HAL-2 promotes homologous pairing during Caenorhabditis elegans meiosis by antagonizing inhibitory effects of synaptonemal complex precursors

Abstract

During meiosis, chromosomes align with their homologous pairing partners and stabilize this alignment through assembly of the synaptonemal complex (SC). Since the SC assembles cooperatively yet is indifferent to homology, pairing and SC assembly must be tightly coordinated. We identify HAL-2 as a key mediator in this coordination, showing that HAL-2 promotes pairing largely by preventing detrimental effects of SC precursors (SYP proteins). hal-2 mutants fail to establish pairing and lack multiple markers of chromosome movement mediated by pairing centers (PCs), chromosome sites that link chromosomes to cytoplasmic microtubules through nuclear envelope-spanning complexes. Moreover, SYP proteins load inappropriately along individual unpaired chromosomes in hal-2 mutants, and markers of PC-dependent movement and function are restored in hal-2; syp double mutants. These and other data indicate that SYP proteins can impede pairing and that HAL-2 promotes pairing predominantly but not exclusively by counteracting this inhibition, thereby enabling activation and regulation of PC function. HAL-2 concentrates in the germ cell nucleoplasm and colocalizes with SYP proteins in nuclear aggregates when SC assembly is prevented. We propose that HAL-2 functions to shepherd SYP proteins prior to licensing of SC assembly, preventing untimely interactions between SC precursors and chromosomes and allowing sufficient accumulation of precursors for rapid cooperative assembly upon homology verification.

Figure 1. Defective nuclear reorganization and homologous pairing in hal-2 mutants.

(A) Chromosome organization in germline nuclei in regions of wild-type (wt) and hal-2 mutant gonads extending from the premeiotic zone (left) through early pachytene (right). In the wild-type gonad, DAPI-stained chromatin appears widely dispersed in premeiotic nuclei and exhibits a highly clustered organization in nuclei within the transition zone (TZ), reflecting clustering of chromosomes during a period of active chromosome mobilization that begins soon after meiotic entry; DAPI signals appear more dispersed in nuclei that have exited the TZ and progressed into pachytene. The hal-2 mutant gonad lacks nuclei with the clustered chromosome configuration characteristic of the TZ, indicating impairment of nuclear reorganization upon meiotic entry. Bar, 10 µm. (B) Left: Wild-type and hal-2 mutant nuclei at mid-pachytene. The wild-type nuclei contain thick parallel tracks of DAPI-stained chromatin corresponding to aligned homolog pairs, whereas the hal-2 mutant nuclei contain disorganized thin DAPI-stained chromatin tracks representing unaligned chromosomes. Right: Each panel shows DAPI-stained chromosomes in a single oocyte at diakinesis, the last stage of meiotic prophase. Whereas 6 bivalents are present in the wild-type oocyte, each representing a homolog pair held together by a chiasma, 12 smaller DAPI-stained bodies (univalents) are observed in the hal-2 mutant oocyte, indicating an absence of chiasmata. Bar, 10 µm. (C) 5S rDNA FISH (left) and immunofluorescence (IF) of X chromosome PC-binding protein HIM-8 (right) in pachtyene nuclei from wild-type and hal-2 germ lines. A single focus or two closely spaced foci is detected in wild-type nuclei, indicating paired homologs. Two widely spaced foci are seen in the hal-2 mutant nuclei, indicating that both the X-PC and 5S rDNA locus were unpaired. Bar, 5 µm.

Figure 2. hal-2 mutants lack multiple markers of PC-mediated chromosomal movement.

(A) IF images of early prophase nuclei stained for PLK-2 and SUN-1 phosphorylated on Ser12 (SUN-1 S12-Pi) in wild-type and hal-2 animals. In the wild-type nuclei, SUN-1 S12-Pi and PLK-2 are concentrated together in bright patches at the nuclear envelope (NE). In the hal-2 nuclei, PLK-2 and SUN-1 S12-Pi are not detected. Bar, 2 µm. (B) IF images of early prophase nuclei stained for ZYG-12::GFP and SUN-1 phosphorylated on Ser8 (SUN-1 S8-Pi) in wild-type and hal-2 animals carrying the zyg-12::gfp transgene. In the wild-type nuclei, SUN-1 S8-Pi and ZYG-12::GFP are concentrated together in bright patches at the NE and are also detected at lower levels throughout the entire NE. In the hal-2 nuclei shown, SUN-1 S8-Pi is not detected and ZYG-12::GFP is dispersed throughout the NE. Bar, 2 µm. (C) IF images of ZIM-2 (chromosome V PC-binding protein) in early prophase nuclei. A single bright focus near the NE is observed in the wild-type nuclei, whereas none is detected in the hal-2 mutant nuclei. Bar, 2 µm.

Figure 3. SYP proteins are loaded onto the axes of unpaired homologs in hal-2 mutants.

(A) Portions of wild-type and hal-2 germ lines extending from premeiotic nuclei (left) to nuclei progressing into meiosis (right), stained with antibodies against HTP-3 (LE component) and SYP-1 (CR protein). In both germ lines, most nuclei in which HTP-3 is detected as discrete axial structures show SYP-1 colocalizing with most HTP-3 stretches, indicating that CR components are competent to load onto the axes of unpaired homologs soon after meiotic entry in the hal-2 mutant. Bar, 5 µm. (B) Co-immunolocalization of LE components (HTP-3 [top] or HIM-3 [bottom]) and CR proteins (SYP-1 [top] or GFP::SYP-3 [bottom]) in wild-type and hal-2 nuclei at mid/late pachytene. LE components and SYP proteins colocalize at the interface between DAPI-stained aligned homologs in wild-type nuclei, whereas SYP proteins colocalize with LE components along the lengths of unpaired chromosomes in hal-2 mutant nuclei. Bar, 2 µm. (C) Localization of LE component HTP-3, CR protein SYP-1 and DAPI-stained chromatin visualized using 3D-SIM; for both genotypes, pairing status of the X chromosomes is indicated by HIM-8 immunostaining (imaged without 3D-SIM). In the wild-type nucleus, two resolvable parallel tracks of HTP-3 flanking a single track of SYP-1 are detected along much of the lengths of the aligned homolog pairs, reflecting the presence of SCs linking pairs of LEs that are separated by a distance of approximately 100 nm , . In the hal-2 nucleus, both SYP-1 and HTP-3 are detected as single tracks at the sister chromatid interface of individual unpaired chromosomes. Bar, 2 µm.

Figure 4. Removal of SYP proteins in hal-2 mutants partially restores chromosome clustering, markers of PC-mediated chromosome movement, and homologous pairing.

(A) IF images of early prophase nuclei stained for PLK-2 and SUN-1 S12-Pi in hal-2 and hal-2; syp-2(ok307) germ lines. Chromosome clustering, SUN-1 Ser12 phosphorylation and NE patches containing PLK-2 and SUN-1 S12-Pi (features that are missing in the hal-2 nuclei) are restored in the hal-2; syp-2 nuclei. Bar, 2 µm. (B) hal-2 and hal-2; syp-2 early prophase nuclei stained with antibodies against ZIM-2. Whereas ZIM-2 foci are not observed in the hal-2 mutant nuclei, bright ZIM-2 foci near the NE are restored in the hal-2; syp-2 double mutants. Bar, 2 µm. (C) Diagram of a hermaphrodite gonad depicting the 5 zones of equal lengths used for quantitation of pairing in Figures 4D–4F. In a wild-type gonad, Zone 1 contains premeiotic nuclei, Zone 2 includes some premeiotic nuclei, the TZ and some early pachytene nuclei, Zone 3 contains early and mid-pachytene nuclei, Zone 4 contains mid and late pachytene nuclei and Zone 5 consists of late pachytene nuclei. (D) Bar graph showing quantitation of pairing at the X-PC, measured as pairing of HIM-8 foci, for Zones 2–5 (the regions with most robust HIM-8 staining). While pairing of HIM-8 foci was abolished in the hal-2 germ lines, high HIM-8 pairing levels were restored in the hal-2; syp-2 double mutant, with pairing levels peaking in Zones 3 and 4 (94%; 87%) at levels approaching those of wild-type and the syp-2 mutant (99%; 98%). Nonetheless pairing levels in hal-2; syp-2 double mutants were significantly lower than those in syp-2 gonads for all zones (Zone 2: p<0.0001, Zone 3: p = 0.012, Zone 4: p = 0.0002, Zone 5: p = 0.0194). Worms used in this analysis carried the zyg-12::gfp transgene. (E) Quantitation of ZIM-2 foci in nuclei from Zone 2, the region in which ZIM-2 staining is most robust in wild-type germ lines. Stacked bar graphs depict the percentage of Zone 2 nuclei with a single ZIM-2 focus (indicating paired V-PCs), two unpaired ZIM-2 foci (indicating unpaired V-PCs) or without any clear ZIM-2 foci (indicative of inactive PCs not engaged in chromosome mobilization). Worms used in this analysis carried the zyg-12::gfp transgene. (F) Bar graph depicting quantitation of pairing levels at the 5S rDNA locus assessed by FISH. A modest partial restoration of pairing at this non-PC locus was observed in the syp-3(ok758); hal-2 double mutant: 5S rDNA pairing levels were effectively abolished in the hal-2 mutant, whereas pairing in syp-3; hal-2 double mutants showed a highly significant improvement over hal-2 only in Zone 2 (p<0.0001). Further, syp-3; hal-2 double mutants displayed significantly lower pairing levels than those in syp-3 gonads for Zones 2–5 (p = 0.0023, p<0.0001, p<0.0001, p<0.0004).

Figure 5. hal-2 mutants exhibit reduced chromosomal localization of HTP-1/2 that is rescued by removal of SYP proteins.

(A) Co-immunolocalization of LE components HTP-3 and HTP-1/2 in pachytene nuclei of indicated genotypes. HTP-1/2 and HTP-3 are detected colocalized along the full lengths of paired and synapsed homologs in wild-type nuclei and along the axes of unpaired chromosomes in nuclei from the late pachytene region in a syp-2 single mutant; HTP-1/2 and HTP-3 IF signals are present in similar ratios in these two genotypes. Images from the hal-2 mutant show that the level of HTP-1/2 relative to HTP-3 was greatly reduced compared to wild-type and syp-2 controls; however, HTP-1/2 localization was not abolished in the hal-2 mutant, as shown in the inset, in which the HTP-1/2 IF signal has been adjusted to highlight the faint HTP-1/2 staining along the axes of the unpaired chromosomes. HTP-1/2 localization in hal-2; syp-2 double mutant nuclei appears similar to the syp-2 single mutant. Bar, 2 µm. (B) Each panel depicts the chromosomes of a single diakinesis oocyte of the indicated genotype, co-stained for HTP-1/2 and HTP-3. HTP-1/2 is detected on the chromosomes in the wild-type and syp-2 control oocytes, but is not detected on the chromosomes in the hal-2 mutant oocyte; HTP-1/2 staining is detected on the chromosomes in the hal-2; syp-2 oocyte, reflecting rescue of the HTP-1/2 localization defect in hal-2 mutants by the removal of SYP proteins. Bar, 2 µm.

Figure 6. hal-2 encodes a protein that concentrates in the nucleoplasm of germ cells.

(A) Schematic diagram of hal-2 gene structure depicting the locations of the nonsense mutation and the 761 bp deletion in the me79 and tm4960 alleles. (B) Western blot of whole worm lysates from the indicated genotypes probed with α-HAL-2 and α-α-TUBULIN (loading control) antibodies. A band corresponding to the expected 36.1 kDa HAL-2 protein is detected in wild-type lysates and is absent from both me79 and tm4960 mutant lysates. (C) Low magnification image of a wild-type whole mount germ line stained with α-HAL-2. HAL-2 staining is first observed in the premeiotic nuclei preceding the TZ, is present in nuclei throughout the TZ and pachytene regions, and progressively weakens during diplotene. Bar, 10 µm. (D) High magnification image of HAL-2 immunolocalization in a portion of a wild-type germ line extending from the premeiotic zone to the TZ; comparison of the DNA staining and HAL-2 staining within the TZ reflects the fact that HAL-2 is concentrated predominantly in the nucleoplasm. Bar, 10 µm. (E) IF images of HAL-2 in wild-type and hal-2 pachytene nuclei. HAL-2 immunostaining does not colocalize with DAPI-stained chromatin and is distributed broadly within the nucleoplasm of wild-type nuclei. (Some short linear stretches are apparent within this nucleoplasmic HAL-2 staining, but they do not correspond to a known structure and their significance is unclear.) HAL-2 immunostaining is absent in hal-2 mutant nuclei, demonstrating the specificity of the antibody and the lack of full-length HAL-2 protein in the mutant. Bar, 10 µm. (F) Immunolocalization of GFP::HAL-2 in pachytene nuclei of worms containing a functional GFP::HAL-2 transgene. GFP::HAL-2 localized to the nucleoplasm of germ cells, in agreement with the localization pattern obtained from α-HAL-2 IF. Bar, 10 µm.

Figure 7. HAL-2 colocalizes with SYP proteins in aggregates that form when SC assembly is prevented.

Co-staining of SYP-1 and HAL-2 in pachytene nuclei of indicated genotypes. In the wild-type nuclei, HAL-2 is localized to the nucleoplasm and exhibits very little colocalization with SYP-1, which is localized at the interface between paired homologs. In contrast, in mutants with severely disrupted SC assembly (cohesin mutant scc-3(ku263) and mutants lacking LE components HTP-3 or HIM-3), SYP-1 does not localize along chromosomes but instead becomes concentrated in nuclear aggregates (presumably polycomplexes); in these mutants, HAL-2 is detected in the nucleoplasm but also colocalizes with SYP-1 in the nuclear aggregates. Further, localization of HAL-2 in nuclear aggregates in scc-3 mutants is SYP-dependent. Bar, 4 µm.

Figure 8. HAL-2 has additional roles in meiosis beyond preventing inappropriate association of SYP proteins with chromosomes.

(A) Immunolocalization of SUN-1 S8-Pi in DAPI-stained gonads of the indicated genotypes, with meiosis progressing from left to right. Panels on the right are zoomed-in images of early (i; blue box) and later (ii; green box) stages of meiotic prophase. In the wild-type gonad, nuclei in the TZ (i) exhibit a clustered chromosome configuration and have multiple SUN-1 S8-Pi NE patches, whereas after progression into pachytene (ii), the chromosomes exhibit a more dispersed distribution within the nuclei and most nuclei have only a single SUN-1 S8-Pi focus remaining on the NE. In the hal-2 mutant gonad, nuclei from both regions (i and ii) have dispersed chromosomes and either lack SUN-1 S8-Pi (i) or have low levels of diffuse SUN-1 S8-Pi NE signal (ii; Figure S10), while in the syp-2 mutant, nuclei from both regions exhibit clustered chromosomes and multiple SUN-1 S8-Pi NE patches, reflecting prolonged persistence of chromosome mobilization. The hal-2; syp-2 double mutant gonad does not exhibit persistent chromosome clustering and multiple persistent SUN-1 S8-Pi NE patches; chromosome clustering is restricted to a short TZ (i), and only a single SUN-1 S8-Pi NE focus persists beyond the TZ, in nuclei with dispersed chromosomes (ii), similar to wild-type gonads. Yellow lines depict the extent of the TZs, with the yellow arrow indicating extension of the TZ in the syp-2 mutant beyond the region shown in the image. The zyg-12::gfp transgene was present in the gonads shown here. Bars, 10 µm. (B) Immunostaining of DNA strand exchange protein RAD-51 in early/mid-pachytene nuclei. While abundant RAD-51 foci are observed in both wild-type nuclei and syp-3 mutant nuclei, very few foci were detected in the hal-2 and syp-3; hal-2 mutants, with most nuclei lacking RAD-51 foci. Bar, 2 µm. (C) Immunostaining of RAD-51 in mid-pachytene nuclei of hal-2 mutants and wild-type controls. Abundant RAD-51 foci are observed in nuclei of hal-2 germ lines exposed to 1 krad of γ irradiation, while few foci are detected in the nuclei of unirradiated hal-2 controls. Bar, 2 µm.