Genetics, Vol. 165, 145-157, September 2003, Copyright © 2003

The Caenorhabditis elegans spe-39 Gene Is Required for Intracellular Membrane Reorganization During Spermatogenesis

Guang-dan Zhua and Steven W. L'Hernaulta
a Program in Genetics and Molecular Biology, Graduate Division of Biological and Biomedical Sciences and Department of Biology, Emory University, Atlanta, Georgia 30322

Corresponding author: Steven W. L'Hernault, Emory University, 1510 Clifton Rd., Atlanta, GA 30322., bioslh{at}biology.emory.edu (E-mail)

Communicating editor: B. J. MEYER


*  ABSTRACT
*TOP
 *ABSTRACT
*MATERIALS AND METHODS
 *RESULTS
 *DISCUSSION
 *LITERATURE CITED

Caenorhabditis elegans spermatid formation involves asymmetric partitioning of cytoplasm during the second meiotic division. This process is mediated by specialized ER/Golgi-derived fibrous body-membranous organelles (FB-MOs), which have a fibrous body (FB) composed of bundled major sperm protein filaments and a vesicular membranous organelle (MO). spe-39 mutant spermatocytes complete meiosis but do not usually form spermatids. Ultrastructural examination of spe-39 spermatocytes reveals that MOs are absent, while FBs are disorganized and not surrounded by the membrane envelope usually observed in wild type. Instead, spe-39 spermatocytes contain many small vesicles with internal membranes, suggesting they are related to MOs. The spe-39 gene was identified and it encodes a novel hydrophilic protein. Immunofluorescence with a specific SPE-39 antiserum reveals that it is distributed through much of the cytoplasm and not specifically associated with FB-MOs in spermatocytes and spermatids. The spe-39 gene has orthologs in Drosophila melanogaster and humans but no homolog was identified in the yeast genome. This suggests that the specialized membrane biogenesis steps that occur during C. elegans spermatogenesis are part of a conserved process that requires SPE-39 homologs in other metazoan cell types.

SPERMATOGENESIS within the testes of multicellular animals is characterized by meiosis that is coordinated with the dramatic reorganization of cytoplasm. The spermatids that result from this process contain a greatly reduced volume of cytoplasm and various specialized membranes and organelles. In the nematode Caenorhabditis elegans, dramatic asymmetric cytoplasmic partitioning accompanies meiosis II, which occurs as spermatids bud from the anucleate residual body (reviewed by L'HERNAULT 1997 Down). During budding, each spermatid receives a haploid nucleus, mitochondria, and specialized organelles called fibrous body-membranous organelle (FB-MO) complexes, while cellular constituents not required for sperm function are placed into the residual body. The FB-MOs are bipartite organelles that contain many membrane and soluble proteins required for spermiogenesis, the process by which spermatids become spermatozoa (ROBERTS et al. 1986 Down).

The FB-MOs are derived from the Golgi apparatus during spermatogenesis and each spermatocyte has many individual FB-MOs (WOLF et al. 1978 Down; ROBERTS et al. 1986 Down). All FB-MOs within a spermatocyte are ultrastructurally similar and their subsequent development appears to be temporally coordinated. Since sperm can be isolated in large quantities and spermatogenesis can be studied using simple in vitro culture conditions (WARD et al. 1981 Down; L'HERNAULT and ROBERTS 1995 Down; MACHACA et al. 1996 Down), it is possible to perform detailed cell biological and biochemical analyses of FB-MO development. FB-MOs are readily visualized by light microscopic techniques and perturbation of their morphogenesis can result in a failure to form spermatids. Such obvious alterations in spermatogenesis can be readily identified, and several mutants that seem to specifically affect FB-MO morphogenesis have been identified. Although FB-MOs are unusual-appearing structures, they have a number of attributes that facilitate study of vesicular transport pathways.

Most previously described mutants affecting FB-MO morphogenesis are sperm-specific in their phenotypic defects, and the cloned members of this group exhibit testis-specific transcription (L'HERNAULT and ARDUENGO 1992 Down; L'HERNAULT et al. 1993 Down; ACHANZAR and WARD 1997 Down). Mutants of two previously described genes, spe-4 (L'HERNAULT and ARDUENGO 1992 Down; ARDUENGO et al. 1998 Down) and spe-5 (MACHACA and L'HERNAULT 1997 Down), show gross defects during early stages of FB-MO morphogenesis. In this article, we present phenotypic and molecular characterization of the spe-39 gene, which is also involved in an early stage of FB-MO morphogenesis. Unlike any previously described mutant, spe-39 mutant spermatocytes never form the MO portion of the FB-MO. Instead, spe-39 mutant spermatocytes accumulate large numbers of vesicles in their cytoplasm. Consistent with prior work, spe-39 mutant spermatocytes usually do not form functional spermatids because this requires normal FB-MOs. This defect causes spe-39 mutants to accumulate terminal spermatocytes that typically contain four haploid nuclei. spe-39 encodes a novel protein and there are orthologs in both the Drosophila melanogaster and the human genomes but no obvious homolog is found in the yeast genome. Although spermatogenesis defects are the most evident zygotic phenotype of spe-39 mutants, its expression is not limited to the testis, suggesting that it has functions outside spermatogenesis.


*  MATERIALS AND METHODS
*TOP
 *ABSTRACT
 *MATERIALS AND METHODS
*RESULTS
 *DISCUSSION
 *LITERATURE CITED

Strains:
C. elegans handling was performed by standard methods (BRENNER 1974 Down) and the wild-type reference strain was var. Bristol (N2). Light microscopic and electron microscopic examination of spermatogenesis in dpy-11(e224) V (BRENNER 1974 Down; BAIRD and EMMONS 1990 Down; KO and CHOW 2002 Down) reveals no detectable defects (Fig 1 and Fig 2), so it was used as a tightly linked (~0.4 cM) morphological marker for spe-39. Other genes and mutations used in this study were as follows: unc-101(sy108) I (LEE et al. 1994 Down); fem-1(hc17ts) IV (NELSON et al. 1978 Down); fem-3(q23gf) IV (BARTON et al. 1987 Down); spe-39(tx12), (eb9), (eb110), and (eb111) V; unc-68(e540) V (BRENNER 1974 Down; MARYON et al. 1996 Down); unc-42(e270) V (BRENNER 1974 Down; BARAN et al. 1999 Down); him-5(e1490) V (HODGKIN et al. 1979 Down); and unc-76(e911) V (HEDGECOCK et al. 1987 Down; BLOOM and HORVITZ 1997 Down). The chromosome V deficiencies nDf18, sDf35 (MCKIM et al. 1988 Down), nDf32 (PARK and HORVITZ 1986 Down), and sDf20 (ROSENBLUTH et al. 1985 Down) were used for mapping. The translocations nT1[unc-?(n754) let-?](IV;V) (also called DnT1; FERGUSON and HORVITZ 1985 Down; TREININ and CHALFIE 1995 Down) and eT1[him-5(e1467)](III;V) (I. GREENWALD and H. R. HORVITZ, unpublished data) were used to balance spe-39 mutations. Except for spe-39 mutants, all strains were obtained from Caenorhabditis Genetics Center (University of Minnesota, Minneapolis).



View larger version (113K):
In this window
In a new window
Download PPT slide
  		Figure 1. Light microscopic phenotypes of wild-type and spe-39 sperm. dpy-11 males were used as the wild-type control for spermatogenesis. (A–C) Dissections of male gonads. (A) The dissected control dpy-11 male gonad released a few spermatocytes of various stages (arrowheads) and a large number of spermatids (arrows). The gonads of dpy-11 spe-39(eb9) (B) and dpy-11 spe-39(tx12) (C) males produced far fewer cells, including some that were spermatid-like (arrows) and some that were arrested at the spermatocyte stage (arrowheads in B, C, and C1). C1 is a magnified view of the arrested spermatocytes in C. (D–H) Stages of wild-type spermatogenesis in dpy-11 males. The DAPI staining of the nuclei (I–M) is shown below the corresponding DIC images (D–H). Meiosis begins with a 4N spermatocyte (D and I). After meiosis I, the spermatocyte can remain undivided (E and J) or divide into two secondary spermatocytes (one shown in F and K), either of which can progress to the budding stage. The dashed line in E outlines the position of a bud that is out of the focal plane. During completion of meiosis II, the condensed, haploid nuclei (J and K) segregate to the budding spermatids (G and L), and what was the central part of the spermatocyte becomes the residual body (E and F). Wild-type spermatids have centrally placed nuclei (G). Activation of a spermatid with Pronase transforms it into a functional spermatozoon (H) that uses its single pseudopod (ps) for crawling motility. (N–Q) DIC images in spe-39 mutants, with corresponding DAPI images (R–T for N–P). N and O are arrested dpy-11 spe-39(eb9) spermatocytes, showing phenotypes also found in the spe-39(tx12) mutant. The arrested spermatocytes either fail to bud spermatids (N) or form abnormally small blebs (O) containing nuclei (S). There are more than four DAPI staining spots (T) in the arrested dpy-11spe-39(tx12) spermatocyte shown in P. (Q) A rarely observed dpy-11 spe-39(eb9) spermatid that has extended projections (arrow) after treatment with Pronase. nu, nucleus; ps, pseudopod; rb, residual body. Bars, 40 µm (A–C and C1) and 5 µm (D–H and N–Q).



View larger version (153K):
In this window
In a new window
Download PPT slide
  		Figure 2. Ultrastructural examination of spermatogenesis. Males of the following genotypes were examined by EM: dpy-11 (A, A1, A2, and B; used as wild-type control), dpy-11 spe-39(eb9) (C, D, and E), and dpy-11 spe-39(tx12) (F, G, H, and I). (A) In wild type, spermatids budding off from the residual body. A condensed nucleus, mitochondria, and fibrous bodies are segregated to a budding spermatid. The fibrous bodies are composed of well-organized fibers (A1) and are surrounded by two layers of membranes (arrow in A2). (B) A wild-type spermatid undergoing activation, showing a smooth layer of perinuclear material (arrowhead), a membranous organelle (mo) ready to fuse with the plasma membrane, the spot of a successful MO fusion (arrow in B), and the extending pseudopod (p). (C–E, G, and H) Phenotypes of spe-39(eb9) and (tx12) arrested spermatocytes that are shared by both mutants include naked FBs (C), the improper segregation of fibrous bodies to the residual body (left arrow in D), a large number of small vesicles (v in E), and disorganized perinuclear material (arrows in G). (H) Magnified view of the small vesicles, showing they are bounded by two layers of membranes (arrows). (F) A large fibrous body in spe-39(tx12) composed of fibers that are not bundled in parallel. (I) Thick-coated vacuoles (tcv), in addition to small vesicles (v), are found in spe-39(tx12) spermatocytes. fb, fibrous body; m, mitochondrion; mo, membranous organelle; n, nucleus; p, pseudopod; rb, residual body; s, spermatid; tcv, thick-coated vacuole; v, vesicle. Bars, 1 µm (A–G and I) and 0.5 µm (H).

The spe-39 mutant alleles tx12 (provided by D. Shakes), eb110, and eb111 were induced by ethyl methanesulfonate (EMS), while spe-39(eb9) was induced with ultraviolet irradiation (provided by J. Ahringer). New spe-39 mutations were induced in cis to dpy-11 by treating dpy-11/+ males with 50 mM EMS for 4 hr (BRENNER 1974 Down). The mutagenized males were mated with dpy-11 spe-39(eb9) unc-42 hermaphrodites and 4000 F1 non-Unc Dpy hermaphrodites were picked to separate plates. Self-sterile hermaphrodites that laid oocytes were identified and crossed with dpy-11 him-5/eT1(him-5) males to create dpy-11 spe-39(candidate)/eT1(him-5). PCR was used to amplify the spe-39 gene from homozygous mutant candidates and this resulted in identification of spe-39(eb110) and (eb111).

Genetic mapping:
In the broods of dpy-11 unc-76/spe-39(eb9) hermaphrodites, 33 of 35 Unc non-Dpy recombinants were Spe, while 37 of 41 Dpy non-Unc recombinants were not Spe. Likewise, among the progeny of dpy-11 unc-42/spe-39(eb9) hermaphrodites, 16 of 18 Unc non-Dpy recombinants were Spe, while only 2 of 14 Dpy non-Unc recombinants were Spe. These data placed spe-39 between dpy-11 and unc-42 and very close to dpy-11.

Light and electron microscopy:
L4 males were picked and raised at 20° in the absence of hermaphrodites for 2 or 3 days. These males were hand-dissected in SM buffer, pH 7.0, adjusted to an osmolarity of 230 with dextrose (MACHACA et al. 1996 Down) and analyzed as described previously (NELSON and WARD 1980 Down); SM is composed of 5 mM HEPES, 1 mM MgSO4, 25 mM KCl, 45 mM NaCl, and 5 mM CaCl2. Spermatids were activated in vitro by adding 0.2 mg/ml Pronase (EMD Biosciences, San Diego) to the dissection medium (WARD et al. 1983 Down). Sperm cytology was examined by differential interference contrast (DIC) through an Olympus BX60 microscope. One µg/ml of 4',6-diamidino-2-phenylindole (DAPI) was added to the dissection medium and the nuclear morphology was visualized under epi-illumination through an appropriate filter. Images were captured with a SensiCam high performance digital camera (Cooke Corp., Auburn Hills, MI) and composed with Image-Pro Plus 4.1 (Media Cybernetics, Silver Spring, MD), VolumeScan 3.0 (VayTek, Fairfield, IA), and Canvas 8 (Deneba Systems, Miami) software.

Samples were prepared for transmission electron microscopy as previously described (SHAKES and WARD 1989A Down) except that dissected worms were embedded in LX112 resin (Ladd Research Industries, Burlington, VT) for thin sectioning.

Nucleic acid methods:
DNA manipulation was performed using standard techniques as described by SAMBROOK et al. 1989 Down. Wizard Plus SV Minipreps DNA purification system (Promega, Madison, WI) was used for plasmid purification. QIAfilter Plasmid Maxi kit (QIAGEN, Valencia, CA) was used for purifying cosmids. DNA was recovered from agarose gels using either the QIAquick gel extraction kit (QIAGEN) or the Geneclean II kit (BIO 101, Vista, CA). The DNA Sequencing and Synthesis Facility at Iowa State University, Ames, Iowa, performed DNA sequencing.

Transgenic complementation:
DNA microinjection was carried out essentially as described previously (MELLO et al. 1991 Down; MELLO and FIRE 1995 Down). Transgenic rescue of the spe-39 spermatogenesis-defective phenotype was used to clone the gene. A group of cosmids that included spe-39 was identified and the interval that contained the gene was located to an individual cosmid and, subsequently, to restriction fragments derived from the rescuing cosmid. Bacterial strains carrying the cosmids were obtained from the Sanger Centre, Cambridge, UK. pPD118.20 (L3785; A. FIRE, S. XU, J. AHNN and G. SEYDOUX, personal communication), which has a chimeric gfp (green fluorescent protein) construct driven by the myo-3 promoter, was included in the injection mixes and used as a marker for identifying transformed animals. Transgenes containing pPD118.20 expressed GFP in the body-wall muscle, which was scored through an Olympus dissecting microscope equipped with an epi-illuminator. dpy-11 spe-39(eb9)/nT1[unc-?(n754) let-?] hermaphrodites were used as the recipients for microinjection, and fertility of transgene-bearing F1 dpy-11 homozygotes was checked to determine whether the transgene included a wild-type copy of spe-39. The injection mixes were buffered with 10 mM Tris-Cl (pH 7.5), 1 mM EDTA, and had a total DNA concentration of 100–200 ng/µl. The mass ratio of testing DNA and pPD118.20 was 1:1–3. Initial experiments indicated that transgenes containing wild-type spe-39 exhibited unstable expression (data not shown). Consequently, injection mixes used for subsequent transformation experiments included ~50% cosmid DNA that lacked spe-39 to increase the sequence complexity of the resulting transgenes. In the germline, complex contextual sequence can cause an increase in the expression of transgenes that would otherwise be epigenetically silenced (KELLY et al. 1997 Down).

cDNA clones:
spe-39 cDNA clones yk398a9, yk499e2, and yk504f7 were obtained from Y. Kohara (National Institute of Genetics, Nishima, Japan) as a {lambda}ZAPII (Stratagene, La Jolla, CA) phage suspension and amplified on lawns of PLK-F' cells grown on LB agar plates supplemented with 2 mM MgSO4 and 0.2% (w/v) glucose. pBluescript clones were excised from {lambda}ZAPII clones in vivo in XL1-Blue cells coinfected with helper phage R408 (RUSSEL et al. 1986 Down).

Antiserum preparation:
Preimmune sera were evaluated by both sperm immunofluorescence and immunoblots of whole-worm lysates (not shown), and rabbits that did not show immunoreactivity in either test were selected for immunization. Zymed Laboratories, South San Francisco, performed peptide syntheses and rabbit immunizations. Several peptides were used for immunization and N8 was found to be the most sensitive antiserum in immunohistochemical studies. Antiserum N8 was raised against a 21-mer peptide that contains residues 4–23 (see Fig 4A) of the deduced SPE-39 sequence plus an N-terminal cysteine that allowed conjugation to keyhole limpet hemocyanin. N8 was purified by its affinity to the same peptide used for immunization. The peptide was coupled directly to the matrix supplied with the SulfoLink kit (Pierce, Rockford, IL) via its N-terminal cysteine and ~2 mg of affinity-purified antibody was recovered from 7 ml of serum.



View larger version (20K):
In this window
In a new window
Download PPT slide
  		Figure 3. Cloning of spe-39. spe-39 was assigned to LG V and further mapped to the genetic interval between the right break points of two deficiencies, nDf32 and nDf18. Terminal arrows point to additional sequences outside the scope of this diagram. Deficiencies are illustrated with lines indicating the deleted regions. Cosmid clones in the region (obtained from the publicly accessible genomic library of Sanger Centre, Cambridge, UK) were tested for their ability to complement spe-39(eb9) by DNA transformation, and ZC404 was found to rescue the self-sterile phenotype of spe-39(eb9). Restriction fragments of ZC404 were tested for spe-39 rescue (indicated by + or -) and the ORF ZC404.3a was identified as the most likely candidate. DNA sequencing revealed that ZC404.3a had a mutation in each of the four spe-39 mutants (see text). The cDNA clone yk504f7 was fully sequenced and found to have the intact 3'-UTR and the complete coding region. CEL04E9 is an EST containing the intact 5'-UTR, including the trans-spliced leader SL1. WormBase (http://www.wormbase.org) describes another ORF, ZC404.3b (not shown), which is an alternatively spliced product of the same gene supported by one cDNA clone, yk887d02 (Y. KOHARA, National Institute of Genetics, Nishima, Japan). This is probably an aberrant splicing event without biological significance, and no spe-39 mutations were found in the putative coding region of this transcript.



View larger version (98K):
In this window
In a new window
Download PPT slide
  		Figure 4. SPE-39 and homologs. The sequences were aligned by the ClustalX program (THOMPSON et al. 1997 Down). (A) Alignment of SPE-39, CG18112 (fruit fly), and FLJ12707 (human). The identical and similar residues are shaded with black and gray, respectively. CG18112 protein sequence is a genomic annotation (GenBank accession no. AAF56946.1). The FLJ12707 protein sequence (GenBank accession no. NP_071350.2) is a National Center for Biotechnology Information (NCBI) predicted reference sequence (RefSeq) based on cDNA data. A line on top of the sequence indicates a spe-39 encoded peptide that was used to prepare antiserum N8. (B) Alignment of the conserved region, using homologous sequences from 10 organisms. The C. elegans (Ce), fruit fly (Dm, D. melanogaster), and human (Hs, Homo sapiens) sequences are from the same sources as in A. The mouse (Mm, Mus musculus) sequence is part of the predicted NCBI RefSeq of the gene AI413782 based on cDNA data (GenBank accession no. NP_598805.1). The other six sequences are from translations of the following ESTs: 17000687311160 (GenBank accession no. BM642896) from African malaria mosquito (Ag, Anopheles gambiae), BP008968 (GenBank accession no. BP008968) from sea squirt (Ci, Ciona intestinalis), fb64f10.y1 (GenBank accession no. AI497371) from zebrafish (Dr, Danio rerio), BJ006387 (GenBank accession no. BJ006387) from Japanese rice fish (Ol, Oryzias latipes), df12d06.y1 (GenBank accession no. BF611820) from African clawed frog (Xl, Xenopus laevis), and AJ444569 (GenBank accession no. AJ444569) from chicken (Gg, Gallus gallus). Positions of amino acids are indicated if the complete protein sequence is available. The motif deduced from this alignment is shown on top of the sequences.

Protein preparation, gel electrophoresis, and immunoblot methods:
Worm lysates for immunoblotting were prepared either by passing worms suspended in SDS-reducing sample buffer (LAEMMLI 1970 Down) through a French press or according to previously described methods (HANNAK et al. 2002 Down). Sperm were isolated from fem-3(q23gf) (BARTON et al. 1987 Down) masculinized hermaphrodites using previously described methods (L'HERNAULT and ROBERTS 1995 Down). Protein samples were resolved in 12% discontinuous SDS-polyacrylamide gels using the Laemmli buffer system (LAEMMLI 1970 Down). Proteins were transferred from gels to Immobilon-P membranes (Millipore, Bedford, MA) and blocked with 5% nonfat dried milk in TBS-T [137 mM NaCl, 0.1% (v/v) Tween 20, 20 mM Tris-Cl, pH 7.6] for 1 hr at room temperature. The membrane was washed with TBS-T and incubated with affinity-purified N8 (at 0.3–0.5 µg/ml in TBS-T) for 1 hr at room temperature. The membrane was then washed with TBS-T and incubated with goat anti-rabbit IgG-HRP conjugate (1:10,000 dilution in TBS-T; Bio-Rad, Hercules, CA) for 1 hr at room temperature followed by more washing with TBS-T. ECL Plus reagents (Amersham Biosciences, Piscataway, NJ) were used for detection.

Immunohistochemistry:
Immunofluorescence microscopy of sperm and male gonad was performed essentially as described previously (MACHACA and L'HERNAULT 1997 Down; ARDUENGO et al. 1998 Down). Males were dissected in SM plus dextrose on Fisherbrand Colorfrost/Plus slides (Fisher Scientific, Pittsburgh), allowed to adhere, and fixed in paraformaldehyde. Excess fluid was removed, a coverslip was placed on the sample, it was quick frozen on dry ice, and the coverslip was removed. After additional rinsing and aldehyde blocking steps, slides were incubated with primary antibodies appropriately diluted in PBS for 1 hr at room temperature. The slides were then washed, incubated with the secondary antibodies and DAPI, washed again, and mounted using the Prolong antifade kit (Molecular Probes, Eugene, OR).

Affinity-purified N8 was used at 1 µg/ml for immunolocalization studies. Culture supernatant containing the monoclonal antibody 1CB4 (OKAMOTO and THOMSON 1985 Down) was used at 1:2000 as a specific marker for MOs in sperm. Fluorescein goat anti-mouse IgG (H + L) and Texas Red-X goat anti-rabbit IgG (H + L) secondary antibodies (Molecular Probes) were used at a 1:1000 dilution (2 µg/ml) in PBS.


*  RESULTS
*TOP
 *ABSTRACT
 *MATERIALS AND METHODS
 *RESULTS
*DISCUSSION
 *LITERATURE CITED

Sterility and ovulation in spe-39 mutants:
All spe-39 alleles were characterized in cis to dpy-11 to facilitate ready identification of homozygous animals. Hermaphrodites that are homozygous for any spe-39 allele are self-sterile and lay oocytes (Table 1). Such self-sterile hermaphrodites produce cross-progeny after mating with wild-type or spe-39/+ males, indicating they have a spermatogenesis-defective (Spe) phenotype (reviewed by L'HERNAULT 1997 Down).


 
View this table:
In this window
In a new window

  		Table 1. Self-fertility, embryonic lethality, and ovulation of spe-39 mutants

Brood sizes of spe-39(tx12), (eb9), or (eb110) average 0–1 when hermaphrodites are grown at either 16° or 25°. In contrast, spe-39(eb111) has a temperature-sensitive Spe phenotype, exhibiting average brood sizes of 41 at 16° and <1 at 25°.

Compared to the sum of progeny and oocytes produced by the dpy-11 control, ovulation of spe-39(eb9), (eb110), or (eb111) hermaphrodites is reduced 2- to 4-fold at 16°, but either slightly reduced or elevated relative to the control at the 25° growth temperature. In contrast, spe-39(tx12) mutants have significantly reduced (~10-fold) ovulation at both growth temperatures. The two weakest spe-39 mutants, eb110 and eb111, produce oocytes that can be easily counted, as is the case for the dpy-11 control and most other spermatogenesis-defective mutants (reviewed by L'HERNAULT 1997 Down). In contrast, oocytes laid by spe-39(tx12) and, occasionally, spe-39(eb9) mutants are distorted and not well separated. It is not known whether this is due to defective oogenesis or crushing of the oocytes after ovulation but, in either case, spe-39 may play a role.

A small percentage (~2% at 16° and ~4% at 25°) of the fertilized eggs laid by dpy-11 control homozygotes arrest development during embryogenesis (Table 1). In spe-39(eb110) and spe-39(eb111) mutants, the rates of lethality are raised to ~29% and ~15%, respectively, of the embryos produced at 16°. Embryonic lethality is not observed in spe-39(tx12) and spe-39(eb9) mutants because these mutants produce oocytes that are usually not fertilized due to the nearly complete penetrance of the spermatogenesis-defective phenotype.

Phenotypic analyses of spermatogenesis in spe-39 mutants:
Spermatogenesis in C. elegans has been described in detail (WOLF et al. 1978 Down; WARD et al. 1981 Down; ROBERTS et al. 1986 Down; L'HERNAULT et al. 1988 Down; L'HERNAULT 1997 Down). In this study, dpy-11 was used as an in cis morphological marker for spe-39 and homozygous dpy-11 males were found to have cytologically wild-type spermatogenesis (Fig 1A and D–M). During meiosis, the 4N nucleus of a primary spermatocyte (Fig 1D and Fig I) divides twice to form four condensed haploid nuclei (Fig 1J). Simultaneous with completing meiosis I, the primary spermatocyte usually divides to give rise to two secondary spermatocytes but cytokinesis is sometimes incomplete. In either case, the 2N nucleus found in each secondary spermatocyte divides into two haploid nuclei (Fig 1E, Fig F, Fig J, and Fig K) and asymmetric partitioning, in which the spermatids bud from a centrally located residual body, accompanies the final cytokinesis (Fig 1E and Fig F). In wild type, the haploid nuclei are always segregated into the budding spermatids (Fig 1J and Fig K). The wild-type spermatid is ~6 µm in diameter and has a centrally located nucleus (Fig 1G and Fig L). Upon ejaculation, the sessile spermatid is activated to become a fully functional spermatozoon. Activation can also occur in vitro by cell surface proteolysis with Pronase (Fig 1H and Fig M) and, like the naturally occurring process, results in extension of a pseudopod to allow motility (ps in Fig 1H; reviewed by L'HERNAULT and ROBERTS 1995 Down).

While a large number of spermatids were released from the dissected wild-type gonad (arrows in Fig 1A), both spe-39(eb9) and spe-39(tx12) mutants showed a dramatic reduction in spermatogenesis. Many fewer spermatid-like cells (arrows in Fig 1B and Fig C) were observed and increased numbers of spermatocyte-like cells were released by the mutants (arrowheads in Fig 1B, Fig C, and C1). Four condensed nuclei are usually observed in these large spermatocyte-like cells, indicating meiosis occurred (Fig 1N and Fig O, with corresponding DAPI images R and S, respectively). The cytokinesis event that normally accompanies meiosis in wild type (Fig 1E and Fig F) does not occur in these spe-39 mutant spermatocytes (Fig 1N and Fig O). Arrested spermatocytes are also observed in spe-4 mutants where they were named terminal spermatocytes (L'HERNAULT and ARDUENGO 1992 Down), a term that is also applicable to many spe-39 defective spermatocytes. Some of these cells attempt budding, but abnormally small buds (Fig 1O) each containing a haploid nucleus (Fig 1S) are the observed result. Occasionally, budding occurs in spe-39 mutants and this results in abnormally small spermatids (arrows in Fig 1B and Fig C, and Fig 7Q and Fig U). Unlike the centrally positioned nuclei in wild-type spermatids (Fig 1G and Fig L), nuclei in spe-39 mutant spermatids are eccentrically located (arrows in Fig 7Q and Fig U). Pronase treatment usually is without effect on spe-39(eb9) spermatids, while it causes wild-type spermatids to extend a pseudopod (Fig 1H). Occasionally, spe-39(eb9) spermatids will extend spike-like projections after Pronase treatment (arrow in Fig 1Q), and this resembles the spike formation that occurs transiently in wild type prior to pseudopodial extension (SHAKES and WARD 1989A Down). In the spe-39(tx12) terminal spermatocytes, more than four DAPI staining spots are occasionally observed (Fig 1P and Fig T), but this feature has not been seen in spe-39(eb9) mutants.



View larger version (40K):
In this window
In a new window
Download PPT slide
  		Figure 5. Immunoblots probed with the anti-SPE-39 antibody N8. Molecular weights of protein standards are shown in daltons. (A) Each lane was loaded with the lysate of 100 adult worms. All worms were in a dpy-11 genetic background and the genotype with respect to spe-39 is indicated. An arrow indicates the wild-type SPE-39 protein. The truncated SPE-39-derived protein encoded by the premature polypeptide chain termination mutant spe-39(eb9) is indicated with an asterisk. Two arrowheads indicate a cross-reacting polypeptide of unknown origin. (B) Each lane was loaded with either the lysate of 100 adult worms of the indicated genotype or 5 x 106 spermatids. fem-1(hc17ts) and fem-3(q23gf) worms were raised at nonpermissive temperature (25°).



View larger version (36K):
In this window
In a new window
Download PPT slide
  		Figure 6. Immunofluorescent localization of SPE-39 in the male gonad. All worms were in a dpy-11 genetic background and the genotype with respect to spe-39 is indicated on A and E. From left to right, each row has a series of four corresponding images: a DIC image, DNA visualized by DAPI staining (blue), staining with the MO-specific monoclonal antibody 1CB4 (green), and staining with the anti-SPE-39 antibody N8 (red). d, distal end of gonad; p, proximal end of gonad. Bars, 50 µm.



View larger version (51K):
In this window
In a new window
Download PPT slide
  		Figure 7. Immunofluorescent localization of SPE-39 during spermatogenesis in males. All worms were in a dpy-11 genetic background and the genotype with respect to spe-39 is indicated on the six leftmost panels. From left to right, each row has a series of four corresponding images: a DIC image, DNA visualized by DAPI staining (blue), staining with the MO-specific monoclonal antibody 1CB4 (green), and staining with the anti-SPE-39 antibody N8 (red). (A–D) Wild-type spermatocyte in the process of spermatid budding. The arrows point to budding spermatids (s) and the anucleate residual body (rb). (E–L) spe-39 mutant spermatocytes. Occasionally, unsuccessful attempts to bud spermatids are observed (arrows in I). (M–P) A wild-type spermatid and (Q–X) spe-39 mutant spermatids. Nuclei are located off center in spe-39 spermatids (arrows in Q and U). Bars, 5 µm.

A major aspect of wild-type spermatogenesis is morphogenesis of a special organelle called the fibrous body-membranous organelle complex (Fig 2A and Fig B). The MOs first appear in spermatogonial cells when meiosis is initiated (WOLF et al. 1978 Down). The major component of each FB is a bundle of parallel fibers of polymerized major sperm protein (MSP; Fig 2A1; also see WARD and KLASS 1982 Down). In wild-type spermatocytes, each MO extends a double-layered membrane envelope (arrow in Fig 2A2) that surrounds its closely associated FB. The FB-MOs, mitochondria, and haploid nuclei all segregate into budding spermatids and eventually occupy most of the cytoplasmic space (Fig 2A). When spermatids have completed their budding from the residual body, the MSP fibers of the FBs depolymerize. After the disassembly of the associated FBs, the MOs move to the periphery of the cytoplasm (Fig 2B). During activation, the MOs fuse with the plasma membrane and release their contents (arrow in Fig 2B), which is a secretory event required for normal spermiogenesis (reviewed by L'HERNAULT 1997 Down).

An ultrastructural analysis of spe-39 mutants reveals defects in both FB-MO morphogenesis and perinuclear morphology. In both spe-39(eb9) (Fig 2C and Fig D) and spe-39(tx12) (Fig 2F), FBs can form, but they apparently are not associated with any organized membranes (fb in Fig 2C and Fig F). We have not observed MO-like structures in spe-39 mutant spermatocytes at any developmental stage, including stages where spermatocytes are in syncitium with the rachis and have nascent FBs (not shown). This suggests that spe-39 is involved in a very early step of FB-MO morphogenesis. Instead of MOs, many small vesicles are present in the cytoplasm of spe-39 terminal spermatocytes (v in Fig 2C) and, occasionally, these small vesicles occupy most of the cytoplasm (v in Fig 2E). A high-magnification view shows that the vesicles have a double-layered membrane envelope (arrows in Fig 2H), which is also a distinctive feature of the wild-type FB-MO (arrow in Fig 2A2). In wild type, the condensed haploid nuclei are surrounded by a smooth layer of perinuclear material (arrowhead in Fig 2B), which most likely contains ribonucleoprotein complexes (WARD et al. 1981 Down). In both spe-39(eb9) (Fig 2E) and spe-39(tx12) (Fig 2G) mutants, the perinuclear halo is absent and, instead, there are electron-dense granules surrounding the nucleus (arrows in Fig 2G). The segregation of FBs is not complete if the spe-39 mutant spermatocyte attempts to form spermatids. Some FBs remain in the residual body (left arrow in Fig 2D), while they are all segregated to the budding spermatids in wild type (Fig 2A). In spe-39(tx12), MSP fibers can aggregate to form large FBs, but these fibers are not bundled in parallel, and they display a disorganized pattern (Fig 2F). Occasionally, larger vacuoles bounded by thick electron-dense coating (tcv, Fig 2I) have been observed in spe-39(tx12).

Cloning of spe-39:
Three-point mapping has placed spe-39 to the right of and very close to dpy-11 on LGV (see MATERIALS AND METHODS; Fig 3). The genetic interval containing spe-39 was defined further by mapping against deficiencies in the region (HODGKIN and MARTINELLI 2001 Down). nDf18 and sDf35 complemented spe-39(eb9), while sDf20 and nDf32 failed to complement spe-39(eb9) for its self-sterile phenotype. This placed spe-39 between the right break points of nDf32 and nDf18 (Fig 3). Since dpy-11 is within the interval deleted by nDf18, and unc-68 is outside the interval deleted by nDf32, spe-39 is located between these two genes. Cosmid clones placed on the physical map in this region (COULSON et al. 1995 Down) were tested by germline transformation for spe-39(eb9) complementation. Transgenes containing the cosmid ZC404 (GenBank accession no. U55363.2) were found to correct the spe-39 self-sterile phenotype. Tests of ZC404-derived restriction fragments indicated that spe-39 is contained in an 8-kb region between a BamHI site and an NdeI site (Fig 3). In this region, three intact open reading frames (ORFs) are predicted by WormBase (http://www.wormbase.org): ZC404.3a, ZC404.3b (see Fig 3 legend), and ZC404.2. ZC404.2 was excluded because a 13.7-kb XbaI fragment, which contains the complete ZC404.2 sequence, did not rescue the self-sterile phenotype of spe-39(eb9) (Fig 3). PCR fragments that covered the complete genomic sequence of ZC404.3a, plus ~640 bp of 5' noncoding region and ~500 bp of 3' noncoding region, were amplified from the spe-39(eb9) mutant and a nonsense mutation was found in the sixth exon. Five cDNA clones have been identified for ZC404.3a (http://www.wormbase.org), and one of them, yk504f7, was completely sequenced and found to contain the intact coding sequence and the complete 3' untranslated region (UTR), including a poly(A) tail. The ZC404.3a-derived expressed sequence tag (EST) CEL04E9 (GenBank accession no. M75836, clone ID cm04e9) has an SL1 sequence trans-spliced (NILSEN 1993 Down) to its 5'-end. The presence of SL1 in cDNA sequence defines the 5'-end of a transcription unit because exons 5' to an SL1 sequence are never observed (reviewed by BLUMENTHAL and STEWARD 1997 Down).

Each of the four spe-39 mutants has a unique mutation in its ZC404.3a gene (Fig 3). spe-39(eb110), spe-39(eb9), and spe-39(eb111) are all nonsense mutations. spe-39(eb110) contains an ochre codon (TAA) that replaces a glutamine codon (CAA) at amino acid position 320. spe-39(eb9) and spe-39(eb111) are both opal (TGA) mutations, changing arginine codons (CGA) at amino acid positions 334 and 393, respectively. spe-39(tx12) is a splice-site mutant, with the conserved acceptor splice site AG changed to AA at the junction of the fourth intron and fifth exon.

spe-39 (ZC404.3a) is predicted to encode a protein of 522 amino acids (GenBank accession no. AAA97964.1) and, according to computational hydropathy analysis (KYTE and DOOLITTLE 1982 Down), the protein has no transmembrane spanning sequences (data not shown). SPE-39 is a novel protein with homologs in D. melanogaster, human (Fig 4A), and mouse (data not shown). The human homolog, FLJ12707, and SPE-39 share 21% identity and 33% similarity. The D. melanogaster homolog is less related to SPE-39, having an identity of 17% with SPE-39 and a similarity of 26%. cDNAs of the human SPE-39 homolog have been isolated from a variety of somatic tissues of endodermal, mesodermal, and ectodermal origins plus whole embryo, testis, and germ cells (http://www.ncbi.nlm.nih.gov/UniGene). Its ubiquity suggests an essential or housekeeping role for this protein. Since there is only one homolog in each organism, they can be considered orthologs, which refer to genetic counterparts derived evolutionarily from a common ancestor through vertical descent (FITCH 1970 Down; KOONIN 2001 Down). No homolog of SPE-39 is found in Saccharomyces cerevisiae or other unicellular organisms, suggesting the function of SPE-39 and its orthologs is specific to multicellular organisms.

In Fig 4A, there is a highly conserved region in the middle of the protein, which is illustrated in detail in Fig 4B. The conserved region is also homologous to translations of ESTs from six other organisms, which are also included in the alignment of Fig 4B. A motif is revealed in the alignment: (LM)-(ED)-x-(FY)-(RK)-S-x-x-(DE)-K-x-x-L-L-x-x-(AL)-(VIM); it is embedded in less-conserved flanking sequences.

Anti-SPE-39 antibody and immunohistochemistry:
The antiserum N8 was prepared against a synthesized peptide that contains amino acids 4–23 of SPE-39 (line over the sequence in Fig 4A). The antiserum was affinity purified and used for immunoblot analysis (Fig 5) and immunofluorescence localization (Fig 6 and Fig 7). In dpy-11 control hermaphrodites, antibody N8 recognizes a single major band around the size predicted by conceptual translation of spe-39 (59.5 kD; arrow in Fig 5A), and this polypeptide is not detectable in either spe-39(eb9) or spe-39(tx12) animals. At this prolonged exposure, a smaller and barely visible polypeptide (asterisk in Fig 5A) is detected only in the lane containing proteins from the spe-39(eb9) nonsense mutant and this is the correct, truncated size (38.4 kD) for SPE-39 in this premature polypeptide chain termination mutant. The spe-39(tx12) mutant has no detectable SPE-39-related band. This prolonged exposure also revealed a polypeptide that is slightly larger than SPE-39 and present in the control and both SPE-39 mutant protein lanes in similar abundance (arrowheads in Fig 5A). This cross-reaction does not affect immunofluorescence localization in sperm and the gonad, because no staining is seen in spe-39 mutants (Fig 6H, Fig 7H, Fig L, Fig T, and Fig X), and the reason this polypeptide shows weak cross-reactivity to an SPE-39 affinity-purified antiserum is not presently understood.

Immunoblot analyses also showed that SPE-39 is expressed in adult N2 hermaphrodites, him-5(e1490) males, 25°-raised fem-1(hc17ts) mutants [which have a female soma and a germline that makes only oocytes (NELSON et al. 1978 Down)], and 25°-raised fem-3(q23gf) mutants [which have a female soma and a germline that makes only sperm (BARTON et al. 1987 Down)]. Bands in Fig 5B are the same polypeptide as the one indicated by the arrow in Fig 5A. The smaller body size of males and/or the absence of oocytes may explain the lower expression of SPE-39 in males compared to worms with a female soma.

The monoclonal antibody 1CB4 (OKAMOTO and THOMSON 1985 Down) was used as a positive control because it specifically labels MOs in sperm. SPE-39 is distributed throughout the whole testis of a dpy-11 male (used as wild-type control; Fig 6, A–D), and there is no obvious abundance difference between the distal or proximal regions of this organ (d and p in Fig 6A). The lack of SPE-39 expression in spe-39(eb9) (Fig 6H) does not obviously affect the expression of the 1CB4 antigen in the proximal part of the testis (Fig 6G). In a wild-type spermatocyte, 1CB4 staining is observed in the budding spermatids, which is consistent with the segregation of the MOs (Fig 7C). In wild-type spermatids, MOs are localized in the periphery of the cytoplasm (Fig 7O). spe-39(eb9) and spe-39(tx12) mutants lack organized MOs, and 1CB4 shows a diffuse staining pattern in both spermatocytes (Fig 7G and Fig K) and spermatids (Fig 7S and Fig W). When budding is attempted by a spe-39(tx12) spermatocyte (arrows in Fig 7I), 1CB4 staining is more concentrated in the periphery of the cytoplasm, but no clear segregation into the buds is seen (Fig 7K). SPE-39 staining shows a diffuse distribution in the cytoplasm in both wild-type spermatocytes and spermatids (Fig 7D). In wild-type spermatocytes at the budding stage, SPE-39 staining is observed in both residual bodies and budding spermatids, but staining is stronger in the residual bodies than in the spermatids (Fig 7D). In wild-type spermatids, staining of SPE-39 is concentrated in the center of the cytoplasm and diminished in the cytoplasmic cortex (Fig 7P), where the MOs are located. Neither spe-39(eb9) nor spe-39(tx12) shows detectable staining for SPE-39 in either their spermatocytes (Fig 7H and Fig L) or their spermatids (Fig 7T and Fig X).


*  DISCUSSION
*TOP
 *ABSTRACT
 *MATERIALS AND METHODS
 *RESULTS
 *DISCUSSION
*LITERATURE CITED

We have presented phenotypic, genetic, and molecular analyses of spe-39, focusing on its role during spermatogenesis. The principal defect shown during spermatogenesis in spe-39 mutants is a failure of FB-MO morphogenesis. The wild-type FB-MO is a compact bipartite structure where membrane from the MO wraps around the bundled fibers of the FB, and each spermatocyte has several dozen FB-MOs (WOLF et al. 1978 Down; WARD et al. 1981 Down). Mutant spe-39 spermatocytes lack the MO portion of the FB-MO and, instead, contain large numbers of small vesicles enveloped by a double-layered membrane. The FBs in spe-39 mutants are not surrounded by membranes and can form unusual structures that differ from the normally well-organized bundles.

Prior work has identified several mutants that affect FB-MO morphogenesis and they share certain phenotypic features with spe-39 mutants. In spe-4 mutants, MOs form but they are severely vacuolated (L'HERNAULT and ARDUENGO 1992 Down), while spe-39 mutants never form MOs. Both spe-4 and spe-39 mutant spermatocytes contain FBs that are not properly enveloped by MO-derived membranes so that they can appear as naked aggregations of fibers. Normally, the FB develops in intimate, continuous association with MO-derived membranes (ROBERTS et al. 1986 Down), but spe-4 and spe-39 mutants show that this association is not necessary for the maintenance of the FBs. This is further supported by analyses of spe-10 mutants, which show premature disassociation of FBs from MOs in their spermatocytes. As a result, MOs segregate to spermatids during budding and naked FBs are left behind in the residual body where they do not immediately disassemble (SHAKES and WARD 1989B Down). In spe-39 mutants, many naked FBs are also observed to reside in the residual body during attempts to form spermatids. These data suggest that association with MOs is required for FBs to segregate properly during spermatid budding.

Several mutants with abnormalities in FB-MO morphogenesis also show defects in spermatid formation. All spe-4 (L'HERNAULT and ARDUENGO 1992 Down; ARDUENGO et al. 1998 Down) and most spe-39 mutant spermatocytes fail to produce spermatids. MOs in spe-5 mutants are vacuolated in a way that is similar to those observed in spe-4 spermatocytes, but MOs still associate with FBs in spe-5 mutants (MACHACA and L'HERNAULT 1997 Down). While functional spermatids sometimes form in spe-5 mutants, spermatocytes are usually terminally arrested and either do not bud or form obviously defective buds that do not become spermatids. Some spe-5 spermatocytes make abnormally small buds (MACHACA and L'HERNAULT 1997 Down) that resemble the small, defective buds sometimes produced by spe-39 mutants. The spermatids occasionally produced by spe-5 mutants are of normal size and have a centrally placed nucleus. In contrast, the occasional spe-39 spermatid is smaller than normal and has a nucleus that is eccentrically placed. Abnormally small spermatids with eccentrically placed nuclei are also found in spe-10 (SHAKES and WARD 1989B Down) and spe-17 (L'HERNAULT et al. 1993 Down) mutants. In addition to the budding of spermatids, spe-10 mutant spermatocytes can form small cytoplasts that contain FBs (SHAKES and WARD 1989B Down); budding of such anucleate structures has not been observed in spe-39 mutant spermatocytes.

In addition to the FB-MOs, the nucleus and its surrounding structures are also defective during spermatogenesis in spe-39 mutants. Usually, meiosis appears normal in the spe-39(eb9) mutant because four condensed nuclei are observed in each arrested spermatocyte. In contrast, spe-39(tx12) mutant spermatocytes can contain more than four chromatin-rich regions. Surrounding the condensed nucleus in wild-type spermatids is a perinuclear halo, which appears as a smooth layer that most likely contains ribonucleoprotein complexes (WARD et al. 1981 Down). Both spe-39(eb9) and spe-39(tx12) mutants lack the perinuclear halo and, instead, have electron-dense granules surrounding the nucleus. Defects in the ultrastructure of perinuclear material are also observed in fer-2, fer-3, fer-4 (WARD et al. 1981 Down), spe-11 (HILL et al. 1989 Down), and spe-4 (L'HERNAULT and ARDUENGO 1992 Down) mutants. In fer-2, fer-3, and fer-4 mutants, the perinuclear halo is absent and is replaced by an accumulation of large tubules. The centriole in spe-11(hc90) spermatids is surrounded by an excessive amount of perinuclear material but the nucleus is associated only with sparse patches of small electron-dense particles. Terminal spe-4(q347) mutant spermatocytes contain four nuclei and electron-dense granules that appear as a "necklace" in thin sections surround each nucleus. Defective perinuclear material also appears as electron-dense granules in spe-39(tx12) terminal spermatocytes, but they are smaller compared to those observed in spe-4(q347). It is not known how the various nuclear and perinuclear defects are related to the sterile phenotypes of these mutants.

Light microscopic and ultrastructural data suggest that the splice acceptor site mutant spe-39(tx12) has the most severe defects in spermatogenesis of all four spe-39 mutants. Homozygous spe-39(tx12) hermaphrodites can only rarely (one progeny found for 16 examined animals) produce self-progeny and this might be explained by the occasional use of AA as a noncanonical splice acceptor site (AROIAN et al. 1993 Down). Alternatively, such weak self-fertility might be an intrinsic feature of the spe-39 zygotic null phenotype. Although some SPE-39 might be present in spe-39(tx12) mutants, it is not detectable by immunoblot analysis, even after prolonged exposure using sensitive enhanced chemiluminescence techniques. Homozygous spe-39(tx12) hermaphrodites produced by heterozygous mothers have generally normal somatic development and are cross-fertile, suggesting the major defects caused by a lack of zygotically expressed SPE-39 occur during spermatogenesis. However, several observations suggest spe-39 also has functions outside spermatogenesis, including: (1) immunoblots reveal that SPE-39 expression occurs in tissues other than the testis; (2) immunostaining indicates that SPE-39 is expressed in most, if not all, cells (not shown); (3) spe-39(tx12) hermaphrodites produce distorted oocytes, suggesting they have defective oogenesis and/or abnormalities in the somatic gonad; and (4) spe-39 mutants can occasionally fertilize a few oocytes and some of the resulting eggs die during embryogenesis. Preliminary RNA interference (FIRE et al. 1998 Down) experiments reveal that prezygotic spe-39 expression plays an important role in both somatic and germline development.

SPE-4 (L'HERNAULT and ARDUENGO 1992 Down; ARDUENGO et al. 1998 Down) and SPE-39 are the only known proteins that affect early stages of FB-MO morphogenesis and for which both mutants and specific antisera exist. Although both SPE-4 and SPE-39 are involved in the morphogenesis of FB-MOs, they have different subcellular distributions. SPE-4 is an integral membrane protein that resides in the MOs (ARDUENGO et al. 1998 Down), while SPE-39 is highly hydrophilic and has a diffuse cytoplasmic distribution in both spermatocytes and spermatids. This indicates that MO morphogenesis requires both MO resident and nonresident proteins. Both SPE-4 and SPE-39 have homologs in humans. The two human presenilins are integral membrane proteins that have homology to SPE-4 and both reside in vesicular compartments of mammalian cells (CHECLER 2001 Down; CUPERS et al. 2001 Down). Presenilin 1 is known to move to the plasma membrane via vesicular trafficking where it participates in the proteolytic processing of target proteins, including the amyloid precursor protein (KAETHER et al. 2002 Down). Like SPE-39, the human ortholog of SPE-39, FLJ12707, has a diffuse cytoplasmic distribution in cultured cells when it is expressed as a fusion to the green fluorescent protein (G.-D. ZHU and S. W. L'HERNAULT, unpublished observations). Unlike the presenilins, the biochemical function of the human SPE-39 ortholog is not presently known.

Wild-type MOs first emerge as vesicular swellings at the edges of Golgi complexes and each MO develops in intimate association with an FB (WOLF et al. 1978 Down). Very little is presently known about how morphogenesis of this structure relates to general models of vesicular trafficking proposed for other eukaryotic cells (reviewed by ROTHMAN and WIELAND 1996 Down; SCHEKMAN and ORCI 1996 Down; SCHMID 1997 Down; JAMIESON 1998 Down; SCALES et al. 2000 Down; MAYER 2002 Down). However, several clues suggest that components of the endocytic pathways are involved in FB-MO formation. The interior of MOs in live spermatids was found to be acidic when cells are stained with a pH-sensitive fluorescent dye known to label lysosomes and other acidic organelles in live mammalian cells and yeast (HILL 2000 Down). Interestingly, spe-5 encodes a B subunit of the vacuolar (H+)-ATPase (P. D. HARTLEY, G.-D. ZHU and S. W. L'HERNAULT, unpublished data). Generally, the vacuolar (H+)-ATPase pumps protons required to acidify the interior of various vacuolar compartments including endosomes, lysosomes, synaptic vesicles, and central vacuoles of plant cells (reviewed by STEVENS and FORGAC 1997 Down). Although more studies are needed to understand the relationship between defects in MO morphogenesis seen in spe-5 mutants and a reduction in the activity of the proton pump, it is tempting to suggest that MO morphogenesis shares some biological machinery with endocytic pathways that involve acidic compartments like endosomes and lysosomes.

The available data indicate that SPE-39 participates in membrane reorganization during spermatogenesis. SPE-39 and its homologous proteins have a conserved region that includes two consecutive leucine residues (LL; Fig 4B). A class of sorting signals used in the endocytic pathways is characterized by a leucine doublet (reviewed by KIRCHHAUSEN 1999 Down). It is not yet known if these two consecutive leucine residues of SPE-39 and its homologs function as a dileucine sorting signal. We are presently combining further genetic analyses of C. elegans SPE-39 with cell biological studies of its ortholog in cultured human cells to gain further insight into the role of this protein. These further studies of SPE-39 and its interactions with known components of various vesicular transport pathways will contribute to our understanding of metazoan-specific aspects of protein sorting.


*  ACKNOWLEDGMENTS

We thank Diane Shakes, Julie Ahringer, the Sanger Centre, Yuji Kohara, and Andy Fire for providing C. elegans mutants or DNA clones. The 1CB4 monoclonal cell line (OKAMOTO and THOMSON 1985 Down) was provided by Julie Ahringer and Jonathan Hodgkin, and its culture supernatant was prepared by Craig Heilman and Alan I. Levey. Wesley C. Lindsey established that eb9 and tx12 were alleles of the same gene. Sheen Scott provided technical assistance with sperm preparation and protein gels, and we also thank Nancy L'Hernault for assistance with electron microscopy. Bill Kelly, Kevin Moses, and Barry Yedvobnick all provided useful comments on the manuscript. We thank Grant MacGregor for useful comments and alerting us to the roles played by dileucine sorting signals in endocytic pathways. The Caenorhabditis Genetics Center provided some nematode strains, and it is funded by the National Institutes of Health National Center for Research Resources. This work was supported by the U.S. Public Health Service through RO1 GM40697 to S.W.L.

Manuscript received February 18, 2003; Accepted for publication May 13, 2003.


*  LITERATURE CITED
*TOP
 *ABSTRACT
 *MATERIALS AND METHODS
 *RESULTS
 *DISCUSSION
 *LITERATURE CITED

ACHANZAR, W. E. and S. WARD, 1997  A nematode gene required for sperm vesicle fusion. J. Cell Sci. 110:1073-1081.[Abstract]

ARDUENGO, P. M., O. K. APPLEBERRY, P. CHUANG, and S. W. L'HERNAULT, 1998  The presenilin protein family member SPE-4 localizes to an ER/Golgi derived organelle and is required for proper cytoplasmic partitioning during Caenorhabditis elegans spermatogenesis. J. Cell Sci. 111:3645-3654.

AROIAN, R. V., A. D. LEVY, M. KOGA, Y. OHSHIMA, and J. M. KRAMER et al., 1993  Splicing in Caenorhabditis elegans does not require an AG at the 3' splice acceptor site. Mol. Cell. Biol. 13:626-637.[Abstract/Free Full Text]

BAIRD, S. E. and S. W. EMMONS, 1990  Properties of a class of genes required for ray morphogenesis in Caenorhabditis elegans.. Genetics 126:335-344.[Abstract]

BARAN, R., R. ARONOFF, and G. GARRIGA, 1999  The C. elegans homeodomain gene unc-42 regulates chemosensory and glutamate receptor expression. Development 126:2241-2251.[Abstract]

BARTON, M. K., T. B. SCHEDL, and J. KIMBLE, 1987  Gain-of-function mutations of fem-3, a sex-determination gene in Caenorhabditis elegans.. Genetics 115:107-119.[Abstract/Free Full Text]

BLOOM, L. and H. R. HORVITZ, 1997  The Caenorhabditis elegans gene unc-76 and its human homologs define a new gene family involved in axonal outgrowth and fasciculation. Proc. Natl. Acad. Sci. USA 94:3414-3419.[Abstract/Free Full Text]

BLUMENTHAL, T., and K. STEWARD, 1997 RNA processing and gene structure, pp. 117–145 in C. elegans II, edited by D. L. RIDDLE, T. BLUMENTHAL, B. J. MEYER and J. R. PRIESS. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY.

BRENNER, S., 1974  The genetics of Caenorhabditis elegans.. Genetics 77:71-94.[Abstract/Free Full Text]

CHECLER, F., 2001  The multiple paradoxes of presenilins. J. Neurochem. 76:1621-1627.[Medline]

COULSON, A., C. HUYNH, Y. KOZONO and R. SHOWNKEEN, 1995 The physical map of the Caenorhabditis elegans genome, pp. 533–550 in Caenorhabditis elegans: Modern Biological Analysis of an Organism, edited by H. F. EPSTEIN and D. C. SHAKES. Academic Press, San Diego.

CUPERS, P., M. BENTAHIR, K. CRAESSAERTS, I. ORLANS, and H. VANDERSTICHELE et al., 2001  The discrepancy between presenilin subcellular localization and gamma-secretase processing of amyloid precursor protein. J. Cell Biol. 154:731-740.[Abstract/Free Full Text]

FERGUSON, E. L. and H. R. HORVITZ, 1985  Identification and characterization of 22 genes that affect the vulval cell lineages of the nematode Caenorhabditis elegans.. Genetics 110:17-72.[Abstract/Free Full Text]

FIRE, A., S. XU, M. K. MONTGOMERY, S. A. KOSTAS, and S. E. DRIVER et al., 1998  Potent and specific genetic interference by double-stranded RNA in Caenorhabditis elegans.. Nature 391:806-811.[Medline]

FITCH, W. M., 1970  Distinguishing homologous from analogous proteins. Syst. Zool. 19:99-113.[Medline]

HANNAK, E., K. OEGEMA, M. KIRKHAM, P. GONCZY, and B. HABERMANN et al., 2002  The kinetically dominant assembly pathway for centrosomal asters in Caenorhabditis elegans is gamma-tubulin dependent. J. Cell Biol. 157:591-602.[Abstract/Free Full Text]

HEDGECOCK, E. M., J. G. CULOTTI, D. H. HALL, and B. D. STERN, 1987  Genetics of cell and axon migrations in Caenorhabditis elegans.. Development 100:365-382.[Abstract]

HILL, D. P., D. C. SHAKES, S. WARD, and S. STROME, 1989  A sperm-supplied product essential for initiation of normal embryogenesis in Caenorhabditis elegans is encoded by the paternal-effect embryonic-lethal gene, spe-11.. Dev. Biol. 136:154-166.[Medline]

HILL, K. L., 2000 Analyses of spermatogenesis and the role of sperm in reproductive interactions in the nematode genus Caenorhabditis. Ph.D. Thesis, Emory University, Atlanta.

HODGKIN, J., and S. MARTINELLI, 2001 2001 Genetic Map of Caenorhabditis elegans. Caenorhabditis Genetics Center, St. Paul.

HODGKIN, J. A., H. R. HORVITZ, and S. BRENNER, 1979  Nondisjunction mutants of the nematode Caenorhabditis elegans.. Genetics 91:67-94.[Abstract/Free Full Text]

JAMIESON, J. D., 1998  The Golgi complex: perspectives and prospectives. Biochim. Biophys. Acta 1404:3-7.[Medline]

KAETHER, C., S. LAMMICH, D. EDBAUER, M. ERTL, and J. RIETDORF et al., 2002  Presenilin-1 affects trafficking and processing of betaAPP and is targeted in a complex with nicastrin to the plasma membrane. J. Cell Biol. 158:551-561.[Abstract/Free Full Text]

KELLY, W. G., S. XU, M. K. MONTGOMERY, and A. FIRE, 1997  Distinct requirements for somatic and germline expression of a generally expressed Caernorhabditis elegans gene. Genetics 146:227-238.[Abstract]

KIRCHHAUSEN, T., 1999  Adaptors for clathrin-mediated traffic. Annu. Rev. Cell Dev. Biol. 15:705-732.[Medline]

KO, F. C. and K. L. CHOW, 2002  A novel thioredoxin-like protein encoded by the C. elegans dpy-11 gene is required for body and sensory organ morphogenesis. Development 129:1185-1194.[Abstract/Free Full Text]

KOONIN, E. V., 2001 An apology for orthologs—or brave new memes. Genome Biol. 2: 1005.1–1005.2.

KYTE, J. and R. F. DOOLITTLE, 1982  A simple method for displaying the hydropathic character of a protein. J. Mol. Biol. 157:105-132.[Medline]

L'HERNAULT, S. W., 1997 Spermatogenesis, pp. 271–294 in C. elegans II, edited by D. L. RIDDLE, T. BLUMENTHAL, B. J. MEYER and J. R. PRIESS. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY.

L'HERNAULT, S. W. and P. M. ARDUENGO, 1992  Mutation of a putative sperm membrane protein in Caenorhabditis elegans prevents sperm differentiation but not its associated meiotic divisions. J. Cell Biol. 119:55-68.[Abstract/Free Full Text]

L'HERNAULT, S. W., and T. M. ROBERTS, 1995 Cell biology of nematode sperm, pp. 273–301 in Caenorhabditis elegans: Modern Biological Analysis of an Organism, edited by H. F. EPSTEIN and D. C. SHAKES. Academic Press, San Diego.

L'HERNAULT, S. W., D. C. SHAKES, and S. WARD, 1988  Developmental genetics of chromosome I spermatogenesis-defective mutants in the nematode Caenorhabditis elegans.. Genetics 120:435-452.[Abstract/Free Full Text]

L'HERNAULT, S. W., G. M. BENIAN, and R. B. EMMONS, 1993  Genetic and molecular characterization of the Caenorhabditis elegans spermatogenesis-defective gene spe-17.. Genetics 134:769-780.[Abstract]

LAEMMLI, U. K., 1970  Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227:680-685.[Medline]