Genetic and Molecular Characterization of the ... - Semantic Scholar

Copyright  1998 by the Genetics Society of America

Genetic and Molecular Characterization of the Caenorhabditis elegans Gene, mel-26, a Postmeiotic Negative Regulator of MEI-1, a Meiotic-Specific Spindle Component M. Rhys Dow and Paul E. Mains Department of Biochemistry and Molecular Biology, University of Calgary, Calgary, Alberta T2N 4N1, Canada Manuscript received May 8, 1997 Accepted for publication May 27, 1998 ABSTRACT We have previously described the gene mei-1, which encodes an essential component of the Caenorhabditis elegans meiotic spindle. When ectopically expressed after the completion of meiosis, mei-1 protein disrupts the function of the mitotic cleavage spindles. In this article, we describe the cloning and the further genetic characterization of mel-26, a postmeiotic negative regulator of mei-1. mel-26 was originally identified by a gain-of-function mutation. We have reverted this mutation to a loss-of-function allele, which has recessive phenotypes identical to the dominant defects of its gain-of-function parent. Both the dominant and recessive mutations of mel-26 result in mei-1 protein ectopically localized in mitotic spindles and centrosomes, leading to small and misoriented cleavage spindles. The loss-of-function mutation was used to clone mel-26 by transformation rescue. As suggested by genetic results indicating that mel-26 is required only maternally, mel-26 mRNA was expressed predominantly in the female germline. The gene encodes a protein that includes the BTB motif, which is thought to play a role in protein-protein interactions.

A

SSEMBLY of the meiotic spindle uses a mechanism distinct from that employed during mitosis. The traditional view of mitotic spindle formation holds that the centrioles, and associated material, nucleate the formation of microtubule asters. The microtubules extend from the asters, particularly toward the chromosomes, ultimately bridging the gap between the poles to form a bipolar spindle. Microtubules may attach to the kinetochores of the chromosomes, or may be cross-linked, possibly by kinesin-like proteins, with microtubules originating from the opposite pole (Hyman and Karsenti 1996; Waters and Salmon 1997). Examination of the formation of meiotic spindles in oocytes of many organisms has demonstrated that a distinct spindle assembly mechanism is used (Schatten et al. 1985; Sawada and Schatten 1988; Gard 1992; Theurkauf and Hawley 1992; Albertson and Thomson 1993). Following nuclear envelope breakdown, short microtubular arrays initially form around the chromosomes. These arrays become larger and appear to be bundled at their minus ends, forming several foci. After a period of time, the foci fuse until the two poles of a meiotic spindle are evident. Thus, oocyte meiotic spindles seem to grow from the inside out, as opposed to being nucleated from external centers (the centrosomes). In some respects, this should not be surprising; it has been known for some time that oocytes usually

Corresponding author: Paul E. Mains, Department of Biochemistry and Molecular Biology, University of Calgary, 3330 Hospital Drive N.W., Calgary, AB, T2N 4N1 Canada. E-mail: [email protected] Genetics 150: 119–128 (September 1998)

lack centrioles (Schatten 1994). While it may be argued that essential components of the centrosome are nonetheless present in the absence of centrioles, proteins thought to play an important role (e.g., g-tubulin) in forming the mitotic spindle are sometimes absent from the poles of acentriolar meiotic spindles (Palacios et al. 1993; Matthies et al. 1996; Merdes and Cleveland 1997). It is also likely that there are additional spindle components unique to acentriolar meiotic spindles. Because of the distinct mechanisms of spindle assembly used during meiosis and mitosis, it is likely that factors required for one type of division could disrupt spindle function if they are incorporated into a spindle of another type. In the nematode Caenorhabditis elegans, the cytoplasm of the egg must support both types of division, within 20 min of each other (Hirsh et al. 1976; Albertson 1984; Kemphues et al. 1986). Clearly some regulatory mechanism is required to ensure that components of the meiotic spindle are removed or inactivated before the onset of mitosis. We have previously described a component unique to the oocyte meiotic spindle (Clark-Maguire and Mains 1994a,b). The maternally-active gene mei-1 belongs to a family of ATPases, members of which play diverse cellular roles with no obvious unifying theme (Confalonieri and Duguet 1995). The precise role of mei-1 protein (MEI-1) in the meiotic spindle is unknown; however, in mei-1 loss-of-function (lf ) mutants, meiotic spindle formation does not advance beyond the initial coalescence of microtubules. In contrast, in embryos from animals bearing a gain-of-function (gf ) allele,

120

M. R. Dow and P. E. Mains

MEI-1 persists after meiosis and is misincorporated into the mitotic spindle and asters. This disrupts rotation of the asters (which accompany the sperm pronucleus) on to the anterior-posterior axis and results in a small, often misoriented spindle. Clearly the regulation of the mei-1 gene is important for the successful completion of both mitosis and meiosis in C. elegans. mel-26 was originally identified in a screen for dominant, temperature-sensitive (ts), maternal-effect embryonic lethal mutations (Mains et al. 1990b). Genetic evidence indicated mel-26(1) limits postmeiotic activity of mei-1: either a gf allele of mei-1 or the mel-26 mutation result in ectopic MEI-1 assembly in mitotic spindles and asters, leading to similar early cleavage defects (ClarkMaguire and Mains 1994b). Furthermore, the mel-26 and mei-1(gf ) mutations enhance one another, suppressors of mei-1(gf ) also suppress mel-26, and mei-1(null) meiotic defects are epistatic to mel-26 mitotic abnormalities (Mains et al. 1990a; Clark-Maguire and Mains 1994b). However, the interpretation that mel-26(1) inhibits postmeiotic mei-1 activity is complicated by the gf nature of the original mel-26 allele. In this article, we describe an apparent null allele of mel-26. The recessive phenotype of this mutation is identical to the dominant phenotype of mel-26(gf ), indicating that the gf allele represents an antimorph. We used the null allele to clone mel-26 by transformation rescue, and we have begun the molecular characterization of this gene. One region of mel-26 protein (MEL26) shows similarity to the recently defined bric a` brac, tramtrack, and Broad Complex gene (BTB) motif that is thought to play a role in protein-protein interactions; otherwise, mel-26 is predicted to encode a novel protein. MATERIALS AND METHODS Strains and culture conditions: Nematode strains were maintained under standard conditions as described by Brenner (1974), and brood analysis was conducted as described by Mains et al. (1990b). Nomenclature follows that of Horvitz et al. (1979). The following genes and alleles were used: Linkage Group I: mei-2(ct102), unc-13(e1091), mei-1(ct46, ct46ct82), daf-8(e1393), unc-29(e1072), mel-26(ct61, ct61sb4, sb45), lin11(e566); Linkage Group III: glp-1(q231); and Linkage Group IV: fem-1(hc17), fem-3(q20). Genetic identification and characterization of mel-26(lf ): mel-26(ct61) is a dominant, ts maternal-effect lethal mutation. To revert ct61 to a recessive lf allele, unc-29 mel-26(ct61)/unc13 daf-8 lin-11 hermaphrodites were mutagenized with ethyl methanesulfonate (EMS) under standard conditions (Brenner 1974) and plated (2 worms/plate) at the permissive temperature of 158. Plates with F1 progeny (z200/plate) were upshifted, before adulthood, to the restrictive temperature of 258 and screened for those showing high numbers of phenotypically wild-type F2 progeny. These plates would represent suppression of the dominant ct61 allele, either by intragenic or extragenic events. The flanking markers on the balancer chromosome allowed the detection of undesired crossover events that had lost ct61, and the daf-8 mutation, which results in ts dauer formation, led to arrest of animals homozygous for the balancer chromosome. From a screen of z7000 F1

animals, four suppressors were identified. Two suppressors were alleles of mei-1 whereas another was an apparent reversion to wild type that was not further characterized. One revertant, designated ct61sb4, represented the desired lf event (see results). This mutation was outcrossed five times to remove extraneous mutations induced by the mutagen. A second screen of z5000 unc-29 mel-26(ct61)/hT2 animals yielded seven suppressors. [hT2 is a reciprocal translocation that suppresses recombination in the desired region (Edgley et al. 1995).] All of the mutations from this second screen resulted in recessive meiotic defects, suggesting that they represent mei-1 or mei-2 alleles. [The high frequency of mei-1 mutations was expected given the previously observed propensity of this gene to mutate to suppression of mei-1(gf) and mel-26 (ClarkMaguire and Mains 1994a).] We were unable to separate ct61 from its cis-linked suppressor sb4, indicating that sb4 is likely an intragenic event. Lin non Daf crossovers were selected from the strain unc-29 mel26(ct61sb4)/unc-13 daf-8 lin-11. Of 80 recombinants, 70 segregated Unc-29, indicating that these crossovers occurred in the 1.7 cM interval between unc-29 and lin-11. None of these was ts, indicating that if sb4 maps to the right of ct61, it is within 0.024 cM. Similarly, we selected Unc non Lin recombinants from the strain daf-8 mel-26(ct61sb4)/unc-29 lin-11. None of 56 recombinants were ts. If sb4 is to the left of ct61, it must be within 0.03 cM. Microscopy and immunofluorescence: Embryos were dissected from gravid hermaphrodites that had been maintained at the restrictive temperature for several hours, mounted on agarose pads, and observed with a Zeiss Axioplan microscope using Nomarski optics. The embryos were flash-photographed with Kodak TechPan film developed at 100 ASA. Fixation for immunofluorescent staining was performed as described by Kemphues et al. (1986). Tubulin localization was determined using a mouse monoclonal antibody to Drosophila a-tubulin (Piperno and Fuller 1985); secondary antibodies were either rhodamine-conjugated goat anti-mouse IgG or fluorescein-conjugated donkey anti-mouse IgG ( Jackson Immunoresearch, West Grove, PA). MEI-1 was visualized as previously described (Clark-Maguire and Mains 1994b), using fluorescein-conjugated goat anti-rabbit IgG or cy3-conjugated donkey anti-rabbit IgG (Jackson Immunoresearch). Chromatin was stained with 49, 6-diamidine-29-phenylindole (DAPI). Photographs were taken with Kodak TechPan film, exposed at 200–400 ASA and developed at 100 ASA. Molecular cloning of mel-26: The correlation between the physical and genetic maps of C. elegans identified cosmids with the potential of encoding the mel-26 gene. DNA from cosmids C08H1, D1004, F25H5, W04H6, W06D4, or ZK858 (10–100 mg/ml) were separately mixed with the plasmid pFR4 [rol6(su1006); 100 mg/ml]. This DNA was injected into the strain daf-8 mel-26(ct61sb4)/unc-13 lin-11. In C. elegans, DNA injected into the hermaphrodite gonad is concatenated into large extrachromosomal arrays that are incorporated into the developing oocytes, and the pRF4 plasmid confers a dominant Rolling phenotype allowing the identification of transformed animals (Mello et al. 1991). Transgenic lines were established, grown at 258, and Daf Rol F1 segregants were down-shifted to 158 to allow exit from dauer. The Rol and non-Rol progeny of these worms were upshifted to 258 as young adults, and hatching levels were compared to identify transgenic rescue of the ct61sb4 recessive ts maternal-effect lethal phenotype. ZK858 was the only cosmid to rescue. To further narrow the location of the sequence encoding mel-26, cosmid DNA was digested with various restriction enzymes to identify those that produced fragments still capable of rescue. The minimally rescuing 7.6-kb XhoI and BglII fragment of ZK858 was subcloned into the XhoI and BamHI sites of pBluescript (Stratagene, La Jolla, CA).

The mel-26 Gene of C. elegans Initial cDNA library screening was conducted using a mixedstage cDNA library in lZAP (Barstead and Waterston 1989) and the smallest rescuing subclone as the probe. After screening approximately 130,000 plaques without success, we turned to a library constructed from embryos containing less than 30 cells (Schauer and Wood 1990). A total of 370,000 plaques were screened from this library; three cDNA clones were isolated. Sequencing of mel-26: Nested deletions of the cDNA and genomic clones were generated using the Erase-a-Base protocol (Promega, Madison, WI). These deletions were sequenced using the ABI PRISM fluorescent cycle-sequencing kit (Perkin Elmer, Foster City, CA). Sequence data were collected for both strands for both the genomic and cDNA clones of mel-26. Mutations of mel-26 were identified using worms homozygous for the allele of interest, which were placed in a PCR tube containing 2.5 ml of worm lysis buffer (50 mm KCl, 2.5 mm MgCl2, 10 mm Tris-HCl (pH 8.3), 0.45% Tween-20, 0.45% NP-40, 0.01% gelatin, 60 mg/ml Proteinase K). Following a 60-min incubation at 608 followed by a 15-min incubation at 958 to eliminate Proteinase K activity, the entire 2.5 ml was used as the template DNA in a PCR reaction. The PCR reactions were carried out in a total volume of 25 ml, containing the 2.5 ml of template, 25 pmol of each primer, and 1 unit of Taq Polymerase in PCR buffer (50 mm Tris-Cl, 1.5 mm MgCl2 and 0.2 mm each dNTP). The PCR parameters consisted of an initial 5 min incubation at 948, 35 cycles of 948 for 40 sec, 558 for 40 sec, and 728 for 4 min and a final 5 min incubation at 728. The positions of the 59 nucleotide [relative to the completed mel-26 sequence (GenBank accession no. U67737)] and the length and direction of the primers used for amplification or sequencing are as follows: primer-17 (419, 18, forward), primer-18 (739, 18, forward), primer-19 (1060, 18, forward), primer-14 (1402, 20, forward), primer-7 (1479, 20, forward), primer-9 (1818, 21, forward), primer-8 (2019, 20, reverse), primer-22 (2108, 23, forward), primer-23 (2908, 19, reverse), primer-13 (3243, 18, reverse), primer-24 (3453, 18, forward), primer-2 (3577, 20, reverse), primer-3 (3866, 20, forward), primer-25 (4028, 20, reverse), primer-16 (4248, 19, forward), primer-12 (4318, 18, reverse), primer-15 (4846, 18, reverse). The entire PCR reaction was loaded on a 1% agarose gel, and the product was purified from the gel using the QiaQuick purification kit (Qiagen, Chatsworth, CA). Approximately 100 ng of the gel-purified DNA was sequenced using the ABI PRISM dye terminator cycle-sequencing protocol (Perkin Elmer) and internal primers. Mutations were confirmed by sequencing an independent PCR product. Analysis of mel-26 mRNA: The MicroFast Track 2.0 kit (Invitrogen, San Diego, CA) was used to isolate poly(A)1 RNA directly from mixed-staged N2 hermaphrodites for RT-PCR. This RNA was nonspecifically reverse transcribed using a poly(dT) primer with an EcoRI adaptor [dGCT GCA GAA TTC GTC GAC (TTT)6] and Superscript II reverse transcriptase (GIBCO BRL, Gaithersburg, MD). The resulting cDNA was RNase H treated to remove the complementary mRNA. A primer complementary to the SL1 trans-splice leader sequence and a gene-specific primer (primer-8) were used in a firstround amplification according to the Superscript II protocol. A second round of PCR, starting with 10% of the first-round product, was conducted using standard PCR conditions. The second-round product was gel-purified, cloned, and sequenced. Wild-type embryos were collected using standard techniques (Sulston and Hodgkin 1988), and RNA was isolated using the method of Meyer and Casson (1986). Adult hermaphrodites carrying ts mutations affecting germline development were used to isolate RNA lacking transcripts specific to the

121 TABLE 1 Genetic analysis of mel-26 mutants Hatching embryos (%)b

Maternal genotypea 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19

ct61/1 ct61/ct61 sb451 sb45/sb45 ct61/sb45 mei-1(ct46) 1/1 1 mei-1(ct46) 1/1 ct61 mei-1(ct46) 1/1 sb45 mei-1(ct46ct82) 1/1 sb45 mei-2(ct102) 1/1 sb45 ct61sb4/1 ct61sb4/ct61sb4 nDf 23/1 nDf 23/ct61sb4 mei-1(ct46) 1/1 ct61sb4 ct61/nDf 23 ct61/ct61sb4 sb45/nDf23 sb45/ct61sb4

158

208

258

97 8 99 20 15 80 27 41 — — 98 8.1 73 2.9 — 7.1 8.3 5.1 7.1

65 1.0 71 2.6 0.7 21 1.4 3.0 — — 99 0.5 73 1.3 15 1.4 1.3 2.1 0.5

6 0 10 0 0 —c — — 94 93 98 0 70 — — 0 0 0 0

a

ct61, ct61sb4 and sb45 are alleles of mel-26. Hatching frequencies in trans to nDf 23 are corrected for the embryonic lethality due to this deficiency, whose values are given in line 13. c Not determined. b

male or female germlines. Animals raised at the nonpermissive temperature carrying a fem-1(hc17) mutation lack the male germline, whereas fem-3(q20 gf ) mutants lack a female germline and glp-1(q231) worms lack undifferentiated germ cells and contain only a few sperm. Poly(A)1 RNA was isolated from these samples using the Micro FastTrack 2.0 kit (Invitrogen). This mRNA was analyzed by Northern blot, using approximately 2 mg of mRNA per gel lane. Relative loading was later examined by comparison of actin levels in the samples (Krause et al. 1989; Krause 1995). Poly(A)1 RNA was isolated from staged gravid N2 and mel26(ct61) hermaphrodites. The relative size and abundance of the mel-26 transcript in these samples was compared by Northern blot. RESULTS

Characterization of dominant and recessive alleles of mel-26: The canonical mel-26 allele, ct61, was identified as a dominant ts maternal-effect embryonic lethal mutation (Mains et al. 1990b). In a more recent screen for dominant ts lethal alleles (Mitenko et al. 1997), we found a second mutation, sb45, whose properties are similar to ct61. Both mutations map to the unc-29 lin11 interval, and we were unable to isolate recombinants between them, which indicated a maximal two-factor distance of 0.02 cM. ct61 and sb45 are qualitatively similar. As shown in Table 1, the ts maternal-effect lethality of both mutations was more severe when homozygous (lines 1–4), and as

122


Figure 1.—Phenotypes of wild-type and mutant mel26 embryos. Nomarski photomicrographs (A, D, G, and J) illustrate the orientation of the spindle in onecell embryos from (A) wildtype, (D) mel-26(ct61), (G) mel-26(sb45), and ( J) mel26(ct61sb4) animals. Triangles denote the spindle poles. Anti-tubulin (B, E, H, and K) and anti-MEI-1 (C, F, I, and L) staining of interphase embryos demonstrate that in (B and C) wild type, MEI-1 is absent from mitotic spindles, but it is localized to the centrosomes in (E and F) mel-26(ct61), (H and I) mel-26(sb45), and (K and L) mel-26(ct61sb4) mutant embryos. The secondary antibody in C, F, H, and L was conjugated to fluorescein; in B, E, and K, the fluorochrome was rhodamine, and in I, cy3 was used. Scale bar, 10 mm.

expected for alleles of the same gene, the mutations failed to complement for this property (line 5, although interpreting this as failure to complement is somewhat ambiguous given the dominant nature of the mutations). Interactions with other genes strengthens the interpretation that ct61 and sb45 are allelic. Like ct61, sb45 was dominantly enhanced by the mei-1(ct46 gf) mutation (lines 6–8). We previously reported that mel26(ct61) was suppressed by dominant suppressors of the gf allele mei-1(ct46 ), such as the dominant-negative allele mei-1(ct46sb82) and the mutation mei-2(ct102) (Mains et al. 1990a). As indicated on lines 9 and 10 of Table 1, sb45 was also suppressed by these mutations. The similar phenotypes of ct61 and sb45 reinforce the proposition that the two mutations are allelic. Embryos from animals of either strain show defects starting with the first mitotic division. The first mitotic spindle of C. elegans is normally oriented along the anterior-posterior axis, slightly posterior of center (Figure 1A). In embryos from both ct61 and sb45 hermaphrodites, the spindle was usually perpendicular to the anterior-posterior axis, at the extreme posterior of the embryo (Figure 1, D and G). The misoriented spindles induced cleavage

furrows along the anterior-posterior axis at right angles to their normal positions. Anterior cytoplasts often formed, and cleavage furrows frequently retracted. As we had previously shown for ct61 (Clark-Maguire and Mains 1994b; Figure 1, E and F), sb45 also resulted in ectopic postmeiotic MEI-1 staining in the spindles and centrosomes (Figure 1, H and I). Thus in all respects, sb45 is qualitatively similar to ct61, and the mutations are very likely allelic. To define the null phenotype of mel-26, we identified an intragenic revertant that converted the dominant gf “poison” characteristic of ct61 into a recessive lf allele (materials and methods). One such mutation, ct61sb4, was found among approximately 12,000 mutagenized chromosomes. As expected for a lf intragenic revertant, sb4 could not be separated from ct61 by recombination, mapping in cis within 0.03 cM (materials and methods). Furthermore, ct61sb4 had neither dominant maternal nor recessive zygotic effects on embryonic viability as the progeny of heterozygous hermaphrodites showed wild-type levels of hatching at all temperatures (Table 1, line 11). ct61sb4 resulted in recessive maternal-effect lethality, showing low levels of

The mel-26 Gene of C. elegans

123

Figure 2.—Mapping and cloning of mel-26. (A) The mel-26 region of chromosome I. Genes used in genetic analysis and mapping of mel-26 alleles are shown. The extent of the deficiency nDf 23 is indicated above the line. (B) The physical map in the mel-26 region, located in the interval between the cloned genes unc-29 and lin11. Thin lines indicate the relative positions of the cosmids injected in this study. Genes and RFLPs localized to the physical map in the region are shown under the thick gray line. (C) Subclones of cosmid ZK858 used to determine the location of the mel-26 coding region. Each subclone was tested for its ability to rescue the maternal-effect lethality of mel-26(ct61sb4) as indicated by the 1 and 2 to the right of each. The mel-26 transcript, as represented by the cDNA clone, is indicated above the minimalrescuing 7.6-kb subclone, with the large and small boxes indicating translated and untranslated regions, respectively. These are shown above a thin line that corresponds to the region of this clone sequenced and deposited in GenBank as accession no. U67737.

hatching at 158 but none at 258 (Table 1, line 12). ct61sb4/ct61sb4 was similar to ct61sb4/nDf 23. (nDf 23 is a deficiency that uncovers mel-26; Figure 2A.) This again suggests the lf nature of the revertant (Table 1, lines 13 and 14). Finally, ct61sb4, unlike its parent, was not a dominant enhancer of mei-1(ct46 gf ) (Table 1, line 15; compare to line 7). If sb4 represents an intragenic lf revertant of mel-26, its behavior in trans to ct61 and sb45 should mimic that of a deficiency. Progeny of both ct61/1 and sb45/1 hermaphrodites showed decreased hatching rates with increased temperature (Table 1, lines 1 and 3), and this was exacerbated to similar extents when the 1 was

replaced with either nDf 23 or ct61sb4 (Table 1, lines 16–19). Thus, in all genetic tests, ct61sb4 behaved as a severe lf allele of mel-26. The recessive embryonic defects of ct61sb4 were identical to those of sb45 and ct61. The early cleavage spindles were misoriented, reduced in size, and contained ectopic MEI-1 (Figure 1, J–L). Because the dominant phenotypes of the gf mutations sb45 and ct61 resemble the recessive phenotypes of the lf mutation ct61sb4, sb45 and ct61 are antimorphic alleles by definition (Muller 1932). This is consistent with previous gene-dosage experiments that indicated that ct61 competed with the wild-type product (Mains et al. 1990b).

124


Molecular identification of mel-26: DNA that included mel-26(1) was identified by transformation rescue of the recessive maternal-effect lethality of ct61sb4 as described in materials and methods. mel-26 was previously mapped to the interval between the cloned genes unc-29 and lin-11 (Figure 2A); cosmids in this region were selected for analysis. Rescue was obtained with cosmid ZK858 but not with cosmids D1004, W06D4, W04H6, F25H5, or C08H1 (Figure 2, B and C). In comparison to mel-26(ct61sb4), which gave z0.01% hatching at 258, mel-26(ct61sb4) transformed with ZK858 showed hatching rates ranging from 2 to 8%. Injection of the 17.2-kb SalI fragment, the 9.7-kb Bgl II fragment, and a plasmid containing the 7.6-kb XhoI/Bgl II fragment all showed hatching above background. Two smaller fragments were not able to rescue the mutation (Figure 2C). Using the 7.6-kb XhoI/Bgl II fragment as a probe, cDNA clones were found in the early embryonic library constructed by Schauer and Wood (1990). Three overlapping clones were isolated, and the longest cDNA (1700 bp) as well as approximately 5000 bp of the rescuing genomic fragment were sequenced. The genomic sequence has been entered into GenBank as accession no. U67737. It is identical to the region around open reading frame (ORF) ZK858.4 in GenBank accession no. Z79759, which represents sequence data obtained by the C. elegans Genome Sequencing Project. The cDNA is derived from six exons, with splice donor and acceptor sites generally matching consensus (GTRART and TTTCAG, respectively; Krause 1995). Sequence comparisons revealed that the C. elegans expressed sequence tag (EST) cm01e12 (Waterston et al. 1992) represents the mel-26 gene. The first 10 nt of cm01e12 (CAAGTTTCAG) match the 39 end of the trans-splice leader SL1 (GGTTTAATTACCCAAGTTT CAG), which is spliced onto the 59 end of many C. elegans genes (Zorio et al. 1994). The mel-26 cDNAs isolated in our library screen did not contain this sequence. To confirm that mel-26 is trans-spliced, RT-PCR was conducted using one primer specific to SL1 and another located in the second exon of mel-26. A fragment of the appropriate size was cloned and sequenced and shown to represent the trans-splicing of SL1 to the 59 end of the mel-26 sequence (data not shown). A poly(A)1 addition site was not contained within the longest cDNA, so the ends of the two remaining cDNA clones were sequenced to identify the point of polyadenylation. The site of polyadenylation is different in the two clones, occurring at positions 4772 and 4800 in GenBank accession no. 67737, respectively. The AAU GAA sequence starting at 4764 is a likely candidate for the polyadenylation signal; the C. elegans consensus is AAUAAA, with 13% having a G at the fourth position (Krause 1995). Sequence analysis of mel-26: The predicted MEL-26 product is 395 amino acids. Database searches using the BLAST set of programs (Altschul et al. 1990) re-

vealed that MEL-26 shows full-length similarity to three C. elegans ORFs, C50C3.8 (III), T16H12.5 (III), and C07D10.2 (II) and the human speckle-type po xvirus and zinc finger (POZ) protein (SPOP). Pairwise sequence comparisons reveal that these sequences show 25–33% identity (39–47% similarity) to MEL-26 over their entire length. While the protein showing the highest identity to MEL-26 is SPOP, it is more likely that the protein encoded by T16H12.5 is the C. elegans homolog of SPOP (Nagai et al. 1997). MEL-26, SPOP, and the three C. elegans ORFs all contain a region of homology to the BTB/POZ motif. Originally identified in the bric a` brac, tramtrack, and Broad-Complex genes of Drosophila (Zollman et al. 1994), the BTB motif has also been described as the POZ motif (Bardwell and Treisman 1994). The BTB-containing proteins showing the highest similarity to MEL-26 are longitudinals lacking from Drosophila and the product of the T8 gene of Myxoma virus. These proteins show 33–34% identity (42–47% similarity) to MEL-26 within the BTB region (Figure 3) but share little or no similarity outside of it. Database searches using MEL-26 sequences outside of the BTB domain identified only SPOP and the three mentioned C. elegans ORFs. Sequencing of ct61sb4 revealed the presence of a stop codon at amino acid position 320 (AGA → TGA), which is not present in either the wild-type or the ct61 sequences. In sb45, a G to A transition in codon 94 (TGT → TAT) predicts a cysteine to tyrosine change (Figure 3). The molecular change present in ct61 has not been identified in sequencing the entire coding region, all introns, and approximately 1 kb of upstream genomic sequence (see discussion). Expression of mel-26 mRNA: Previous genetic analysis of mel-26 mutations indicated a strict maternal requirement for gene activity, suggesting that it is required only in the female germline. The mel-26(ct61) temperaturesensitive period begins at the one-cell stage and extends to the onset of gastrulation (approximately 2 hr postfertilization; Mains et al. 1990b). Northern blot analysis was conducted to determine the relative abundance of the mel-26 message in various tissues using strains carrying mutations affecting development of the male, female, or both germlines. A single transcript of 1700 nt, as expected from the sequence, was enriched in the female germline and embryos. The mutation glp1(q231) disrupts mitotic proliferation of the germline; the few germ cells that are produced (z15) develop into sperm cells (Austin and Kimble 1987). Even though there was more mRNA in the glp-1 sample, it is clear that the relative level of mel-26 was lower than in wild-type gravid hermaphrodites (Figure 4). Germlines of fem-1(lf ) worms are female, with oocytes but no sperm, while fem-3(gf ) show the opposite phenotype. Comparisons of the relative abundance of the mel-26 message in these strains indicated that mel-26 mRNA is expressed in the developing oocytes (Figure 4). Al-

Figure 3.—Alignment and comparison of the inferred protein sequence of mel-26 and related proteins. Alignment of the sequences was performed using the Genetic Corporation Group (Madison, WI) Pileup program; the freeware program Boxshade was used to illustrate the sequence similarity. Black and gray shading indicate regions of identity and similarity, respectively, to the MEL-26 sequence. The BTB region is underlined. The positions and codon changes of sb45 and sb4 are indicated.

126


Figure 4.—Northern blot analysis of the mel-26 mRNA. The abundance of mel-26 message in poly(A)1 RNA from wild-type embryos and gravid hermaphrodites is compared to samples obtained from adult hermaphrodites mutant for fem-1 (lacking sperm and fertilized eggs), fem-3(gf ) (lacking oocytes and fertilized embryos), and glp-1 (lacking both male and female germline). The blot was probed with the mel-26 full length cDNA and again with actin to show relative loading levels. The transcript size is 1700 nt.

though a high level of mel-26 mRNA accumulated in fem-1(lf ) animals, no mel-26 message was evident in the fem-3(gf ) sample. The message is present in fertilized embryos, though the abundance of the message may be decreasing relative to gravid hermaphrodites. There is no indication that the mel-26(ct61) allele has any effect on the size or abundance of the mel-26 message (data not shown). DISCUSSION

Characterization of the mei-1 gene of C. elegans has demonstrated that it is an essential component of the spindle during oocyte meiosis (Clark-Maguire and Mains 1994a,b). Absence of mei-1 leads to the failure of meiotic spindle formation, whereas the abnormal persistence of mei-1 activity into the subsequent mitotic divisions disrupts the proper functioning of the mitotic spindles. Therefore, mei-1 must function at meiosis and be inactivated before the first mitotic cleavage. Genetic analysis of the mel-26 gene indicates that it acts as the postmeiotic negative regulator of mei-1, ensuring that mei-1 activity does not interfere with mitotic divisions. To better understand the role of mei-1 and mel-26 in meiotic cell division, we undertook the cloning and further genetic characterization of the mel-26 gene. The interpretation that mel-26 represents a postmeiotic inhibitor of mei-1 was complicated by the gf nature of the original mel-26 allele, ct61. The genetic work we describe in this article confirms the interpretation that mel-26(1) does indeed inhibit mei-1 function. We identified a severe lf mutation of mel-26, ct61sb4, as an intragenic revertant of ct61, and we also describe a new dominant allele mel-26(sb45). The behaviors of the dominant

mutations in trans to either a deficiency or ct61sb4 were very similar, indicating that ct61sb4 is near null (Table 1). The phenotypes of the dominant mutations are virtually identical to the recessive defects of ct61sb4, indicating that ct61 and sb45 are antimorphs (Muller 1932). All three mutations are defective in preventing ectopic MEI-1 expression in the mitotic spindles and asters of the newly fertilized embryo, which consequently results in small, misoriented spindles (Figure 1). Therefore, the genetic properties of mel-26 are consistent with the interpretation that mel-26 wild-type product inhibits postmeiotic mei-1 activity. We cloned the mel-26 gene by transformation rescue of the recessive maternal-effect lethality of mel-26(ct61sb4). One coding region and its corresponding cDNAs were identified in the minimal rescuing fragment. Northern blot analysis of this transcript indicates that it is most highly expressed in the female germline, consistent with genetic analysis of mel-26, which indicated only maternal gene function. The predicted product of the mel-26 gene is a member of a family of related C. elegans genes, showing full-length similarity to the deduced products of three genes predicted by the C. elegans Genomic Sequencing Project, T16H12.5, C07D10.2, and C50C3.8 (Figure 3). (These ORFs are not located near any genes known to interact with mel-26 or mei-1.) The recently identified SPOP—identified as the major antigen in an autoimmune serum generating a speckled nuclear pattern by indirect immunofluorescent staining—also shows full-length similarity to MEL26; however, the product of the ORF T16H12.5, rather than MEL-26, likely represents the C. elegans homolog of SPOP, as extended regions of amino acid identity have been noted throughout the protein (Nagai et al. 1997). MEL-26, SPOP, and the three C. elegans ORFs all contain a motif with some similarity to the BTB motif (Zollman et al. 1994). As mentioned in results, this motif was first identified in the products of the Drosophila transcription factor genes bric a` brac, tramtrack, and Broad-Complex (Zollman et al. 1994), but this domain has since been found in several other types of proteins and in a number of organisms (Albagli et al. 1995). Many of the proteins seem to interact with DNA, or actin, though the role of the BTB motif in this binding is unclear. Chen et al. (1995) suggested that the BTB domain of the bric a` brac protein forms an a-helical structure that mediates dimerization. They used sitedirected mutagenesis of residues within the BTB domain to demonstrate a sequence-specific interaction between molecules. It was also shown that mutation of two charged residues within the BTB domain abolished interaction with a wild-type BTB domain; however, the double mutant was able to bind to a second double mutant BTB domain. The two antimorphic alleles of mel-26 interfere with

The mel-26 Gene of C. elegans

the function of the wild-type allele. One simple model to explain this observation is that MEL-26 normally forms a multimeric complex, mediated by the BTB domain, and incorporation of mutant MEL-26 inactivates the complex. The strong lf allele ct61sb4 was isolated as a revertant of the dominant-negative ct61 allele and results in the loss of the C-terminal 75 amino acids; this may indicate that the C-terminal region of the protein, which does not include the BTB domain, plays a role in the formation or function of the putative multimeric complex. It is notable that ct61sb4 was found at the relatively low frequency of 1/12,000 mutagenized chromosomes. The forward mutation rate after standard EMS mutagenesis in C. elegans is usually in the range of 1/1000 to 1/5000 (Brenner 1974; Anderson and Brenner 1984; Park and Horvitz 1986; Rogalski and Riddle 1988). The two antimorphic alleles, ct61 and sb45, were found among 24,000 mutagenized chromosomes (Mains et al. 1990b; Mitenko et al. 1997), the same frequency as sb4. This finding may indicate that mel-26 has as many EMSmutable sites disposing it to mutate to antimorphic as to null alleles. We have been unable to identify the molecular lesion responsible for the ct61 allele, although we have sequenced the entire coding region of the gene, all introns, and 1 kb of promoter sequence. We have also compared the size and abundance of the mel-26 transcript and looked for the presence of genomic rearrangements in a ct61 background and have found no changes relative to wild type (data not shown). There are at least three possible explanations for this failure to find the mutation. First, it is formally possible that there are two closely linked loci that interact genetically [i.e., mel-26(ct61) and sup(sb4)] and that we have cloned sup(sb4) but not mel-26. sb45 results in a sequence change in the same gene as sb4 and would hence be an allele of sup(sb4) rather than mel-26. This scenario is unlikely, however. ct61 and sb45 are very similar to one another in that both are dominant ts maternal-effect lethal mutations and result in similar mitotic defects. Furthermore, both mutations respond in the same fashion when in trans to either ct61sb4 or a deficiency of the region, both ct61 and sb45 enhance mei-1(gf ), and both are suppressed by mei-1 suppressors (Table 1). In addition, we were unable to isolate crossovers between ct61 and either sb4 or sb45. Thus, if ct61 represents a gene separate from sb4 or sb45, it must be within 0.03 cM of these mutations (see materials and methods). Based on the estimates of Barnes et al. (1995), this genetic distance would correspond to a region of approximately 45 kb or less. We examined the C. elegans Genome Sequencing Consortium’s sequence of cosmids in this region for duplications of the mel-26 sequence, and for other candidates for interacting genes, but found none. The second possibility for our failure to find the ct61 lesion is that the ct61 allele is due to a promoter muta-

127

tion 59 of the region we sequenced. A precedent exists in C. elegans for dominant ts promoter mutations (Perry et al. 1994); however, these mutations were hypermorphic, whereas ct61 is antimorphic. The third possibility is that there is an additional exon, which may be used in a class of transcripts, that has not been identified. We are currently addressing these possibilities. Characterization of mel-26 suggests that the ct61sb4 allele is a null by both genetic and molecular criteria. At the same time, mel-26(ct61sb4) homozygotes exhibit a low, but not zero, level of hatching at 158. This may represent residual gene activity. Alternatively, it may indicate that mel-26 is not absolutely essential at lower temperatures. This possibility, in turn, could indicate that the temperature sensitivity of the ct61 and sb45 alleles does not reflect a temperature-sensitive change in the mutant mel-26 protein products but rather the disruption of a process that is innately temperaturesensitive—that is, the elimination of mei-1(1) activity. It is unclear how MEL-26 acts to inhibit MEI-1 activity following meiosis. Other than the BTB domain, there are no known motifs in the predicted protein to suggest a biological activity. Future analysis of the cellular distribution of MEL-26 may provide some clues. We thank Debra Zimmerman, Tom Clandinin, and Nuhzat Averill for early work on this project and members of the McGhee and Mains laboratories for many useful discussions. We also thank Mark Edgley and Ann Rose for providing strains. Some strains were obtained from the Caenorhabditis Genetics Center, funded by the National Institutes of Health Center for Research Resources. This work was supported by studentships from the Natural Sciences and Engineering Research Council of Canada and the Alberta Heritage Foundation for Medical Research to M.R.D. and by grants from the Medical Research Council of Canada and the Alberta Heritage Foundation for Medical Research to P.E.M.

LITERATURE CITED Albagli, O., P. Dhordain, C. Deweindt, G. Lecocq and D. Leprince, 1995 The BTB/POZ domain: a new protein-protein interaction motif common to DNA- and actin-binding proteins. Cell Growth Differ. 6: 1193–1198. Albertson, D. G., 1984 Formation of the first cleavage spindle in nematode embryos. Dev. Biol. 101: 61–72. Albertson, D. G., and J. N. Thomson, 1993 Segregation of holocentric chromosomes at meiosis in the nematode, Caenorhabditis elegans. Chromosome Res. 1: 15–26. Altschul, S. F., W. Gish, W. Miller, E. W. Myers and D. J. Lipman, 1990 Basic local alignment search tool. J. Mol. Biol. 215: 403– 410. Anderson, P., and S. Brenner, 1984 A selection for myosin heavy chain mutants in the nematode Caenorhabditis elegans. Proc. Natl. Acad. Sci. USA 81: 4470–4474. Austin, J., and J. Kimble, 1987 glp-1 is required in the germ line for regulation of the decision between mitosis and meiosis in C. elegans. Cell 51: 589–599. Bardwell, V. J., and R. Treisman, 1994 The POZ domain: a conserved protein-protein interaction motif. Genes Dev. 8: 1664– 1677. Barnes, T. M., Y. Kohara, A. Coulson and S. Hekimi, 1995 Meiotic recombination, noncoding DNA and genomic organization in Caenorhabditis elegans. Genetics 141: 159–179. Barstead, R. J., and R. H. Waterston, 1989 The basal component

128


of the nematode dense-body is vinculin. J. Biol. Chem. 264: 10177–10185. Brenner, S., 1974 The genetics of Caenorhabditis elegans. Genetics 77: 71–94. Chen, W., S. Zollman, J. L. Couderc and F. A. Laski, 1995 The BTB domain of bric a brac mediates dimerization in vitro. Mol. Cell Biol. 15: 3424–3429. Clark-Maguire, S., and P. E. Mains, 1994a mei-1, a gene required for meiotic spindle formation in Caenorhabditis elegans, is a member of a family of ATPases. Genetics 136: 533–546. Clark-Maguire, S., and P. E. Mains, 1994b Localization of the mei1 gene product of Caenorhabditis elegans, a meiotic-specific spindle component. J. Cell Biol. 126: 199–209. Confalonieri, F., and M. Duguet, 1995 A 200-amino acid ATPase module in search of a basic function. BioEssays 17: 639–650. Edgley, M. L., D. L. Baillie, D. L. Riddle and A. M. Rose, 1995 Genetic balancers, pp. 147–184 in Caenorhabditis elegans: Modern Biological Analysis of an Organism, edited by H. F. Epstein and D. C. Shakes. Academic Press, Inc., San Diego. Gard, D., 1992 Microtubule organization during maturation of Xenopus oocytes: assembly and rotation of the meiotic spindles. Dev. Biol. 151: 516–530. Hirsh, D., D. Oppenheim and M. Klass, 1976 Development of the reproductive system of Caenorhabditis elegans. Dev. Biol. 49: 200– 219. Horvitz, H. R., S. Brenner, J. Hodgkin and R. K. Herman, 1979 A uniform genetic nomenclature for the nematode Caenorhabditis elegans. Mol. Gen. Genet. 175: 129–133. Hyman, A. A., and E. Karsenti, 1996 Morphogenetic properties of microtubules and mitotic spindle assembly. Cell 84: 401–410. Kemphues, K. J., N. Wolf, W. B. Wood and D. Hirsh, 1986 Two loci required for cytoplasmic organization in early embryos of Caenorhabditis elegans. Dev. Biol. 113: 449–460. Krause, M., M. Wild, B. Rosenzweig and D. Hirsh, 1989 Wildtype and mutant actin genes in Caenorhabditis elegans. J. Mol. Biol. 208: 381–392. Krause, M., 1995 Transcription and translation, pp. 483–512 in Caenorhabditis elegans: Modern Biological Analysis of an Organism, edited by H. F. Epstein and D. C. Shakes. Academic Press, Inc., San Diego. Mains, P. E., K. J. Kemphues, S. A. Sprunger, I. A. Sulston and W. B. Wood, 1990a Mutations affecting the meiotic and mitotic divisions of the early Caenorhabditis elegans embryo. Genetics 126: 593–605. Mains, P. E., I. A. Sulston and W. B. Wood, 1990b Dominant maternal-effect mutations causing embryonic lethality in Caenorhabditis elegans. Genetics 125: 351–369. Matthies, H. J. G., H. B. Mcdonald, L. S. B. Goldstein and W. E. Theurkauf, 1996 Anastral meiotic spindle morphogenesis: Role of the non-claret disjunctional kinesin-like protein. J. Cell Biol. 134: 455–464. Mello, C. C., J. M. Kramer, D. Stinchcomb and V. Ambros, 1991 Efficient gene transfer in C. elegans: extrachromosomal maintenance and integration of transforming sequences. EMBO J. 10: 3959–3970. Merdes, A., and D. W. Cleveland, 1997 Pathways of spindle pole formation: different mechanisms; conserved components. J. Cell Biol. 138: 953–956. Meyer, B. J., and L. P. Casson, 1986 C. elegans compensates for the difference in X chromosome dosage between the sexes by regulating transcript levels. Cell 47: 871–881.

Mitenko, N. L., J. R. Eisner, J. R. Swiston and P. E. Mains, 1997 A limited number of Caenorhabditis elegans genes are readily mutable to dominant, temperature-sensitive maternal-effect embryonic lethality. Genetics 147: 1665–1674. Muller, H. J., 1932 Further studies on the nature and causes of gene mutations. Proc. 6th Int. Congr. Genet. 1: 213–255. Nagai, Y., T. Kojima, Y. Muro, T. Hachiya, Y. Nishizawa et al., 1997 Identification of a novel nuclear speckle-type protein, SPOP. FEBS Let. 418: 23–26. Palacios, M. J., H. C. Joshi, C. Simerly and G. Schatten, 1993 g-tubulin reorganization during mouse fertilization and early development. J. Cell Sci. 104: 383–389. Park, E.-C., and H. R. Horvitz, 1986 Mutations with dominant effects on the behavior and morphology of the nematode Caenorhabditis elegans. Genetics 113: 821–852. Perry, M. D., C. Trent, B. Robertson, C. Chamblin and W. B. Wood, 1994 Sequenced alleles of the Caenorhabditis elegans sexdetermination gene her-1 include a novel class of conditional promoter mutations. Genetics 138: 317–327. Piperno, G., and M. T. Fuller, 1985 Monoclonal antibodies specific for an acetylated form of a-tubulin recognize the antigen in cilia and flagella from a variety of organisms. J. Cell Biol. 101: 2085–2094. Rogalski, T. M., and D. L. Riddle, 1988 A Caenorhabditis elegans RNA polymerase II gene, ama-1 IV, and nearby essential genes. Genetics 118: 61–74. Sawada, T., and G. Schatten, 1988 Microtubules in ascidian eggs during meiosis, fertilization, and mitosis. Cell Motil. Cytoskel. 9: 219–230. Schatten, G., 1994 The centrosome and its mode of inheritance: the reduction of the centrosome during gametogenesis and its restoration during fertilization. Dev. Biol. 165: 299-335. Schatten, G., C. Simerly and H. Schatten, 1985 Microtubule configurations during fertilization, mitosis, and early development in the mouse and the requirement for egg microtubulemediated motility during mammalian fertilization. Proc. Nat. Acad. Sci. USA 82: 4152–4156. Schauer, I. E., and W. B. Wood, 1990 Early C. elegans embryos are transcriptionally active. Develop. 110: 1303–1317. Sulston, J., and J. Hodgkin, 1988 Methods, pp. 587–606 in The Nematode Caenorhabditis elegans, edited by W. B. Wood. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY. Theurkauf, W. E., and R. S. Hawley, 1992 Meiotic spindle assembly in Drosophila females: behavior of nonexchange chromosomes and the effects of mutations in the nod kinesis-like protein. J. Cell Biol. 116: 1167–1180. Waters, J. C., and E. D. Salmon, 1997 Pathways of spindle assembly. Curr. Opin. Cell Biol. 9: 37–43. Waterston, R., C. Martin, M. Craxton, C. Huynh, A. Coulson et al., 1992 A survey of expressed genes in Caenorhabditis elegans. Nat. Genet. 1: 114–123. Zollman, S., D. Godt, G. G. Prive, J. L. Couderc and F. A. Laski, 1994 The BTB domain, found primarily in zinc finger proteins, defines an evolutionarily conserved family that includes several developmentally regulated genes in Drosophila. Proc. Natl. Acad. Sci. USA 91: 10717–10721. Zorio, D. A. R., N. N. Cheng, T. Blumenthal and J. Spieth, 1994 Operons as a common form of chromosomal organization in C. elegans. Nature 372: 270–272. Communicating editor: R. K. Herman