kep1 interacts genetically with dredd/Caspase-8, and kep1 mutants alter the balance of dredd isoforms

Abstract

The Drosophila kep1 gene encodes an RNA binding protein related to the murine QUAKING apoptotic inducer. We have previously shown that kep1 can induce apoptosis when transfected into different cell lines. To better define the role of Kep1 in apoptosis, we generated kep1 null flies. These flies were viable, but females displayed reduced fertility, with approximately half of the eggs laid from kep1− homozygotes failing to hatch. In addition, loss of kep1 suppressed GMR-rpr-mediated apoptosis in the Drosophila eye, and kep1 mutant flies had increased susceptibility to Escherichia coli infection. We found that Kep1 bound dredd RNA in vitro, and that extracts prepared from kep1 mutant ovaries had markedly reduced proteolytic cleavage activity toward the caspase-8 target substrate IETD-7-amino-4-trifluoromethyl coumarin. We observed increased levels of the β isoform of dredd mRNA in kep1 mutants as compared with wild-type. Taken together, our results suggest that Kep1 regulates apoptosis by influencing the processing of dredd RNA.

Programmed cell death, or apoptosis, regulates the destruction of unnecessary cells in metazoan development, and alterations in the apoptotic program have been implicated in pathogenesis (for review, see ref. 1). KH domain-containing RNA binding proteins have been proposed as a family of apoptotic inducers (2). For example, the mouse quaking (QKI) protein and the Drosophila Kep1 and Sam50 proteins are potent inducers of cell death in the absence of any other signal (2–4). Expression of QKI causes apoptosis in transfected cells (4), and the C-terminal 14 aa of the QKI-7 isoform are necessary and sufficient to induce apoptosis (2). Kep1 (also referred to as Qkr58E-3; ref. 5), Sam50, and QKI all belong to the STAR protein family that contains a GSG domain (3, 6, 7). Like other single KH domain-containing proteins (3, 4, 8, 9), Kep1 can form homodimers and can bind homopolymeric RNA in vitro. In addition, lysates obtained from S2 cells transfected with Kep1 can induce isolated HeLa nuclei to show morphological characteristics indicative of apoptosis (3). Other KH domain-containing proteins also play a role in apoptosis. Induction of MCG10, an RNA binding protein containing two KH domains, can suppress proliferation of cells by inducing apoptosis and cell cycle arrest in G₂/M (10). Similarly, overexpression of the Drosophila homologue of the fragile X protein (Dfmr1) leads to apoptotic cell loss in all adult tissues examined (11). Finally, a proteomic analysis of Jurkat cells identified 21 proteins, of which 15 contained RNA binding motifs, that are modified during Fas-induced apoptosis, suggesting an important role for RNA binding proteins in the apoptotic process (12).

Many cells undergo apoptosis during embryogenesis and metamorphosis of Drosophila (13–15). Central to the apoptotic program are enzymes collectively referred to as caspases, which are synthesized as inactive proenzymes. These zymogens are cleaved to generate the catalytically active molecules that initiate a proteolytic cascade. Loss of the caspase CED-3 in the nematode Caenorhabditis elegans leads to a loss of most developmental programmed cell death (16). The functions of three other caspases encoded in the C. elegans genome have not yet been established. Several genes encoding caspases are present in the Drosophila genome: Dcp1 (17), Damm (18), decay (19), and Ice (20), which encode effector caspases with short prodomains, and dream (21), dredd (22), and Dronc (23), which encode initiator caspases with long prodomains. The initiator caspases are cleaved to their active forms, which in turn cleave effector caspases and other molecules.

In Drosophila, all embryonic cell death is under the control of three closely linked genes, reaper (rpr), grim, and head involution defect (hid), that all fall within a chromosomal deletion Df(3R)H99 (H99; refs. 24–26). These genes induce expression of genes encoding caspases, including dredd. In oogenesis, dredd mRNA is expressed beginning at stage 10 in nurse cells and the developing oocyte. dredd mRNA accumulates at high levels in stage-12 to -13 egg chambers, coincident with nurse cell apoptosis. dredd is also expressed in developing embryos, with its expression preceding the death of the cell in which it is found. dredd mRNA is essentially absent in H99 embryos, and mutations that remove dredd are capable of suppressing GMR-rpr-induced apoptosis in the Drosophila eye (22).

dredd produces at least four different mRNAs (22). The δ and γ isoforms encode full-length Dredd proteins that differ by the presence of an additional 6 aa in the δ isoform. The protein encoded by the α isoform has most of its prodomain deleted. The β isoform of dredd mRNA differs from the others because of the inclusion of intron II, which can result in either the translation of a protein that contains all of the prodomain but lacks most of the catalytic domain, or possibly a protein that contains only the caspase domain. Evidence from cDNA sequencing projects suggests that the β isoform normally exists in several tissues (27).

In addition to its role in apoptosis, dredd is involved in the Drosophila antibacterial response (28, 29). In dredd mutants, the expression of all genes that code for peptides with antibacterial activity following Gram-negative bacterial infection is repressed (29). Activation of the Drosophila immune response is preceded by the proteolytic cleavage of the NF-κB factor Relish. The cleavage of Relish is independent of the proteosome but has an absolute requirement for Dredd (30). These data suggest either a direct role for Dredd in the cleavage of Relish or in the participation in a proteolytic cascade that results in Relish activation.

Although kep1 is an inducer of apoptosis, how it regulates cell death and its broader cellular role are unknown. To investigate the function of kep1 in Drosophila development, we generated kep1− null flies. We found that kep1− null flies are viable but display reduced fertility. Furthermore, we show that the Kep1 protein can bind dredd mRNA in vitro. Mutations in dredd and kep1 behave in a similar fashion when infected with Escherichia coli. Loss of either dredd or kep1 could also suppress the GMR-rpr-induced apoptosis in the Drosophila eye. We conclude that Kep1 is a positive regulator of apoptosis in Drosophila that exerts its effect in part through dredd.

Methods

P-Element Mobilization.

To create a null allele of kep1, excision lines were generated by the introduction of a transposase source into EP2493/CyO (obtained from Todd Laverty, Berkeley Drosophila Genome Project, Berkeley, CA). EP2493/CyO males were crossed to w;Bic-D^PA66 Su(Bic-D^PA66) cn/CyO;Δ2-3 Sb/TM3 Ser virgin females (31). EP2493/CyO;Δ 2–3 Sb/+ males were individually crossed with w;P[w+]/CyO virgin females. Individual white-eyed males that have lost the P element, EP2493*/CyO, were crossed to P[w+]/CyO virgin females, and white-eyed balanced progeny were crossed to obtain individual balanced lines. Excision lines were screened by Southern analysis using the complete kep1 cDNA as a probe.

Fertility Testing.

Drosophila lines were tested for fertility by crossing to Oregon-R (OreR) flies. Flies of the appropriate genotype were introduced into cages and were subjected to a 24-h prelay on apple juice plates. Eggs were subsequently collected for a 24-h period and were scored 30 h later according to the number and phenotype of unhatched progeny.

Characterization of the KEX15 Excision.

Genomic DNA was isolated from homozygous kep1 flies and used as a template for PCR. Oligonucleotides 5′-CTCAGTAAACTCGCTTGG-3′ and 5′-CCTAATACTAGTTAAAGCC-3′ were designed to amplify between positions 89907 and 91364 of the BAC clone BACR48M13 (coordinates are those of GenBank accession no. AC005714). The product isolated from the PCR was cloned into the EcoRV site of pBlueScript II and sequenced using T7 and T3 sequencing primers (Sheldon Biotechnology Centre, McGill University; ref. 3).

Generation of Transgenic Flies.

Myc-kep1 was excised from the pBlueScript II vector (3) by using a KpnI/XbaI digest and cloned into the KpnI/XbaI sites of pCasper4. A 1.45-kb genomic DNA fragment containing all of the sequence upstream of kep1 was amplified from a yw background by using oligonucleotides 5′-AAATCTAGACCTAATACTAGTTAAAGCC-3′ and 5′-AAATCTAGATTTTATCATAACAATCTTTAC-3′. The PCR fragment was digested with XbaI and cloned upstream of kep1 in the myc-kep1 pCasper4 construct. Transgenic flies were generated using standard microinjection techniques (32) in a yw recipient, using the pTurbo plasmid as a transposase source.

UV Crosslinking Assay.

Radiolabeled RNA was generated from expressed sequence tag GH10971 by using the T7 RNA polymerase from PCR products generated using T7 sequencing primer and 5′-CATATTATCCCTGCTAAGATTGGC-3′ (primer 01–58) or 5′-CCACATTGTATCCCATCGAGG-3′ (primer 01–59) or using oligonucleotides designed to amplify intron II containing a T7 RNA polymerase site at the 5′ end oligonucleotide 5′ T7-CGGAATGTTAGCAGGGATAATATG-3′ (primer 01–165) and 5′-GTGGAGAAGGGCAATGTTGTACTTC-3′ (primer 01–166). RNAs were incubated on ice for 30 min with HeLa cell extracts obtained after transfection of myc-kep1 or myc-Sam50. Samples were UV-crosslinked in a Stratalinker (Stratagene) for 20 min on ice, followed by RNase A treatment before Myc immunoprecipitation. Samples were run on a 10% PAGE.

Caspase-8 Assays.

Dredd activity was assessed using the Apo-alert kit (CLONTECH) following the manufacturer's protocol. The equivalent of two pairs of ovaries from well fed flies were used per assay. We followed the conversion of the caspase-8 7-amino-4-trifluoromethyl coumarin (AFC) substrate (IETD-AFC) to AFC over a time course of 1 h at 25°C. All assays were performed in duplicates on three separate days by using independent ovary extracts. Experiments were designed to maintain the rate of conversion of the substrate throughout the time course of the experiment within the linear range of the instrument (Fluoromark PM microplate reader, Bio-Rad).

RT-PCR and the Generation and Screening of Ovary cDNA Libraries.

Libraries were generated using the Stratagene cDNA synthesis kit according to the manufacturer's protocol. cDNAs were cloned into the EcoRI/XhoI site of the Uni Zap vector by using poly(A)+ RNA obtained from 2- to 3-day-old, well fed females. Approximately 750,000 plaque-forming units were screened for each library by using a 500-bp fragment obtained from the 5′ end of the dredd EST LP05556. We isolated eight cDNA clones (four from the yw library, and four from the kep1− library), and all were completely sequenced. All eight cDNA clones obtained had different 5′ ends and were therefore independent clones. For RT-PCR experiments, 2 pmol of oligonucleotide 5′-GTCGATGTGCTGGTCCGGCG-3′ (primer 01–60) was used to prime the reverse transcription reaction by using 1 μg of poly(A)⁺ mRNA isolated from yw or kep1− ovaries. One-tenth of each reverse transcription reaction was used in each of the subsequent PCR by using the cycling conditions 94°C for 30 s, 60°C for 30 s, and 74°C for 45 s over 30 cycles employing Vent polymerase (New England Biolabs). Oligonucleotide pairs were either 5′-ATGGCCGGATCAAACCTGTTG-3′ (primer 01–55), located in exon 1, and 5′-GGAGAAGGGCAATGTTGTACTTC-3′ (primer 01–57), located in intron II, or primers 01–55 and 01–59 located in exon 2. One-tenth of each PCR was loaded onto a 1.5% agarose gel.

E. coli Infections of dredd and kep1 Mutant Flies.

Two- to 4-day-old female flies were pricked with a 30-gauge needle dipped in a concentrated solution of E. coli (5 ml of bacteria grown to OD 0.6 were recovered and suspended in 500 μl of media). Flies were allowed to recover for 8 h at room temperature and all subsequent values were relative to the number of flies alive at this time point. Flies were incubated at 29°C and, after daily examination for survivors, were transferred to fresh vials. Counts were taken every day for 7 days.

Electron Microscopy.

Flies were dissected and placed onto a platform. Scans were performed on unfixed tissues by using the S-3000N (Hitachi, Tokyo) scanning electron microscope at the Institute of Materials and Intelligent Systems, University of Sherbrooke. Scans were at ×200 magnification, using a 3-kV beam.

Results

Generation of kep1 Mutants.

The line EP2493 (33) carries an EP element inserted 67 bp 5′ from the predicted transcriptional start site of kep1. We used this line to generate kep1 knockout flies by using imprecise P-element excision. Seventy-seven lines were initially generated, of which seven lines contained a rearrangement in their genome as assessed by Southern blot analysis and genomic PCR (data not shown). Sequence analysis confirmed that one of the above lines, which we termed kep1−, contained a 1.4-kb deletion spanning exons 1–5 of the kep1 gene (Fig. 1A).

Figure 1.

Characterization of kep1− mutants. (A) Schematic diagram of the arrangement of kep1 and two neighboring genes in region 58E of chromosome 2 in wild-type (Top) and in the kep1− deletion (Middle). Note that much of the kep1 gene is deleted in the kep1− deletion. The wild-type and kep1− genomic sequences are given at the Bottom. PCR was used to amplify the region surrounding the P-element excision site in kep1− homozygotes, the amplified fragment was sequenced, and its sequence was compared with that of a BAC clone that did not contain the original P element (BACR48M13). The start of the kep1 cDNA in BACR48M13 is at position 89990. The sequence of the remaining fragment of the P element, after the imprecise excision event, is underlined. (B) Phenotypes of eggs laid by kep1− mutant females; details and frequencies of the different phenotypes are described in Table 1.

kep1 Mutant Females Have Reduced Fertility and Produce Defective Eggs.

kep1− homozygous flies were viable and displayed no obvious morphological phenotype. We compared the fertility of kep1− females to heterozygous and wild-type controls, and found that the percentage of eggs hatching from kep1−/CyO females was indistinguishable from that of wild-type controls (Table 1). In contrast, homozygous kep1− females displayed reduced fertility, with only 46% of the kep1− eggs hatching. This reduced fertility was rescued with the introduction of a kep1 transgene. Female kep1− homozygotes carrying two copies of the kep1 transgene had fertility similar to the OreR controls (Table 1). The unhatched eggs produced by homozygous kep1− females fell into three major phenotypic classes (Fig. 1B). Eggs belonging to the first class (17.7%) were short and fat in appearance but had the proper amount of dorsal appendage material (Table 2). The second class (7%) included eggs that were normal in size but had little or no dorsal appendage material. The remaining 75.3% had a normal appearance (Table 2).

Table 1.

Fertility of kep1− flies

Females	Males	Total eggs	Hatched	% Hatching
OreR	OreR	164	151	92
kep1−/CyO	OreR	128	120	94
kep1−/kep1−	OreR	261	121	46
P[kep1+]; −/−	OreR	1,102	941	85

Female and male flies of the genotypes listed above were introduced into cages and were allowed to lay eggs for a 24-h period. Eggs were scored 30 h later for their ability to hatch. 

Table 2.

Phenotypes of eggs produced by kep1− flies

Egg phenotype	OreR		kep1−/CyO		kep1−/kep1−
	No. of eggs	%	No. of eggs	%	No. of eggs	%
Normal	803	100	771	96.2	310	75.3
Short eggs	0	0	18	2.2	73	17.7
DA defects	0	0	13	1.6	29	7.0

Flies of the genotypes listed above were introduced into cages and allowed to lay eggs on apple juice plates supplemented with live yeast (two changes, 1 h each). Eggs were then collected for 1 h and scored for their phenotypes 8 h later. The data above are the cumulative numbers obtained from egg collections carried out on three consecutive days. DA, dorsal appendage. 

Kep1 Interacts in Vitro with dredd mRNA.

To identify potential RNA targets for Kep1, we used an in vitro RNA crosslinking assay to test whether Kep1 could associate with several mRNAs that encode proteins involved in the apoptotic process. We found that dredd mRNA bound myc-tagged Kep1, whereas the mRNAs encoding the effector caspase Dcp1 and the Drosophila inhibitor of apoptosis 2 (diap2; ref. 34) could not be crosslinked to the Kep1 protein (Fig. 2A and data not shown). The interaction observed between Kep1 and dredd mRNA was specific for Kep1, because a closely related Drosophila KH domain-containing protein Sam50 failed to bind dredd mRNA (Fig. 2 A and B).

Figure 2.

Kep1 protein can be UV crosslinked to dredd mRNA. (A) Radiolabeled RNAs (refer to Methods) were incubated in the presence of HeLa cell extracts transfected with either myc-Sam50 (lanes 1, 2, 5, 6, 9, and 10) or myc-kep1 (lanes 3, 4, 7, 8, 11, and 12). These extracts were immunoprecipitated with control serum (odd-numbered lanes) or 9E10 anti-Myc antiserum (even-numbered lanes). (B) Myc immunoblot using 9E10 anti-Myc antiserum showing the presence of the transfected proteins in both extracts. (C) Organization of part of the dredd gene. Exons are shown as dark boxes; arrows indicate the relative position of the oligonucleotides used in the RT-PCR experiments to investigate the nature of the different dredd mRNA isoforms in various mutants. (D) RT-PCR were fractionated on a 1.5% agarose gel. Lanes indicate the source of the mRNA used in the reverse transcription reactions in the presence (+RT) or absence of reverse transcriptase.

The β Isoform of dredd mRNA Is More Abundant in kep1 Ovaries.

To test the possibility that Kep1 has an effect on dredd RNA metabolism, we probed Northern blots with the 3′-untranslated region of dredd that detects all known dredd isoforms. We initially focused on RNA isolated from ovaries because dredd is expressed in this tissue (22). A single band of ≈1.6 kb was observed in extracts from wild-type and kep1− ovaries consistent with the size of mRNAs encoding the β, δ, or γ, but not the α, isoforms of dredd (data not shown). To determine whether the ratio of mRNAs encoding these dredd isoforms is altered in kep1− ovaries, we generated cDNA libraries from kep1− homozygous ovaries and wild-type controls. Four independent dredd clones were isolated from the kep1− library: two encoded the δ isoform of dredd and two encoded the truncated β isoform. In contrast, all four independent clones recovered from a wild-type library encoded the δ isoform of dredd. These data suggest that kep1 activity influences the ratio between the β and δ isoforms. To extend this result, the isoforms of dredd mRNA were examined in wild-type and kep1− mutant populations by using an RT-PCR assay. The relative positions of the oligonucleotides used in the RT-PCR experiments are illustrated in Fig. 2C. RT-PCR products corresponding to mRNAs encoding the β and δ isoforms were present in RNA isolated from kep1− ovaries, whereas only the mRNA that encodes the δ isoform was observed in wild-type samples (Fig. 2D). These data indicate that kep1 activity negatively influences the accumulation of the β isoform of dredd.

Reduced Dredd/Caspase-8 Activity in kep1−/− Ovaries.

Dredd contains a long prodomain typical of a group III apical caspase (35). Because Dredd is most closely related to caspase-8 (36), we measured caspase-8 activity in Drosophila ovaries obtained from different genotypes by using a fluorometric-based assay that relies on the ability to cleave the nonfluorometric substrate IETD-AFC to free the fluorescent AFC moiety. As expected, mutations in dredd reduced caspase-8 activity relative to the wild-type control (Fig. 3, B118/B118 vs. OreR). However, this activity was still above background levels (≈3,000 RFU), indicating that our assay is not strictly specific for Dredd, but also detects additional group III caspase activities that can cleave the IETD sequence (35). We observed a dramatic decrease of IETDase activity in kep1− homozygous ovary extracts (kep1−/−), even more pronounced than for the dredd allele tested. Introduction of two copies of the kep1 transgene into the kep1− homozygous background restored the IETDase activity to near wild-type control levels (P[kep1];kep1−). Ovary extracts obtained from kep1− heterozygotes had an intermediate level of IETDase activity. These data indicate that IETDase activity in ovaries depends on the level of Kep1.

Figure 3.

IETDase activity in ovary extracts. Extracts were prepared and IETDase activity was measured as described in Methods. Results are the average readings from duplicate samples from one representative experiment. RFU, relative fluorescent units. OreR was used as the wild-type control, and other extracts were analyzed from kep1− heterozygotes (kep1−/CyO) and homozygotes (kep1−/−), homozygotes for two alleles of dredd (EP1412/EP1412, B118/B118), and ovaries obtained from flies carrying two copies of the kep1 transgene in a kep1 mutant homozygous background (P[kep1];kep1−/−).

kep1 Mutant Flies Have Increased Susceptibility to E. coli Infection.

To investigate whether kep1− and dredd mutants have other common phenotypes, we examined whether kep1− flies were also susceptible to Gram-negative bacterial infection, as are dredd flies (28, 29). We challenged flies by pricking them with a needle dipped in either a concentrated solution of E. coli or a solution of PBS as a control. Under the conditions used, control flies showed a slight decrease in viability after an E. coli insult (Fig. 4A), similar to that observed for injection with PBS alone (Fig. 4B). In contrast, dredd mutants showed a dramatic decrease in viability after E. coli infection, with only 10% of the flies still alive by the end of the observation period. kep1 mutants also showed a decrease in viability, but this decrease was not as dramatic as that obtained for dredd. The survival rates for kep1 mutants were 30% and 40% for homozygotes and heterozygotes, respectively. Unlike OreR or kep1− heterozygous flies, homozygous kep1− flies, and flies of both alleles of dredd, show a decrease in viability after PBS injections (Fig. 4B). This decrease is possibly a result of a bacterial infection being established at the wound site.

Figure 4.

Susceptibility of flies to E. coli infection. Flies were challenged with E. coli or subjected to a mock infection as described in Methods. OreR was used as the wild-type control, and survival was also scored for kep1− heterozygotes (kep1−/CyO) and homozygotes (kep1−/−), and homozygotes for two alleles of dredd (EP1412/EP1412, B118/B118).

The kep1 Mutation Suppresses rpr-Induced Apoptosis in the Eye.

dredd mRNA accumulation requires the activity of the genes present at the H99 locus, including rpr. Overexpression of rpr in photoreceptor precursor cells by using the GMR-rpr transgene leads to severe apoptosis that depends on dredd activity; so, a dredd mutation suppresses this phenotype (22). We assessed whether mutations in kep1− were also capable of suppressing GMR-rpr-mediated apoptosis by using the same experimental system. Scanning electron microscopic images obtained from unfixed wild-type eyes showed normal morphology (Fig. 5A). In contrast, in kep1− mutants we sometimes observed an outgrowth of extra cells (Fig. 5B). We crossed flies carrying the GMR-rpr transgene with flies carrying a mutation either in dredd or kep1. As previously reported, GMR-rpr apoptosis was partially suppressed when both copies of dredd were removed (Fig. 5F), and we observed that the kep1 mutation also suppressed GMR-rpr-mediated apoptosis (Fig. 5E). dredd has been shown to be unable to suppress GMR-hid apoptosis (22); consistent with this, we were unable to detect any suppression of GMR-hid apoptosis with either dredd or kep1 mutations (data not shown).

Figure 5.

Scanning electron micrographs of compound eyes. Flies of various genotypes were visualized by scanning electron micrographs as described in Methods. (A) yw. (B) kep1−/kep1−. (C) B118/B118. (D) GMR-rpr/+. (E) kep1−/kep1−;GMR-rpr/+. (F) B118/B118;GMR-rpr/+. The arrow illustrates an example of cell overgrowth occasionally seen in kep1−/kep1− mutant eyes.

Discussion

Our study of kep1− mutant flies confirms the in vitro data previously obtained that link kep1 to the apoptotic process, and implicates Kep1 in dredd RNA processing. The β isoform of dredd that accumulates in kep1− mutants encodes a truncated protein that may have a similar function to the decoy caspase-8 molecules identified in vertebrates (28). These molecules are similar in sequence to caspase-8, except that they lack essential catalytic residues and therefore may compete with caspase-8 for sites on the receptor. Regulation of dredd RNA processing may be an alternative strategy that Drosophila has evolved to establish an additional level of control of caspase-8 activity. The protein produced by the β isoform of dredd may act in a dominant-negative fashion, either by binding to dFADD and down-regulating a Fas-like signaling pathway, or at the apoptosome through its ability to interact with Dark (37). A recent report (38) implicates dFADD in regulating the response to infection by Gram-negative bacteria, which, taken together with our results, suggests that a dominant-negative isoform of Dredd could operate at the level of receptor binding.

Although Dredd is the Drosophila caspase most similar to caspase-8, because of its long prodomain and DED sequences, our assays of caspase-8 activity indicate that other Drosophila caspases possess IETDase activity as well, and these caspases also appear to be regulated by Kep1. Dronc, another group III caspase with substantial IETDase activity despite its active site, which is very divergent from that of Dredd (39), may be responsible for this additional activity. Dronc is induced by ecdysone (23, 40), and ecdysone inhibits apoptosis of nurse cells (41, 42); so in developing ovaries, Dronc, although present, must be in an inactive form. In our ovary extracts, however, we may have diluted out any short-lived inhibitory proteins, for example Diap1, which can bind Dronc (43). Although there is no reported evidence for alternative splicing of dronc, it will be informative to determine whether Kep1 can bind dronc or other Drosophila caspase mRNAs, and to examine whether any alternative transcripts are present in kep1− mutants.

blastp analysis identifies the closest mammalian homologue of Kep1 as the SLM2 protein (44), also referred to as Étoile (45). We have observed that transfection of SLM2 into NIH 3T3 cells results in activation of apoptosis, which can be inhibited by cotransfection of the Drosophila inhibitor of apoptosis DIAP1 (M.D.F. and S.R., unpublished data). Recent work (46) supports the involvement of SLM2 in splice site selection, because rat SLM2 was shown to interact with various proteins involved in alternative splicing, including SRp30c and the splicing associated factor YT521-B. SLM2 can also influence the splicing pattern, in a concentration-dependent manner, of the CD44v5, human transformer-2β and τ minigenes in cotransfection experiments (46). Alternative splicing of the genes encoding the caspase Ich-1 (47), Bcl-X (48), CED-4 (49), caspase 9 (50), and interleukin-1β converting enzyme (51) have also been reported, although the RNA binding proteins involved have not been identified. In all of these cases, the spliced isoforms show the opposite apoptotic functions of their parental products. Thus, regulation of RNA processing may be a general mechanism of regulating the balance between active and repressor forms of proteins involved in the apoptotic pathway, and Kep1 may belong to a family of RNA binding proteins that influence expression of these genes at the level of RNA processing.

Abbreviations

AFC

7-amino-4-trifluoromethyl coumarin

OreR

Oregon-R