DEVELOPMENTAL DYNAMICS 241:505–521, 2012

a RESEARCH ARTICLE

TEG-1 CD2BP2 Regulates Stem Cell Proliferation and Sex Determination in the C. elegans Germ Line and Physically Interacts With the UAF-1 U2AF65 Splicing Factor

Chris Wang,1 Laura Wilson-Berry,2 Tim Schedl,2 and Dave Hansen1*

Background: For a stem cell population to exist over an extended period, a balance must be maintained between self-renewing (proliferating) and differentiating daughter cells. Within the Caenorhabditis ele- gans germ line, this balance is controlled by a genetic regulatory pathway, which includes the canonical Notch signaling pathway. Results: Genetic screens identified the gene teg-1 as being involved in regulat- ing the proliferation versus differentiation decision in the C. elegans germ line. Cloning of TEG-1 revealed that it is a homolog of mammalian CD2BP2, which has been implicated in a number of cellular processes, including in U4/U6.U5 tri-snRNP formation in the pre-mRNA splicing reaction. The position of teg-1 in the genetic pathway regulating the proliferation versus differentiation decision, its single mutant phenotype, and its enrichment in nuclei, all suggest TEG-1 also functions as a splicing factor. TEG-1, as well as its human homolog, CD2BP2, directly bind to UAF-1 U2AF65, a component of the U2 auxiliary factor. Conclusions: TEG-1 functions as a splicing factor and acts to regulate the proliferation versus meiosis decision. The interaction of TEG-1 CD2BP2 with UAF-1 U2AF65, combined with its previously described function in U4/U6.U5 tri-snRNP, suggests that TEG-1 CD2BP2 functions in two distinct locations in the splicing cascade. Developmental Dynamics 241:505–521, 2012. VC 2012 Wiley Periodicals, Inc.

Key words: CD2BP2; TEG-1; stem cell proliferation; pre-mRNA splicing; UAF-1; U2AF65

Key findings: Developmental Dynamics TEG-1 inhibits stem cell proliferation in the C. elegans germ line. TEG-1 likely functions as a splicing factor. TEG-1 CDBP2 physically interacts with UAF-1 U2AF65.

Accepted 3 January 2012

INTRODUCTION is tied to their ability to produce to make the desired tissue. Maintain- both self-renewing and differentiating ing the balance between self-renewing Stem cells provide the material neces- sary for tissue generation during devel- daughter cells. Self-renewing daughter and differentiating daughter cells is opment, and tissue replacement and cells retain the stem cell characteristics key to proper stem cell function since a repair in adult animals. The capacity of of the parental cell; therefore, a pool of disruption in this balance prevents stemcellstocontributetotheseproc- stem cells is maintained over time. The long term tissue generation; too little esses over an extended time period differentiating daughter cells are used self-renewal results in the eventual

1University of Calgary, Department of Biological Sciences, Alberta, Calgary, Canada 2Department of Genetics, Washington University School of Medicine, St. Louis, Missouri Grant sponsor: Canadian Institutes of Health Research (CIHR); Natural Sciences and Engineering Research Council of Canada (NSERC); National Institutes of Health (NIH); Grant number: GM63310. *Correspondence to: Dave Hansen, Department of Biological Sciences, University of Calgary, Calgary, Alberta, Canada T2N 1N4. E-mail: [email protected] DOI 10.1002/dvdy.23735 Published online 30 January 2012 in Wiley Online Library (wileyonlinelibrary.com).

VC 2012 Wiley Periodicals, Inc. 506 WANG ET AL.

depletion of the stem cell population, decrease, allowing cells to enter mei- proliferative fate, resulting in a com- while too little differentiation results in otic prophase. The first obvious indica- plete germline tumor (Berry et al., not enough cells being made for proper tion that cells have entered into mei- 1997; Hansen et al., 2004a). Other tissue generation. Understanding how otic prophase is when the DNA takes genes have been identified that func- this balance between self-renewal and on a crescent moon shape as chromo- tion with the GLP-1/Notch signaling differentiation is maintained is an somes pair and synapse (Francis et al., and gld-1/gld-2 pathways, or parallel essential step in understanding stem 1995a; Dernburg et al., 1998; Mac- to them, to regulate the balance cell biology and function. Queen and Villeneuve, 2001). Cells between proliferation and differentia- The germ line of the nematode continue to move proximally while tion. These genes function in many Caenorhabditis elegans has emerged transiting through the various stages different cellular processes. For exam- as a powerful model to study how a of meiotic prophase, eventually form- ple, mutations that decrease the func- stem cell population maintains the ing fully differentiated gametes tion of the proteasome enhance over- balance between self-renewal and dif- (sperm or oocytes depending on the proliferation mutants, indicating a ferentiation (Hansen and Schedl, age of the hermaphrodite). function to promote meiotic develop- 2006; Kimble and Crittenden, 2007). The GLP-1/Notch signaling path- ment (MacDonald et al., 2008). As In the C. elegans hermaphrodite, the way promotes the proliferative fate, another example, the cell-cycle regu- germ line is contained in two gonad at least in part, by inhibiting the lators Cyclin E (CYE-1) and CDK2 arms that meet at a common uterus. activities of two redundant meiotic (CDK-2) promote the proliferative Populations of mitotically dividing promoting genetic pathways, referred fate (Fox et al., 2011; Jeong et al., (proliferating) cells exist at the ends to as the gld-1 and gld-2 pathways 2011). A number of splicing factors of the two gonad arms distal from the (Kadyk and Kimble, 1998) (Fig. 1B). have also been identified as being uterus (Fig. 1A). In the adult, each Within the gld-1 pathway are the gld- involved in regulating the prolifera- population consists of 200–250 cells 1 and nos-3 genes (Francis et al., tion versus differentiation decision (Lamont et al., 2004; Killian and Hub- 1995b; Hansen et al., 2004b); with (Belfiore et al., 2004; Mantina et al., bard, 2005), with cells progressing gld-1 encoding a KH domain contain- 2009; Kerins et al., 2010; Zanetti through the cell cycle approximately ing RNA binding translational regula- et al., 2011). Genetically, these splic- every 6.5–8 hr (Fox et al., 2011). At tor (Jones and Schedl, 1995), and nos- ing factors appear to function largely least some of these proliferating cells 3 encoding a protein with similarity in the gld-1 pathway, functioning are considered stem cells. Cells are to Drosophila Nanos (Kraemer et al., downstream of GLP-1/Notch signal- kept in the proliferative state due to 1999), also a translational regulator. ing (Fig. 1B). A screen of splicing fac- their close proximity to the Distal Tip Within the gld-2 pathway are the gld- tors found that proteins functioning Cells (DTCs), with a single DTC cap- 2 and gld-3 genes, encoding the cata- throughout the splicing cascade par- ping the distal end of each gonad arm. lytic portion of a poly(A) polymerase ticipate in the proliferation/differen- The proliferation promoting influence (Kadyk and Kimble, 1998), and a tiation decision, suggesting that an of the DTC is primarily accomplished homolog to Drosophila BicC, respec- overall decrease in splicing or spliceo- through the conserved GLP-1/Notch tively (Eckmann et al., 2002). The some activity contributes to over-pro-

Developmental Dynamics signaling pathway (Hansen and spatial organization of the activity of liferation (Kerins et al., 2010). Schedl, 2006; Kimble and Crittenden, the GLP-1/Notch signaling pathway Pre-mRNA splicing involves the re- 2007; Racher and Hansen, 2010). The relative to the activities of the gld-1 moval of introns from the pre-mRNA DSL (Delta/Serrate/LAG-2) LAG-2 and gld-2 pathways is key to main- between the exons prior to nuclear and APX-1 ligands are expressed on taining the balance between prolifera- export and translation. The splicing the surface of the DTC (Henderson tion and differentiation. Loss or reaction is accomplished through the et al., 1994; Tax et al., 1994; Nadara- reduction of GLP-1/Notch signaling in activity of the spliceosome, which is a jan et al., 2009). When they bind the the distal end results in ectopic acti- large RNA-Protein complex that is GLP-1/Notch receptor (Yochem and vation of the gld-1 and gld-2 path- assembled in a stepwise manner Greenwald, 1989), which is expressed ways in this region, causing cells to (Wahl et al., 2009; Valadkhan and on the surface of the germ cells (Crit- enter meiosis prematurely and a loss Jaladat, 2010). Some of the earliest tenden et al., 1994), the cleaved intra- of the proliferative population (Austin steps in spliceosome assembly include cellular portion of GLP-1 (referred to and Kimble, 1987; Eckmann et al., recognition of the 50 and 30 splice sites as GLP-1(INTRA)) is thought to enter 2004; Hansen et al., 2004b). Con- on the pre-mRNA by protein or RNA/ the nucleus and bind to the CSL versely, ectopic activation of GLP-1/ protein complexes. The U1 snRNP (CBF1/Suppressor of Hairless/LAG-1) Notch signaling beyond its normal particle, consisting of the U1 snRNA LAG-1 transcription factor (Christen- region of activity, or a combined and a number of proteins, recognizes sen et al., 1996; Mumm and Kopan, reduction in the activities of the and binds to the 50 splice site (Sera- 2000) and the SEL-8/LAG-3 co-activa- redundant gld-1 and gld-2 pathways, phin and Rosbash, 1989; Michaud tor (Doyle et al., 2000; Petcherski and results in excess proliferation (Berry and Reed, 1991). The U2 Auxiliary Kimble, 2000). This complex then acti- et al., 1997; Kadyk and Kimble, 1998; Factor (U2AF) recognizes and binds vates genes necessary for the prolifer- Pepper et al., 2003; Hansen et al., to the 30 splice site (Wahl et al., 2009; ative fate. As cells move along the 2004b). Indeed, in certain mutant Valadkhan and Jaladat, 2010). gonad arm, away from the DTC, Notch gene and/or allele combinations, all Human U2AF consists of two pro- signaling levels are thought to cells within the gonad arm have the teins, U2AF65 and U2AF35 (Zamore Developmental Dynamics i.1. Fig. i.2. Fig. itledo h oa,the gonad, the of end distal the as gld-2 down- to redundant referred two pathways, in stream genes of inhibit activities intermediates, the of path- number signaling a Notch through way, core the of components shown are are entry black. in meiotic proliferation promote prolifer- and/or inhibit promote that ation genes while green, that in shown Genes versus line. the genes proliferation in the decision showing differentiation regulating in pathway involved genetic simplified bar Scale first oocytes. forming then eventually sperm, and zone, in transition meiosis the entering gonad proximally, the move of Cells end prolifer-arm. distal or the meiotic mitotic in existing with cells mark left, ative the and to to is cells, Distal (red) cells. cycling antibodies mitotically antibodies anti-HIM-3 mark anti-REC-8 to (green) visualize morphology, to (blue) DAPI nuclear with stained arm gonad dissected hermaphrodite stage) larval fourth the the line. in determination germ sex and ance 1. Fig. rlfrto.Aatdfo asnadSchedl (2006). and inhibiting Hansen from and/or Adapted proliferation. promoting (differentiation) thereby entry meiotic active, become pathways h eei eainhp ewe genes the between in elegans determination C. sex relationships regulating in involved genetic the dl emprdtspoueoocytes. produce (2006). Schedl and Ellis hermaphrodites from Adapted that such development adult during black. made is then in switch and shown a sperm, that are produce genes first fate while Hermaphrodites female blue, the in shown promote are fate male vrl ieo h itlmttczn sstatisti- is in zone larger mitotic distal cally the of size overall with loehne h vrpoieaiephenotype of over-proliferative the enhances (red). also observed into are entered have meiosis that cells of n clusters although phenotype, over-proliferation this show of arms teg-1(oz230) found gonad are the the cells of throughout mitotic ends however, proximal arms); the gonad arm in gonad observed is the sperm of with cross-reactivity part anti-REC-8 some distal (although the to restricted n ftegndamta r ioi (anti- bars Error mitotic test. are anti-HIM-3( that nuclei, arm gonad distal REC-8( the the from of diameters cell end measured of as number mutant, the single by either in than mals n phe- over-proliferation (100% notype, this (E) show arms arms gonad gonad of the and of distal regions the proximal both in proliferation extensive ing teg-1(oz230) 2(q497); ( n niHM3atbde rd omr meiotic Both mark cells. to cells, (red) antibodies mitotic anti-HIM-3 mark and with anti- to morphology, (green) stained nuclear antibodies REC-8 show were to arms (blue) DAPI Gonad adult fourth stage. the from past larval day arms back- one animals gonad genetic hermaphrodite different Dissected in grounds. phenotype erative 2. Fig. B nml,gona 15 at grown animals, ) gld-2(q497) unc-32(e189) ahas sclsmv wyfo the from away move cells As pathways. þ C: gen n itlt n meiotic any to distal and [green] ) h rlfrto/ifrnito bal- proliferation/differentiation The teg-1(oz230) a-,glp-1 lag-2, > ipiidgntcptwyshowing pathway genetic simplified A A: emln.Gnsta rmt the promote that Genes line. germ 0.Alsriswr lomarked also were strains All 20). teg-1(oz230) ( C idtp dl oedypast day (one adult wild-type A l-(r0g)teg-1(oz230) glp-1(ar202gf) ¼ nml 10 fgndarms gonad of (100% animals ) igemtns() with (D), mutants single 1SD.n cl bar Scale . þ ;[red]). ); ,and nacsteover-prolif- the enhances obemtnsdisplay- mutants double ¼ ( A ,so proliferation show C, lag-1 and ) 0gndarms. gonad 20 P ¼ D,E .elegans C. gld-1 < hc encode which , ¼ 20 : glp-1(ar202gf) glp-1(ar202gf) 9.7 20 teg-1(oz230) m .elegans C. and gld-1 m. m m. 10-5 F: > gld-2 germ B: gld- and ani- The 20), A t - 508 WANG ET AL.

and Green, 1989), which in C. elegans found that TEG-1 binds UAF-1, the balance between proliferation and dif- are UAF-1 and UAF-2, respectively. worm homolog of the U2AF65 splicing ferentiation (Kadyk and Kimble, U2AF65 binds to a polypyrimidine factor, and that this interaction is con- 1998), and since the activities of both tract that is found at the 30 end of served in humans. Therefore, TEG-1 the gld-1 and gld-2 pathways must be introns, while U2AF35 binds to the 30 CD2BP2 likely is involved in two dis- reduced in order for a germline splice site (Wahl et al., 2009; Valad- tinct steps in the splicing reaction. tumour to be formed, this screen was khan and Jaladat, 2010). Although in specifically designed to identify genes humans the polypyrimidine tract can that function in the gld-1 genetic vary somewhat in sequence and posi- RESULTS pathway, although mutations in genes 0 tion relative to the 3 splice site, in C. teg-1 Alleles Identified in Two that function upstream of Notch sig- elegans the conserved UUUUCAG/R Over-Proliferation Enhancer naling, or in a parallel pathway, may sequence is found immediately adja- also cause over-proliferation in a gld- 0 Screens cent to the 3 splice site (Blumenthal 2(q497) background. Since gld- and Steward, 1997). Both UAF-1 and In an effort to better understand how 2(q497) homozygous animals are ster- UAF-2 bind to this conserved sequence the balance between proliferation and ile due to defects in gametogenesis (Zorio and Blumenthal, 1999a). Bind- differentiation is controlled in the (Kadyk and Kimble, 1998), we muta- ing of U2AF to the 30 splice site helps C. elegans germ line, we sought to genized gld-2(q497) homozygotes that in the recruitment of the U2 snRNP identify genes involved in regulating were fertile because they carried the (Wahl et al., 2009; Valadkhan and this process. To identify these genes, gaDp1 free duplication, which con- Jaladat, 2010). In addition to recogni- we performed two genetic screens for tains a wild-type copy of gld-2. The tion of the 50 and 30 splice sites by U1 mutations that enhance germline progeny (F2) of cloned F1 animals and U2AF, respectively, SF1/BBP over-proliferation. In the first screen, were screened for those with germline (SFA-1 in C. elegans; Mazroui et al., we mutagenized animals that con- over-proliferation. These tumorous 1999), binds to the branch point (Abo- tained the glp-1(oz112oz120) weak animals presumably lost the gaDp1 vich and Rosbash, 1997; Berglund gain-of-function allele (Berry et al., free duplication carrying the wild- et al., 1997), which is just upstream 1997). These animals are relatively type copy of gld-2, and were homozy- from the polypyrimidine tract. Subse- wild-type with respect to germline gous for the newly generated muta- quent steps in spliceosomal assembly proliferation at 15C, with only 0.05% tion. Since these animals were sterile, include replacement of SF1/BBP at of glp-1(oz112oz120) gonad arms con- we recovered the induced mutation the branch point by U2, recruitment of taining an over-proliferative germ from siblings that carried gaDp1[gld- the U4/U6.U5 tri-snRNP, and disasso- line (4/8,000 gonad arms) (Berry, 2(þ)], and were either homozygous or ciation of U1 and U2 from the spliceo- 1998). We screened for recessive heterozygous for the new mutation. some. This stepwise assembly contrib- enhancers of the over-proliferative We called genes identified in this utes to the activated spliceosome phenotype by cloning F1 progeny of screen syt (synthetic tumorous). We (complex B*), which allows for the two mutagenized animals and screening previously reported two genes identi- transesterification reactions of intron for those that showed 25% of prog- fied in this screen; syt-1, which is

Developmental Dynamics removal to occur. The first transesteri- eny (F2) with tumorous germ lines. allelic to nos-3 (Hansen et al., 2004b), fication reaction results in a free 50 From screening 8,200 haploid and pas-5, which encodes a subunit of exon and the 50 end of the intron genomes, we identified 12 enhancing the proteasome (MacDonald et al., attached to the branch site in a lariat mutations, with the strongest 7 corre- 2008). From this screen, we also iso- formation. The second transesterifica- sponding to 4 complementation lated the teg-1(oz230) allele; there- tion reaction ligates the two exons and groups. We have named these teg- fore, we isolated two alleles of teg-1, liberates the intron (Wahl et al., 2009; 1(oz189), teg-2(oz192, oz194, oz216, each from independent screens with Valadkhan and Jaladat, 2010). oz218), teg-3(oz190), and teg-4(oz210) different genetic backgrounds. Here we demonstrate that reduction (tumourous enhancer of glp- in teg-1 activity causes a disruption in 1(oz112oz120)). We have previously Analysis of teg-1 the balance between proliferation and reported the cloning of teg-4, which Over-Proliferation differentiation in the C. elegans germ encodes a homolog of the human line. In two different sensitized genetic SAP130 splicing factor (a subunit of Since teg-1 mutant alleles were iso- backgrounds, teg-1 mutants were iden- the SF3b splicing complex) (Mantina lated from two different screens, we tified by their ability to enhance over- et al., 2009). teg-2 and teg-3 are in know that teg-1 mutants can cause/ proliferation, resulting in a germ line various stages of characterization, enhance over-proliferation in differ- tumor. We identified teg-1 as encoding mapping, and cloning, which we will ent sensitized genetic backgrounds. a homolog of proteins from yeast to discuss elsewhere. Here we report the To more closely examine the nature of humans, including CD2BP2 and yeast characterization and cloning of teg-1. these over-proliferative tumors, we Lin1p, which have been implicated in In the second screen, we identified stained dissected gonads with numerous cellular functions, including mutations that have a synthetic over- markers for mitotic and meiotic cells; splicing. Within the splicing reaction, proliferation phenotype with the gld- anti-REC-8 antibodies mark mitotic CD2BP2 has been suggested to be 2(q497) allele. Since gld-2 pathway zone cells (including those in pre-mei- involved in U4/U6.U5 tri-snRNP for- genes function redundantly with gld- otic S-phase) and anti-HIM-3 antibod- mation (Laggerbauer et al., 2005). We 1 pathway genes in regulating the ies mark meiotic prophase cells TEG-1 CD2BP2 INHIBITS STEM CELL PROLIFERATION 509

(Zetka et al., 1999; Pasierbek et al., of these genes, Y47D3A.27, contained rich sequences (Nishizawa et al., 2001; Hansen et al., 2004a). In per- premature stop codons associated 1998; Freund et al., 1999). Alignment forming the characterization, we used with each of the two teg-1 alleles (Fig. of proteins from numerous species the glp-1(ar202gf) allele for the sensi- 3). An integrated transgene contain- revealed a sequence signature for the tized background to be consistent ing the Y47D3A.27 gene rescues the domain; W-X-Y-X6-11-G-P-F-X4-M-X2- with the analyses of other over-prolif- mutant phenotype of teg-1(oz230) (see W-X3-G-Y-F, where X denotes any eration enhancing mutations (Hansen below). Additionally, antibodies raised amino acid (Kofler and Freund, 2006). et al., 2004a,b; MacDonald et al., against the predicted Y47D3A.27 pro- TEG-1 does contain some sequence 2008; Mantina et al., 2009). We found tein, which detects a protein of the similarity to the GYF signature in its that glp-1(ar202gf) teg-1(oz230) germ correct size in wild-type extracts, fails carboxy end; however, it does not com- lines were over-proliferative, with to detect a protein in teg-1(oz230) mu- pletely match the signature (Fig. 3). excess REC-8(þ) HIM-3() cells as tant extracts (see below). Together, Therefore, based solely on sequence compared to wild type or either single these data support teg-1 as being similarity to the characterized GYF mutant (Fig. 2). However, the excess Y47D3A.27. domain, it is not clear if TEG-1 con- proliferation was primarily found in Sequencing of a cDNA correspond- tains this functional protein interac- the distal and proximal regions of the ing to teg-1, yk82g9, as well as tion domain in its carboxy terminus gonad, with meiotic cells found in sequencing RT-PCR generated cDNA (see below). between. The average size of the dis- using primers that are predicted to tal mitotic zone in glp-1(ar202gf) teg- amplify the entire coding region, 1(oz230) double mutant animals was revealed a gene model differing TEG-1 Controls Germline 35 cell diameters from the distal end, slightly from the predicted gene Development in a Similar while the average sizes for the teg- model for Y47D3A.27 (http:// Fashion to Splicing Factors 1(oz230) and glp-1(ar202gf) single ws220.wormbase.org/). Sequencing of mutants were 14 and 22 cell diame- these cDNAs suggests a gene model As mentioned, TEG-1 is homologous ters, respectively (Fig. 2C, F). There- that contains five exons and is pre- to CD2BP2 homologs, which have fore, the double mutant displays a dicted to produce a protein of 353 been suggested to have many func- late-onset-tumorous phenotype in amino acids (as compared to the pre- tions. Perhaps the best-characterized which the size of the distal mitotic dicted gene model that includes six function of these homologs is their zone is larger than wild type, which exons and a 370–amino acid protein; role in pre-mRNA splicing, with has a mitotic zone 20 cell diameters Fig. 3). The oz230 and oz189 alleles CD2BP2 functioning in the formation in length (Crittenden et al., 1994; contain premature stop codons at of the U4/U6.U5 tri-snRNP (Lagger- Hansen et al., 2004a). amino acid positions 53 and 142, bauer et al., 2005). Interestingly, As has been observed with other respectively. Therefore, oz189 and splicing factors have been shown to be mutations that are synthetic tumor- oz230 are likely strong loss-of-func- involved in regulating the prolifera- ous with gld-2(q497), gld-2(q497); teg- tion or null alleles. tion versus differentiation decision in 1(oz230) animals, although over-pro- BLAST alignment of the predicted the C. elegans germ line (Puoti and

Developmental Dynamics liferative, have some cells that enter TEG-1 protein identifies homologous Kimble, 1999, 2000; Belfiore et al., meiotic prophase (Fig. 2E). The pro- proteins from a variety of species, 2004; Konishi et al., 2008; Mantina portion of cells that appear meiotic including CD2BP2 from humans, et al., 2009; Kerins et al., 2010; differ from animal to animal; how- DmLIN1 from Drosophila mela- Zanetti et al., 2011); therefore, teg-1’s ever, all gonad arms have some cells nogaster, and Lin1p from Saccharo- involvement in this decision could be that have entered into meiosis (anti- myces cerevisiae (Fig. 3). Although through a role in pre-mRNA splicing. HIM-3(þ)). Therefore, the combined conserved regions are found over the However, the overall low level of simi- removal of gld-2 and teg-1 function entire lengths of the proteins, the larity between TEG-1 and its homo- does not completely shift the balance level of similarity is not overly high; logs, as well as the other cellular func- towards proliferation. TEG-1 is 42% similar and 31% identi- tions for CD2BP2 homologs, prevent cal to human CD2BP2, 42% similar us from assuming that TEG-1 has a teg-1 Encodes a Homolog of and 31% identical to Drosophila role in splicing. We reasoned that if DmLIN1, and 26% similar and 15% TEG-1 is functioning as a splicing fac- Human CD2BP2 identical to yeast Lin1p. CD2BP2 tor, and if it is a loss of this function To determine how teg-1 may be func- homologs have been suggested to in teg-1 mutants that contributes to tioning at the molecular level, we function in a number of processes, the over-proliferation defect, teg-1 mapped and cloned the gene. teg- including pre-mRNA splicing, DNA mutants should show similar pheno- 1(oz230) was localized to a 0.4-cM replication, chromosome segregation, types and epistatic relationships as region between spe-16 and dpy-18 on and the mammalian immune other splicing factor mutants. To test linkage group III by standard three- response (Nishizawa et al., 1998; this, we first made double mutants factor mapping. Single nucleotide Freund et al., 1999). CD2BP2 homo- with teg-1(oz230) and mutants of the polymorphism and deficiency map- logs do not have any obvious motifs, genes functioning in the gld-1 and ping further narrowed the critical other than a GYF motif in the carboxy gld-2 pathways regulating the prolif- region to 25 kb containing three end. The GYF domain was first iden- eration versus meiotic entry decision. genes. Sequencing revealed that one tified in CD2BP2, and binds proline- As with other splicing factor mutants, 510 WANG ET AL.

Fig. 3. TEG-1 is homologous to proteins from yeast to humans. A: Sequencing of the yk82g9 cDNA, analysis of contig MM454_contig00336, as well as sequencing reverse transcribed PCR amplification products, revealed a gene model for teg-1 that includes five exons. Boxes repre- Developmental Dynamics sent exons, while joining lines represent introns. Blackened portions of the exons represent non-coding regions of the exons (50 and 30 UTRs). From the A in the initiator ATG, the first exon includes nucleotides from position 1–138, exon two 239–544, exon three 1,948–2,220, exon four 3,106–3,315 (the gene model contained in wormbase places exon four at 3,106–3,323), exon five 4,218–4,352 (Note: only includes coding por- tions of exons, including taa stop codon). 50RACE and 30RACE revealed a 7-bp 50UTR and a 479-bp 30UTR, and also revealed that the teg-1 transcript is SL1 trans-spliced. The positions of the mutations associated with the two alleles of teg-1 are shown, with the oz230 mutation changing AAG to UAG (amber) at codon 53 (position 257 on genomic DNA from ATG). The oz189 mutation changes UGG to UGA (opal) at codon 142 (position 526 on genomic DNA from ATG). B: An alignment of amino acid sequences between the TEG-1 predicted protein and homologs from Saccharomyces cerevisiae, Drosophila melanogaster, and humans. Alignment was performed using ClustalW, with similar amino acids shown with black text on a grey background, and identical amino acids shown with white text on a black background. Amino acids were considered similar or identical if three or four of the four proteins contained a similar or identical amino acid at a given position. The thick black underline at the end of the proteins represents the regions deleted in the DGYF constructs. Asterisks are below each of the amino acids in CD2BP2 that are part of the GYF signature (W-X-Y-X6-11-G-P-F-X4-M-X2-W-X3-G-Y-F) (Kofler and Freund, 2006). The predicted locations of the premature stop codons introduced by the oz189 and oz230 alleles are shown. TEG-1 is 14.8, 30.9, and 30.7% identical to ScLin1p, DmLIN1, and HsCD2BP2, respectively; and 25.6, 42.2, and 41.7% similar to these same proteins. Genbank accession numbers: ScLin1p (NP_012026.1), DmLin1 (Q9VKV5), and HsCD2BP2 (NP_006101.1).

teg-1(oz230) forms synthetic tumors which is also consistent with what germ line, gld-2(q497); glp-1(q175) with both of the gld-2 pathway genes has been found with other splicing teg-1(oz230) animals are not tumor- (gld-2 and gld-3), but not with the factor mutants (Mantina et al., 2009; ous, but rather display the under-pro- gld-1 pathway genes (gld-1 and nos-3; Kerins et al., 2010). liferative Glp phenotype associated Table 1), suggesting that teg-1 likely Additionally, we also observed a dif- with glp-1(q175) (Table 1). Similar functions in the gld-1 pathway. Addi- ference in epistasis with gld-2 and epistasis results were observed for tionally, the gld-3(q730); teg-1(oz230) gld-3, suggesting that their functions the teg-4 and prp-17 splicing factors tumor is epistatic to glp-1(q175null) are not equivalent in the gld-2 path- (Mantina et al., 2009; Kerins et al., (Table 1), suggesting that teg-1 func- way. While gld-3(q730); glp-1(q175) 2010). Tumor formation in gld- tions downstream of Notch signaling, teg-1(oz230) animals have a tumorous 3(q730); glp-1(q175) teg-1(oz230), TEG-1 CD2BP2 INHIBITS STEM CELL PROLIFERATION 511

gld-3 reduces the activity of the gld-2 TABLE 1. Epistasis Analysis of teg-1 With Genes Regulating the pathway to a greater extent than the Proliferation Vs. Differentiation Decision removal of gld-2 activity, or that gld-3 has functions in addition to its role in Proliferation vs. differentiation the gld-2 pathway. In any event, the a Genotype phenotype (20 C) fact that a synthetic tumor is formed teg-1(oz189 and oz230) Wild-type in gld-3(q730); teg-1(oz230) animals, gld-1(q485) Wild-typeb even in the absence of GLP-1/Notch c nos-3(oz231) Wild-type signaling, strongly supports the gld-2(q497) Wild-typed e placement of teg-1 downstream of gld-3(q730) Wild-type GLP-1/Notch signaling. Additionally, glp-1(q175) Glpf the fact that teg-1 shows the same gld-2(q497) gld-1(q485) Tumorousg gld-1(q485); teg-1(oz230) Wild-typeh complex epistatic relationships as the nos-3(oz231); teg-1(oz230) Wild-typei teg-4 and prp-17 splicing factors sup- gld-2(q497); teg-1(oz230) Tumorousj ports teg-1 functioning as a splicing gld-3(q730); teg-1(oz230) Tumorous factor. gld-2(q497) gld-1(q485); glp-1(q175) Tumorousk l gld-2(q497); glp-1(q175) teg-1(oz230) Glp Characterization of teg-1 gld-3(q730); glp-1(q175) teg-1(oz230) Tumorousm Single Mutants a 100% of animals show the phenotype described, unless otherwise stated. For each To further compare teg-1(0) pheno- genotype, a minimum of 30 gonad arms was analyzed. b types with those observed in other (Francis et al., 1995a). gld-1(q485) XX animals have tumorous germ lines; however, splicing factor mutants, we analyzed this has to do with female germ cells entering into meiotic prophase, but failing to teg-1’s single mutant phenotypes. As progress through meiosis. These cells then return to mitosis, forming a tumor. The initial entry into meiosis in these animals is very similar to wild-type; therefore, with other splicing factor mutants they do not have a defect in the balance between proliferation and meiotic entry. (Belfiore et al., 2004; Mantina et al., c(Hansen et al., 2004b). 2009; Kerins et al., 2010; Zanetti d(Kadyk and Kimble, 1998). et al., 2011), teg-1 has a role in germ- e(Eckmann et al., 2004). line sex determination; loss of teg-1 f(Austin and Kimble, 1987). Germ line proliferation abnormal. The proliferative pop- function results in a cold sensitive ulation of cells is depleted, resulting in gonads containing only a limited number of Mog (masculinization of the germ differentiated cells. In glp-1 animals, only 16 sperm are formed per gonad arm. g line) phenotype. At 15 C, teg-1 her- (Kadyk and Kimble, 1998). maphroditic animals fail to switch hVab progeny from gld-1(q485)/hT2; vab-7(e1562) teg-1(oz230) were examined by from the production of sperm to Nomarski optics and dissected gonads were analyzed using fluorescent microscopy. Both XX and XO animals were examined. oocytes, resulting in over 600 sperm 6 ¼ iUnc progeny from nos-3(oz231); unc-32(e189) teg-1(oz230) were examined by made per gonad arm (632 156, n 5), Developmental Dynamics Nomarski optics. as compared to 150 in wild-type ani- jVab or Unc self or cross progeny from gld-2(q497)/hT2; vab-7(e1562) teg-1(oz189 or mals (Hirsh et al., 1976) (Fig. 4). We oz230)/hT2 or gld-2(q497)/hT2; unc-32(e189) teg-1(oz189 or oz230)/hT2 were also found that fog-1(q241), fog- examined by Nomarski optics. Also, dissected gonads were DAPI stained and ana- 3(q443), and fem-3(e1996) are all epi- lyzed using fluorescent microscopy. static to teg-1(oz230) (Table 2), sug- k(Kadyk and Kimble, 1998). l gesting that teg-1 likely functions Premature entry to meiosis phenotype similar to a glp-1(null) phenotype. Unc prog- upstream of these genes. Epistasis of eny from gld-2(q497)/hT2; unc-32(e189) glp-1(q175 and q172) teg-1(oz230)/hT2 germline sex determination genes were examined by Nomarski optics and dissected gonads were analyzed using fluo- rescent microscopy. with other splicing factors, as well as mUnc progeny from gld-3(q730)/mIn1; unc-32(e189) glp-1(q175) teg-1(oz230)/hT2 somatic reporter assays with fem-3, were examined by Nomarski optics and dissected gonads were analyzed using fluo- suggest that the splicing factors could rescent microscopy. be functioning on the fem-3 gene, ulti- mately regulating its translation (Graham and Kimble, 1993; Graham et al., 1993; Ellis and Kimble, 1995; with a complete absence of GLP-/ 1/Notch signalling, if teg-1 functions Gallegos et al., 1998; Kerins et al., Notch signaling, strongly supports redundantly with another gene/ 2010). A complexity to this is that we the placement of teg-1 downstream of pathway in inhibiting proliferation, found that both fog-2(null) and tra- Notch signaling in the genetic path- for example, if removing teg-1 func- 2(gain-of-function) are, at least par- way regulating the proliferation tion does not completely remove the tially, epistatic to teg-1(oz230) (Table versus differentiation decision. The activity of the gld-1 pathway. That 2). However, the epistasis with fog-2 absence of tumor formation in gld-2(q497); glp-1(q175) teg-1(oz230) is incomplete (88% of teg-1(oz230); gld-2(q497); glp-1(q175) teg-1(oz230) animals are Glp, while gld-3(q730); fog-2(q71) animals are Fog [Feminiza- animals could still be consistent with glp-1(q175) teg-1(oz230) animals are tion of the germ line], with the teg-1 functioning downstream of GLP- tumorous, suggests that removal of remaining animals showing some 512 WANG ET AL.

teg-1 single mutants also display other phenotypes. At 20C, teg-1(0) ani- mals are sterile, producing sperm and small, abnormally shaped oocytes (Fig. 4), while at 25C, they produce func- tional sperm and oocytes, and therefore are fertile, provided that their mother had one wild-type copy of teg-1 (in other words, mþz animals are fertile). Homozygous progeny of homozygous mothers (mz), grown at 25C, are sterile with no differentiated germ cells being formed (Fig. 4). Therefore, in addition to the cold-sensitive Mog phe- notype, which is consistent with TEG-1 functioning as a splicing factor, TEG-1 also has functions in germline prolifer- ation and oogenesis.

TEG-1 Is Enriched in the Nucleus Splicing occurs in the nucleus of cells; therefore, most splicing factors are nu- clear enriched. To determine if TEG-1 is also nuclear enriched, supporting a role for TEG-1 as a splicing factor, as well as to gain additional insight into TEG-1 function, we raised antibodies against TEG-1 (see Experimental Pro- cedures section). These antibodies are specific to TEG-1 as they detect an 65-kD protein on Western blots that is absent in teg-1(oz230) homozygotes (Fig. 5). Staining revealed that TEG-1 is enriched in the nuclei of all germ- Fig. 4. teg-1(oz230) single mutant animals display temperature-specific germline phenotypes. Developmental Dynamics line and somatic cells of the gonad Germline phenotypes of teg-1(oz230) homozygous animals were analyzed using Nomarski optics in adult animals one day past the fourth larval stage. Animals were grown at (B)15C, (C)20C, (Fig. 5); however, staining above back- and (D, E)25C. A wild-type animal (A) grown at 20C is provided for comparison. For each ani- ground was also detected in the cyto- mal, only one gonad arm is shown, with the distal arm on top, and the distal end to the left. The plasm of the gonad and intestine, sug- gonad arm is outlined with a dashed line, from the distal end to the spermatheca. Strains were gesting that TEG-1 is also present in grown as teg-1(oz230)/hT2 heterozygotes at the specified temperature for multiple generations the cytoplasm (Fig. 5). Within the prior to analyzing teg-1(oz230) homozygous animals for their germ line phenotype, except for teg-1(oz230 m-z-) animals (E), in which the mother of the analyzed animal was also homozygous nuclei, TEG-1 is found throughout the for teg-1(oz230). B: Gonads in animals grown at 15C are somewhat smaller and exclusively nucleoplasm, but appears to be absent produce sperm, even in older animals. C: At 20C, animals produce sperm and smaller mis- in the regions containing DNA. TEG-1 shapen oocytes; however, they are sterile. D: At 25 C, animals produce sperm and oocytes, as levels are not uniform throughout the well as viable progeny; however, the progeny (E) are sterile, with no obvious oocyte production, although a limited number of sperm are observed. For each temperature >100 animals were an- cells of the gonad. TEG-1 levels alyzed, with all animals grown at the same temperature exhibiting similar phenotypes. Actual ge- appear highest in nuclei of the notype unc-32(e189) teg-1(oz230). Scale bar ¼ 20 mm. oocytes, the somatic Distal Tip Cell (DTC), and sheath cells. TEG-1 sperm production); therefore, teg-1 Therefore, even at 15C, knocking enrichment in the nucleus is consist- could still function with the other out teg-1 function likely does not ent with a role in splicing. splicing factors in regulating fem-3, completely block the sex determin- and the epistasis with fog-2(null) and ing signaling pathway at the posi- TEG-1 Physically Interacts tra-2(gf) could reflect a limited role tion where teg-1 functions. An With Known Splicing Factor for teg-1 in regulating fem-3. It must increase in a feminization signal UAF-1 be noted that teg-1’s role in sex deter- upstream of teg-1,suchaswithfog- mination is not essential; homozy- 2(null) and tra-2(gf),couldstill The genetic and expression analyses gotes are not Mog at 25C, even result in sufficiently decreased ac- described above suggest that TEG-1 though the teg-1(oz189 & oz230) al- tivity of the fem genes so as to is functioning, at least in part, as a leles are likely null (see below). cause feminization. splicing factor. Additionally, the TEG-1 CD2BP2 INHIBITS STEM CELL PROLIFERATION 513

These reciprocal co-IP experiments TABLE 2. Interaction of teg-1 With Genes in the Germ Line provide strong evidence that at least Sex Determination Pathway a proportion of TEG-1 and UAF-1

exist in a complex within the cell; Genotype Germline phenotype (15 C) however, this does not demonstrate teg-1(oz189 and oz230) Moga that these two proteins directly inter- b c fog-1(q241) Fog act. To test if the TEG-1–UAF-1 inter- fog-2(q71)d Foge f action is direct, we expressed both fog-3(q443) Fog proteins in bacterial cells, where fem-3(e1996)g Fog mRNA splicing does not occur and tra-2(e2020gf)h Fog fog-1(q241); teg-1(oz230)i Fog splicing factors are not present. teg-1(oz230); fog-2(q71)j Fog Therefore, if TEG-1 and UAF-1 are fog-3(q443); teg-1(oz230)k Fog part of the same multi-protein com- teg-1(oz189); fem-3(e1996)l Fog plex, the other proteins of the complex tra-2(e2020gf); teg-1(oz230)m Fog likely do not exist within the bacterial cell, and only direct interactions aMasculinization of the germ line. would be detected. We found that bac- b(Barton and Kimble, 1990). terially expressed TEG-1 and UAF-1 cFeminization of the germ line. co-IP (Fig. 6B), suggesting that the d (Schedl and Kimble, 1988). two proteins directly bind. Further- e XX only; XO animals are wild type. more, when we expressed a truncated f(Ellis and Kimble, 1995). version of TEG-1, which lacks a g(Hodgkin, 1986). h(Doniach, 1986). potential (although quite degenerate) iAll Vab Unc progeny of fog-1(q241) unc-13(e51)/ hT2; vab-7(e1562) teg-1(oz230)/ GYF protein binding motif (called hT2 were Fog. TEG-1DGYF; Fig. 3), an interaction jfog-2(q71) rol-9(sc148)/ unc-51(e369) fog-2(q71) males mated with vab-7(e1562) teg- between TEG-1DGYF and UAF-1 was 1(oz230)/hT2 hermaphrodites. L4 hermaphrodites cloned and put at 15C. Of those not detected (Fig. 6B). The lack of that segregated both Rol and Vab animals, 21/24 Rol Vab progeny (vab-7(e1562) interaction between TEG-1DGYF and teg-1(oz230); fog-2(q71) rol-9(q71) contained small oocytes and no sperm; 3/24 con- UAF-1 provides an important nega- tained sperm. 21/21 non-Rol non-Unc Vab animals (teg-1(oz230); fog-2(q71)/þ) were tive control, suggesting that the bac- Mog. terial TEG-1–UAF-1 co-IPs are spe- kAll Unc progeny of teg-1(oz230)/hT2; unc-13(e51) fog-3(oz230) were Fog. cific. Furthermore, the lack of lvab-7(e1562) teg-1(oz189); unc-24(e138) fem-3(e1996) dpy-20(e1282) females were obtained from an unbalanced strain. detectable binding between TEG- D mtra-2(e2020) males were mated with vab-7(e1562) teg-1(oz230)/hT2 purged her- 1 GYF and UAF-1 suggests that the maphrodites. Single male progeny were then mated with single female progeny. degenerate GYF domain is necessary From matings that failed to segregate hT2, L4 (XX) Vab animals were cloned and for the TEG-1–UAF-1 interaction.

Developmental Dynamics scored the next day. 19/25 animals had small oocytes and no sperm. The remain- While reducing uaf-1 activity through ing six animals were Mog (presumably tra-2(þ)/tra-2(þ)). RNAi results in an embryonic lethal phenotype (Zorio and Blumenthal, 1999b), partially reducing uaf-1 activ- human and yeast TEG-1 homologs wild-type protein (Fig. 5). After TEG- ity through RNAi in a glp-1(gf) back- physically interact with known splic- 1 and its associated proteins were ground results in a significant ing factors (see Discussion), although IP’d, we subjected the samples to increase in proliferation in the germ the overall similarity between TEG-1 mass-spectometry analysis to deter- line (Table 3), suggesting that, like and its homologs is not overly high, so mine the identities of the interacting teg-1, uaf-1 is also involved in regulat- it cannot be assumed that TEG-1 proteins, and compared these proteins ing the proliferation versus differen- would have the same binding part- to control IPs. UAF-1, a conserved tiation decision in the C. elegans germ ners in C. elegans (Fig. 3). To increase splicing factor that is homologous to line. Additionally, since a complete our understanding of how TEG-1 may mammalian U2AF65, was identified reduction of teg-1 activity in teg-1 null function in the splicing reaction, we in two independent TAP::TEG-1 IPs, mutants still allows for animals to sought to identify proteins with which but was not detected in either of the reach adulthood, while severely it binds. To accomplish this, we two control IPs. A full description of reducing uaf-1 activity results in em- expressed a TAP-tagged version of proteins found to bind TEG-1 will be bryonic lethality, suggests that the TEG-1 in animals and immunopreci- described elsewhere. In order to verify TEG-1 function may be modulatory. pitated (IP’d) TAP::TEG-1 and associ- the TEG-1–UAF-1 interaction, we ated proteins. Importantly, this performed reciprocal IPs from whole The TEG-1–UAF-1 Interaction tagged version of TEG-1 rescues teg- worm lysates. UAF-1 was co-IP’d with Is Conserved in Humans 1(oz230) mutant phenotypes, and is TEG-1 (using TEG-1-specific antibod- expressed at a similar level as endoge- ies), and TEG-1 was co-IP’d with An interaction between homologs of nous TEG-1, suggesting that TAP::- UAF-1 (using UAF-1-specific antibod- TEG-1 and UAF-1 has not been previ- TEG-1 is functioning similarly to the ies) (Fig. 6A). ously reported in any system. Given 514 WANG ET AL.

Fig. 5. TEG-1 is nuclear enriched in all gonadal cells. A: A dissected gonad arm from a wild-type (N2) adult hermaphrodite, one day past the fourth larval stage grown at 20C, stained with DAPI (blue) to visualize nuclear morphology, and anti-TEG-1 specific antibodies (red). The distal end of the gonad is to the left. Highest levels of TEG-1 enrichment are observed in nuclei of the distal tip cell (asterisk), sheath cells (arrowhead), and oocytes (large nuclei in the proximal half of the gonad arm). The inset shows a blow-up of a portion of the pachytene region of the gonad arm, with the DAPI and anti-TEG-1 channels merged. The regions of the nuclei occupied by DNA and TEG-1 appear to be mutually exclusive. In- tensity comparisons of oocyte and intestine cytoplasm in wild-type and teg-1(oz230) animals demonstrates that some TEG-1 is present in the cytoplasm (oocyte intensity measurements: 1,695 6 428 [n¼10] in N2 vs. 4566258 [n¼10] in teg-1(oz230); Intestine intensity measurements: 1,504 6 218 [n¼15] in N2 vs. 596 6 233 [n¼17] in teg-1(oz230)). Scale bar ¼ 20 mm. B: Western blot showing relative size of the TEG-1 protein, and the specificity of the anti-TEG-1 antibodies. A TAP-tag (HA-His8x-TEV-Myc) was fused to the N-terminus of teg-1 genomic DNA to generate an integrated transgenic line. Proteins at 65 kDa (endogenous TEG-1) and at 72 kDa (TAP-TEG-1) are detected in the extract of the transgenic strain. The comparable expression levels between these two proteins indicates that the TAP-TEG-1 is expressed at a level similar to endogenous TEG-1. Approximately 80 adult hermaphorodites from each C. elegans strain were loaded onto each lane of the SDS-PAGE gel. The wild-type and teg-1(oz230) strains are also marked with unc-32(e189). Paramyosin, detected with MH16 antibodies, was used as a loading control.

the overall low level of similarity CD2BP2, the human homolog of TEG- the GFP-negative control (Fig. 6C). We between TEG-1 and its homologs (Fig. 1, was transformed into Hela cells, also detected the interaction by per- 3), it is possible that the TEG-1–UAF- followed by IP with anti-GFP antibod- forming the reciprocal IP; immunopre- 1 interaction is unique to nematodes, ies (Fig. 6C). As a control, a construct cipitate obtained using U2AF65 anti- and not present in other systems. To encoding only GFP was transformed bodies contained GFP::CD2BP2, as

Developmental Dynamics test if the TEG-1–UAF-1 interaction into cells and IP’d. Using antibodies detected using anti-GFP antibodies. is present in humans, we tested if the specific to U2AF65, the human homo- Additionally, CD2BP2 and U2AF65 same interaction could be detected in log of UAF-1, we detected U2AF65 co- expressed in bacterial cells were co- HeLa cells. A GFP-tagged version of IP with GFP::CD2BP2, but not with IP’d, suggesting that the interaction between the two proteins is direct (Fig. 6D). Together, these results dem- TABLE 3. Interaction of uaf-1 With glp-1(ar202gf) onstrate that the TEG-1–UAF-1 inter- action, which we first identified in Germline phenotype C. elegans, is also present in humans. at 18C (%) b c a Genotype Wild type Mog Pro n DISCUSSION rrf-1d; glp-1(ar202gf); pL4440(RNAi)e 96 0 4 54 rrf-1; uaf-1(RNAi)f 49 51 0 49 We isolated teg-1 alleles in two differ- rrf-1; glp-1(ar202gf); uaf-1(RNAi) 32 15 53 34 ent mutant screens, each designed to identify genes that function in the aNumber of gonad arms. proliferation versus differentiation bMasculinization of the germ line. decision in the C. elegans germ line. cProximal proliferation. Gonads contain over-proliferative tumor in the proximal We found that, in addition to its role in end of the gonad. the proliferation versus differentiation d rrf-1(pk1417). decision, teg-1 also functions in regulat- epL4440 is an empty vector, used for control RNAi. f ing germline sex determination, germ- RNAi was performed by placing gravid hermaphrodites on RNAi plates and scoring line proliferation, and oogenesis. Clon- the progeny, ensuring that progeny were not exposed to RNAi until after hatching. ing of teg-1 revealed that it encodes a If uaf-1 RNAi was performed by placing L4 animals on the RNAi plate, this resulted in most of the progeny arresting as embryos, and a very small brood size. homolog of human CD2BP2, and that it is enriched in the nucleoplasm of TEG-1 CD2BP2 INHIBITS STEM CELL PROLIFERATION 515

splicing factors that resulted in these phenotypes are found throughout the various steps of the splicing reaction. Therefore, it is not just one step of the splicing reaction that can be dis- rupted to cause these phenotypes, nor is it likely that a subset of splicing factors have been co-opted into a func- tion separate from the other splicing factors. Rather, it appears as though a general disruption in splicing, or the splicing machinery, results in the Mog and over-proliferation pheno- types. Attempts at identifying poten- tial target mRNAs whose mis-splicing results in a Mog or over-proliferation phenotype have thus far been unsuc- cessful (Puoti and Kimble, 1999; Bel- fiore et al., 2004), although defects in splicing due to a reduction in one of these factors has recently been described (Zanetti et al., 2011). As we have previously suggested, a failure Fig. 6. TEG-1 directly interacts with UAF-1 in C. elegans extracts, and this interaction is con- served in humans. A: Reciprocal co-immunoprecipitations (IPs) were performed using UAF-1 spe- in detecting causative splicing defects cific antibodies (left) and TEG-1 specific antibodies (right), followed by SDS electrophoresis and could be due to the reduction in prop- Western blotting. In the UAF-1 IP, TEG-1-specific antibodies detect a band of the correct size, erly spliced product (or accumulation whereas an IP using non-specific pre-immunized rabbit serum did not result in detection of TEG- of mis-spliced product) being too small 1. In the TEG-1 IP, a band of the correct size was detected with UAF-1-specific antibodies, but not when pre-immune serum was used to perform the IP. B: E. coli expressed GST-TEG-1 and to detect using conventional techni- 6X-His-UAF-1 co-IP. 6X-His-UAF-1 was expressed in E. coli cells with GST, GST-TEG-1, or GST- ques, or that the proper targets may TEG-1 with the GYF domain removed (GST-TEG-1(DGYF); Fig. 3B). When UAF-1-specific antibod- not have yet been tested (Mantina ies were used to IP, only GST-TEG-1(full length) was co-IP’d (top). Likewise, when GST-specific et al., 2009). An alternate explanation antibodies were used to IP, only the strain containing GST-TEG-1(full length) was able to co-IP could be that the reduction in the UAF-1, as detected by UAF-1-specific antibodies. C: Co-IP of U2AF65 and CD2BP2 from HeLa extracts. GFP or GFP-CD2BP2 were transiently expressed in HeLa cells and proteins were IP’d splicing machinery that causes the using anti-GFP-specific antibodies (left) or anti-U2AF65-specific antibodies (right). U2AF65 and over-proliferation phenotype may not U2AF35 were detected in the anti-GFP IP on a Western blot, while the negative control, Hsp90 be due to a defect in splicing, but heat shock protein, was not co-IP’d. Likewise, GFP-CD2BP2 was co-IP’d with U2AF65, while rather that a reduction in splicing fac- GFP and Hsp90 were not (right). U2AF35 was co-IP’d with U2AF65, whether using lysate from Developmental Dynamics GFP-expressing cells or GFP-CD2BP2-expressing cells. D: E. coli expressed GST-CD2BP2 and tor activity could interfere with some U2AF65 co-IP. U2AF65 was expressed in E. coli cells with GST, GST-CD2BP2(full length) or GST- other aspect of RNA metabolism, CD2BP2(DGYF). Only GST-CD2BP2(full length) was co-IP’d with U2AF65. Likewise, when IP was which results in over-proliferation. performed using anti-GST antibodies, U2AF65 was only detected when lysate from cells express- ing GST-CD2BP2(full length) was used; U2AF65 could not be detected when lysate from cells expressing GST or GST-CD2BP2(DGYF) was used for the IP. Function of TEG-1/CD2BP2 in the Splicing Reaction cells. We also demonstrated that TEG- including masculinization of the germ The splicing of introns from pre- 1 physically interacts with UAF-1, the line, and over-proliferation in sensi- mRNAs is a multistep process that C. elegans homolog of human U2AF65. tized genetic backgrounds (Puoti and involves many protein and RNA com- Finally, we demonstrated that this is a Kimble, 2000; Belfiore et al., 2004; plexes. The U2 Auxiliary Factor conserved interaction, with the human Konishi et al., 2008; Mantina et al., (U2AF) is involved in the formation of homologsofTEG-1andUAF-1inter- 2009; Kerins et al., 2010; Zanetti the splicing commitment complex, acting in human cells, revealing a et al., 2011). Indeed, a comprehensive including being necessary for proper potential new role for TEG-1 CD2BP2 examination of splicing factors and U2 snRNP binding to the pre-mRNA in the splicing reaction. the ability of a reduction in their (Ruskin et al., 1988; Zamore and activity to cause over-proliferation in Green, 1989). U2AF binds to the 30 Splicing and the Balance different genetic backgrounds has splice site of the intron, which Between Proliferation and recently been performed (Kerins involves U2AF65, the large subunit of et al., 2010). Of 114 splicing factors U2AF, binding to a polypyrimidine Differentiation tested by RNAi, 11 disrupted germ- tract, and the small U2AF subunit, A number of genes encoding splicing line sex determination resulting in a U2AF35, binding to the AG/ splice factors have been identified that, Mog phenotype, while 31 enhanced site (Merendino et al., 1999; Wu et al., when mutated, result in similar the over-proliferation phenotype of a 1999; Zorio and Blumenthal, 1999a; germline phenotypes in C. elegans, glp-1(gf) mutant. Importantly, the Wahl et al., 2009; Valadkhan and 516 WANG ET AL.

Jaladat, 2010). In vertebrates, both 52K) was detected in experiments in the 30 splice site, with UAF-1 binding the length of the polypyrimidine tract, which Complexes A and B were iso- to the analogous UUUUCAG/R and the distance between the tract lated and their protein constituents sequence adjacent to the 30 splice site and the splice site, can vary between determined through proteomic analy- in C. elegans. In mammalian cells, introns (Wahl et al., 2009; Valadkhan sis (Hartmuth et al., 2002; Deckert three proteins of the U5 snRNP and Jaladat, 2010). However, in C. ele- et al., 2006). If CD2BP2(U5-52K) is replace U2AF in this region during gans most introns have a highly con- only present in the U5 snRNP prior to spliceosome assembly (Chiara et al., served consensus, UUUUCAG/R, with U4/U6.U5 tri-snRNP formation, and if 1997). Even though these three U5 no spacing between the tract and the the tri-snRNP is pre-assembled prior proteins (110, 116, and 220 kDa) are splice site (Zhang and Blumenthal, to becoming part of the spliceosome, not the same U5 snRNP proteins with 1996; Blumenthal and Steward, 1997; then CD2BP2(U5-52K) should never which CD2BP2(U5-52K) interacts (15 Kent and Zahler, 2000). Both UAF-1, be part of the spliceosome as it is and 102 kDa), it is possible that func- the C. elegans homolog of U2AF65 being assembled on the pre-mRNA, tions that CD2BP2(U5-52K) provides (Zorio et al., 1997), and UAF-2, the C. including as part of Complex A or while binding to the U5 snRNP and to elegans homolog of U2AF35 (Zorio and Complex B. Furthermore, the U4/ UAF-1 U2AF65 are related. For exam- Blumenthal, 1999b), bind to this con- U6.U5 tri-snRNP has not yet become ple, it is possible that TEG-1 sensus sequence (Zorio and Blumen- part of the splicesome during Complex CD2BP2(U5-52K) is involved in assist- thal, 1999a), and are likely responsible A formation. Therefore, even if ing both UAF-1 U2AF65 and the U5 for the high degree of sequence conser- CD2BP2(U5-52K) does have some af- snRNP to bind to the 30 end of the vation in the consensus C. elegans 30 finity for U5 after tri-snRNP forma- intron. It is also possible that TEG-1 splice site (Hollins et al., 2005). tion, it should not be isolated with CD2BP2(U5-52K) is involved in facili- Homologs of TEG-1 have been impli- Complex A. Our identification of TEG- tating the replacement of UAF-1 cated in splicing. Human CD2BP2 co- 1CD2BP2(U5-52K)asbindingto U2AF65 by the U5 snRNP. Differenti- purifies with the U5 snRNP, binding UAF-1 U2AF65 is consistent with its ating between these and other possi- the 15- and 102-kDa proteins; there- presence in Complexes A and B, and bilities will require studying the tim- fore, CD2BP2 has also been called suggests an additional function for ing of the TEG-1/CD2BP2 interactions U5-52K (Bach et al., 1989; Behrens TEG-1 CD2BP2(U5-52K). with UAF-1/U2AF65 and the U5 and Luhrmann, 1991; Laggerbauer We do not know when or why TEG- snRNP during spliceosome assembly. et al., 2005). During the stepwise 1 CD2BP2(U5-52K) binds to UAF-1 assembly of the spliceosome, the U2AF65 during spliceosome assem- U5 snRNP associates with the U4/U6 bly. However, U2AF65 has been EXPERIMENTAL di-snRNP to form the U4/U6.U5 tri- shown to bind other proteins during PROCEDURES snRNP (Behrens and Luhrmann, spliceosome assembly. For example, C. elegans Growth Conditions 1991). Importantly, even though in both yeast and humans U2AF65 and Strains CD2BP2(U5-52K) binds to the U5 binds to the branch point binding pro- snRNP, it is not found in the U4/ tein SF1, presumably to aid in com- Maintenance and manipulation of C.

Developmental Dynamics U6.U5 tri-snRNP, suggesting that mitment complex formation (Abovich elegans was performed as previously CD2BP2(U5-52K) may be involved in and Rosbash, 1997). Additionally, described (Brenner, 1974). Bristol N2 the assembly of the tri-snRNP, but dis- SF3b155, which is part of the U2 was the wild-type strain used and the sociates from U5 during tri-snRNP snRNP, binds to U2AF65 as the U2 following mutants were used in this formation. (Laggerbauer et al., 2005). snRNP replaces SF1 at the branch study. LGI: fog-1(q241), gld-2(q497), The ortholog of TEG-1 in Saccharomy- point (Gozani et al., 1998). Two-hybrid rrf-1(pk1417), gld-1(q485), fog-3 ces cerevisiae, Lin1p, also binds to the analysis also identified multiple other (q443); LGII: tra-2(e2020gf), gld-3 U5 snRNP, suggesting that this role in U2AF65 splicing factor binding part- (q730), nos-3(oz231); LGIII: unc-32 tri-snRNP assembly is conserved ners (Prigge et al., 2009); however, the (e189) glp-1(ar202gf), glp-1(oz112oz (Bialkowska and Kurlandzka, 2002). functions of these interactions are 120), spe-16(hc54), teg-1(oz189), teg-1 We propose that the interaction we unknown. Therefore, U2AF65 binds (oz230), dpy-18(e364) LGIV: fem-3 have detected between TEG-1 different proteins as formation of the (e1996) LGV fog-2(q71). CD2BP2(U5-52K) and UAF-1 U2AF65 spliceosome progresses. TEG-1 reveals a function for TEG-1 CD2BP2(U5-52K) binding to UAF-1 Mapping teg-1 CD2BP2(U5-52K) that is distinct from U2AF65 may be another such interac- its role in tri-snRNP formation. tion functioning in spliceosome forma- Using standard three-factor mapping, Indeed, proteomic analyses of Com- tion. Although we do not yet know we mapped both the over-prolifera- plex A and Complex B support a func- what the consequences of this interac- tion and Mog phenotypes of teg- tion for TEG-1 CD2BP2 that is inde- tion may be, it is intriguing that both 1(oz189) to a 0.5 cM region between, pendent of tri-snRNP formation. The UAF-1 U2AF65 and the U5 snRNP, or very close to, spe-16 and dpy-18 on U4/U6.U5 tri-snRNP binds to the spli- with which TEG-1 CD2BP2(U5-52K) linkage group III. We further nar- ceosome as a preassembled complex was previously shown to bind (Lagger- rowed the critical region through a (Bindereif and Green, 1987; Cheng bauer et al., 2005), associate with the combination of single nucleotide poly- and Abelson, 1987; Konarska and 30 splice site. As mentioned, U2AF65 morphism (SNP) mapping and defi- Sharp, 1987). However, CD2BP2(U5- binds to the polypyrimidine tract near ciency mapping. The right boundary TEG-1 CD2BP2 INHIBITS STEM CELL PROLIFERATION 517

of the critical region was determined described (Jones et al., 1996). Briefly, XbaI and HinDIII sites and used to through SNP mapping using teg- dissected gonads were fixed with 3% transform the E. coli HT115(DE3) 1(oz189) dpy-18(e364)/HA-8. The paraformaldehyde for 10 min, fol- strain, which were able to produce cold-sensitive Mog phenotype was lowed by fixation/permeabilization dsRNA (Timmons and Fire, 1998). mapped by picking Dpy non-Teg with 20C, 100% methanol for > 30 (Mog) animals. A recombinant was min. The gonads were blocked in 3% Transgenic Array and obtained that retained N2 DNA from BSA. Affinity purified rat anti-TEG-1 Integration dpy-18 to a SNP in T28D6.5, making (N-terminal) antibodies were used at this SNP the right boundary of the 1:500 dilution, anti-REC-8 antibodies The TAP-tag teg-1 construct critical region (SNP between were used at 1:200 dilution, and anti- (pDH122) was injected into N2 her- 11337877 and 11339172-wormbase HIM-3 antibodies (Zetka et al., 1999) maphrodite gonads, together with freeze WS220; information provided were used at 1:500 dilution. Nuclear pJM67 (an elt-2::GFP construct by Kate Hill). The left boundary of the DNA was visualized by DAPI. Images kindly provided by J. M. McGhee, critical region was narrowed by map- were taken using a Zeiss (Zeiss, University of Calgary) and the pRF4 ping the endpoint of the deficiency Thornwood, NY) Imager Z1 micro- plasmid as visible transformation tDf2, which complements teg-1(oz189). scope equipped with an Axiocam markers (Mello et al., 1991; Fukushige We found that tDf2 deletes part of MRM digital camera. et al., 1998). After following the select- Y47D3A.25, but not the gene immedi- able co-injection markers for two gen- ately to the right, Y47D3A.27. There- Plasmid Construction erations, the extrachromosomal arrays fore, the left boundary of the critical were randomly integrated into the ge- region is in Y47D3A.25. These left and TAP-tag teg-1 (pDH122) nome by gamma-irradiation (137Cs) right boundaries narrowed the critical construction. with a total dose of 3,000 rads. Two region to a portion of chromosome III The teg-1 promoter region, a 522-bp independent integrants, designated containing only three genes: fragment containing sequences down- ugIs3 [tap::teg-1] (strain XB307) and Y47D3A.27, Y47D3A.28, and T28D6.6. stream of rab-35 stop codon and ugIs4 [tap-teg-1] (strain XB308), were Sequencing revealed that Y47D3A.27 upstream of teg-1 start codon, was isolated and out-crossed three times. contained nonsense mutations corre- PCR amplified (teg-1 Pro F primer: The integrated array corresponding to sponding to both the teg-1(oz189) and tctgaaaattgaacaatttgtg; teg-1 Pro R ugIs3 was crossed into teg-1(oz230) teg-1(oz230) alleles. Sequencing of the primer: tttttggctgaaaatgagtgaa) and and was able to rescue the sterile phe- yk82g9 cDNA, as well as RT-PCR sub-cloned into the pBluescript II KS notype of teg-1(oz230) allele. amplified cDNA (using primers R10- (þ) vector (Stratagene, La Jolla, CA) CCGAAACGCGCCGTTTCTTT and at NotI and BamHI sites. The tandem Preparation of C. elegans R11-TTACAAATAAAGCTCGAAATCA) affinity purification (TAP) tag encodes Extracts was used to determine the inton/exon for HA-8xHis-TEV-Myc epitopes and boundaries. was amplified by PCR (TAP F primer: To generate sufficient starting mate- atgtacccatacgatgtcccaga; TAP R pri- rial for purification processes, proteo- Developmental Dynamics Generation of Anti-TEG-1 mer:caaatcctcctcgctgatc) using the mic analyses, and immunoprecipita- Antibodies pSB_GW::TAG vector as template tion studies, the wild-type or (Polanowska et al., 2004; Walhout XB307(ugIs3) transgenic line from 20 Two synthetic peptides, one corre- and Boulton 2006). The amplified cleared, 35-mm plates was used to in- sponding to the TEG-1 amino-terminal TAP-tag was then sub-cloned into the oculate 250 ml S medium supple- sequence (CVKERPGNADEDEEEKK vector containing the promoter at the mented with a E. coli OP50 pellet con- KFHT), and one to the carboxyl-termi- BamHI and EcoRI sites, where the centrated from 2 L of LB broth nal sequence (CAGAPMEQEEEEAD EcoRI site was filled-in with Klenow (Sulston and Hodgkin, 1988). The DSVKWEY), were conjugated by Key- worm culture was grown for 3–5 days enzyme. The teg-1 coding region was hole-limpet hemocyanin (KLH) and amplified, using teg-1 Gen F primer at 20 C on a shaking platform (250 injected into rats and rabbits for anti- (ccgaaacgcgccgtttctt) and teg-1 Gen R rpm) until a freshly cleared culture body production by Covance (Covance primer (tgatgttgggttgcagaagcct), and was generated. The live worms were Inc.,Princeton,NJ).Antibodiesfrom sub-cloned into the EcoRV site of the harvested by sucrose gradient fol- each species of the antisera were affin- pCRII-TOPO vector (Invitrogen, lowed by three washes with 0.1 M ity purified against the corresponding Carlsbad, CA). The sequence contain- NaCl (Polanowska et al., 2004). The peptides coupled to the SulfoLink Cou- ing both the teg-1 promoter and the pellet was flash frozen in liquid nitro- pling Gel (Pierce, Rockford, IL) accord- TAP-tag was sub-cloned into the KpnI gen and stored at 80 C. The flash- ing to the manufacturer’s instruction. and SpeI sites of this pCRII-TOPO frozen worm pellet was resuspended vector, resulting in the completed res- in 5 volumes of CSK lysis buffer (100 Gonad Dissections and cuing construct (pDH122). mM PIPES, pH 6.0, 100 mM NaCl, 3 Indirect Immunofluorescence mM MgCl2, 1 mM EGTA, 1 mM dithi- othreitol [DTT], 0.3 M sucrose, 0.5% Staining pL4440-uaf-1 Construct (pDH182). Triton X-100, 1 mM PMSF, and Gonad dissections and antibody stain- The uaf-1 cDNA sequence (368–951) EDTA-free complete protease inhibi- ing were performed as previously was sub-cloned into pL4440 vector at tor tablet [Roche, Montreal, QC]) 518 WANG ET AL.

(Polanowska et al., 2004). The worm TEG-1 antibody data. Common pro- Mini-Trans Blot Cell (Bio-Rad). The suspension was disrupted by French teins found between two independent membrane was blocked with 5% Car- Press at 20,000 psi and the soluble affinity purified experiments were an- nation milk and incubated overnight lysate was obtained by centrifugation alyzed for their mRNA expression pat- at 4C with either: (1) affinity purified at 10,000g for 20 min. The protein terns using the NEXTDB in situ data- TEG-1 (N-terminal) antibodies at concentration was determined by Bio- base to determine if they were germ 1:1,000; (2) anti-UAF-1 antibodies Rad Protein Assay (Bio-Rad, Hercu- line expressed (Kohara, 2001). (Zorio and Blumenthal, 1999a) at les, CA) according to the manufac- 1:20,000; (3) anti-U2AF65 antibodies turer’s instruction. Co-Immunoprecipitation (Sigma) at 1:2,000; (4) anti-U2AF35 antibodies (Novus Biologicals; Little- (Co-IP) Affinity Purification of ton, CO) at 1:2,000; (5) anti-GFP For UAF-1 co-IP, extracts from 1.5 ml (Rockland Immunochemicals, Gil- TAP-TEG-1 of packed N2 pellet was obtained as bertsville, PA) at 1:2,000; (6) anti- A two-step affinity purification was described above. For each immuno- Hsp90a (Stressgen, San Diego, CA) at used to prepare samples for mass precipitation, 10 mg of the extract 1:1,000; or (7) anti-paramyosin spectrometry identification. To pre- was incubated overnight at 4C with (MH16, hybridoma cell line from De- pare pre-immune and anti-TEG-1 of 20 mg of pre-immune sera, anti- velopmental Studies Hybridoma antibodies-coupled protein A beads, UAF-1 antibodies, or affinity purified Bank, University of Iowa, Iowa City, 0.5 ml of agarose protein A beads rat anti-TEG-1 (C-terminal) antibod- IA) tissue culture supernatant at 1:1, (Sigma, St. Louis, MO) were incu- ies coupled to protein G beads in TBS (137 mM NaCl, 25 mM Tris- bated with 0.5 ml of each pre-immune (Sigma). The beads were washed HCl, pH 7.2, 2.5 mM KCl)-5% milk so- sera or affinity purified rabbit anti- three times with CSK lysis buffer and lution. The membrane was incubated TEG-1 (C-terminal) antibodies over- resuspended in 2 SDS-PAGE load- 2 hr at room temperature with HRP- night at 4C. The pre-immune and ing buffer. Bound proteins were ana- conjugated secondary antibodies antibody-coupled beads were cross- lyzed by immunoblots. (Jackson Laboratories, Bar Harbor, linked as described in Polanowska ME) diluted at 1:5,000 and then et al. (2004). Approximately 6–8 ml of visualized by SuperSignal WestPico Bacterial Pull-Down packed XB307(ugIs3) worm pellet substrate (Thermo Scientific, Wal- was lysed in CSK lysis buffer as Experiments tham, MA) or Amersham ECL described above. The extract was af- TEG-1 (2-353) and UAF-1 (1-496) Advance kit (GE Healthcare, Wauke- finity purified through a 2-ml Ni-NTA were sub-cloned into pGEX-4T1 (GE sha, WI). Images were collected on agarose column (Qiagen, German- Healthcare, Waukesha, WI) and pET- Kodak (Rochester, NY) BioMax MR town, MD). After washing the column 28a (Novagen, Darmstadt, Germany), films. with 5 volumes of PBS containing 10 respectively. A truncated teg-1 lacking mM imidazole, the coupled proteins the GYF domain (amino acid residues HeLa Tissue Culture and were eluted with 10 ml of 100 mM im- 296–353) was also sub-cloned into Transfection Developmental Dynamics idazole in PBS (137 mM NaCl, 10 mM pGEX-4T1. Both GST- and His6x-tag Na2PO4-12H2O, 2.5 mM KCl, 1.8 mM constructs were co-expressed in the E. HeLa-CCL2 cell line (a gift from M. J. KH2PO4, pH 7.2). In the second coli BL21(DE3) strain (Tolia and Lohka, University of Calgary) was immunoaffinity step, the imidazole Joshua-Tor, 2006). For pull-down grown in Dulbecco’s modified Eagle’s eluates were equally divided and assays, His6x-UAF-1 fusion proteins medium supplemented with 3.7 g/ml batch absorbed with pre-immune or that bind to GST-TEG-1 recombinant NaHCO3, pH 7.2, 10% fetal bovin se- anti-TEG-1 antibody-coupled protein proteins were co-immunoprecipitated rum, 4 mM L-glutamine, 75 mg/ml Aat4C for 2–3 hr. Beads were then by immobilizing GST, GST-TEG-1, Penicillin and Streptomycin at 37 C transferred to a Poly-Prep column and GST-TEG-1DGYF on glutathione with 5% CO2. The plasmids express- (Bio-Rad) and washed three times agarose beads (GE Healthcare, Wau- ing GFP-CD2BP2 (Kofler et al., 2004) with 6 volumes of PBS. The bound kesha, WI). On the other hand, GST were prepared by Qiagen EndoFree proteins were eluted three times with recombinant proteins that interact maxi plasmid kit (Qiagen) and trans- m 100 l/fraction of 0.1 M glycine, pH with the His6x-UAF-1 fusion proteins fected into HeLa-CCL2 using Lipo- 2.5, and then neutralized with 1 M were purified on Ni-NTA beads (Qia- fectamine 2000 (Invitrogen). At 24 hr Tris, pH 8.0. The protein samples gen). The beads were washed five after transfection, cells were washed were concentrated by speed-vac, times with PBS containing 0.3 M three times with ice-cold PBS and lysed resolved on 12.5% Laemmli SDS- NaCl. Proteins were eluted by adding in HEPES lysis buffer (40 mM HEPES, PAGE gels, and then stained with Bio- 2 SDS-PAGE loading buffer into the pH 7.5, 120 mM NaCl, 1 mM EDTA, Safe Coomassie (Bio-Rad). Bands were beads and analyzed by immunoblots. 1% Triton X-100, 10 mM pyrophos- excised and samples were analyzed by phate, 50 mM NaF, 1.5 mM Na3VO4, the SAMS Centre for Proteomics at and protease inhibitor cocktail [Roche]) Immunoblotting University of Calgary by LC-MS/MS on ice for 30 min with intermittent vor- for protein identification. Peptide hits Proteins were resolved on a 10% SDS- texing. Whole cell lysate was centri- that were found in the pre-immune PAGE gel and transferred to Trans- fuged at 10,000g for 10 min at 4Cand control were subtracted from the anti- Blot nitrocellulose membrane using the protein concentration of the soluble TEG-1 CD2BP2 INHIBITS STEM CELL PROLIFERATION 519

fraction was determined. For each of spermatogenesis in the germ line of gans initiates by a conserved mecha- immunoprecipitation, 2 mg of the Caenorhabditis elegans. Genetics 125: nism and is dispensable for homologous extract was pre-cleared with the pro- 29–39. chromosome synapsis. Cell 94:387–398. Behrens SE, Luhrmann R. 1991. Immu- Doniach T. 1986. Activity of the sex- tein A beads and then incubated over- noaffinity purification of a [U4/U6.U5] determining gene tra-2 is modulated to night at 4 Cwith20mg of either anti- tri-snRNP from human cells. Genes allow spermatogenesis in the C. elegans U2AF65 monoclonal antibodies (Sigma) Dev 5:1439–1452. hermaphrodite. Genetics 114:53–76. or anti-GFP antibodies (Rockland Belfiore M, Pugnale P, Saudan Z, Puoti A. Doyle TG, Wen C, Greenwald I. 2000. 2004. Roles of the C. elegans cyclophi- SEL-8, a nuclear protein required for Immunochemicals) that were coupled lin-like protein MOG-6 in MEP-1 bind- LIN-12 and GLP-1 signaling in Caeno- to protein A beads. The beads were ing and germline fates. Development rhabditis elegans. Proc Natl Acad Sci subsequently washed five times with 131:2935–2945. USA 97:7877–7881. the lysis buffer and dissolved in 2 Berglund JA, Chua K, Abovich N, Reed Eckmann CR, Kraemer B, Wickens M, SDS-PAGE loading buffer for immuno- R, Rosbash M. 1997. The splicing factor Kimble J. 2002. GLD-3, a bicaudal-C BBP interacts specifically with the pre- homolog that inhibits FBF to control blotting analysis. mRNA branchpoint sequence germline sex determination in C. ele- UACUAAC. Cell 89:781–787. gans. Dev Cell 3:697–710. ACKNOWLEDGMENTS Berry LW. 1998. Regulation of the Mi- Eckmann C, Crittenden SL, Suh N, Kimble totic/Meiotic cell fate decision in Caeno- J. 2004. GLD-3 and control of the mitosis/ We thank Kate Hill for providing valu- rhabditis elegans. In: Department of meiosis decision in the germline of Caeno- able information on single nucleotide Genetics. St. Louis: Washington Univer- rhabditis elegans. Genetics 168:147. polymorphisms in the teg-1 region. We sity School of Medicine. p 327. Ellis R, Schedl T. 2006. Sex determination thank Pavel Pasierbek and Joseph Berry LW, Westlund B, Schedl T. 1997. in the germ line WormBook, ed. The Germ-line tumor formation caused by C. elegans research community, Worm- Loidl for anti-REC-8 antibodies, Moni- activation of glp-1, a Caenorhabditis Book, doi/10.1895/wormbook.1.82.1, http:// que Zetka for anti-HIM-3 antibodies, elegans member of the Notch family of www.wormbook.org. and Tom Blumenthal for UAF-1 anti- receptors. Development 124:925. Ellis RE, Kimble J. 1995. The fog-3 gene bodies and advice. We thank Ted Han- Bialkowska A, Kurlandzka A. 2002. Pro- and regulation of cell fate in the germ sen for generating the two TEG-1- teins interacting with Lin 1p, a putative line of Caenorhabditis elegans. Genetics link between chromosome segregation, 139:561–577. specific peptides, Christian Freund for mRNA splicing and DNA replication in Fox PM, Vought VE, Hanazawa M, Lee providing the GFP::CD2BP2 con- Saccharomyces cerevisiae. Yeast 19: MH, Maine E, Schedl T. 2011. Cyclin E struct, Adrian Krainer for the pET9c- 1323–1333. and CDK-2 regulate proliferative cell U2AF65 construct, and Carrie She- Bindereif A, Green MR. 1987. An ordered fate and cell cycle progression in the C. manko for anti-Hsp90a antibodies and pathway of snRNP binding during elegans germline. Development 138: mammalian pre-mRNA splicing com- 2223–2234. use of her tissue culture facility. We plex assembly. EMBO J 6:2415–2424. Francis R, Barton MK, Kimble J, Schedl thank members of the Hansen lab, and Blumenthal T, Steward K. 1997. RNA T. 1995a. gld-1, a tumor suppressor anonymous reviewers, for critical processing and gene structure. In: Rid- gene required for oocyte development in reading of the manuscript. Some nem- dle DL, Blumenthal T, Meyer B, Priess Caenorhabditis elegans. Genetics 139: JR, editors. C. elegans II. Cold Spring 579–606. atode strains used in this work were Harbor, NY: Cold Spring Harbor Labo- Francis R, Maine E, Schedl T. 1995b. provided by the Caenorhabditis ratory Press. p 117–145. Analysis of the multiple roles of gld-1 Genetics Center, which is funded by Brenner S. 1974. The genetics of Caeno- in germline development: interactions Developmental Dynamics the NIH National Center for Research rhabditis elegans. Genetics 77:71–94. with the sex determination cascade and Resources (NCRR). This work was Cheng SC, Abelson J. 1987. Spliceosome the glp-1 signaling pathway. Genetics assembly in yeast. Genes Dev 1: 139:607–630. funded by a grant from the National 1014–1027. Freund C, Dotsch V, Nishizawa K, Rein- Institutes of Health (NIH Grant Chiara MD, Palandjian L, Feld Kramer herz EL, Wagner G. 1999. The GYF do- GM63310) to T.S., and grants from the R, Reed R. 1997. Evidence that U5 main is a novel structural fold that is 0 Canadian Institutes of Health snRNP recognizes the 3 splice site for involved in lymphoid signaling through Research (CIHR) and the Natural Sci- catalytic step II in mammals. EMBO J proline-rich sequences. Nat Struct Biol 16:4746–4759. 6:656–660. ences and Engineering Research Christensen S, Kodoyianni V, Bosenberg Fukushige T, Hawkins MG, McGhee JD. Council of Canada (NSERC) to D.H. M, Friedman L, Kimble J. 1996. lag-1, a 1998. The GATA-factor elt-2 is essential gene required for lin-12 and glp-1 signal- for formation of the Caenorhabditis ele- ing in Caenorhabditis elegans,ishomolo- gans intestine. Dev Biol 198:286–302. REFERENCES gous to human CBF1 and Drosophila Gallegos M, Ahringer J, Crittenden S, Su(H). Development 122:1373–1383. Kimble J. 1998. Repression by the 30 Abovich N, Rosbash M. 1997. Cross-intron Crittenden SL, Troemel ER, Evans TC, UTR of fem-3, a sex-determining gene, bridging interactions in the yeast com- Kimble J. 1994. GLP-1 is localized to relies on a ubiquitous mog-dependent mitment complex are conserved in the mitotic region of the C. elegans control in Caenorhabditis elegans. mammals. Cell 89:403–412. germ line. Development 120:2901. Embo J 17:6337–6347. Austin J, Kimble J. 1987. glp-1 is Deckert J, Hartmuth K, Boehringer D, Gozani O, Potashkin J, Reed R. 1998. A required in the germ line for regulation Behzadnia N, Will CL, Kastner B, potential role for U2AF-SAP 155 inter- of the decision between mitosis and Stark H, Urlaub H, Luhrmann R. 2006. actions in recruiting U2 snRNP to the meiosis in C. elegans. Cell 51:589–599. Protein composition and electron mi- branch site. Mol Cell Biol 18:4752–4760. Bach M, Winkelmann G, Luhrmann R. croscopy structure of affinity-purified Graham PL, Kimble J. 1993. The mog-1 1989. 20S small nuclear ribonucleopro- human spliceosomal B complexes iso- gene is required for the switch from tein U5 shows a surprisingly complex lated under physiological conditions. spermatogenesis to oogenesis in Caeno- protein composition. Proc Natl Acad Sci Mol Cell Biol 26:5528–5543. rhabditis elegans. Genetics 133:919–931. USA 86:6038–6042. Dernburg AF, McDonald K, Moulder G, Graham PL, Schedl T, Kimble J. 1993. Barton MK, Kimble J. 1990. fog-1, a reg- Barstead R, Dresser M, Villeneuve AM. More mog genes that influence the ulatory gene required for specification 1998. Meiotic recombination in C. ele- switch from spermatogenesis to 520 WANG ET AL.

oogenesis in the hermaphrodite germ rhabditis elegans. Dev Dyn 239: Mello CC, Kramer JM, Stinchcomb D, line of Caenorhabditis elegans. Dev 1555–1572. Ambros V. 1991. Efficient gene transfer Genet 14:471–484. Killian DJ, Hubbard EJ. 2005. Caeno- in C. elegans: extrachromosomal main- Hansen D, Hubbard EJ, Schedl T. 2004a. rhabditis elegans germline patterning tenance and integration of transforming Multi-pathway control of the prolifera- requires coordinated development of the sequences. Embo J 10:3959–3970. tion versus meiotic development deci- somatic gonadal sheath and the germ Merendino L, Guth S, Bilbao D, Martinez sion in the Caenorhabditis elegans line. Dev Biol 279:322–335. C, Valcarcel J. 1999. Inhibition of msl-2 germline. Dev Biol 268:342–357. Kimble J, Crittenden SL. 2007. Controls splicing by Sex-lethal reveals interac- Hansen D, Schedl T. 2006. The regulatory of germline stem cells, entry into meio- tion between U2AF35 and the 30 splice network controlling the proliferation- sis, and the sperm/oocyte decision in site AG. Nature 402:838–841. meiotic entry decision in the Caeno- Caenorhabditis elegans. Annu Rev Cell Michaud S, Reed R. 1991. An ATP-inde- rhabditis elegans germ line. Curr Top Dev Biol 23:405–433. pendent complex commits pre-mRNA to Dev Biol 76:185–215. Kofler MM, Freund C. 2006. The GYF do- the mammalian spliceosome assembly Hansen D, Wilson-Berry L, Dang T, main. FEBS J 273:245–256. pathway. Genes Dev 5:2534–2546. Schedl T. 2004b. Control of the prolifer- Kofler M, Heuer K, Zech T, Freund C. Mumm JS, Kopan R. 2000. Notch signal- ation versus meiotic development deci- 2004. Recognition sequences for the ing: from the outside in. Dev Biol 228: sion in the C. elegans germline through GYF domain reveal a possible spliceoso- 151–165. regulation of GLD-1 protein accumula- mal function of CD2BP2. J Biol Chem Nadarajan S, Govindan JA, McGovern M, tion. Development 131:93–104. 279:28292–28297. Hubbard EJ, Greenstein D. 2009. Hartmuth K, Urlaub H, Vornlocher HP, Kohara Y. 2001. [Systematic analysis of MSP and GLP-1/Notch signaling coor- Will CL, Gentzel M, Wilm M, Luhrmann gene expression of the C. elegans ge- dinately regulate actomyosin-dependent R. 2002. Protein composition of human nome]. Tanpakushitsu Kakusan Koso cytoplasmic streaming and oocyte prespliceosomes isolated by a tobramy- 46:2425–2431. growth in C. elegans. Development 136: cin affinity-selection method. Proc Natl Konarska MM, Sharp PA. 1987. Interac- 2223–2234. Acad Sci USA 99:16719–16724. tions between small nuclear ribonucleo- Nishizawa K, Freund C, Li J, Wagner G, Henderson ST, Gao D, Lambie EJ, Kimble protein particles in formation of Reinherz EL. 1998. Identification of a J. 1994. lag-2 may encode a signaling spliceosomes. Cell 49:763–774. proline-binding motif regulating CD2- ligand for the GLP-1 and LIN-12 recep- Konishi T, Uodome N, Sugimoto A. 2008. triggered T lymphocyte activation. Proc tors of C. elegans. Development 120: The Caenorhabditis elegans DDX-23, a Natl Acad Sci USA 95:14897–14902. 2913–2924. homolog of yeast splicing factor PRP28, Pasierbek P, Jantsch M, Melcher M, Hirsh D, Oppenheim D, Klass M. 1976. is required for the sperm-oocyte switch Schleiffer A, Schweizer D, Loidl J. Development of the reproductive system and differentiation of various cell types. 2001. A Caenorhabditis elegans cohe- of Caenorhabditis elegans. Dev Biol 49: Dev Dyn 237:2367–2377. sion protein with functions in meiotic 200–219. Kraemer B, Crittenden S, Gallegos M, chromosome pairing and disjunction. Hodgkin J. 1986. Sex determination in Moulder G, Barstead R, Kimble J, Genes Dev 15:1349–1360. the nematode C. elegans: analysis of Wickens M. 1999. NANOS-3 and FBF Pepper AS, Killian DJ, Hubbard EJ. tra-3 suppressors and characterization proteins physically interact to control 2003. Genetic analysis of Caenorhabdi- of fem genes. Genetics 114:15–52. the sperm-oocyte switch in Caenorhab- tis elegans glp-1 mutants suggests re- Hollins C, Zorio DA, MacMorris M, Blu- ditis elegans. Curr Biol 9:1009–1018. ceptor interaction or competition. menthal T. 2005. U2AF binding selects Laggerbauer B, Liu S, Makarov E, Vorn- Genetics 163:115–132. for the high conservation of the C. ele- locher HP, Makarova O, Ingelfinger D, Petcherski AG, Kimble J. 2000. LAG-3 is gans 30 splice site. RNA 11:248–253. Achsel T, Luhrmann R. 2005. The a putative transcriptional activator in Jeong J, Verheyden JM, Kimble J. 2011. human U5 snRNP 52K protein the C. elegans Notch pathway. Nature Cyclin E and Cdk2 control GLD-1, the (CD2BP2) interacts with U5–102K 405:364–368. Developmental Dynamics mitosis/meiosis decision, and germline (hPrp6), a U4/U6.U5 tri-snRNP bridg- Polanowska J, Martin JS, Fisher R, Scopa stem cells in Caenorhabditis elegans. ing protein, but dissociates upon tri- T, Rae I, Boulton SJ. 2004. Tandem PLoS genetics 7:e1001348. snRNP formation. RNA 11:598–608. immunoaffinity purification of protein Jones AR, Schedl T. 1995. Mutations in Lamont LB, Crittenden SL, Bernstein D, complexes from Caenorhabditis elegans. gld-1, a female germ cell-specific tumor Wickens M, Kimble J. 2004. FBF-1 and Biotechniques 36:778–780, 782. suppressor gene in Caenorhabditis ele- FBF-2 regulate the size of the mitotic Prigge JR, Iverson SV, Siders AM, gans, affect a conserved domain also region in the C. elegans germline. Dev Schmidt EE. 2009. Interactome for aux- found in Src-associated protein Sam68. Cell 7:697–707. iliary splicing factor U2AF(65) suggests Genes Dev 9:1491–1504. MacDonald LD, Knox A, Hansen D. 2008. diverse roles. Biochim Biophys Acta Jones AR, Francis R, Schedl T. 1996. Proteasomal regulation of the pro- 1789:487–492. GLD-1, a cytoplasmic protein essential liferation vs. meiotic entry decision in Puoti A, Kimble J. 1999. The Caenorhab- for oocyte differentiation, shows stage- the C. elegans germ line. Genetics 180: ditis elegans sex determination gene and sex-specific expression during Cae- 905–920. mog-1 encodes a member of the DEAH- norhabditis elegans germline develop- MacQueen AJ, Villeneuve AM. 2001. Nu- Box protein family. Mol Cell Biol 19: ment. Dev Biol 180:165. clear reorganization and homologous 2189–2197. Kadyk LC, Kimble J. 1998. Genetic regu- chromosome pairing during meiotic pro- Puoti A, Kimble J. 2000. The hermaphro- lation of entry into meiosis in Caeno- phase require C. elegans chk-2. Genes dite sperm/oocyte switch requires the rhabditis elegans. Development 125: Dev 15:1674–1687. Caenorhabditis elegans homologs of 1803–1813. Mantina P, Macdonald L, Kulaga A, Zhao PRP2 and PRP22. Proc Natl Acad Sci Kent WJ, Zahler AM. 2000. Conservation, L, Hansen D. 2009. A mutation in teg-4, USA 97:3276–3281. regulation, synteny, and introns in a which encodes a protein homologous to Racher H, Hansen D. 2010. Translational large-scale C. briggsae–C. elegans the SAP130 pre-mRNA splicing factor, control in the C. elegans hermaphrodite genomic alignment. Genome Res 10: disrupts the balance between prolifera- germ line. Genome 53:83–102. 1115–1125. tion and differentiation in the C. elegans Ruskin B, Zamore PD, Green MR. 1988. Kerins JA, Hanazawa M, Dorsett M, germ line. Mech Dev 126:417–429. A factor, U2AF, is required for U2 Schedl T. 2010. PRP-17 and the pre- Mazroui R, Puoti A, Kramer A. 1999. snRNP binding and splicing complex mRNA splicing pathway are preferen- Splicing factor SF1 from Drosophila assembly. Cell 52:207–219. tially required for the proliferation ver- and Caenorhabditis: presence of an N- Schedl T, Kimble J. 1988. fog-2, a sus meiotic development decision and terminal RS domain and requirement germ-line-specific sex determination germline sex determination in Caeno- for viability. RNA 5:1615–1631. gene required for hermaphrodite TEG-1 CD2BP2 INHIBITS STEM CELL PROLIFERATION 521

spermatogenesis in Caenorhabditis ele- Wahl MC, Will CL, Luhrmann R. 2009. The meiosis, and splice site selection. Dev gans. Genetics 119:43–61. spliceosome: design principles of a Biol 354:232–241. Seraphin B, Rosbash M. 1989. Identifica- dynamic RNP machine. Cell 136:701– Zetka MC, Kawasaki I, Strome S, Muller F. tion of functional U1 snRNA-pre-mRNA 718. 1999. Synapsis and chiasma formation in complexes committed to spliceosome as- Walhout AJM, Boulton SJ. 2006. Bio- Caenorhabditis elegans require HIM-3, a sembly and splicing. Cell 59:349–358. chemistry and molecular biology. In: meiotic chromosome core component that Sulston J, Hodgkin J. 1988. Methods. In: Community TCeR, editor. WormBook. functions in chromosome segregation. Wood WB, editor. The nematode Caeno- [http://www.wormbook.org]. Genes Dev 13:2258–2270. rhabditis elegans. Cold Spring Harbor, Wu S, Romfo CM, Nilsen TW, Green MR. Zhang H, Blumenthal T. 1996. Functional NY: Cold Spring Harbor Laboratory. 1999. Functional recognition of the 30 analysis of an intron 30 splice site in p 587–606. splice site AG by the splicing factor Caenorhabditis elegans. RNA 2: Tax FE, Yeargers JJ, Thomas JH. 1994. U2AF35. Nature 402:832–835. 380–388. Sequence of C. elegans lag-2 reveals a Yochem J, Greenwald I. 1989. glp-1 and Zorio DA, Blumenthal T. 1999a. Both sub- cell-signalling domain shared with lin-12, genes implicated in distinct cell- units of U2AF recognize the 30 splice Delta and Serrate of Drosophila. Na- cell interactions in C. elegans, encode site in Caenorhabditis elegans. Nature ture 368:150–154. similar transmembrane proteins. Cell 402:835–838. Timmons L, Fire A. 1998. Specific interfer- 58:553–563. Zorio DA, Blumenthal T. 1999b. U2AF35 ence by ingested dsRNA. Nature 395:854. Zamore PD, Green MR. 1989. Identifica- is encoded by an essential gene clus- Tolia NH, Joshua-Tor L. 2006. Strategies tion, purification, and biochemical char- tered in an operon with RRM/cyclophi- for protein coexpression in Escherichia acterization of U2 small nuclear lin in Caenorhabditis elegans. RNA 5: coli. Nature methods 3:55–64. ribonucleoprotein auxiliary factor. Proc 487–494. Valadkhan S, Jaladat Y. 2010. The spli- Natl Acad Sci USA 86:9243–9247. Zorio DA, Lea K, Blumenthal T. 1997. ceosomal proteome: at the heart of Zanetti S, Meola M, Bochud A, Puoti A. Cloning of Caenorhabditis U2AF65: an the largest cellular ribonucleoprotein 2011. Role of the C. elegans U2 snRNP alternatively spliced RNA containing a machine. Proteomics 10:4128–4141. protein MOG-2 in sex determination, novel exon. Mol Cell Biol 17:946–953. Developmental Dynamics