and
SMA and raise several new lines of inquiry.
What factors regulate Cajal body homeostasis?
The study of Cajal body function and formation is really in its infancy. After lying
dormant for nearly a century, the advent of immunofluorescence has allowed
identification of a plethora of CB localizing proteins (Matera, 1999a; Matera and
Shpargel, 2006). Only in the last five years has CB function in RNP biogenesis begun to
take shape. CBs now appear to be an efficiency center for snRNP, snoRNP, and telomerase RNP maturation and assembly (Jady et al., 2003; Jady et al., 2006; Matera and
Shpargel, 2006; Tomlinson et al., 2006; Verheggen et al., 2002). In fact the loss of CBs in coilin knockout mice has a pronounced effect on organismal viability and fertility
(Tucker et al., 2001). However these nuclear structures are quite mysterious. CBs are not present in all tissues and cell types, and CBs exhibit sudden formation and directed movements (Platani et al., 2002; Young et al., 2000). With this in mind, the factors
89 controlling CB formation, movement, and response to environmental cues is not well
understood.
Coilin was initially designated the CB marker protein because
immunofluorescence studies localize the protein almost exclusively to CBs (Andrade et
al., 1991). Our laboratory subsequently has demonstrated that coilin is required for
formation of a “true” CB in its entirety (Tucker et al., 2001). Mouse embryonic fibroblast cell lines (MEFs) derived from coilin knockout mice contain “residual” CBs. In the absence of coilin, the snoRNP, snRNP/scaRNP, and SMN complex dissociate into separate nuclear bodies (Jady et al., 2003; Tucker et al., 2001). Thus it seems that coilin is the glue that holds the CB together.
I have now discovered in Chapter AI that the coilin Carboxy-terminus (C-
terminus) controls CB number. Elimination of the coilin C-terminus renders the protein
unregulated and permits excessive formation of CBs. In contrast, mutation of C-terminal
serine and threonine residues to constitutively unphosphorylated states inhibits CB
formation. Therefore, while the oligomerization domain of coilin resides in the Amino-
terminus of the protein, phosphorylation of the C-terminus regulates accessibility of the
oligomerization domain. However, in coilin knockout MEFs, snRNPs and scaRNPs can
still aggregate together in nuclear foci (Jady et al., 2003). This suggests that some
secondary structural component persists in the absence of coilin. In Chapter II, I observed
that RNAi of SMN and snurportin lead to severe snRNP biogenesis defects and loss of
CB homeostasis. Therefore CBs breakdown when levels of nascent snRNPs entering the
nucleus are severely compromised. Consistently, inhibiting the snRNP biogenesis
pathway with Leptomycin-B also breaks down CBs, while overexpression of snRNP
90 proteins induces the formation of CBs (Carvalho et al., 1999; Sleeman et al., 2001). This also helps to explain why CBs tend to be more prevalent in cell types with elevated metabolic activity. Overall, the snRNPs themselves or a yet unidentified protein associated with snRNPs provides some scaffolding activity for nuclear body formation.
Coilin on the other hand organizes the CB subcomponents into a common nuclear body and regulates number and size of CBs.
Gemin proteins function in snRNP assembly, but by what mechanism?
In Chapter II, I delve into the requirement of individual SMN complex proteins in snRNP
Sm core formation. Remarkably, SMN and Gemins2-4 are conclusively required for efficient Sm core assembly. However we still don’t comprehend the exact mechanism of
SMN complex action in this process (see model diagrams in Fig. 4-1). The SMN complex binds directly to snRNAs while individual Gemin proteins can directly interact with distinct subsets of Sm proteins (Gubitz et al., 2004). So it is possible that Gemin proteins aid in recruitment and proper alignment of Sm proteins into the Sm core structure (Fig. 4-
1A). Additionally, Gemin6 and Gemin7 dimerize forming a tertiary structure that resembles Sm fold domains on Sm proteins (Ma et al., 2005). Conceivably, this structure may provide a scaffold for Sm protein attachment, molding them into the appropriate core structure (Fig. 4-1B). Alternatively, the Gemin6-Gemin7 structure may be able to compete with Sm proteins in binding to other components of the SMN complex. This competition could provide a mechanism to free Sm proteins from the SMN complex permitting assembly on the snRNA. Gemin3 and Gemin4 also associate in their own subcomplex outside of the SMN complex (Mourelatos et al., 2002). Examining the
91 crystal structure of these additional subgroupings will provide insight into whether this is
a common theme among Gemins.
Gemin3 is a member of the DEAD box family of RNA helicases. It has ATP-
dependent RNA helicase activity (Yan et al., 2003). However, RNA helicase activity is not restricted to unwinding double-stranded RNA. RNA helicases are also known to remove proteins from single-stranded RNA (Jankowsky, 2005), so Gemin3 could function in separating the SMN complex off the snRNA to allow for appropriate Sm protein core attachment (Fig. 4-1C). Additionally, Gemin3 ATPase activity may act as a structural switch, conforming the SMN complex to permit snRNP assembly Fig. 4-1A,C).
It will be important to examine the phenotypic effects in detail when Gemin3 or SMN is knocked down by RNAi. Specifically, tagged snRNA purification in the absence of
Gemin3 should pull out an intermediate snRNP/SMN complex providing insight into
Gemin3’s mechanistic action. Furthermore, RNAi of Gemin3 followed by rescue utilizing exogenous Gemin3 constructs can test the requirements of various Gemin3 domains in snRNP assembly. For example, after Gemin3 RNAi, rescue with a Gemin3 construct mutated for ATPase activity, RNA helicase activity, Gemin4 binding, or Sm protein binding could identify what Gemin3 function is required in snRNP assembly.
92
93 Figure 4-1: Models for SMN complex function in snRNP Sm core assembly. (A) SMN
complex members interact with different subsets of Sm proteins. These interactions may
organize Sm proteins into proper conformations prior to assembly on the snRNA. A
molecular switch, such as ATP hydrolysis by the Gemin3 RNA helicase, may alter SMN
complex structure to facilitate Sm protein assembly onto the snRNA. (B) The Gemin6-
Gemin7 heterodimer structurally resembles Sm-fold motifs of Sm proteins. Therefore, the
Gemin6-Gemin7 dimer may initiate a scaffold for Sm protein attachment and pre-
formation of the Sm ring. Gemin6-Gemin7 dimers most closely resemble the SmB/D3
heterodimer, so replacement of these Gemins with SmB/D3 may drive assembly of the
pre-formed Sm ring onto the snRNA. (C) The SMN complex may play an active role in
snRNP assembly. The ATP hydrolysis and helicase activity of Gemin3 may mediate
SMN complex removal from the snRNA and/or active assembly of Sm proteins onto the
snRNA.
Are Gemins required for additional snRNP biogenesis steps?
Following Sm core formation, the SMN complex persists on the snRNP throughout
downstream cytoplasmic steps (Narayanan et al., 2004a; Narayanan et al., 2002). SMN
directly interacts with trimethylguanosine synthase 1 (Tgs1) enzyme that
hypermethylates the snRNA cap (Mouaikel et al., 2003). SMN complex function in
snRNA cap hypermethylation can be tested in vitro. A radiolabeled snRNA,
precomplexed with Sm proteins, can be incubated with siRNA transfected cellular lysate.
In vitro cap hypermethylation can be assayed by immunoprecipitation using a TMG cap specific antibody.
94 The SMN complex even remains with the snRNP through nuclear import. SMN
itself interacts directly with importin-β, however the entire complex is required for snRNP import (Narayanan et al., 2004a; Narayanan et al., 2002). Interestingly, Gemin4
protein sequence annotation predicts that a putative nuclear localization signal resides in
the N-terminus of the protein. In Chapter AII, exogenous Gemin4 localization is almost
entirely nuclear. Furthermore, Gemin4 overexpression enhances both Gemin3 and SMN
nuclear localization. Additional experiments will be required to define if Gemin4 is
indeed required for SMN complex and snRNP import in vivo. Conceivably, digitonin
permeabilized import assays may be the best way to test endogenous import activity.
Fluorescently labeled snRNPs or SMN complex members could be incubated in Gemin4
siRNA treated cellular lysate and tested for nuclear uptake. In the absence of Gemin4, the
import of these substrates may be kinetically hindered or be completely absent.
Can Gemin snRNP biogenesis defects promote Spinal Muscular Atrophy?
Upon establishing that several Gemin proteins are required for snRNP biogenesis, in
Chapter III, I examined mutant Gemin3 Drosophila lines to better comprehend the impact
this pathway has in SMA. Strikingly, Drosophila Gemin3 mutants phenocopy several
Smn mutations. Mutations in both genes elicit defects in larval viability, pupation, motor
function, and neuronal pathfinding. Furthermore, Drosophila Gemin3, Smn, and SmD2
mutations genetically interact by enhancing defects in neuromuscular development and
wing formation of the adult fly. Because Gemin3 and Smn mutants genetically interact
with SmD2, snRNP biogenesis defects most likely have some involvement in the
95 pathogenesis of their mutant phenotypes. Therefore, loss of SMN complex function in
snRNP biogenesis does contribute to SMA like phenotypes.
While Drosophila Gemin3 functions in snRNP assembly, it is still possible that
novel tissue specific functions may contribute to the observed neuronal phenotypes in
Gemin3 mutants. SMN itself has been implicated in neuronal growth cone transport of
mRNAs for neurite pathfinding and extension (Rossoll et al., 2003). More recently,
localization studies have placed Gemin2, Gemin6, Gemin7, and unrip together with SMN
in growth cones, so it is certainly possible that Gemin3 may also reside within this
complex (Sharma et al., 2005). In Chapter AII, expression of exogenous Gemin3 targets
to the Golgi complex along with SMN. As packaging of vesicles destined for neuronal
growth cones originates from the Golgi complex, Gemin3 could potentially aid in
delivery of the SMN complex to these neuronal structures. RNAi of Gemin3 in neuronal
cell types could probe this possibility.
Alternatively, Gemin3 may exhibit muscle specific functions. RNAi of SMN in
myoblast muscle precursors inhibits myoblast fusion and myotube formation (Shafey et
al., 2005). Intriguingly, Drosophila SMN and Gemin3 both interact with Rac2 by high-
confidence yeast-2-hybrid analysis (Fly GRID database). Rac2 is a GTPase capable of controlling changes in cellular shape. Drosophila mutations in Rac2 exhibit defects in
myoblast fusion, perhaps through a cellular pathway involving dSMN and dGemin3
(Hakeda-Suzuki et al., 2002). Additionally, SMN has distinct localization patterns to neuromuscular junctions and Z-discs of muscle sarcomeres (Fan and Simard, 2002;
Rajendra et al., manuscript in preparation). Thus, a thorough characterization of Gemin3 neuronal and muscular localization patterns will be critical in determining the exact mode
96 of pathogenesis in Gemin3 mutant larvae. Ultimately, a combination of defects in both
snRNP biogenesis and tissue specific functions may underlie the Gemin3 and Smn mutant phenotypes.
Which SMN functions are required for appropriate neuromuscular development?
While other snRNP biogenesis components such as Gemin3 and SmD2 are required for
neuromuscular development, the question remains: Are SMN loss of function phenotypes
due to snRNP biogenesis defects? As Smnex33 flies interact genetically with SmD2
mutants, it is apparent that snRNP biogenesis is at least partially responsible for Smn
mutant phenotypes (Chapter III). Consistently, SMA Type I point mutations in SMN
were defective in snRNP assembly assays (Chapter II). In recent work, knockdown of
Smn and Gemin2 in zebrafish and frogs lead to motor neuron degeneration (Winkler et al., 2005). These phenotypes were rescued by supplementation with purified, assembled snRNPs. Therefore, bypassing the requirement of SMN in snRNP biogenesis can, at some level, prevent SMA-like disease progression.
My focus of study has centered on the hypothesis that mutation in other snRNP
biogenesis pathway components will phenocopy Smn mutant phenotypes. Despite
proving this hypothesis true, Smn mutant phenotypes may still originate from a
combination of tissue specific and snRNP biogenesis defects. In future research, we will
need to pick apart these various SMN functions. We will need to generate point or
domain mutations in SMN that specifically eliminate one function without disrupting
another. For example, can a mutation that specifically perturbs snRNP assembly without disrupting neuronal growth cone transport generate SMA phenotypes? However this plan
97 may prove troublesome, because SMN is very sensitive to mutagenesis and in several
cases, the mutant protein gets degraded. Alternatively, while my thesis has described genetic interactions between Smn and snRNP biogenesis components, future work can explore genetic interactions between Smn and genes important for neuronal pathfinding
(such as profilin) or muscle cell fusion (such as Rac2).
Does the primary SMA defect originate in neurons or muscles (or both)?
I feel the most intriguing question in SMA biology pertains to defining the causative
tissue and pathway that creates the primary defect in human SMA patients. These
questions will require clarification so that drug or gene therapy delivery schemes can be
optimized to target the appropriate site of action. Our laboratory has made great strides in
generating Smnex33, an adult Drosophila model for SMA. This hypomorphic enhancer
mutation produces an adult neuromuscular phenotype (Rajendra et al., manuscript in
preparation). However, one caveat for this model is that the mutation is essentially tissue
and temporal specific because reduction in dSMN only occurs in the adult thorax.
Alternatively, the human SMA condition stems from diminished SMN levels throughout
the entire body. The mouse SMA Type I model, Smn knockout expressing a human
SMN2 transgene, is the more accurate representation of the human allelic combination
(Monani et al., 2000). Furthermore, mice and humans have much more conservation of
physiology. Therefore it will be more useful to use the mouse SMA type I model to
elucidate the tissues and pathways of SMA origin. Tissue specific Smn knockout mice
have already demonstrated that complete elimination of Smn from either neurons or
muscles can each individually lead to neuromuscular phenotypes (Cifuentes-Diaz et al.,
98 2001; Frugier et al., 2000). But these experiments are misleading because the true disease
condition arises from a hypomorphic loss of SMN. This point is further emphasized by
the fact that complete knockout of Smn from the liver leads to developmental liver defects (Vitte et al., 2004). Therefore, complete Smn knockout in any tissue would
probably cause tissue specific phenotypes because of a complete loss of snRNP
biogenesis in those cell types.
To probe hypomorphic tissue specificity of SMA, I propose to create a
conditional SMN2 knock-in mouse (Fig. 4-2A). This allele will introduce exons 6-8 of
SMN2 downstream of the entire endogenous Smn locus (the Smn-loxP-SMN2 allele).
Under normal conditions this allele will express a full-length Smn transcript. However,
Cre-recombinase will recombine the loxP sites to create a hybrid locus fusing Smn exons
1-5 with SMN2 exons 6-8 (Fig. 4-2A). This should recreate the same hypomorphic situation observed in SMA patients, but this condition can be induced in a tissue specific manner. This targeting design is feasible because the exon 5-6 junction is conserved between Smn and SMN2, so there will be no frameshift disruption of the protein sequence. Furthermore, expression of SMN2 in the mouse preserves the gene’s splicing defect and hypomorphic SMN production (Monani et al., 2000). Excessive copies of the
SMN2 transgene can completely rescue Smn knockout phenotypes, thus the mouse and human proteins appear to be interchangeable.
The Smn-loxP-SMN2 knock-in line can be crossed to mice expressing Cre- recombinase in tissue specific patterns (Fig. 4-2B). Crossing CMV-Cre (expressed in all tissues) into Smn-loxP-SMN2-/- mice will induce hypomorphic SMN production in all tissues. These mice should resemble Smn-/-; SMN2 transgenic mice that exhibit early
99 postnatal motor neuron degeneration, muscular atrophy, and lethality. Now how does this phenotype compare to Smn-loxP-SMN2-/- mice expressing Cre-recombinase in neurons
(driven by the Rat Neuron Specific Enolase promoter-NSE) or in skeletal muscle (driven by the skeletal muscle α-actin promoter)? Will hypomorphic SMN expression from these individual tissues recapitulate the early postnatal phenotypes or will this require hypomorphic SMN production from both sides of the neuromuscular junction? These mice can also be used to define the timecourse of the neuromuscular phenotypes. By collecting E14-E18 embryonic lumbar preparations, immunofluorescense for pre-synaptic and post-synaptic markers (such as acetylcholine receptor) can identify when the motor neuron degeneration phenotypes occur. Do the motor neurons fail to form neuromuscular junctions, or do they form a synapse that is subsequently lost leading to motor neuron degeneration?
Depending on which tissue is responsible for the hypomorphic SMA phenotype, a tissue specific SMN transgene can be designed to rescue mutant phenotypes. The conditional GFP-SMN transgene (GFP-loxP-Smn) (Fig. 4-2A) is designed to express
GFP from an Smn promoter when placed in a wild-type background. But when crossed to
NSE-Cre, the loxP sites will bring Smn in frame with the GFP in neurons only.
Therefore, if neurons appear to be the cell type responsible for the hypomorphic SMN phenotype, then Smn-loxP-SMN2-/- mice that have undergone Cre-mediated recombination in their germline can be crossed onto a NSE-Cre; GFP-loxP-Smn background to see if Smn expression in neurons alone can rescue a ubiquitous SMN hypomorph.
100 Finally, the Smn-loxP-SMN2-/- mice can be utilized to assess if the phenotypes
generated by hypomorphic SMN production in a specific tissue are due to defects in
snRNP biogenesis. Through a search of the international gene trap consortium database, several gene trap ES cell lines are available for snRNP biogenesis components. Genetraps disrupting snRNA export (PHAX), the SMN complex (SMN, Gemin2,4,5), Sm proteins
(SmB, D1, D2, D3, LSm1-7), and snRNP import (snurportin) are all available.
Furthermore, a Gemin3 genetrap may be available from Lexicon Genetics. Smn-loxP-
SMN2 -/-; tissue specific Cre mice can be crossed onto a background that is heterozygous for a given snRNP biogenesis genetrap to test for enhancement of mutant phenotypes. It will be fascinating to study if any of these components can modify SMA severity.
101
102 Figure 4-2: Experimental analysis of tissue specificity in SMA. (A) Design to create
tissue specific SMN hypomorphic mice or tissue specific transgenic GFP-Smn mice.
Allele 1: exons 6-8 of SMN2 can be introduced downstream of the entire endogenous
Smn locus by gene targeting. This allele (Smn-loxP-SMN2) will express wild-type Smn.
Allele 2: strategic placement of loxP sites will allow for Cre-mediated recombination to
create a hybrid Smn (exons 1-5) and SMN2 (exons 6-8) gene. While allele 1 produces
full-length Smn, allele 2 will alter splicing of the gene to resemble SMN2 gene products
(only 20% of full length SMN will be produced). Allele 3: the GFP-loxP-Smn transgene will express GFP from the Smn promoter in a wild-type background. Allele 4: the loxP sites of GFP-loxP-Smn will fuse Smn with GFP when recombined by Cre. (B)
Schematics of genetic crosses. Smn-loxP-SMN2-/- hypomorphic SMN production can be
driven throughout the entire organism by CMV-Cre, throughout neurons by the Rat
(Neuron Specific Enolase) NSE-Cre, throughout skeletal muscle by α-actin-Cre, or throughout liver tissue by α-fetoprotein-Cre. Phenotypes generated by the organism wide
Smn-loxP-SMN2-/- hypomorph can then be tested for rescue by the GFP-loxP-Smn transgene expressing GFP-Smn in tissue specific patterns.
Why does snRNP biogenesis have such a profound effect on motor neurons?
The evidence available from this thesis indicates that snRNP biogenesis is a factor in
neuromuscular development. At first glance it is difficult to comprehend how loss of a
ubiquitous housekeeping function can display such a dramatic tissue specific phenotype.
However, SMA phenotypes are hypomorphic, not null mutations. So compared to other cell types, neurons must have some elevated sensitivity to perturbed snRNP biogenesis
103 levels, but why (see models for SMA tissue specific pathogenesis in Fig. 4-3)? One possible explanation is that the perturbation in snRNP biogenesis created by mutation of the SMN complex and/or SmD2 may be enhanced in neurons due to tissue specific factors (Fig. 4-3A). For example in neurons, SmB is replaced by SmN (McAllister et al.,
1988). This type of tissue specific factor may kinetically alter the snRNP biogenesis pathway or splicing itself, rendering neurons more susceptible to snRNP biogenesis defects. Motor neurons have an incredible metabolism and produce excessive amounts of specialized mRNAs (such as for axon pathfinding). As a result, lower rates of splicing could have a more pronounced effect on this cell type because these mRNAs can’t be spliced with the required efficiency (Fig. 4-3B). Additionally, neurons in general display an abundance of alternative splicing. Hence, these particular transcripts, already compromised with suboptimal splice sites, would be most susceptible to problems created by impaired spliceosome production (Fig. 4-3C). It is entirely possible that a combination of factors promote the motor neuron pathology. The inefficient snRNP production may sensitize all cell types, but motor neurons may have secondary problems in neuronal transport or muscle cell fusion that compound the cellular trauma above a particular threshold (Fig. 4-3D). The main feature that distinguishes motor neurons is the neuromuscular junction. Therefore snRNP biogenesis defects may compound with problems in neuromuscular junction formation and maintenance to incite SMA phenotypes (Fig. 4-3D). In fact, mouse models of SMA exhibit motor neuron degeneration with no loss of sensory neurons, so the neuron-muscle interplay may be a major pathogenic factor (Jablonka et al., 2006). Finally, motor neurons may have evolved a specialized snRNP biogenesis function (Fig. 4-3E). Splicing can even take place in
104 neuronal dendrites (Glanzer et al., 2005). Thus axonal localization of the SMN complex may aid in snRNP transport to regionalized high concentrations of splicing in axons.
Finally, SMN and Sm proteins could conceivably assemble onto some novel type of RNP required for growth cone maintenance and extension (Fig. 4-3F). While all of these models highlight possible explanations for neuronal pathogenesis in SMA, the same explanations could be applied to muscle tissue, as it is not known which tissue leads to
SMA pathology (or if defects in both tissues lead to SMA). The experiments mentioned earlier, designed to pick apart SMN function in various cellular pathways, will be instrumental in deciphering between these hypothesized modes of pathogenesis.
While my work has advanced our understanding of snRNP biogenesis and SMA, the disease is incredibly complex and will require many more dedicated researchers to unravel its true nature. The questions and experiments that I have proposed will be critical in deciphering SMA pathogenesis. In due time we will have a greater understanding of the disease and hopefully this information will ultimately aid in SMA patient therapy.
105
Figure 4-3: Models for SMA tissue specific pathogenesis. Neurons (or muscle tissue) may be particularly susceptible to SMN complex and/or SmD2 mutant snRNP biogenesis defects because: (A) Tissue specific splicing factors (such as SmN which replaces SmB in neurons) may alter the kinetics of snRNP biogenesis or splicing to enhance neuronal defects in this pathway. (B) Neurons transcribe and splice particular mRNAs (such as growth cone mRNAs) at very high levels. These mRNAs may be particularly susceptible to snRNP biogenesis defects. (C) Neurons have numerous alternatively spliced mRNAs.
106 These alternative transcripts have weak splice sites and may be most sensitive to snRNP
biogenesis levels. (D) Neurons or skeletal muscle may exhibit SMA pathology if the
initial snRNP biogenesis defect is compounded by a lack of tissue specific functions
performed by SMN (or the SMN complex). (E) These tissues may have developed tissue
specific snRNP biogenesis and splicing functions. For example, snRNPs may need to be assembled and transported to neuronal growth cones for specialized splicing events. (F)
As mutation of the SMN complex and SmD2 yield neuromuscular phenotypes, these
proteins may alternatively be assembled into a novel type of RNP with specialized
neuronal or muscular functions.
107 Appendix Chapter I
Control of Cajal body number is mediated by the coilin C-terminus
Karl B. Shpargel, Jason K. Ospina, Karen E. Tucker, A. Gregory Matera* and Michael D. Hebert#
Department of Genetics Case Western Reserve University School of Medicine Cleveland, OH 44106-4955 USA
#Present address: Dept. of Biochemistry Univ. of Mississippi Medical Center Jackson, MS 39216-4505
*Corresponding author Tel: 216-368-4922 Fax: 216-368-0491 Email: [email protected]
Note: This chapter is a manuscript published in the Journal of Cell Science 116(Pt. 2): 303-312.
108 Abstract Cajal bodies (CBs) are nuclear suborganelles implicated in the posttranscriptional
maturation of small nuclear and small nucleolar RNAs. The number of CBs displayed by
various cell lines and tissues can be quite variable and factors that control CB numbers within a given cell have yet to be described. In this report, we show that specific regions within the C-terminus of coilin, the CB marker protein, are responsible for regulating the number of nuclear foci. Despite the fact that the coilin N-terminal domain is responsible for its self-oligomerization activity, truncation or mutation of predicted sites of phosphorylation in the conserved C-terminal region leads to a striking alteration in the number of nuclear bodies. Similarly, coilin constructs from various species display differential propensities to form nuclear foci when expressed in heterologous backgrounds. We mapped the domain responsible for this variability to the coilin C- terminus utilizing chimeric proteins. Furthermore, the activities responsible for regulating coilin self-association must reside in the nucleus, as constructs lacking critical nuclear localization sequences fail to form foci in the cytoplasm. Factors controlling the putative signal transduction cascade that phosphorylates coilin are also discussed. The results point to a model whereby phosphorylation of the coilin C-terminus regulates the availability of the N-terminal, self-interaction domain.
109 Introduction It is increasingly evident that the nucleus, like the cytoplasm, contains a wide variety of subdomains that perform discrete functions. These nuclear substructures are dynamic and their presence likely reflects underlying cellular processes (Dundr and Misteli, 2001;
Matera, 1999a; Spector, 2001). For example, the most easily recognizable nuclear domain, the nucleolus, marks the site of ribosome synthesis, and possibly other cellular events (Andersen et al., 2002; Olson et al., 2000). The functions of other nuclear domains, such as Cajal bodies (CBs) and Gemini bodies (gems) are less clear, although these domains have been implicated in aspects of small nuclear ribonucleoprotein biogenesis (Gall, 2000; Matera, 1999b). Recent work (Darzacq et al., 2002; Jady and
Kiss, 2001) has revealed the existence of small RNAs (scaRNAs) that are localized specifically in the Cajal body and are important for the posttranscriptional modification of U snRNAs, thus strengthening the role for CBs in snRNP biogenesis. Similarly, nuclear bodies in yeast and mammals are thought to be important for maturation of U snRNAs (Mouaikel et al., 2002; Verheggen et al., 2002). Additionally, the gem contains the survival of motor neurons protein, SMN, which is the protein mutated in patients with
Spinal Muscular Atrophy (Lefebvre et al., 1995; Liu and Dreyfuss, 1996). SMN is thought to play a crucial role in cytoplasmic Sm protein assembly onto snRNAs (Fischer et al., 1997; Meister et al., 2000; Meister et al., 2001). A nuclear role for the SMN complex is less-well defined, but the available evidence points to snRNP recycling
(Pellizzoni et al., 1998).
In addition to their ability to move throughout the nucleus, another fascinating feature of nuclear bodies is their capacity to form within the nuclear milieu without apparent support structures (Muratani et al., 2002; Platani et al., 2002). Several nuclear
110 domains, such as Sam68 bodies, PML bodies, Cajal bodies, and gems contain ‘signature’
proteins (Sam68, PML, coilin and SMN, respectively) that are used as markers for each domain. Characterization of these marker proteins has revealed that they each share the ability to self-associate (Chen et al., 1997; Hebert and Matera, 2000; Lorson et al., 1998;
Perez et al., 1993). Indeed, self-association may be a common theme utilized by nuclear
body marker proteins to provide a scaffold upon which other components of the
respective domain can then coalesce (Hebert and Matera, 2000; Misteli, 2001). Despite
elucidation of marker proteins and characterization of a myriad of other protein and RNA
constituents in nuclear bodies, understanding of mechanisms that regulate the size and
number of these nuclear domains is lacking. Such mechanisms must exist and are likely
propagated either through intrinsic properties of the marker proteins (e.g. structure or
expression level) or extrinsic factors that affect protein-protein interaction. Interestingly,
Lamond and co-workers recently showed that transient upregulation of SmB protein in
cells that normally do not display Cajal bodies promotes the nucleation of these structures
(Sleeman et al., 2001). This finding supports the idea that Cajal bodies play a role in
some aspect of snRNP metabolism, and suggests snRNP expression levels can affect the
localization of coilin.
Post-translational modification of marker proteins may trigger conformational changes that facilitate self- or other protein interactions crucial for nuclear body
assembly. PML bodies, for example, require the SUMO modification of the PML protein
in order to properly form (Ishov et al., 1999; Müller et al., 1998). Additionally, the
phosphorylation state of Sam68 affects its ability to self-oligomerize (Chen et al., 1997)
and thus may affect its capacity to form nuclear bodies. Likewise, coilin
111 hyperphosphorylation, which occurs during mitosis (Carmo-Fonseca et al., 1993), results
in a reduction in self-association (Hebert and Matera, 2000). Cajal bodies disassemble
during mitosis and reform at early- to mid-G1 phase (Andrade et al., 1993; Carmo-
Fonseca et al., 1993). Thus, the phosphorylation status of coilin might influence Cajal
body formation. We have also shown that phosphorylation can control localization by
exposing or sequestering a putative nucleolar localization signal (Hebert and Matera,
2000). This idea has precedence, as the function of the retinoblastoma protein is thought
to be controlled by a similar process of sequential phosphorylation events (Harbour et al.,
1999).
We have recently shown that symmetrical dimethylation of arginines within the
coilin RG box motif is vital for the in vivo incorporation of the SMN complex into CBs
(Hebert et al., 2002a). Without this modification, the SMN complex fails to localize in
CBs and forms gems. Therefore, coilin methylation, like phosphorylation, can affect
nuclear body formation and composition.
We are interested in understanding how nuclear body size and number are
regulated. In this report, we set out to identify regions of coilin, if any, that are
responsible for controlling these processes. Surprisingly, we found that coilin proteins
from human, mouse and frog displayed different localization patterns when expressed in,
both HeLa cells and mouse embryonic fibroblasts (MEFs). Additional experiments using
MEFs derived from coilin knockout embryos showed that specific residues within the
carboxy-terminus of coilin are important for the proper regulation of CB numbers.
Finally, although the coilin self-association domain resides in the amino-terminus of the protein, point mutations and small truncations in the distal C-terminus affect its ability to
112 form or target to CBs. We conclude that coilin contains an intrinsic nuclear body formation potential, shared among other nuclear body marker proteins, but is subject to increasing layers of regulation from frog, to mouse, to human.
113 Materials and Methods
Constructs and Mutagenesis
GFP-tagged full-length human and mouse coilin constructs were cloned into pEGFP vectors (Clontech) as described (Hebert and Matera, 2000; Tucker et al., 2001). Chimeric constructs were generated by cloning the human coilin N-terminus or C-terminus into the corresponding portion of the GFP-mouse coilin construct utilizing a conserved HindIII site at 1434-bp in the mouse coilin mRNA sequence (Genbank Accession: 7710007). All truncation constructs, deletions, and point mutations were generated by QuickChange mutagenesis (Stratagene). Constructs were verified by sequencing. Primer sequences can be provided upon request. SMN, fused to a myc-tag, was a kind gift from G. Dreyfuss
(University of Pennsylvania).
Cell Culture, Transfection, and Immunofluorescence
The mouse embryonic fibroblast cell line was established from coilin -/- knockout mice as described (Tucker et al., 2001). The coilin knockout MEF or HeLa cells were grown in
DMEM (GIBCO BRL), supplemented with 10% FBS (GIBCO BRL). Cells were grown to subconfluency on chambered slides (Nunc) and transfected for 24 hours. MEF cells were transfected using LippofectAMINE (GIBCO BRL); HeLa cells were transfected using SuperFect (Quiagen) as directed. Cells were fixed in 4% paraformaldehyde, extracted in 0.5% Triton X-100, and processed for microscopy as previously described
(Frey and Matera, 1995) Immunofluorescence was carried out with myc (1:40; Santa
Cruz Biotechnology), R288 (1:100;(Andrade et al., 1993), and R508 (1:200; (Chan et al.,
1994b) antibodies.
114 Results
The coilin self-interaction domain affects epitope recognition.
Previously, we showed that the amino terminus of human coilin encodes a self-
association domain that mediates targeting to Cajal bodies (Hebert and Matera, 2000).
Constructs lacking this domain fail to accumulate in CBs (Hebert and Matera, 2000).
Subsequently, we analyzed one of these constructs, myc-coilin(94-576), with two different coilin polyclonal antibodies, R508 (Chan et al., 1994a) and R288(Andrade et al.,
1993). Since these antisera were each raised against C-terminal regions of coilin, both of the epitopes should be present in the N-terminal deletion construct (Fig. A1-1A).
However, despite high levels of coilin(94-576) expression in transfected HeLa cells (as assessed by antibodies to the myc-tag), antibody R508 does not detect the mutant protein
(Fig. A1-1B, top panels). Indeed, the signal obtained from R508 was derived almost entirely from the endogenous coilin, whereas the ectopically expressed myc-coilin(94-
576) was not detected. On the other hand, nucleoplasmic staining of coilin(94-576) was readily observed by staining with R288 (Fig. A1-1B, bottom panels). Thus, although coilin(94-576) contains the epitopes for both antibodies, only R288 reacts with this mutant protein. Notably, R508 was raised against a peptide corresponding to the extreme
C-terminus of coilin, whereas R288 was raised against the entire C-terminal domain (Fig.
A1-1A). We therefore conclude that the presence of the coilin self-association domain, which resides in the N-terminus, can affect C-terminal epitope recognition.
115
Figure A1-1: Deletion of the coilin self-association domain affects epitope recognition.
(A) Schematic of human coilin showing the location of the self-interaction/CB targeting domain (Hebert and Matera, 2000), two nuclear localization signals (NLS), nucleolar localization signal (NoLS) (Hebert and Matera, 2000), and the RG box which mediates interaction with SMN (Hebert et al., 2001). The regions of coilin that were used to generate anti-coilin antibodies R288 (Andrade et al., 1993) and R508 (Chan et al., 1994a) are also indicated. (B) Coilin lacking the self-association domain is not recognized by anti-coilin Ab R508. HeLa cells transfected with myc-tagged coilin lacking the first 93 amino acids were subject to staining with anti-myc (left panels) or anti-coilin (right panels) antibodies.
116 The carboxy-terminus of coilin regulates Cajal body number.
In addition to finding that coilin is a self-associating protein and that this domain affects
C-terminal epitope recognition, we also have shown that fusion of GFP to the coilin C- terminus results in formation of numerous foci (Fig. A1-2)(Hebert and Matera, 2000).
Normal coilin localization (diffuse nucleoplasmic plus bright foci) is obtained upon fusion of GFP to its N-terminus (Fig. A1-2). However, fusion of the GFP moiety to the
C-terminus results not only in numerous coilin foci, but a reduction in the nucleoplasmic pool of protein (Fig. A1-2). Given the results in Figure 1, it seemed plausible that fusion of GFP to the C-terminus of coilin produced an aberrantly-folded protein, culminating in a large increase in the number of coilin foci. In other words, the coilin C-terminus regulates CB number. This hypothesis is supported by a truncation mutant, coilin(1-481), that produces the same phenotype observed for the coilin-GFP fusion (Fig. A1-2).
Furthermore, larger C-terminal truncations of coilin result in varied patterns of localization, from nucleolar accumulations to large nucleoplasmic inclusions, called pseudo-CBs (Bohmann et al., 1995; Hebert and Matera, 2000). These results suggest that, in vivo, the coilin C-terminus might serve to mask the N-terminal region and that fusion of this domain to GFP or outright truncation of C-terminal residues could unmask the self-association domain.
In order to test this idea further, we wanted to determine whether formation of coilin foci could take place in the cytoplasm or if it were restricted to the nucleus. If foci formation were limited to the nucleus, it would imply that important interactions or modifications (e.g., phosphorylation) are localized within this cellular compartment.
Alternatively, formation of foci in the cytoplasm would indicate either that coilin
117 modifiers exist in the cytoplasm, or that foci formation is an intrinsic property of the
coilin N-terminus. The nuclear localization signals (NLSs) reside within the loosely
conserved internal region that links the more highly conserved N- and C-terminal
domains (Bohmann et al., 1995). We therefore created internal deletions spanning coilin
residues 106-234. In the GFP-coilin background, the ∆NLS protein shows exclusively
cytoplasmic localization, lacking foci, in HeLa cells (Fig. A1-2, lower left panel).
Surprisingly, endogenous coilin localization is not altered in cells expressing the GFP-
coilin∆NLS (data not shown). Thus, it is possible that deletion of the NLS region may affect the self-association activity of coilin. To test this, and to verify that coilin foci formation is limited to the nucleus, we took advantage of the coilin construct that generates a plethora of dots, coilin-GFP. HeLa cells transfected with coilin∆NLS-GFP typically displayed 2-5 foci in the nucleus with additional bright staining throughout the cytoplasm (Fig. A1-2, lower middle panel). Thus despite the lack of an NLS, a small amount of coilin∆NLS-GFP protein was imported into the nucleus, perhaps by binding to endogenous coilin molecules. In order to test this idea, we transfected coilin∆NLS-GFP into coilin knockout cells (Tucker et al., 2001), and found that, indeed, the construct was completely restricted to the cytoplasm in all cells (Fig. A1-2, lower right panel). It should be noted that HeLa cells expressing coilin∆NLS-GFP displayed a markedly reduced number of foci, compared to the parental construct (Fig. A1-2, compare upper middle and lower middle panels). Nonetheless, the ability to form bodies in the nucleus and not in the cytoplasm verifies the idea that the subcellular localization of coilin is important for their formation.
118
Figure A1-2: Coilin mutants display unregulated nuclear body formation. Upper panels:
HeLa cells were transfected with GFP fused to the amino-terminus of coilin (left), GFP
fused to the carboxy-terminus of coilin (middle), or a GFP-tagged truncation of coilin (1-
482, right). Note that several out-of-focus dots appear blurred. Lower panels: Deletion of
the nuclear localization signals (residues 106-234) in the GFP-coilin background results
in exclusively cytoplasmic localization without foci in HeLa cells (left). Deletion of the
NLSs in the coilin-GFP background results in a similar pattern, with the exception that
nuclear foci are detected. Although the bulk of the fluorescence was cytoplasmic, two
populations of cells were observed, some with faint nuclear foci others with brighter foci
(middle). When transfected into coilin knockout MEFs, this same coilin∆NLS-GFP construct localized solely to the cytoplasm (right). The GFP-coilin(1-482) foci colocalized with other CB markers (SMN and fibrillarin) when transfected into HeLa cells (data not shown).
119 Coilins from different species demonstrate specific nuclear body regulatory potentials.
When transiently expressed in HeLa cells at low- to medium-levels, human GFP-coilin properly localizes to the nucleoplasm and the CB. However, high levels of expression result in a dramatic reorganization of ectopic coilin, endogenous coilin, and other constituents of the CB, such as SMN, to the nucleoplasm (Hebert and Matera, 2000).
High levels of overexpression does not alter the overall nuclear architecture since other nuclear structures, such as PML bodies, are unaffected (Hebert and Matera, 2000).
Furthermore, the GFP moiety is not responsible for this phenotype because coilin constructs with a myc-tag display the same effect (our unpublished observations). In contrast to the results using the human coilin constructs in HeLa cells, high levels of
GFP-mouse coilin overexpression does not abolish foci formation. Instead, cells expressing high levels of mouse GFP-coilin displayed the unregulated pattern of coilin foci observed for human coilin(1-481) or coilin-GFP (Figs. A1-2 and A1-3). Similarly, frog YFP-coilin also displayed a multitude of foci when expressed in HeLa cells (Fig.
A1-3). Surprisingly, transfection of these three coilins into a murine embryonic fibroblast
(MEF) cell line lacking endogenous coilin (Tucker et al., 2001) revealed additional species-specific regulatory differences. For example, human coilin failed to form CBs in most of these MEFs, even in those with low expression levels (Fig. A1-3). Contrastingly, mouse coilin readily formed foci in MEF cells when expressed at low levels. However,
MEF cells expressing high levels of mouse coilin showed only nucleoplasmic staining.
Thus expression of heterologous coilins in human cells resulted in dysregulation of CB numbers. When mouse coilin, for example, was expressed in murine cells, the number of coilin foci did not increase. Moreover, the number of coilin foci was completely
120 unregulated when frog coilin was expressed at high levels in both cell types (Fig. A1-3).
This was most easily observed upon expression of frog coilin in the knockout MEFs. In
this background, large coilin aggregates were observed, even in the cytoplasm.
Furthermore, the pool of nucleoplasmic frog coilin in MEF cells was virtually non- existent.
In summary, human coilin is able to effectively target to pre-existing CBs in HeLa
cells when expressed at low levels but is unable to target, and indeed disrupts CBs, at
high expression levels. In MEFs cells lacking endogenous coilin, the human protein is
ineffective at nucleating CBs (Fig. A1-3). Mouse coilin is able to effectively form CBs in
mouse cells at low expression levels but, like human coilin in HeLa cells, becomes
diffusely nucleoplasmic at higher levels of expression. In the HeLa background, however,
mouse coilin displayed numerous foci. Similar findings were observed for frog coilin in
both HeLa and MEF cells. At least for mouse and frog coilin proteins, the expression
level was an important determinant for nuclear body formation. Human coilin, however,
appears to have additional constraints regarding its ability to form nuclear bodies.
121
Figure A1-3: Coilins from different species display variation in nuclear body formation.
HeLa cells (left two columns) or mouse embryonic cells lacking endogenous coilin (right
two columns) were transfected with FP-tagged human, mouse, or frog coilin.
Representative cells of both low and high expressers are shown. The nucleus is defined
by a dotted line for MEF cells transfected with FP-frog coilin.
Chimeric coilin constructs delineate the Cajal body regulatory region.
We next set out to better define the region within coilin that governs CB formation and
number. Given that truncation or fusion of the C-terminus resulted in formation of numerous coilin foci (Fig. A1-2), we speculated that this region might play an important regulatory role. These findings, coupled with the observation that compared to the human protein, mouse coilin has an increased potential for nuclear foci formation in HeLa cells
122 (Fig. A1-3) led us to postulate that the C-terminus of mouse coilin imparts differential regulatory constraints. To test this idea, we generated chimeric constructs of human and mouse coilin by swapping the C-terminal 96 amino acids (Fig. A1-4). Following expression in coilin knockout MEFs, we assessed their capacities to form nuclear foci
(Fig. A1-4). As previously observed (Tucker et al., 2001), cells expressing moderate levels of mouse coilin typically display foci (87%), whereas those expressing human coilin rarely display foci (7%; Fig. A1-4). Furthermore, the number of foci per cell was greatly reduced (0.09 for human vs. 2.84 for mouse). Strikingly, replacement of the mouse C-terminus with the corresponding human portion of coilin reduced both the frequency of cells displaying foci and the numbers of foci per cell (Fig A1-4, compare
87% for mouse vs. 1% for the mouse/human chimera). Conversely, replacement of the human C-terminus with the mouse region increased the number of cells that displayed coilin foci by 10-fold (Fig. A1-4, 7% for human vs. 70% for human/mouse). We conclude that the C-terminal 96 amino acids of human coilin down-regulates the capacity for nuclear body formation, whereas the corresponding region in mouse coilin increases this capability.
123 Figure A1-4: Schematic of human and mouse coilin. Chimeric constructs we generated
by use of a conserved HindIII site, essentially swapping the amino acids downstream of
this site. Numbers of foci per cell and percentages of cells with foci are upon transfection
into coilin knockout MEFs are displayed to the right of the representative construct.
Critical serine residues in the C-terminus of coilin control nuclear body formation.
Despite the apparent phenotypic differences in nuclear body-forming potential, an
alignment of the various coilin C-termini demonstrates a high level of identity in this
region (Fig. A1-5A). Interestingly, frog coilin shows several residues that are
substantially changed compared to the corresponding amino acids in human and mouse
(e.g., frog contains DEE at 499-501 compared to NGA in human). These differences,
although slight, may account for the observed localization phenotypes. Curiously, the
biggest differences between human and mouse are in the very C-terminal, serine rich, portion. Given that human coilin phosphorylation occurs exclusively on serine residues
(Carmo-Fonseca et al., 1993), and that several of the serines in its C-terminus are predicted to be phosphorylated (NetPhos, Technical University of Denmark;
http://genome.cbs.dtu.dk/services/NetPhos/), we decided to systematically generate
mutations in this region and assess their nuclear body-forming potentials. The mutations
were each produced in GFP-coilin backgrounds and the constructs were subsequently
transfected into HeLa cells or coilin knockout MEFs. A summary of the mutations and
their proclivities to form nuclear bodies in MEF cells is shown in Figure A1-5B. The mutation of arginine 561 in human coilin to a stop codon resulted in a variable phenotype. In HeLa cells, human coilin(1-560) was mostly nucleoplasmic with nuclear
124 aggregates, although a small percentage of cells (< 10%) appeared normal (not shown).
In coilin knockout MEFs, human coilin(1-560) mirrors what is observed in HeLa, with the exception that no cells displayed coilin foci.
The corresponding truncation mutation in mouse coilin (arginine 557 to stop) showed a dramatic change in localization when expressed in HeLa cells compared to full- length mouse coilin (Fig. A1-5C). Most HeLa cells transfected with mouse coilin(1-556) displayed primarily nucleoplasmic staining, even at lower expression levels.
Additionally, high expression levels of mouse coilin(1-556) resulted in nucleoplasmic staining and corresponding disassembly of endogenous CBs (Fig A1-5C, arrow). The overexpression of mouse coilin(1-556) in HeLa cells is reminiscent of the expression pattern seen for the high expression of human coilin in HeLa cells, which likewise results in nucleoplasmic staining and subsequent disruption of CBs (Hebert and Matera, 2000).
Curiously, overexpression of wildtype mouse coilin in HeLa cells also reduced the amount of endogenous coilin in CBs, despite the observed foci present for mouse coilin
(Fig. A1-5C, top panels, arrow). Thus we conclude that removal of the C-terminal 13 amino acids of mouse coilin effectively makes it behave like the human protein. In support of this idea, MEF cells transfected with mouse coilin(1-556) no longer displayed
CBs, but showed nucleoplasmic staining (Fig. A1-5B). Therefore, these 13 amino acids affect not only the ability to target to preformed CBs, but the capacity to form foci in the absence of endogenous coilin.
Larger C-terminal truncations of both mouse (1-477) and human (1-481) coilin resulted in formation of multiple foci in MEF cells (Fig. A1-5B). However, smaller C- terminal deletions (human: 1-516 and 1-560; mouse: 1-512, 1-535 and 1-556) lost the
125 ability to form coilin foci when transfected into MEF cells. Curiously, a construct lacking the C-terminal 5 residues, mouse(1-564), was capable of generating nuclear bodies whereas mouse(1-556) did not, thus narrowing a region that may structurally inhibit nuclear body formation to residues 478-563 in mouse coilin (Fig. A1-5B).
Given the serine-rich composition of the C-terminal 13 residues in mouse coilin
(5 out of 13), we speculated that the phosphorylation states of these residues might play a role in the ability to form foci. In particular, we were interested in the last two serines of mouse coilin (S567 and S568), which are conserved in human coilin (Fig. A1-5A).
Mutation of these serines, along with the C-terminal threonine in mouse coilin SST to
AGA, resulted in a remarkable mislocalization both in MEF and HeLa cells (Fig. A1-
5B,C). Indeed, mouse coilin (AGA) and (1-556) showed virtually identical overexpression localization patterns in HeLa (Fig. A1-5C). These findings led us to speculate that S567 and S568 in the mouse protein need to be phosphorylated to properly form or target to CBs. When the corresponding serines were mutated to aspartates (SST to DDT), mimicking a constitutively phosphorylated state, the ability to form foci in the absence of endogenous coilin and to produce numerous coilin foci in HeLa cells was retained (Fig. A1-5B,C). However, mutation of mouse residues S567 or S568 (to either D or A) individually had little effect on the localization pattern (Figure A1-5B, data not shown). Point mutagenesis of other residues suspected to affect the folding of the C- terminus also failed to produce notable changes in coilin sublocalization compared to the parental construct (Fig. A1-5B). Thus mutation of both serines 567 and 568 together had an effect whereas no effect was observed when mutated singly.
126 We speculated that the mutants that cannot form nuclear bodies (mouse 1-556 and
SST to AGA) were somehow deficient in coilin self-association. Coimmunoprecipitation assays using these constructs, however, also failed to show any differences in the abilities of the mutated coilins to interact with coilin or other components of the CB, such as SMN
(data not shown). Therefore, it is possible that the mutations in the mouse coilin background specifically affect the ability to target or form CBs. Our inability to identify human coilin constructs, other than coilin-GFP and GFP-coilin(1-481), that form foci in
MEF cells (Fig. A1-5B) suggests the existence of additional regulatory controls for human coilin.
127
128
Figure A1-5: Mutations in mouse coilin affect nuclear body formation. (A) Alignment of
the C-termini of human, mouse, and frog coilin. The HindIII site used in generating the coilin chimeras is shown. (B) Summary of the mutations generated in this study. All constructs are fused to the C-terminus of GFP, with the exception of human coilin-GFP.
The constructs were transfected into MEF cells lacking endogenous coilin (Tucker et al.,
2001) and scored for their ability to generate foci. Note that human constructs coilin-GFP
and (1-481) as well as mouse construct (1-477) produce numerous unregulated foci. (C)
HeLa cells were transfected with GFP-tagged wild-type mouse coilin or mutations of
129 mouse coilin. Endogenous coilin (right column) was detected by using an antibody
specific to human coilin (R508). Arrows mark the transfected cells, with concurrent
disruption of Cajal bodies (right panels). Arrowheads show non-transfected cells, which
display normal coilin localization in CBs and the nucleoplasm (right column). Foci
formed by GFP-mouse coilin-SST to DDT were able to recruit SMN and fibrillarin,
whereas loss of coilin foci upon overexpression of GFP-mouse coilin-SST to AGA also
correlated with a loss of SMN foci (data not shown).
Other nuclear body proteins display unregulated foci formation.
To generalize the idea that, apart from coilin, proteins considered markers for different
nuclear bodies can induce foci formation upon overexpression, we surveyed the literature. Notably, various mutants of Sam68, the marker protein for Sam68 nuclear
bodies, appear to have lost the ability to regulate nuclear body formation (Chen et al.,
1999). Cells transfected with wild-type Sam68 typically display 2-5 nuclear bodies, along
with a diffuse nucleoplasmic staining. However, mutants such as Sam68∆L1 produce
numerous foci (up to 30) along with a concurrent loss of nucleoplasmic staining (Chen et
al., 1999). Several human coilin mutants also display a similar phenotype (Fig. A1-2).
PML, which is considered to be the marker protein for PML bodies, also generates more
PML bodies when overexpressed in HeLa cells and larger (but fewer) PML bodies when
overexpressed in knockout MEFs (Ishov et al., 1999; Mu et al., 1997).
On the other hand, overexpression of wildtype SMN, which is vital for gem
formation, reportedly does not change the number of gems in transfected cells (Pellizzoni
et al., 1998). We have found, however, that high expression of myc-tagged SMN, both in
130 HeLa cells and coilin knockout MEFs, results in formation of numerous foci (Fig. A1-6).
The additional foci are present both in the nucleus and the cytoplasm. The phenotype is
similar to that observed upon ectopic expression of the SMN∆N27 mutant protein
(Pellizzoni et al., 1998). The main difference is that the enlarged nuclear foci seen with
SMN∆N27 (Pellizzoni et al., 1998) were not observed in cells overexpressing the wildtype protein. The large cytoplasmic blobs (Fig. A1-6A, bottom) were observed only at the higher expression levels. In HeLa cells, SMN overexpression produces nuclear foci which sometimes contain coilin and can be considered CBs, and other foci that lack coilin and thus can be considered gems (Fig. A1-6A, arrowheads). In summary, we conclude that the overall level of SMN expression (Fig. A1-6) and the extent of coilin dimethylation (Hebert et al., 2002a) are major determinants in gem formation.
131
Figure A1-6: Ectopically expressed SMN produces numerous SMN foci. (A) HeLa cells expressing high levels of myc-tagged SMN were visualized by anti-myc antibodies
(green). Coilin was detected in the same cell with R508 (in red) and the merged image is shown (right panel). Arrows denote Cajal bodies, which contain both SMN and coilin whereas arrowheads mark some of the SMN gems, which are present both in the cytoplasm and the nucleus. The bottom panels display some of the large SMN foci that are observed upon high expression of SMN. (B) MEF coilin knockout cells transiently
132 transfected with myc-SMN were detected by anti-myc antibodies. Arrowheads demarcate some of the foci, including large aggregates, formed in high expressing cells (right).
133
Discussion
In this report, we demonstrate that the Cajal body marker protein, coilin, contains a
domain capable of regulating the number of nuclear bodies per cell. This domain is located in the C-terminus of the protein and contains putative phosphoserine residues.
One of the most interesting aspects of this study is the finding that heterologous coilin proteins display differential nuclear body regulatory potentials in different cellular backgrounds. For example, mouse coilin produces numerous foci in HeLa cells when expressed at high levels, but fails to form foci when expressed at high levels in murine cells. However, frog coilin produces numerous foci both in HeLa and murine cells, and can even generate cytoplasmic bodies.
Interestingly, frog coilin does not have a serine-rich C-terminus as do the mouse
and human proteins (Fig. A1-5A). Furthermore, unlike the human protein (Carmo-
Fonseca et al., 1993), Xenopus coilin is not highly phosphorylated (M. Bellini, personal
communication). It is possible, therefore, that frog coilin is “locked” in a conformation
that renders it competent for nuclear body formation and thus is insensitive to
phosphorylation. Consequently, frog coilin would be expected to readily form nuclear
foci (Fig. A1-3). Human and mouse coilin appear to have greater regulatory constraints
with regard to their ability to form nuclear bodies. This idea is supported by the human
coilin∆NLS mutants (Fig. A1-2), which are incapable of forming cytoplasmic foci. These
results suggest that factors responsible for phosphorylating coilin are located in the
nucleus.
134 Given that alterations in coilin phosphorylation levels can affect both self- association and nucleolar localization (Hebert and Matera, 2000; Lyon et al., 1997;
Sleeman et al., 1998), it is likely that this modification plays an important role in CB- targeting as well. We showed that, unlike human coilin, a chimeric human/mouse construct can readily form foci in murine cells lacking endogenous coilin. A reciprocal construct containing mouse coilin with the human tail displayed a phenotype similar to that of the human protein expressed in the mouse cells (Fig. A1-4). Considering that mutation of the coilin N-terminal self-association domain alters C-terminal epitope recognition by the anti-coilin Ab R508, we speculated that self-association would be regulated via the C-terminal tail. Coimmunoprecipitation experiments, however, failed to show any change in coilin self-association among the various mutants. Furthermore, the observation that mouse coilin(1-556) and coilin(AGA) disrupt the endogenous coilin in
HeLa cells supports the notion that self-interaction is not affected (Fig. A1-5C). While it is conceivable that self-interaction is required for nuclear body formation, additional cellular factors (e.g., high levels of Sm proteins; (Sleeman et al., 2001) and/or modifications of coilin may be necessary. Since mutation of the mouse coilin C-terminal residues to mimic constitutively phosphorylated serines (SST to DDT) results in wildtype levels of foci formation, we conclude that these sites are normally phosphorylated. Since kinase recognition sites are often dependent on previous phosphorylation events, the phosphorylation of these residues could facilitate additional serine phosphorylation elsewhere within the coilin protein. The fact that mutation of individual S567 or S568 serines to mimic an unphosphorylated state had no effect on mouse coilin localization emphasizes the complexity of coilin phosphorylation. Furthermore, the inability of
135 human coilin to form foci at high expression levels in HeLa cells (or to form any foci at
all in MEF cells) may be due to titration of the putative coilin kinases. We have shown
that coilin is a substrate for the kinases CDK2-cyclinE (Liu et al., 2000)and casein kinase
II (Hebert and Matera, 2000), however, phosphatases that modulate the phosphorylation
state of coilin have not been described. Current studies are focused on identification of
residues within coilin that are post-translationally modified, and thus alter its
conformation, localization and function.
As an alternative model, it is possible that mutation of the coilin C-terminus may
not affect phosphorylation, but rather, alter a binding site between coilin and other
proteins. The C-terminus may be needed to interact with another protein essential for CB
formation, and overexpression of coilin may titrate out this integral component thus
dispersing CBs. Another such example exists whereby overexpression of the immediate-
early protein IE1 leads to a disruption in PML body formation by interacting with the
major structural factor, PML (Ahn et al., 1998). Additionally, coilin itself may encode the
structural integrity of the CB, and the mouse coilin C-terminal mutants deficient in CB
formation may simply yield an alteration in coilin structure rather than phosphorylation,
leading to CB dispersal. Future experiments will need to address whether the implicated
residues are indeed phosphorylated in vivo and if differences in phosphorylation levels
can account for CB number differences between cell lines.
In conclusion, various nuclear body marker proteins share a common capacity for
self-association and nuclear body formation. However, simply making more protein is not always sufficient to generate additional nuclear foci, especially in the case of the
Cajal body. Interestingly CBs can be induced to form in cell lines that do not typically
136 possess them by overexpression of Sm proteins (Sleeman et al., 2001). Given that post- translational modification of coilin plays a vital role in Cajal body composition (Hebert et
al., 2002a) and formation (this work), we are especially interested in assessing the
modification status of coilin in the nucleoplasmic vs. CB compartments. Recently-
developed methods for CB purification (Platani et al., 2002) should greatly aid such an
endeavor.
137 Appendix Chapter II
Exogenous Gemin4 expression enhances Gemin3 and SMN nuclear localization
Karl B. Shpargel and A. Gregory Matera
Department of Genetics Case Western Reserve University School of Medicine Cleveland, OH 44106-4955 USA
*Corresponding author Tel: 216-368-4922 Fax: 216-368-0491 Email: [email protected]
Note: This chapter documents experiments that analyze exogenous Gemin3 and Gemin4 localization patterns.
138 Abstract
Spinal Muscular Atrophy (SMA) is a neurodegenerative autosomal recessive disease that results in a loss of spinal motor neurons in the anterior horn of the lower spinal cord. The gene encoding the Survival of Motor Neurons protein (SMN) is mutated in over 95% of
SMA cases. The underlying problem that leads to SMA pathology may stem from the most extensively characterized function of SMN in assembling the Sm core onto the U snRNAs that mediate splicing. SMN is a component of a large macromolecular complex that mediates this process. The SMN complex has been purified for the identification of its components. One component of this complex, Gemin3, is a member of the DEAD box family of RNA helicases, and could be necessary to mediate the protein-snRNA interactions and structural rearrangements needed for Sm core assembly onto the snRNA.
In this study we analyze localization patterns of Gemin3 and Gemin4 to gain insight into their function. Gemin3 over-expression retains SMN to the cytoplasm with localization to the Golgi complex. Alternatively, Gemin4 appears to be involved in Gemin3 and SMN import into the nucleus as both proteins exhibit increased nuclear localization with
Gemin4 over-expression.
139 Materials and Methods
DNA Constructs
Gemin3 was PCR amplified from cDNA and cloned into eGFP-N1. Gemin4 was PCR amplified from cDNA and cloned into eGFP-C1 or pCMV-myc. Myc-SPN was cloned as described (Narayanan et al., 2002)
Transfection and cellular treatments
HeLa cells were transfected with superfect as directed (Invitrogen). 20µM brefeldin-A was incubated in cell culture media overnight to disrupt the Golgi complex.
140
Results and Discussion
In an initial attempt to characterize Gemin3 and Gemin4, we subcloned their cDNAs into expression vectors for localization studies. Typically these proteins localize diffusely throughout the cytoplasm with concentration in nuclear Cajal bodies (Charroux et al.,
1999; Charroux et al., 2000). Remarkably, even moderate Gemin3-GFP expression localizes in a peri-nuclear pattern that resembles the Golgi complex (Fig. A2-1, top panel). Gemin3-GFP interacts with SMN, drawing it into this organelle. However, overexpression of myc-SMN displays a distinct phenotype, drawing Gemin3 into large cytoplasmic aggregates (Fig. A2-1, second panel). Intriguingly, the Golgi like localization pattern fails to co-localize with other snRNP biogenesis components (myc-
SPN, third panel) or miRNP components (myc-eIF2C2, fourth panel). Therefore this localization pattern appears to be specific for the SMN complex.
141
142 Figure A2-1: Gemin3 overexpression localizes SMN to a Golgi like complex. Gemin3-
GFP is expressed alone, or co-transfected with myc-SMN (second panel), myc-SPN
(snurportin, third panel), or myc-eIF2C2 (fourth panel). Thus, Golgi localization patterns
are specific for SMN complex members.
To definitively determine if Gemin3-GFP localizes to the Golgi complex, we
performed immunofluorescence for a Golgi specific antibody, Golgin (Fig. A2-2, top
panel). While Gemin3-GFP and Golgin overlap in the Golgi complex, they fail to
completely co-localize. Therefore, Gemin3 and Golgin may be contained within slightly different Golgi compartments. Furthermore, ablation of the Golgi complex with brefeldin-A also dissociates the Gemin3-GFP signal (Fig. A2-2, bottom panel). Thus
Gemin3 does in fact localize to the Golgi complex.
143 Figure A2-2: Golgi dissociation alters Gemin3 localization. Overexpression of Gemin3-
GFP overlaps with a Golgi protein, Golgin (top panel), and brefeldin-A treatment dissociates the Golgi complex and Gemin3 localization (bottom panel).
While Gemin3 overexpression retains SMN in the cytoplasmic Golgi complex, overexpression of GFP-Gemin4 localizes almost entirely in the nucleus (Fig. A2-3).
Furthermore GFP-Gemin4 overexpression drives SMN into the nucleus (Fig. A2-3, top two panels) Additionally, while Gemin3-GFP localizes intensely to the Golgi complex, co-expression with myc-Gemin4 draws the Gemin3 into the nucleus (Fig. A2-3, bottom panel). Therefore, Gemin4 is capable of importing Gemin3 and SMN into the nucleus.
144
Figure A2-3: Gemin4 overexpression imports SMN and Gemin3 into the nucleus. Cells transfected with GFP-Gemin4 exhibit elevated levels of SMN in the nucleus (top panels).
Co-transfection of Gemin3-GFP with myc-Gemin4 alleviates the Gemin3 Golgi localization and imports the protein into the nucleus (bottom panel).
145 In conclusion, Gemin3 overexpression draws SMN into the Golgi complex.
Alternatively, Gemin4 overexpression imports both SMN and Gemin3 into the nucleus.
Interestingly, the entire SMN complex is required for snRNP import (Narayanan et al.,
2004b). Thus it is possible that Gemin4 may function in this import pathway. Analysis of
Gemin4 sequence predicts that a nuclear localization signal resides in the Amino- terminus of the protein. Future experiments will be required to determine if the NLS of
Gemin4 is functional, and if endogenous Gemin4 participates in snRNP import.
146 Bibliography
Ahn, J. H., Brignole, E. J., 3rd and Hayward, G. S. (1998). Disruption of PML subnuclear domains by the acidic IE1 protein of human cytomegalovirus is mediated through interaction with PML and may modulate a RING finger-dependent cryptic transactivator function of PML. Mol Cell Biol 18, 4899-913.
Andersen, J. S., Lyon, C. E., Fox, A. H., Leung, A. K., Lam, Y. W., Steen, H., Mann, M. and Lamond, A. I. (2002). Directed proteomic analysis of the human nucleolus. Curr Biol 12, 1-11.
Andrade, L. E. C., Chan, E. K. L., Raska, I., Peebles, C. L., Roos, G. and Tan, E. M. (1991). Human autoantibody to a novel protein of the nuclear coiled body: Immunological characterization and cDNA cloning of p80 coilin. J. Exp. Med. 173, 1407-1419.
Andrade, L. E. C., Tan, E. M. and Chan, E. K. L. (1993). Immunocytochemical analysis of the coiled body in the cell cycle and during cell proliferation. Proc. Natl. Acad. Sci. USA 90, 1947-1951.
Baccon, J., Pellizzoni, L., Rappsilber, J., Mann, M. and Dreyfuss, G. (2002). Identification and characterization of Gemin7, a novel component of the survival of motor neuron complex. Journal of Biological Chemistry 277, 31957-62.
Bohmann, K., Ferreira, J. and Lamond, A. (1995). Mutational analysis of p80 coilin indicates a functional interaction between coiled bodies and the nucleolus. J. Cell Biol. 131, 817-831.
Brahe, C. and Bertini, E. (1996). Spinal muscular atrophies: recent insights and impact on molecular diagnosis. J Mol Med 74, 555-62.
Brahms, H., Meheus, L., de Brabandere, V., Fischer, U. and Luhrmann, R. (2001). Symmetrical dimethylation of arginine residues in spliceosomal Sm protein B/B' and the Sm-like protein LSm4, and their interaction with the SMN protein. Rna 7, 1531- 42.
Brahms, H., Raymackers, J., Union, A., de Keyser, F., Meheus, L. and Luhrmann, R. (2000). The C-terminal RG dipeptide repeats of the spliceosomal Sm proteins D1 and D3 contain symmetrical dimethylarginines, which form a major B-cell epitope for anti-Sm autoantibodies. J Biol Chem 275, 17122-9.
Bühler, D., Raker, V., Luhrmann, R. and Fischer, U. (1999). Essential role for the tudor domain of SMN in spliceosomal U snRNP assembly: implications for spinal muscular atrophy. Hum Mol Genet 8, 2351-7.
147 Burghes, A. H. M. (1997). When is a deletion not a deletion? When it is converted. Am. J. Hum. Genet. 61, 9-15.
Campbell, L., Potter, A., Ignatius, J., Dubowitz, V. and Davies, K. (1997). Genomic variation and gene conversion in Spinal Muscular Atrophy: Implications for disease process and clinical phenotype. Am. J. Hum. Genet. 61, 40-50.
Carissimi, C., Baccon, J., Straccia, M., Chiarella, P., Maiolica, A., Sawyer, A., Rappsilber, J. and Pellizzoni, L. (2005). Unrip is a component of SMN complexes active in snRNP assembly. FEBS Lett 579, 2348-54.
Carissimi, C., Saieva, L., Baccon, J., Chiarella, P., Maiolica, A., Sawyer, A., Rappsilber, J. and Pellizzoni, L. (2006). Gemin8 is a novel component of the survival motor neuron complex and functions in small nuclear ribonucleoprotein assembly. J Biol Chem 281, 8126-34.
Carmo-Fonseca, M., Ferreira, J. and Lamond, A. I. (1993). Assembly of snRNP-containing Coiled Bodies Is Regulated in Interphase and Mitosis - Evidence that the Coiled Body Is a Kinetic Nuclear Structure. J. Cell Biol. 120, 841-852.
Cartegni, L. and Krainer, A. R. (2002). Disruption of an SF2/ASF-dependent exonic splicing enhancer in SMN2 causes spinal muscular atrophy in the absence of SMN1. Nat Genet 30, 377-84.
Carvalho, T., Almeida, F., Calapez, A., Lafarga, M., Berciano, M. T. and Carmo-Fonseca, M. (1999). The spinal muscular atrophy disease gene product, SMN: A link between snRNP biogenesis and the Cajal (coiled) body. J. Cell Biol. 147, 715-728.
Chan, E. K. L., Takano, S., Andrade, L. E. C., Hamel, J. C. and Matera, A. G. (1994a). Structure, expression and chromosomal localization of the human p80-coilin gene. Nucleic Acids Res. 22, 4462-4469.
Chan, E. K. L., Tan, E. M., Ward, D. C. and Matera, A. G. (1994b). Human 60-kD SS-A/Ro ribonucleoprotein autoantigen gene localized to 1q31 by flourescence in situ hybridization. Genomics 23, 298-300.
Chan, Y. B., Miguel-Aliaga, I., Franks, C., Thomas, N., Trulzsch, B., Sattelle, D. B., Davies, K. E. and van den Heuvel, M. (2003). Neuromuscular defects in a Drosophila survival motor neuron gene mutant. Hum Mol Genet 12, 1367-76.
Charroux, B., Pellizzoni, L., Perkinson, R. A., Shevchenko, A., Mann, M. and Dreyfuss, G. (1999). Gemin3. A novel dead box protein that interacts with SMN, the spinal muscular atrophy gene product, and is a component of gems. J Cell Biol 147, 1181-94.
148 Charroux, B., Pellizzoni, L., Perkinson, R. A., Yong, J., Shevchenko, A., Mann, M. and Dreyfuss, G. (2000). Gemin4: A novel component of the SMN complex that is found in both gems and nucleoli. J. Cell Biol. 148, 1177-86.
Chen, T., Boisvert, F. M., Bazett-Jones, D. P. and Richard, S. (1999). A role for the GSG domain in localizing Sam68 to novel nuclear structures in cancer cell lines. Mol. Biol. Cell 10, 3015-33.
Chen, T., Damaj, B. B., Herrera, C., Lasko, P. and Richard, S. (1997). Self- association of the single-KH-domain family members Sam68, GRP33, GLD-1, and Qk1: role of the KH domain. Mol Cell Biol 17, 5707-18.
Cifuentes-Diaz, C., Frugier, T., Tiziano, F. D., Lacene, E., Roblot, N., Joshi, V., Moreau, M. H. and Melki, J. (2001). Deletion of murine SMN exon 7 directed to skeletal muscle leads to severe muscular dystrophy. Journal of Cell Biology 152, 1107- 14.
Clermont, O., Burlet, P., Benit, P., Chanterau, D., Saugier-Veber, P., Munnich, A. and Cusin, V. (2004). Molecular analysis of SMA patients without homozygous SMN1 deletions using a new strategy for identification of SMN1 subtle mutations. Hum Mutat 24, 417-27.
Coovert, D., Le, T., McAndrew, P., Strasswimmer, J., Crawford, T., Mendell, J., Coulson, S., Androphy, E., Prior, T. and Burghes, A. (1997). The survival motor neuron protein in spinal muscular atrophy. Hum. Mol. Genet. 6, 1205-1214.
Cougot, N., van Dijk, E., Babajko, S. and Seraphin, B. (2004). 'Cap-tabolism'. Trends Biochem Sci 29, 436-44.
Cusco, I., Barcelo, M. J., del Rio, E., Baiget, M. and Tizzano, E. F. (2004). Detection of novel mutations in the SMN Tudor domain in type I SMA patients. Neurology 63, 146-9.
Darzacq, X., Jady, B. E., Verheggen, C., Kiss, A. M., Bertrand, E. and Kiss, T. (2002). Cajal body-specific small nuclear RNAs: a novel class of 2'-O- methylation and pseudouridylation guide RNAs. Embo J 21, 2746-56.
DiDonato, C. J., Chen, X. N., Noya, D., Korenberg, J. R., Nadeau, J. H. and Simard, L. R. (1997). Cloning, characterization, and copy number of the murine survival motor neuron gene: homolog of the spinal muscular atrophy-determining gene. Genome Res. 7, 339-352.
Dundr, M. and Misteli, T. (2001). Functional architecture in the cell nucleus. Biochem J 356, 297-310.
149 Elbashir, S. M., Harborth, J., Lendeckel, W., Yalcin, A., Weber, K. and Tuschl, T. (2001). Duplexes of 21-nucleotide RNAs mediate RNA interference in cultured mammalian cells. Nature 411, 494-8.
Fan, L. and Simard, L. R. (2002). Survival motor neuron (SMN) protein: role in neurite outgrowth and neuromuscular maturation during neuronal differentiation and development. Hum Mol Genet 11, 1605-14.
Feng, W., Gubitz, A. K., Wan, L., Battle, D. J., Dostie, J., Golembe, T. J. and Dreyfuss, G. (2005). Gemins modulate the expression and activity of the SMN complex. Hum Mol Genet 14, 1605-11.
Fischer, U., Liu, Q. and Dreyfuss, G. (1997). The SMN-SIP1 complex has an essential role in spliceosomal snRNP biogenesis. Cell 90, 1023-1029.
Fornerod, M., Ohno, M., Yoshida, M. and Mattaj, I. W. (1997). CRM1 is an export receptor for leucine-rich nuclear export signals [see comments]. Cell 90, 1051-60.
Frey, M. R. and Matera, A. G. (1995). Coiled Bodies Contain U7 Small Nuclear RNA and Associate with Specific DNA Sequences in Interphase Cells. Proc. Natl. Acad. Sci. U.S.A. 92, 5915-5919.
Friesen, W. J., Massenet, S., Paushkin, S., Wyce, A. and Dreyfuss, G. (2001). SMN, the product of the spinal muscular atrophy gene, binds preferentially to dimethylarginine-containing protein targets. Mol Cell 7, 1111-7.
Frugier, T., Nicole, S., Cifuentes-Diaz, C. and Melki, J. (2002). The molecular bases of spinal muscular atrophy. Curr Opin Genet Dev 12, 294-8.
Frugier, T., Tiziano, F. D., Cifuentes-Diaz, C., Miniou, P., Roblot, N., Dierich, A., Le Meur, M. and Melki, J. (2000). Nuclear targeting defect of SMN lacking the C-terminus in a mouse model of spinal muscular atrophy. Hum Mol Genet 9, 849-58.
Gall, J. G. (2000). Cajal Bodies: The First 100 Years. Ann. Rev. Cell and Dev. Biol. 16, 273-300.
Gates, J., Lam, G., Ortiz, J. A., Losson, R. and Thummel, C. S. (2004). rigor mortis encodes a novel nuclear receptor interacting protein required for ecdysone signaling during Drosophila larval development. Development 131, 25-36.
Giot, L., Bader, J. S., Brouwer, C., Chaudhuri, A., Kuang, B., Li, Y., Hao, Y. L., Ooi, C. E., Godwin, B., Vitols, E. et al. (2003). A protein interaction map of Drosophila melanogaster. Science 302, 1727-36.
150 Girard, C., Mouaikel, J., Neel, H., Bertrand, E. and Bordonne, R. (2004). Nuclear localization properties of a conserved protuberance in the Sm core complex. Exp Cell Res 299, 199-208.
Glanzer, J., Miyashiro, K. Y., Sul, J. Y., Barrett, L., Belt, B., Haydon, P. and Eberwine, J. (2005). RNA splicing capability of live neuronal dendrites. Proc Natl Acad Sci U S A 102, 16859-64.
Golembe, T. J., Yong, J. and Dreyfuss, G. (2005). Specific sequence features, recognized by the SMN complex, identify snRNAs and determine their fate as snRNPs. Mol Cell Biol 25, 10989-1004.
Gubitz, A. K., Feng, W. and Dreyfuss, G. (2004). The SMN complex. Exp Cell Res 296, 51-6.
Gubitz, A. K., Mourelatos, Z., Abel, L., Rappsilber, J., Mann, M. and Dreyfuss, G. (2002). Gemin5, a novel WD repeat protein component of the SMN complex that binds Sm proteins. J Biol Chem 277, 5631-6.
Hahnen, E., Schonling, J., Rudnik-Schoneborn, S., Raschke, H., Zerres, K. and Wirth, B. (1997). Missense mutations in exon 6 of the survival motor neuron gene in patients with spinal muscular atrophy (SMA). Hum Mol Genet 6, 821-5.
Hakeda-Suzuki, S., Ng, J., Tzu, J., Dietzl, G., Sun, Y., Harms, M., Nardine, T., Luo, L. and Dickson, B. J. (2002). Rac function and regulation during Drosophila development. Nature 416, 438-42.
Harbour, J. W., Luo, R. X., Dei Santi, A., Postigo, A. A. and Dean, D. C. (1999). Cdk phosphorylation triggers sequential intramolecular interactions that progressively block Rb functions as cells move through G1. Cell 98, 859-69.
Hebert, M. D. and Matera, A. G. (2000). Self-association of coilin reveals a common theme in nuclear body localization. Mol Biol Cell 11, 4159-71.
Hebert, M. D., Shpargel, K. B., Ospina, J. K., Tucker, K. E. and Matera, A. G. (2002b). Coilin methylation regulates nuclear body formation. Dev Cell 3, 329-37.
Hebert, M. D., Szymczyk, P. W., Shpargel, K. B. and Matera, A. G. (2001). Coilin forms the bridge between Cajal bodies and SMN, the spinal muscular atrophy protein. Genes & Development 15, 2720-9.
Helmken, C., Hofmann, Y., Schoenen, F., Oprea, G., Raschke, H., Rudnik- Schoneborn, S., Zerres, K. and Wirth, B. (2003). Evidence for a modifying pathway in SMA discordant families: reduced SMN level decreases the amount of its interacting partners and Htra2-beta1. Hum Genet 114, 11-21.
151 Hesser, B. A., Henschel, O. and Witzemann, V. (2006). Synapse disassembly and formation of new synapses in postnatal muscle upon conditional inactivation of MuSK. Mol Cell Neurosci 31, 470-80.
Huber, J., Cronshagen, U., Kadokura, M., Marshallsay, C., Wada, T., Sekine, M. and Luhrmann, R. (1998). Snurportin1, an m3G-cap-specific nuclear import receptor with a novel domain structure. Embo J 17, 4114-26.
Ishov, A. M., Sotnikov, A. G., Negorev, D., Vladimirova, O. V., Neff, N., Kamitani, T., Yeh, E. T., Strauss, J. F., 3rd and Maul, G. G. (1999). PML Is Critical for ND10 Formation and Recruits the PML-interacting Protein Daxx to this Nuclear Structure When Modified by SUMO-1. J Cell Biol 147, 221-234.
Izaurralde, E., Kutay, U., von Kobbe, C., Mattaj, I. W. and Görlich, D. (1997). The asymmetric distribution of the constituents of the Ran system is essential for transport into and out of the nucleus. Embo J 16, 6535-47.
Izaurralde, E., Lewis, J., Gamberi, C., Jarmolowski, A., McGuigan, C. and Mattaj, I. W. (1995). A cap-binding protein complex mediating U snRNA export. Nature 376, 709-12.
Izaurralde, E., Lewis, J., McGuigan, C., Jankowska, M., Darzynkiewicz, E. and Mattaj, I. W. (1994). A nuclear cap binding protein complex involved in pre- mRNA splicing. Cell 78, 657-68.
Jablonka, S., Holtmann, B., Meister, G., Bandilla, M., Rossoll, W., Fischer, U. and Sendtner, M. (2002). Gene targeting of Gemin2 in mice reveals a correlation between defects in the biogenesis of U snRNPs and motoneuron cell death. Proceedings of the National Academy of Sciences of the United States of America 99, 10126-31.
Jablonka, S., Karle, K., Sandner, B., Andreassi, C., von Au, K. and Sendtner, M. (2006). Distinct and overlapping alterations in motor and sensory neurons in a mouse model of spinal muscular atrophy. Hum Mol Genet 15, 511-8.
Jablonka, S., Schrank, B., Kralewski, M., Rossoll, W. and Sendtner, M. (2000). Reduced survival motor neuron (Smn) gene dose in mice leads to motor neuron degeneration: an animal model for spinal muscular atrophy type III. Hum Mol Genet 9, 341-6.
Jady, B. E., Darzacq, X., Tucker, K. E., Matera, A. G., Bertrand, E. and Kiss, T. (2003). Modification of Sm small nuclear RNAs occurs in the nucleoplasmic Cajal body following import from the cytoplasm. EMBO Journal 22, 1878-88.
Jady, B. E. and Kiss, T. (2001). A small nucleolar guide RNA functions both in 2'-O-ribose methylation and pseudouridylation of the U5 spliceosomal RNA. EMBO Journal 20, 541-51.
152 Jady, B. E., Richard, P., Bertrand, E. and Kiss, T. (2006). Cell cycle- dependent recruitment of telomerase RNA and Cajal bodies to human telomeres. Mol Biol Cell 17, 944-54.
Jankowsky, E. (2005). Biophysics: helicase snaps back. Nature 437, 1245.
Kambach, C., Walke, S., Young, R., Avis, J. M., de la Fortelle, E., Raker, V. A., Luhrmann, R., Li, J. and Nagai, K. (1999). Crystal structures of two Sm protein complexes and their implications for the assembly of the spliceosomal snRNPs. Cell 96, 375-87.
Kiss, T. (2004). Biogenesis of small nuclear RNPs. J Cell Sci 117, 5949-51.
Lefebvre, S., Burglen, L., Reboullet, S., Clermont, O., Burlet, P., Viollet, L., Benichou, B., Cruaud, C., Millasseau, P., Zeviani, M. et al. (1995). Identification and characterization of a spinal muscular atrophy-determining gene. Cell 80, 155-65.
Lefebvre, S., Burlet, P., Liu, Q., Bertrandy, S., Clermont, O., Munnich, A., Dreyfuss, G. and Melki, J. (1997). Correlation between severity and SMN protein level in spinal muscular atrophy. Nat. Genet. 16, 265-269.
Liu, J. L., Hebert, M. D., Ye, Y., Templeton, D. J., Kung, H. J. and Matera, A. G. (2000). Cell cycle-dependent localization of the CDK2-cyclin E complex in Cajal (coiled) bodies. J. Cell Sci. 113 ( Pt 9), 1543-52.
Liu, J. L., Murphy, C., Buszczak, M., Clatterbuck, S., Goodman, R. and Gall, J. G. (2006). The Drosophila melanogaster Cajal body. J Cell Biol 172, 875-84.
Liu, Q. and Dreyfuss, G. (1996). A novel nuclear structure containing the survival of motor neurons protein. EMBO J. 15, 3555-3565.
Liu, Q., Fischer, U., Wang, F. and Dreyfuss, G. (1997). The spinal muscular atrophy disease gene product, SMN, and its associated protein SIP1 are in a complex with spliceosomal snRNP proteins. Cell 90, 1013-1021.
Lobo, S. M. a. H., N. (1994). Transcription of snRNA promoters by RNA polymerases II and III. In Transcription, Mechanisms and Regulation, (ed. R. C. C. a. J. W. Conaway), pp. 127-159. New York: Raven Press, Ltd. Lorson, C. L. and Androphy, E. J. (1998). The domain encoded by exon 2 of the survival motor neuron protein mediates nucleic acid binding [published erratum appears in Hum Mol Genet 1998 Oct;7(11):1831]. Hum Mol Genet 7, 1269-75.
Lorson, C. L. and Androphy, E. J. (2000). An exonic enhancer is required for inclusion of an essential exon in the SMA-determining gene SMN. Hum Mol Genet 9, 259-65.
153 Lorson, C. L., Hahnen, E., Androphy, E. J. and Wirth, B. (1999). A single nucleotide in the SMN gene regulates splicing and is responsible for spinal muscular atrophy. Proc Natl Acad Sci U S A 96, 6307-11.
Lorson, C. L., Strasswimmer, J., Yao, J. M., Baleja, J. D., Hahnen, E., Wirth, B., Le, T., Burghes, A. H. and Androphy, E. J. (1998). SMN oligomerization defect correlates with spinal muscular atrophy severity. Nature Genet 19, 63-66.
Lyon, C. E., Bohmann, K., Sleeman, J. and Lamond, A. I. (1997). Inhibition of protein dephosphorylation results in the accumulation of splicing snRNPs and coiled bodies within the nucleolus. Exp. Cell Res. 230, 84-93.
Ma, Y., Dostie, J., Dreyfuss, G. and Van Duyne, G. D. (2005). The Gemin6- Gemin7 heterodimer from the survival of motor neurons complex has an Sm protein-like structure. Structure 13, 883-92.
Matera, A. G. (1998). Of coiled bodies, gems, and salmon. J. Cell. Biochem. 70, 181-192.
Matera, A. G. (1999a). Nuclear bodies: multifaceted subdomains of the interchromatin space. Trends Cell Biol 9, 302-9.
Matera, A. G. (1999b). RNA splicing: More clues from Spinal Muscular Atrophy. Curr. Biol. 9, R140-R142.
Matera, A. G. and Shpargel, K. B. (2006). Pumping RNA: nuclear bodybuilding along the RNP pipeline. Curr Opin Cell Biol.
Mattaj, I. W. (1986). Cap trimethylation of U snRNA is cytoplasmic and dependent on U snRNP protein binding. Cell 46, 905-11.
McAllister, G., Amara, S. G. and Lerner, M. R. (1988). Tissue-specific expression and cDNA cloning of small nuclear ribonucleoprotein-associated polypeptide N. Proc Natl Acad Sci U S A 85, 5296-300.
McConnell, T. S., Lokken, R. P. and Steitz, J. A. (2003). Assembly of the U1 snRNP involves interactions with the backbone of the terminal stem of U1 snRNA. Rna 9, 193-201.
McWhorter, M. L., Monani, U. R., Burghes, A. H. and Beattie, C. E. (2003). Knockdown of the survival motor neuron (Smn) protein in zebrafish causes defects in motor axon outgrowth and pathfinding. Journal of Cell Biology 162, 919-31.
154 Meister, G., Buhler, D., Laggerbauer, B., Zobawa, M., Lottspeich, F. and Fischer, U. (2000). Characterization of a nuclear 20S complex containing the survival of motor neurons (SMN) protein and a specific subset of spliceosomal Sm proteins. Hum Mol Genet 9, 1977-86.
Meister, G., Buhler, D., Pillai, R., Lottspeich, F. and Fischer, U. (2001). A multiprotein complex mediates the ATP-dependent assembly of spliceosomal U snRNPs. Nat Cell Biol 3, 945-9.
Meister, G., Eggert, C. and Fischer, U. (2002). SMN-mediated assembly of RNPs: a complex story. Trends in Cell Biology 12, 472-8.
Misteli, T. (2001). Protein dynamics: Implications for nuclear architecture and gene expression. Science 291, 843-847.
Monani, U. R., Lorson, C. L., Parsons, D. W., Prior, T. W., Androphy, E. J., Burghes, A. H. and McPherson, J. D. (1999). A single nucleotide difference that alters splicing patterns distinguishes the SMA gene SMN1 from the copy gene SMN2. Hum. Mol. Genet. 8, 1177-83.
Monani, U. R., Sendtner, M., Coovert, D. D., Parsons, D. W., Andreassi, C., Le, T. T., Jablonka, S., Schrank, B., Rossol, W., Prior, T. W. et al. (2000). The human centromeric survival motor neuron gene (SMN2) rescues embryonic lethality in Smn(-/-) mice and results in a mouse with spinal muscular atrophy. Hum Mol Genet 9, 333-9.
Mouaikel, J., Narayanan, U., Verheggen, C., Matera, A. G., Bertrand, E., Tazi, J. and Bordonne, R. (2003). Interaction between the small-nuclear-RNA cap hypermethylase and the spinal muscular atrophy protein, survival of motor neuron. EMBO Rep 4, 616-22.
Mouaikel, J., Verheggen, C., Bertrand, E., Tazi, J. and Bordonne, R. (2002). Hypermethylation of the cap structure of both yeast snRNAs and snoRNAs requires a conserved methyltransferase that is localized to the nucleolus. Mol Cell 9, 891-901.
Mourelatos, Z., Dostie, J., Paushkin, S., Sharma, A., Charroux, B., Abel, L., Rappsilber, J., Mann, M. and Dreyfuss, G. (2002). miRNPs: a novel class of ribonucleoproteins containing numerous microRNAs. Genes & Development 16, 720-8.
Mu, Z. M., Le, X. F., Vallian, S., Glassman, A. B. and Chang, K. S. (1997). Stable overexpression of PML alters regulation of cell cycle progression in HeLa cells. Carcinogenesis 18, 2063-9.
Müller, S., Matunis, M. J. and Dejean, A. (1998). Conjugation with the ubiquitin-related modifier SUMO-1 regulates the partitioning of PML within the nucleus. Embo J 17, 61-70.
155 Muratani, M., Gerlich, D., Janicki, S. M., Gebhard, M., Eils, R. and Spector, D. L. (2002). Metabolic-energy-dependent movement of PML bodies within the mammalian cell nucleus. Nat Cell Biol 4, 106-10.
Narayanan, U., Achsel, T., Lührmann, R. and Matera, A. G. (2004b). Coupled in vitro import of U snRNPs and SMN, the spinal muscular atrophy protein. Mol Cell 16, 223-34.
Narayanan, U., Ospina, J. K., Frey, M. R., Hebert, M. D. and Matera, A. G. (2002). SMN, the spinal muscular atrophy protein, forms a pre-import snRNP complex with snurportin1 and importin beta. Hum Mol Genet 11, 1785-95.
Nesic, D., Tanackovic, G. and Kramer, A. (2004). A role for Cajal bodies in the final steps of U2 snRNP biogenesis. J Cell Sci 117, 4423-33.
Ohno, M., Segref, A., Bachi, A., Wilm, M. and Mattaj, I. W. (2000). PHAX, a mediator of U snRNA nuclear export whose activity is regulated by phosphorylation. Cell 101, 187-98.
Olson, M. O., Dundr, M. and Szebeni, A. (2000). The nucleolus: an old factory with unexpected capabilities. Trends Cell Biol 10, 189-96.
Pearn, J. (1978). Incidence, prevalence, and gene frequency studies of chronic childhood spinal muscular atrophy. J Med Genet 15, 409-13.
Pearn, J. (1980). Classification of spinal muscular atrophies. Lancet 8174, 919- 922.
Pellizzoni, L., Charroux, B. and Dreyfuss, G. (1999). SMN mutants of spinal muscular atrophy patients are defective in binding to snRNP proteins. Proc Natl Acad Sci U S A 96, 11167-72.
Pellizzoni, L., Kataoka, N., Charroux, B. and Dreyfuss, G. (1998). A novel function for SMN, the spinal muscular atrophy disease gene product, in pre-mRNA splicing. Cell 95, 615-624.
Pellizzoni, L., Yong, J. and Dreyfuss, G. (2002). Essential role for the SMN complex in the specificity of snRNP assembly. Science 298, 1775-9. Perez, A., Kastner, P., Sethi, S., Lutz, Y., Reibel, C. and Chambon, P. (1993). PMLRAR homodimers: distinct DNA binding properties and heteromeric interactions with RXR. Embo J 12, 3171-82.
Platani, M., Goldberg, I., Lamond, A. I. and Swedlow, J. R. (2002). Cajal body dynamics and association with chromatin are ATP-dependent. Nat Cell Biol 4, 502- 8.
156 Plessel, G., Fischer, U. and Lührmann, R. (1994). m3G cap hypermethylation of U1 small nuclear ribonucleoprotein (snRNP) in vitro: evidence that the U1 small nuclear RNA-(guanosine-N2)-methyltransferase is a non-snRNP cytoplasmic protein that requires a binding site on the Sm core domain. Mol Cell Biol 14, 4160-72.
Rajendra, T. K., Gonsalvez, G. B., Shpargel, K. B., Salz, H. K., and Matera, A. G. A Drosophila model of Spinal Muscular Atrophy reveals a sarcomeric function for SMN in striated muscle. manuscript in preparation.
Ramon y Cajal, S. R. (1903). Un sencillo metodo de coloracion selectiva del reticulo protoplasmico y sus efectos en los diversos organos nerviosos de vertebrados y invertebrados. Trab. Lab. Invest. Biol. (Madrid) 2, 129-221.
Rossoll, W., Jablonka, S., Andreassi, C., Kroning, A. K., Karle, K., Monani, U. R. and Sendtner, M. (2003). Smn, the spinal muscular atrophy-determining gene product, modulates axon growth and localization of beta-actin mRNA in growth cones of motoneurons. J Cell Biol 163, 801-12.
Rossoll, W., Kroning, A. K., Ohndorf, U. M., Steegborn, C., Jablonka, S. and Sendtner, M. (2002). Specific interaction of Smn, the spinal muscular atrophy determining gene product, with hnRNP-R and gry-rbp/hnRNP-Q: a role for Smn in RNA processing in motor axons? Hum Mol Genet 11, 93-105.
Schaffert, N., Hossbach, M., Heintzmann, R., Achsel, T. and Luhrmann, R. (2004). RNAi knockdown of hPrp31 leads to an accumulation of U4/U6 di-snRNPs in Cajal bodies. Embo J 23, 3000-9.
Schrank, B., Gotz, R., Gunnersen, J. M., Ure, J. M., Toyka, K. V., Smith, A. G. and Sendtner, M. (1997). Inactivation of the survival motor neuron gene, a candidate gene for human spinal muscular atrophy, leads to massive cell death in early mouse embryos. Proc. Natl. Acad. Sci. USA 94, 9920-9925.
Seipelt, R. L., Zheng, B., Asuru, A. and Rymond, B. C. (1999). U1 snRNA is cleaved by RNase III and processed through an Sm site-dependent pathway. Nucleic Acids Research 27, 587-95.
Selenko, P., Sprangers, R., Stier, G., Buhler, D., Fischer, U. and Sattler, M. (2001). SMN tudor domain structure and its interaction with the Sm proteins. Nat Struct Biol 8, 27-31.
Shafey, D., Cote, P. D. and Kothary, R. (2005). Hypomorphic Smn knockdown C2C12 myoblasts reveal intrinsic defects in myoblast fusion and myotube morphology. Exp Cell Res 311, 49-61.
157 Sharma, A., Lambrechts, A., Hao, L. T., Le, T. T., Sewry, C. A., Ampe, C., Burghes, A. H. and Morris, G. E. (2005). A role for complexes of survival of motor neurons (SMN) protein with gemins and profilin in neurite-like cytoplasmic extensions of cultured nerve cells. Exp Cell Res.
Shpargel, K. B. and Matera, A. G. (2005). Gemin proteins are required for efficient assembly of Sm-class ribonucleoproteins. Proc Natl Acad Sci U S A 102, 17372- 7.
Shpargel, K. B., Ospina, J. K., Tucker, K. E., Matera, A. G. and Hebert, M. D. (2003). Control of Cajal body number is mediated by the coilin C-terminus. J Cell Sci 116, 303-12.
Sleeman, J., Lyon, C. E., Platani, M., Kreivi, J.-P. and Lamond, A. I. (1998). Dynamic interactions between splicing snRNPs, coiled bodies and nucleoli revealed using snRNP protein fusions to the green fluorescent protein. Exp. Cell Res. 243, 290- 304.
Sleeman, J. E., Ajuh, P. and Lamond, A. I. (2001). snRNP protein expression enhances the formation of Cajal bodies containing p80-coilin and SMN. J Cell Sci 114, 4407-19.
Sleeman, J. E. and Lamond, A. I. (1999a). Newly assembled snRNPs associate with coiled bodies before speckles, suggesting a nuclear snRNP maturation pathway. Curr Biol 9, 1065-74.
Sleeman, J. E. and Lamond, A. I. (1999b). Newly assembled snRNPs associate with coiled bodies before speckles, suggesting a nuclear snRNP maturation pathwayy. Curr. Biol. 9, 1065-1074.
Spector, D. L. (2001). Nuclear domains. J Cell Sci 114, 2891-3.
Spradling, A. C., Stern, D., Beaton, A., Rhem, E. J., Laverty, T., Mozden, N., Misra, S. and Rubin, G. M. (1999). The Berkeley Drosophila Genome Project gene disruption project: Single P-element insertions mutating 25% of vital Drosophila genes. Genetics 153, 135-77.
Sun, Y., Grimmler, M., Schwarzer, V., Schoenen, F., Fischer, U. and Wirth, B. (2005). Molecular and functional analysis of intragenic SMN1 mutations in patients with spinal muscular atrophy. Hum Mutat 25, 64-71.
Talbot, K. and Davies, K. E. (2001). Spinal muscular atrophy. Semin Neurol 21, 189-97.
158 Thibault, S. T., Singer, M. A., Miyazaki, W. Y., Milash, B., Dompe, N. A., Singh, C. M., Buchholz, R., Demsky, M., Fawcett, R., Francis-Lang, H. L. et al. (2004). A complementary transposon tool kit for Drosophila melanogaster using P and piggyBac. Nat Genet 36, 283-7.
Tomlinson, R. L., Ziegler, T. D., Supakorndej, T., Terns, R. M. and Terns, M. P. (2006). Cell cycle-regulated trafficking of human telomerase to telomeres. Mol Biol Cell 17, 955-65.
Tucker, K. E., Berciano, M. T., Jacobs, E. Y., LePage, D. F., Shpargel, K. B., Rossire, J. J., Chan, E. K., Lafarga, M., Conlon, R. A. and Matera, A. G. (2001). Residual Cajal bodies in coilin knockout mice fail to recruit Sm snRNPs and SMN, the spinal muscular atrophy gene product. J Cell Biol 154, 293-307.
Verheggen, C., Lafontaine, D. L., Samarsky, D., Mouaikel, J., Blanchard, J. M., Bordonne, R. and Bertrand, E. (2002). Mammalian and yeast U3 snoRNPs are matured in specific and related nuclear compartments. Embo J 21, 2736-45.
Vitte, J. M., Davoult, B., Roblot, N., Mayer, M., Joshi, V., Courageot, S., Tronche, F., Vadrot, J., Moreau, M. H., Kemeny, F. et al. (2004). Deletion of murine Smn exon 7 directed to liver leads to severe defect of liver development associated with iron overload. Am J Pathol 165, 1731-41.
Wan, L., Battle, D. J., Yong, J., Gubitz, A. K., Kolb, S. J., Wang, J. and Dreyfuss, G. (2005). The survival of motor neurons protein determines the capacity for snRNP assembly: biochemical deficiency in spinal muscular atrophy. Mol Cell Biol 25, 5543-51.
Will, C. L. and Luhrmann, R. (2001). Spliceosomal UsnRNP biogenesis, structure and function. Current Opinion in Cell Biology 13, 290-301.
Winkler, C., Eggert, C., Gradl, D., Meister, G., Giegerich, M., Wedlich, D., Laggerbauer, B. and Fischer, U. (2005). Reduced U snRNP assembly causes motor axon degeneration in an animal model for spinal muscular atrophy. Genes Dev 19, 2320- 30.
Wirth, B. (2000). An update of the mutation spectrum of the survival motor neuron gene (SMN1) in autosomal recessive spinal muscular atrophy (SMA). Hum Mutat 15, 228-37.
Yan, X., Mouillet, J. F., Ou, Q. and Sadovsky, Y. (2003). A novel domain within the DEAD-box protein DP103 is essential for transcriptional repression and helicase activity. Molecular & Cellular Biology 23, 414-23.
159 Yong, J., Golembe, T. J., Battle, D. J., Pellizzoni, L. and Dreyfuss, G. (2004). snRNAs contain specific SMN-binding domains that are essential for snRNP assembly. Mol Cell Biol 24, 2747-56.
Yong, J., Pellizzoni, L. and Dreyfuss, G. (2002). Sequence-specific interaction of U1 snRNA with the SMN complex. EMBO Journal 21, 1188-96.
Young, P. J., Le, T. T., Dunckley, M., Nguyen, T. M., Burghes, A. H. and Morris, G. E. (2001). Nuclear gems and Cajal (coiled) bodies in fetal tissues: nucleolar distribution of the spinal muscular atrophy protein, SMN. Exp Cell Res 265, 252-61.
Young, P. J., Le, T. T., thi Man, N., Burghes, A. H. and Morris, G. E. (2000). The relationship between SMN, the spinal muscular atrophy protein, and nuclear coiled bodies in differentiated tissues and cultured cells. Exp Cell Res 256, 365-74.
160