BIR-1, a Caenorhabditis elegans homologue of Survivin, regulates transcription and development

Marta Kostrouchova*†, Zdenek Kostrouch†‡, Vladimir Saudek§, Joram Piatigorsky¶, and Joseph Edward Rall†ʈ

*Laboratory of Molecular Biology and Genetics and ‡Laboratory of Molecular Pathology, Institute of Inherited Metabolic Disorders, First Faculty of Medicine, Charles University, CZ-128 01 Prague 2, Czech Republic; †Diabetes Branch, National Institute of Diabetes and Digestive and Kidney Diseases, and ¶Laboratory of Molecular and Developmental Biology, National Eye Institute, National Institutes of Health, Bethesda, MD 20892

Contributed by Joseph Edward Rall, February 7, 2003 bir-1,aCaenorhabditis elegans inhibitor-of-apoptosis gene homol- phorylation of histone H3 on serine 10 (15), a phosphorylation ogous to Survivin is organized in an operon with the transcription that is involved in promoter activation (6, 9, 10, 13). Together, cofactor C. elegans SKIP (skp-1). Because genes arranged in oper- these data suggested that BIR-1 may have a transcriptional ons are frequently linked functionally, we have asked whether function unrelated to cell division. BIR-1 also functions in transcription. bir-1 inhibition resulted in In this study, we report evidence that BIR-1 is a transcriptional multiple developmental defects that overlapped with C. elegans regulator for numerous target genes. We show that the loss of SKIP loss-of-function phenotypes: retention of eggs, dumpy, move- function of bir-1 results in phenotypes that partially overlap with ment defects, and lethality. bir-1 RNA-mediated interference de- CeSKIP and CHR3 (nhr-23) loss of function. We also show that creased expression of several gfp transgenes and the endogenous several transgenes (elt-2::gfp, hlh-1::gfp, and hlh-2::gfp) and two genes dpy-7 and hlh-1. Immunoblot analysis revealed decreased endogenous genes (dpy-7 and hlh-1) are inhibited as a conse- phosphoacetylated histones in bir-1 RNA-mediated interference- quence of bir-1 loss of function. Western blot analysis using treated worms. In a heterologous transfection system, BIR-1 aug- antibodies against phosphoacetylated histone H3 (S10-P K14-Ac ments hormone-regulated transcription and has an addi- and K9-Ac S10-P) shows that bir-1 inhibition in worms negatively tive effect with SKIP. These results show that BIR-1 functions in the affects phosphoacetylation of histone H3. Finally, we demon- regulation of transcription and development. strate that BIR-1 acts as coactivator in a heterologous transfec- tion system, and bir-1-transfected cells have increased phos- odulation of specific gene transcription depends on or- phoacetylated histones H3. Thus our results show that BIR-1, a Mchestrated events involving the modification of chromatin member of the inhibitor-of-apoptosis gene family, functions in and cooperation between RNA polymerase II and supporting transcription regulation and is involved in regulation of devel- factors. A large number of events participate in this process opment. including the assembly of RNA polymerase II active complex on particular promoters, cooperation and modification of TFII Methods transcription factors, and interactions with numerous cofactors Strains. The following strains expressing GFP fusion proteins (for review see ref. 1). Modification of chromatin proteins such were used: JM 63 (elt-2::gfp) integrated line (20, 21); PD7963 as acetylation, methylation, and phosphorylation further influ- (hlh-1::gfp) integrated line; PD8097 (hlh-2 ::gfp) (22, 23); ences transcription activation or repression and represents an- egl-15::gfp (24), Nde::gfp (25), and hlh-8::gfp (26); his2B::gfp (27), other level of regulation (2–6). Regulation of gene transcription a gift from G. Seydoux (Johns Hopkins University, Baltimore); includes complex chromatin reorganization. During this process, and PD 6904 (let-858::gfp), a gift from W. Kelly (Emory Uni- histones in promoter regions as well as proteins involved in versity, Atlanta). As a wild type, C. elegans Bristol strain was used transcription complex are modified covalently. Modification of and maintained as described (28). histones H3 and H4 includes methylation, acetylation, and phosphorylation of N-terminal residues. A growing number of bir-1 RNA-Mediated Interference (RNAi). The RNAi was introduced cofactors and specific enzymes with methyltransferase, acetyl- by feeding bacteria producing double-stranded (ds)RNA to transferase, and aminotransferase activity were shown recently worms (feeding method) and microinjection of dsRNA into the also to precede acetylation and contribute to the activation of gonad of young adult hermaphrodites (microinjection method) transcription (3, 7, 8). Phosphorylation of histones is part of (29, 30). mitotic chromosome condensation but was shown recently also The full-length bir-1 genomic sequence was amplified by PCR to precede acetylation and contribute to the activation of and cloned into the L4440 vector. The plasmid (4851) was transcription (9–13). transformed into HT115 Escherichia coli and induced by isopro- The Caenorhabditis elegans baculoviral inhibitor-of-apoptosis ␤ repeat protein 1 (bir-1) gene in C. elegans encodes a homolog of pyl- -D-galactoside as described (30). The same construct was the human gene Survivin (14, 15). Survivin in mammals func- used for preparation of dsRNA and used for microinjections as tions as an inhibitor of apoptosis (16). In C. elegans, bir-1 has only described (29, 31). The injected larvae were transferred into new been linked thus far to spindle midzone formation and cytoki- plates at 12-h intervals. The embryos were scored after 12 h of nesis (14, 15). Our previous work noted that bir-1 is expressed incubation, and the larvae were followed for the next 3–7 days. from an operon with the transcription cofactor SKI-binding protein (SKIP; skp-1) (17). In C. elegans, genes expressed from Abbreviations: BIR-1, Caenorhabditis elegans baculoviral inhibitor-of-apoptosis repeat an operon are often functionally linked (18, 19), suggesting that protein 1; SKIP, SKI-binding protein; CeSKIP, C. elegans SKIP; RNAi, RNA-mediated inter- bir-1 may also have a transcriptional function related to C. ference; ds, double-stranded; mRXR␣, murine retinoid X receptor ␣; THR, thyroid hormone elegans SKIP (CeSKIP). CeSKIP is indispensable for C. elegans receptor; DTC, distal tip cell; T3, ; TSA, Trichostatine A. development, and its loss-of-function phenotype partially over- §Present address: Lead Generation Informatics, Avensis Pharma, Vitry-sur-Seine F-94400, France. laps with that of nuclear hormone receptor CHR3 (nhr-23) (17). ʈTo whom correspondence should be addressed at: Laboratory of Molecular Biology and CeSKIP and bir-1 are expressed strongly during embryonic Genetics, Institute of Inherited Metabolic Disorders, First Faculty of Medicine, Charles Univer- development and continue to be expressed in certain postmitotic sity, Ke Karlovu 2, CZ-128 01 Prague 2, Czech Republic. E-mail: marta.kostrouchova@ cells to adulthood (17). Furthermore, bir-1 can mediate phos- lf1.cuni.cz.

5240–5245 ͉ PNAS ͉ April 29, 2003 ͉ vol. 100 ͉ no. 9 www.pnas.org͞cgi͞doi͞10.1073͞pnas.0730770100 Downloaded by guest on October 1, 2021 The reporter genes elt-2::gfp, hlh-1::gfp, hlh-2::gfp, hlh-8::gfp, with ratios: 400 ng of reporter plasmid, 150 ng of human THR␤1, his2B::gfp, and let-858::gfp were used to assay the expression of mRXR␣, flag-SKIP, and bir-1, and 15 ng of control cytomega- transgenes. Animals were fed bacteria producing bir-1 dsRNA or lovirus-Renilla luciferase vector. Empty vectors and plasmid transformed with empty vector. For transcription induction, the pBluescript SK(Ϫ) were used to substitute for receptor, bir-1,or plates contained 4 mM isopropyl-␤-D-galactoside, and both flag-SKIP. Plasmid pBluescript was also used to maintain the cultures were induced with 4 mM isopropyl-␤-D-galactoside for total amount of DNA constant (1 ␮g or 1,115 ng in some 4 h. Approximately 50 animals from each transgenic line were experiments) and 3 ␮l of FuGENE. Where possible, all plasmids screened for expression of particular transgene with an Olympus were mixed first as a master mix and split after mixing to (Tokyo) BX60 microscope. minimize inequalities. All experiments were done at least twice and in triplicate or quadruplicate. After preparation of trans- Gene-Expression Assays. The animals were fed with bacteria fection suspensions and appropriate incubations, the medium producing bir-1 dsRNA or empty vector. Approximately 10,000 with hormone-depleted serum and with gentamycin was added, worms were used for each experiment. The worms were grown mixed, and 1 ml of transfection medium was overlaid over cells. from synchronized L1 stage for one or two generations on 2% After a 24-h incubation, the medium was removed and ex- agarose plates and collected as young adults just when some changed for fresh medium supplemented with 10Ϫ6 M triiodo- started to lay embryos. RT-PCR was performed from total RNA thyronine (T3, Sigma) dissolved in ethanol as 10Ϫ3 M stock as described (32, 33). The dpy-7, elt-2, and hlh-1 genes were solution or with ethanol only. Trichostatine A (TSA) was assayed. These experiments were done independently three dissolved also in ethanol and used at final concentration 10 nM. times in duplicate, and each was assayed by PCR several times. Cells were harvested 24 h later and used for the dual luciferase For Western blot analysis, synchronized L1 stage animals were method (Promega) as recommended by the manufacturer. The grown for two generations on nematode growth medium plates human telomerase RNA ␤1, mRXR␣, F2 thyroid-responsive and collected as young adults. Protein extracts were prepared by construct, and DR4 construct were kind gifts from Paul Yen, and using standard protocol with Laemmli buffer. Protein content ME-TRE was a kind gift from Keiko Ozato (National Institute was estimated by using the BCA protein assay kit (Pierce), and of Child Health and Human Development, National Institutes of 20 ␮g of protein were loaded per lane. For immunodetection of Health, Bethesda). bir-1 was amplified from total RNA prepared phosphoacetylated histones, antibodies specific for S10-P from mixed stages of C. elegans and cloned into pCDNA3 K14-Ac (07-081) (Upstate Biotechnology, Charlottesville, VA) expression vector. and K9-Ac S10-P (9711S) (Cell Signaling Technology, Beverly, For immunodetection of phosphoacetylated histones in bir-1- MA) were used at a 1:1,000 dilution. Anti-rabbit secondary transfected cells, MCF-7 cells were transfected with bir-1 antibodies labeled by peroxidase (Sigma) were used at a 1:5,000 pCDNA3 or empty pCDNA3 vector. Ratios were as for tran- dilution. The ECL plus chemiluminescent system (Amersham scription assay [150 ng of bir-1 pCDNA3 or empty vector, Pharmacia) was used for detection. Films were scanned and pBluescript SK(Ϫ] to fill the DNA content to 1 ␮g per 35-mm analyzed by using the program NIH IMAGE (www.nih.gov). petri dish). Cells were incubated for 24 h and processed for Western blot analysis. Immunohistochemistry. Worms were kept on plates seeded with bacteria producing bir-1 dsRNA or empty plasmid. Larvae and Results young adults were transferred onto glass slides coated with C. elegans bir-1 Loss of Function Leads to Developmental Defects polylysine, frozen on dry ice, and fixed shortly by methanol and Independent of Its Known Cell-Division Functions. Loss of bir-1 acetone. Antibodies against phosphoacetylated histones were activity in C. elegans by RNAi leads to early embryonic arrest due used at a 1:100 dilution and secondary antibody labeled by to cell-division defects (14, 15). To bypass this lethality and assay fluorescein at a 1:200 dilution. Worms were photographed by BIR-1 functions postembyronically, we treated larvae with bir-1 using an Olympus BX60 microscope. RNAi by using a bacterial feeding method. Several developmen- tal defects were observed in treated animals (Fig. 1). bir-1 RNAi BIOLOGY Bir-1 Overexpression and Ectopic Expression. bir-1 genomic se- resulted in an egg-laying (Egl) phenotype in 80% of young adult DEVELOPMENTAL quence was amplified by PCR and cloned into heat-shock animals (n ϭ 3,408) that became nearly fully penetrant in older promoter vector pPD 49.83 (a kind gift from Andrew Fire, adults. bir-1 RNAi-treated hermaphrodites laid 50% the normal Carnegie Institution of Washington, Baltimore). Transgenic number of embryos compared with control animals (n ϭ 20 for lines were prepared by microinjection, and overexpression was each group). The vulva and egg-laying muscles were formed induced by incubating animals at 34°Cfor2h. normally suggesting a neuronal defect as the cause of the Egl phenotype in these bir-1 RNAi adults (Fig. 2 B–D). Cell Cultures and Transient Transfections. bir-1 cDNA was cloned A short and fat dumpy (Dpy) phenotype was also observed in into pCDNA3 expression vector (Invitrogen). Flag-SKIP was a a fraction (n ϭ 992, 29%) of larvae at L3 and L4 stage and young gift from Diane Hayward (Johns Hopkins School of Medicine, adult animals when L1 larvae were treated with bir-1 RNAi (Fig. Baltimore) (34). Human telomerase RNA ␤1 and murine reti- 1A). If larvae L2 were treated with bir-1 RNAi, the effect was noid X receptor ␣ (mRXR␣) were gifts from Paul Yen (National less pronounced and visible later during adulthood (n ϭ 208, Institute of Diabetes and Digestive and Kidney Diseases, 26%). In animals showing the pronounced Dpy phenotype the National Institutes of Health). MCF-7 (human breast cancer movement was also affected. The worms showed slow uncoor- cell line) and cos-7 (green monkey kidney cells) cells were dinated (Unc) movement. Non-Dpy animals also had movement obtained from the American Tissue Culture Collection and defects after bir-1 RNAi that consisted of larger-than-normal maintained as recommended. Briefly, minimum essential me- amplitudes of body curvature. dium was supplemented with 10% FCS (Life Technologies, There was also a germ-line defect in 20% of bir-1 RNAi- Grand Island, NY) and gentamycin (100 ␮g͞ml) and incubated treated animals (n ϭ 680). Wild-type adult hermaphrodites have in 5% CO2 atmosphere at 37°C. For transfections, medium was two germ-line-filled gonad arms arranged symmetrically above stripped from thyroid hormone receptor (THR) by using a the vulva. During the L2–L4 larval stages, the gonads form by Bio-Rad AG 1-X8 (200–400 mesh) anion-exchange resin simi- elongation of the germ-line and somatic gonad primordium. At larly as described (35). Transfections were performed by using the leading edge of each elongating gonad arm is a distal tip cell FuGENE (Roche) as recommended: the proportions were cal- (DTC), the migration path of which dictates the shape of the culated as for three 24-well plate wells (one 35-mm petri dish) gonad. After the L3 molt, the DTCs U-turn and continue to grow

Kostrouchova et al. PNAS ͉ April 29, 2003 ͉ vol. 100 ͉ no. 9 ͉ 5241 Downloaded by guest on October 1, 2021 composing egg laying and anal muscles were normally developed and observed on transgenic lines expressing CeSKIP::gfp (Fig. 2). Interestingly, the DTCs did not have defects in cell division as might be expected from the known BIR-1 function in cytokine- sis, suggesting that only a DTC migration defect was responsible for the abnormal gonads. There were two additional defects observed in bir-1 RNAi- treated animals that were present less frequently. In 10% (n ϭ 344) of bir-1 RNAi-treated animals, there was an intestinal defect consisting of gut distention. A similar percentage of animals had a protruding vulva (Pvul) phenotype (Fig. 1A).

bir-1 Inhibition Negatively Affects Gene Expression. If bir-1 is re- quired for gene expression, we reasoned that it might be possible to see changes in transgene expression after bir-1 RNAi. To test this, we used strains expressing GFP fusion proteins from specific promoters and assayed expression after feeding animals bacteria producing bir-1 dsRNA. As reported previously, bir-1 RNAi of adults results in 100% embryonic lethality at the 30-cell (or less) stage. To assay transgene expression in these arrested embryos, we used hlh-2::gfp, which would be expected to be expressed in all cells at this stage. The effect of bir-1 RNAi on embryos was observed on progeny of hermaphrodites fed by bacteria producing dsRNA. In contrast to the ubiquitous ex- pression observed in untreated embryos, hlh-2::gfp was observed in only four to six cells of bir-1 RNAi-arrested embryos. It should be noted that the cells of arrested embryos were not normal, and the nucleus was irregular in morphology, suggesting that loss of Fig. 1. Developmental defects caused by bir-1 inhibition. (A) An affected hlh-2::gfp expression was secondary to early cytokinesis defects animal with dumpy (short and thick) phenotype after treatment with bir-1 reported previously (14, 15). RNAi. Protrusion of the egg-laying organ is indicated by an arrow. (B) Reten- Because of problems with interpretation of gene expression in tion of embryos in the uterus of an adult hermaphrodite treated by bir-1 RNAi. arrested embryos, we focused on postembryonic expression in Note the multinucleate embryos. (C) Defective growth of germ line in bir-1 bir-1 RNAi-treated animals. We used elt-2::gfp and hlh-1::gfp RNAi-treated animals: The germ line makes two turns (arrows), and its growth transgenes to assay larval expression. elt-2::gfp is an intestinal- is elongated. The position of the DTC is marked (arrowhead). specific reporter gene that we used as a marker for transcription in developed larvae. In controls the transgene is expressed in nuclei of all 20 intestinal cells. In bir-1 RNAi experiments, the in opposite directions, resulting in two symmetrically reflexed fluorescence was visible in four to six nuclei in the proximal part gonads. The gonads in bir-1 RNAi-treated animals often made of the intestine and in one or two nuclei in the distal part of the multiple turns (usually three) during elongation (Fig. 1C). The intestine. In the remaining cells the fluorescence was decreased elongation defect appeared to be stochastic, and the growth of strongly or disappeared, which is consistent with the inhibition only one gonad arm was affected in any given animal. Cells of transcription (Fig. 3 A and B). The hlh-1::gfp is expressed in all body wall muscles during larval development and in adults (Fig. 3C). The expression of this transgene decreased in bir-1 RNAi-treated animals (Fig. 3D), although not as dramatically as seen with elt-2::gfp.

bir-1 Inhibition Affects the Expression of Endogenous Genes. To determine whether the expression of endogenous genes could be affected by bir-1 RNAi we treated Ϸ10,000 larvae with bacteria expressing bir-1 RNAi and tested the expression of the endog- enous genes hlh-1 and dpy-7. The expression of dpy-7 decreased Ͼ50% compared with controls, and hlh-1 expression was also lowered in bir-1 RNA-treated animals (approximately one-fifth of control animals). As a complementary approach to bir-1 inhibition by RNAi, we Fig. 2. bir-1 RNAi does not influence the expression of CeSKIP::gfp.(A)An wanted to know whether overexpression of bir-1 would have any animal expressing the CeSKIP::gfp transgene after treatment with bir-1 RNAi. effect in C. elegans. We prepared transgenic lines carrying bir-1 The expression of CeSKIP is not decreased by bir-1 RNAi. The strong expression regulated by a heat-shock promoter and examined animals after is in pharynx (arrowhead), neurons, epidermal cells (arrows) in head region, heat shock for any visible phenotypes. There were no visible and in ventral neuronal cord. (B) Distal part of the same animal showing phenotypic changes in worms overexpressing bir-1. expression of CeSKIP::gfp in neurons, epidermal cells, and intestinal muscles. Note the properly developed intestinal muscles expressing CeSKIP::gfp (ar- BIR-1 Augments Transcription from Thyroid Hormone-Regulated Pro- row). (C) The presence of Egl phenotype and multinucleate embryos in the moters. same animal as shown in A documenting the developmental defects caused by Our results in C. elegans suggested that bir-1 activity was bir-1 RNAi (Nomarski optics). The arrow indicates the position of the egg- required for normal gene expression and development. To laying organ shown in D.(D) The strong expression of CeSKIP::gfp in ventral provide additional evidence that BIR-1 could indeed affect neuronal cord (arrows) and in egg-laying muscles (arrowhead), which are also transcription we used a cell culture system in which gene normally developed. expression could be assayed more easily. Because no immortal

5242 ͉ www.pnas.org͞cgi͞doi͞10.1073͞pnas.0730770100 Kostrouchova et al. Downloaded by guest on October 1, 2021 Fig. 3. The effect of bir-1 inhibition on gene expression. (A) A part of the body of L4 larva showing the expression of elt-2::gfp in intestinal cells. The bulbus of pharynx (arrowhead) and normal expression of intestinal cells are marked. (B) Decreased fluorescence of the elt-2::gfp transgene is shown in most intestinal cells of a similar larva L4 treated by bir-1 RNAi. The expression of elt-2::gfp is not changed in six nuclei in proximal part of the intestine and in one cell in distal part (arrowheads). (C) The normal expression of hlh-1::gfp transgene in body wall muscles (arrows) in head and proximal part of the body of a young adult hermaphrodite. (D) Expression of the hlh-1::gfp transgene in an animal treated with bir-1 RNAi is decreased markedly.

Fig. 4. bir-1 augments thyroid hormone-dependent transcription in a het- cell line exists for C. elegans, we opted for a mammalian erologous transfection system. (A) Cotransfection of bir-1 enhances basal as cell-culture assay with THRs regulating the expression of a well as thyroid hormone-induced transcription in MCF-7 cells. Experiments luciferase reporter gene. THRs are powerful transcription fac- cotransfected with empty vector are indicated by light columns, and experi- tors that both activate or repress the responsive promoters. Both ments cotransfected with bir-pCDNA3 are indicated by dark columns. Exper- activation (in the presence of the ligand T3) and repression are iments 1–8 are transfected with a reporter (ME-TRE-luc) and control reporter cytomegalovirus-Renilla luciferase: 1–4 are treated with vehicle, 5–8 are easily detectable in transfection experiments (36, 37). MCF-7 treated with thyroid hormone, 1 and 5 are cotransfected with empty vectors (human breast cancer cell line) and cos-7 (green monkey kidney pCDNA3, 2 and 6 are cotransfected with human telomerase RNA ␤1 (hTR␤1), cells) cells have very low levels of endogenous functional THRs 3 and 7 are cotransfected with mRAR␣, and 4 and 8 are cotransfected with in Western blot analysis (data not shown). For transfection human telomerase RNA ␤1 and mRXR␣. Results are in relative luciferase units experiments, we amplified bir-1 from C. elegans total RNA by related to control Renilla luciferase expression. BIR-1 (dark columns) aug- PCR and cloned it into a pCDNA3 expression vector. ments basal as well as thyroid hormone-induced transcription. (B) BIR-1 has a We assayed thyroid hormone responsiveness in this cell line in stimulatory effect on thyroid hormone-induced transcription in MCF-7 cells treated with the histone deacetylase inhibitor TSA. Cells are transfected as the absence or presence of cotransfected C. elegans bir-1.Inthe described for A. Note that thyroid hormone stimulation of transcription is MCF-7 cancer cell line, cotransfections of bir-1 increased both conserved on TSA background, but the values (related to the same amount of

͞ BIOLOGY basal as well as T3 THR-induced transcription from thyroid- transfected Renilla vector) are Ϸ20 times higher. (C) The effect of bir-1 and

responsive promoters (Fig. 4A). The basal transcription increase flag-human SKIP are additive in THR-dependent transcription. Experiments DEVELOPMENTAL was Ϸ20–50%. The positive effect of BIR-1 was similar in 1–8: MCF-7 cells are transfected with a reporter (ME-TRE-luc), control reporter T3͞THR-induced transcription. BIR-1 enhanced also TSA- cytomegalovirus-Renilla luciferase, human THR␤1, and mRXR␣. Cells are induced and TSA͞thyroid hormone-induced transcription (Fig. treated with vehicle in the experiments shown in lines 1–4 and with thyroid 4B). TSA is a strong histone deacetylase inhibitor that was shown hormone in 5–8; 1 and 5 are cotransfected with empty vector (pCDNA3), and 2 and 6 are cotransfected with Bir-1 pCDNA3; 3 and 7 are cotransfected with to increase transcription from viral as well as thyroid hormone- flag-SKIP, and 4 and 8 are cotransfected with Bir-1 pCDNA3 and flag-SKIP. (D regulated promoters (38). To test whether SKIP cooperates with and E) The positive effect of BIR-1 and SKIP and their additive effect on thyroid BIR-1 in transfection experiments, we used a human homologue hormone-induced transcription in cos-7 cells. Experiments are as described for cloned in expression vector with flag sequence (a kind gift from C. Cells are treated with vehicle (D)or10nMTSA(E). The additive effect of Diane Hayward) that was shown to function in transfection BIR-1 and flag-SKIP is apparent in basal and T3- and T3͞TSA-dependent experiments (34) and was more likely to interact with mamma- transcription. lian cofactors in transcription assay. SKIP had an additional positive effect on T3-induced transcription (Fig. 4C). The tran- scription stimulation effect of BIR-1 and SKIP and their additive peptide used for generation of purchased antibodies is identical effect was observed also in cos-7 cells (Fig. 4 D and E). with the C. elegans sequence of histone H3. Western blot analysis of C. elegans protein extracts revealed one major band with a BIR-1 Affects Phosphoacetylation of Histones H3. Because BIR-1 was mass of 15 kDa recognized by both antibodies at reducing shown to increase mitotic phosphorylation of histone H3 on conditions. Staining of fixed specimens showed strong nuclear serine 10 and phosphorylation of this residue was linked recently staining of dividing cells of embryos and larvae. bir-1 RNAi to promoter activation, we assayed whether bir-1 expression animals caused a decrease of phosphoacetylated histones rec- affects phosphoacetylation of histone H3 by using specific ognized by both antibodies. The decrease in immunohistochem- antibodies. To assay histone H3 phosphoacetylation we used ical staining of fixed animals was mimicked by Western blot commercially available antibodies (directed against S10-P analysis of protein extracts from worms treated by bir-1 RNAi K14-Ac H3 and K9-Ac S10-P H3). Histones H3 are conserved (Fig. 5C, lanes 2 and 4). between C. elegans and mammals. The sequence covering the To determine whether bir-1 overexpression affects the level of

Kostrouchova et al. PNAS ͉ April 29, 2003 ͉ vol. 100 ͉ no. 9 ͉ 5243 Downloaded by guest on October 1, 2021 tion, and CeSKIP RNAi does not lead to defects in cytokinesis that are observed on embryos developing in bir-1 RNAi-treated animals. Moreover, CeSKIP RNAi caused embryonic arrest at the time of gastrulation (at the 50- to 100-cell stage) and strong molting defects in larvae but no defects in cell division. In contrast, bir-1 RNAi caused early embryonic arrest at the two-cell stage, and the arrested embryos were multinucleated. In RNAi experi- ments using bacteria expressing ds bir-1 RNAi, the embryonic arrest occurred later at the 16- or 30-cell stage. Keeping with this, bir-1 RNAi did not alter the pattern of expression of CeSKIP::gfp from the transgene containing complete bir-1 genomic sequence (Fig. 2 A, B, and D). This indicates that phenotypic changes caused by bir-1 and CeSKIP RNAi are specific for both of these genes. The developmental defects resulting from bir-1 RNAi cannot be attributed to defects in cell division, which does not occur in the affected cells during later larval stages. The Egl phenotype was observed in animals that had properly developed egg-laying organs, egg-laying and uterine muscles, and neurons. The defect is most likely of neuronal origin as that for the defective movement. Both defects are characteristic for both bir-1 and CeSKIP loss of function. The developmental changes are highly penetrant but the growth of germ line, an organ with intensive cell divisions, M lineage development, oocytes, sperm development, and sper- matheca are not reduced. One possibility is that chromatin Fig. 5. BIR-1 affects phosphoacetylation of histones H3. (A) Immunodetec- proteins are different, as is the case for sperm chromatin (39, 40). tion of S10-P K14-Ac histone H3 in L4 larva. Most nuclei are labeled. (B) Another possibility is that the transcription function of BIR-1 is Decrease of labeling for S10-P K14-Ac histone H3 in bir-1 RNAi-treated larva more sensitive to bir-1 RNAi at the level used in our experiments. L4. (C) Western blot analysis of phosphoacetylated histones H3. Larvae L1–L4 Nevertheless, the germ lines (the proximal and distal parts of were treated by bir-1 RNAi (lanes 2 and 4) by feeding with bacteria producing gonads) are developing normally. The proper development of dsRNA or carrying empty vector (lanes 1 and 3). Protein extracts were sepa- rated by PAGE, blotted on membrane, and immunoanalyzed by using anti- DTCs was also observed in Nomarski optics. This is in contrast body against phosphoacetylated histones H3 [lanes 1 and 2 for S10-P K14-Ac to embryos developing in bir-1 RNAi-affected animals defective H3 (07-081) and lanes 3 and 4 for K9-Ac S10-P H3 (9711S)]. There is a decrease in cell division as described elsewhere (14, 15) and observed in in phosphoacetylated histone in bir-1 RNAi-treated worms. Lanes 5–8, West- our experiments. ern blot analysis of protein extracts from transfected MCF-7 cells. Cells were transfected with bir-1 pCDNA3 (lanes 6 and 8) or empty vector (lanes 5 and 7), BIR-1 Is Involved in Regulation of Transcription by RNA Polymerase II. harvested 24 h later, and processed for Western blot by using antibodies We have found that CHR3 (nhr-23) (31, 33), CeSKIP (17), and against S10-P K14-Ac H3 (lanes 5 and 6) and K9-Ac S10-P H3 (lanes 7 and 8). bir-1 loss of function have partially overlapping phenotypes. BIR-1 leads to an increase in phosphoacetylated histones H3 (lanes 2 and 4). (D) CeSKIP loss of function leads to marked decrease of RNA Densitometric analysis of results. polymerase II-dependent transcription of specific genes (17). dpy-7, a collagen gene, was found to be inhibited in CHR3 phosphoacetylated histones positively, we transfected MCF-7 (nhr-23) loss of function (33) by RNAi. We show here that bir-1 cells with bir-1 and assayed the level of phosphoacetylated RNAi leads to decreased transcription of some (but not all) histones H3 by Western blot. Cells transfected with bir-1 had an transgenes (elt-2::gfp and hlh-1::gfp) that were found to be CeSKIP increased level of immunoreactive phosphoacetylated histones decreased also by RNAi. These results indicate that bir-1 may be required for activities H3 recognized by both antibodies (Fig. 5C, lanes 6 and 8). The of the same promoters that depend on CeSKIP and CHR3 positive effect was observed also on cos-7 and NIH 3T3 cells (nhr-23) and suggest that not all promoters are equally sensitive transfected with BIR-1 (data not shown). to BIR-1. Discussion Moreover, BIR-1 expression led to stimulation of T3- dependent as well as basal transcription from thyroid hormone- bir-1 Loss of Function Triggers Developmental Changes Not Related to responsive promoters. The effect was further additive with Cell Division. Previous studies have shown that bir-1 is required in transcription stimulation by the histone deacetylase inhibitor early development for proper cytokinesis (14, 15). Loss of BIR-1 TSA. These data suggest that BIR-1 effect and histone acety- activity in the embryo leads to early lethality resulting from the lation are cooperative. effects of abnormal cell divisions. In this study, we show that bir-1 is also required for functions not related to cell division, most BIR-1 Affects Phosphoacetylation of Histones H3. Complex covalent likely transcriptional regulation of some genes in C. elegans. Loss modification of histones is part of the transcription activation of bir-1 activity postembryonically results in multiple develop- process. Histone modifications include methylation, acetylation, mental defects and a decrease of histone H3 phosphoacetylation. and phosphorylation. Although histone acetylation is a well RNAi delivered by feeding bacteria producing dsRNA specific established component of transcription activation, phosphory- for bir-1 leads to strong phenotypes (Egl, dpy, and larval lation and specific methylation of histones have been shown lethality) that overlaps with CeSKIP loss of function (17). recently to be part of specific transcription activation (refs. 6, 9, However, some characteristic phenotypic changes of bir-1 and and 11; for review see refs. 2, 4, and 7). Phosphorylation of serine CeSKIP loss of function differ. bir-1 RNAi does not result in 10 of histone H3 is linked to acetylation of lysine 14 in Gcn5- molting defects that are characteristic for CeSKIP loss of func- regulated promoters (10) and even facilitates the lysine 14

5244 ͉ www.pnas.org͞cgi͞doi͞10.1073͞pnas.0730770100 Kostrouchova et al. Downloaded by guest on October 1, 2021 acetylation (10, 11). Interestingly, BIR-1 was shown to phos- another domain frequently found in inhibitor-of-apoptosis pro- phorylate serine 10 of histone H3 in connection with aurora teins. The simplest members of the family such as BIR-1 or kinase AIR-2 (15). Several H3 serine 10-specific kinases have Survivin are composed of just one single BIR domain. The C. been identified including Ip11͞aurora family kinases and Snf1 elegans genome contains two Bir genes; the second one, bir-2, (12, 41). Snf1 was shown to activate transcription of INO1 by codes for a protein with a tandem of two BIR domains. In phosphorylation of serine 10 of histone H3, which in turn contrast to pronounced phenotypes obtained by bir-1 RNAi, facilitates acetylation of H3 on lysine 14 (12). Snf1 also regulates BIR-2 cannot substitute the function of BIR-1. This differs from Adr1-dependent transcription by directly enhancing Adr1 bind- the finding that BIR-1 function can be rescued by the human ing to chromatin (42). Phosphoacetylation of histone H3 on homologue Survivin (14). serine 10 and lysine 14 is part of protein kinase A-mediated gene In conclusion, we show that BIR-1 functions in transcription activation by follicle-stimulating hormone in granulosa cells of regulation. It is also possible that BIR-1 contributes to specific the ovary (43). histone phosphorylation during the transcription activation In the present study, bir-1 RNAi led to a decrease of both process. S10-P K14-Ac and K9-Ac S10-P phosphoacetylated histones H3 in worms. By immunohistochemistry we observed that the We thank Dr. Michael Krause (National Institute of Diabetes and Digestive and Kidney Diseases, National Institutes of Health) for decrease of the level of phosphoacetylated histones H3 caused support, advice, and critical reading of the manuscript and constructs by bir-1 RNAi was not only linked to cell division but also was hlh-1, hlh-2, and hlh-8; Dr. Paul Yen for constructs DR-4-luc and F2-luc; visible in nondividing cells. In addition, MCF-7, cos-7, and NIH Dr. Andrew Fire for vector L4440 and pPD 49.83; Dr. Diane Hayward 3T3 cells transfected with bir-1 had increased levels of phos- for flag-SKIP; Dr. Keiko Ozato for ME-TRE-luc; and Dr. Jacob Robbins phoacetylated histones. Thus, phosphoacetylation of histones (National Institute of Diabetes and Digestive and Kidney Diseases, H3 on serine 10 is likely to be directly involved in the effect of National Institutes of Health) for critical reading of the manuscript. Z.K. bir-1 on transcription. is supported by Grant Agency of the Czech Republic Grant 304-00-1128 and Ministry of Education of the Czech Republic Grant 11110000-3, and Members of the inhibitor-of-apoptosis family usually contain M.K. is supported by Grant Agency of the Czech Republic Grant several repeats of BIR domain and a RING domain. The general 304-02-1347. Z.K. and M.K. are associates at the Diabetes Branch, function of the proteins with the RING domain is probably E3 National Institute of Diabetes Digestive and Kidney Diseases, National ubiquitin ligase activity. The caspase recruitment domain is Institutes of Health.

1. Ito, M. & Roeder, R. G. (2001) Trends Endocrinol. Metab. 12, 127–134. 23. Krause, M., Park, M., Zhang, J. M., Yuan, J., Harfe, B., Xu, S. Q., Greenwald, 2. Felsenfeld, G., Clark, D. & Studitsky, V. (2000) Biophys. Chem. 86, I., Cole, M., Paterson, B. & Fire, A. (1997) Development (Cambridge, U.K.) 124, 231–237. 2179–2189. 3. Jenuwein, T. & Allis, C. D. (2001) Science 293, 1074–1080. 24. DeVore, D. L., Horvitz, H. R. & Stern, M. J. (1995) Cell 83, 611–620. 4. Berger, S. L. & Felsenfeld, G. (2001) Mol. Cell 8, 263–268. 25. Harfe, B. D. & Fire, A. (1998) Development (Cambridge, U.K.) 125, 421–429. 5. Crosio, C., Fimia, G. M., Loury, R., Kimura, M., Okano, Y., Zhou, H., Sen, S., 26. Corsi, A. K., Kostas, S. A., Fire, A. & Krause, M. (2000) Development Allis, C. D. & Sassone-Corsi, P. (2002) Mol. Cell. Biol. 22, 874–885. (Cambridge, U.K.) 127, 2041–2051. 6. Li, J., Lin, Q., Yoon, H. G., Huang, Z. Q., Strahl, B. D., Allis, C. D. & Wong, 27. Praitis, V., Casey, E., Collar, D. & Austin, J. (2001) Genetics 157, 1217–1226. J. (2002) Mol. Cell. Biol. 22, 5688–5697. 28. Brenner, S. (1974) Genetics 77, 71–94. 7. Berger, S. L. (2002) Curr. Opin. Genet. Dev. 12, 142–148. 29. Fire, A., Xu, S., Montgomery, M. K., Kostas, S. A., Driver, S. E. & Mello, C. C. 8. Hasan, S. & Hottiger, M. O. (2002) J. Mol. Med. 80, 463–474. (1998) Nature 391, 806–811. 9. Nowak, S. J. & Corces, V. G. (2000) Genes Dev. 14, 3003–3013. 30. Timmons, L. & Fire, A. (1998) Nature 395, 854. 10. Lo, W. S., Trievel, R. C., Rojas, J. R., Duggan, L., Hsu, J. Y., Allis, C. D., 31. Kostrouchova, M., Krause, M., Kostrouch, Z. & Rall, J. E. (1998) Development Marmorstein, R. & Berger, S. L. (2000) Mol. Cell 5, 917–926. (Cambridge, U.K.) 125, 1617–1626. 11. Cheung, P., Tanner, K. G., Cheung, W. L., Sassone-Corsi, P., Denu, J. M. & 32. Johnstone, I. L. & Barry, J. D. (1996) EMBO J. 15, 3633–3639. Allis, C. D. (2000) Mol. Cell 5, 905–915. 33. Kostrouchova, M., Krause, M., Kostrouch, Z. & Rall, J. E. (2001) Proc. Natl.

12. Lo, W. S., Duggan, L., Tolga, N. C., Emre, Belotserkovskya, R., Lane, W. S., Acad. Sci. USA 98, 7360–7365. BIOLOGY Shiekhattar, R. & Berger, S. L. (2001) Science 293, 1142–1146. 34. Zhou, S., Fujimuro, M., Hsieh, J. J., Chen, L., Miyamoto, A., Weinmaster, G. DEVELOPMENTAL 13. Lefebvre, B., Ozato, K. & Lefebvre, P. (2002) EMBO Rep. 3, 335–340. & Hayward, S. D. (2000) Mol. Cell. Biol. 20, 2400–2410. 14. Fraser, A. G., James, C., Evan, G. I. & Hengartner, M. O. (1999) Curr. Biol. 35. Samuels, H. H., Stanley, F. & Casanova, J. (1979) 105, 80–85. 9, 292–301. 36. Petty, K. J., Desvergne, B., Mitsuhashi, T. & Nikodem, V. M. (1990) J. Biol. 15. Speliotes, E. K., Uren, A., Vaux, D. & Horvitz, H. R. (2000) Mol. Cell 6, Chem. 265, 7395–7400. 211–223. 37. Tomura, H., Lazar, J., Phyillaier, M. & Nikodem, V. M. (1995) Proc. Natl. Acad. 16. Ambrosini, G., Adida, C. & Altieri, D. C. (1997) Nat. Med. 3, 917–921. Sci. USA 92, 5600–5604. 17. Kostrouchova, M., Housa, D., Kostrouch, Z., Saudek, V. & Rall, J. E. (2002) 38. Jones, K. E., Brubaker, J. H. & Chin, W. W. (1994) Endocrinology 134, 543–548. Proc. Natl. Acad. Sci. USA 99, 9254–9259. 39. Lewis, J. D. & Ausio, J. (2002) Biochem. Cell Biol. 80, 353–361. 18. Blumenthal, T. & Spieth, J. (1996) Curr. Opin. Genet. Dev. 6, 692–698. 40. Sassone-Corsi, P. (2002) Science 296, 2176–2178. 19. Blumenthal, T., Evans, D., Link, C. D., Guffanti, A., Lawson, D., Thierry-Mieg, 41. Hsu, J. Y., Sun, Z. W., Li, X., Reuben, M., Tatchell, K., Bishop, D. K., J., Thierry-Mieg, D., Chiu, W. L., Duke, K., Kiraly, M. & Kim, S. K. (2002) Grushcow, J. M., Brame, C. J., Caldwell, J. A., Hunt, D. F., et al. (2000) Cell Nature 417, 851–854. 102, 279–291. 20. Fukushige, T., Hawkins, M. G. & McGhee, J. D. (1998) Dev. Biol. 198, 286–302. 42. Young, E. T., Kacherovsky, N. & Riper, K. V. (2002) J. Biol. Chem. 277, 21. Fukushige, T., Hendzel, M. J., Bazett-Jones, D. P. & McGhee, J. D. (1999) Proc. 38095–38103. Natl. Acad. Sci. USA 96, 11883–11888. 43. Salvador, L. M., Park, Y., Cottom, J., Maizels, E. T., Jones, J. C., Schillace, 22. Krause, M., Harrison, S. W., Xu, S. Q., Chen, L. & Fire, A. (1994) Dev. Biol. R. V., Carr, D. W., Cheung, P., Allis, C. D., Jameson, J. L. & Hunzicker-Dunn, 166, 133–148. M. (2001) J. Biol. Chem. 276, 40146–40155.

Kostrouchova et al. PNAS ͉ April 29, 2003 ͉ vol. 100 ͉ no. 9 ͉ 5245 Downloaded by guest on October 1, 2021