We assessed the expression of TRIP13 in a panel of adult mouse tissues. TRIP13 was primarily expressed in testis but detected at low levels in ovary and liver (Figure 1A). In the testis, TRIP13 was prominent in the cytoplasm of primary spermatocytes, especially leptotene and zygotene cells (Figure 1B). Given the expression of TRIP13 in spermatocytes, we examined its localization on meiotic chromosomes by immunostaining of surface spread nuclei of prophase I spermatocytes (Figure 1C). Previously, TRIP13 was reported to localize to telomeres in spermatocytes (Gomez et al., 2019). Indeed, we found that TRIP13 localized to telomeres from leptotene to diplotene cells (Figure 1C). Strikingly, we found that, in addition to telomeres, TRIP13 localized to the synaptonemal complex (SC) in early pachytene spermatocytes but disappeared from the SC in the mid to late pachytene spermatocytes (Figure 1C). The synaptonemal complex consists of two lateral elements and one central element. The central region is comprised of the transverse filament protein SYCP1 and central element proteins, which appear upon synapsis. Given the novel localization of TRIP13 to the SC, we performed confocal immunofluorescent microscopy with super-resolution deconvolution to determine the location of TRIP13 within the SC. In contrast with the lateral element marker SYCP3, TRIP13 localized to the SC central region (Figure 1D). We examined the localization of TRIP13 in oocytes. TRIP13 localized strongly to meiotic telomeres in all stages of prophase I, but not as filaments on the SC (Figure 1 - figure supplement 1). These findings are consistent with the genetic requirement of TRIP13 in meiosis in both sexes (Li and Schimenti, 2007; Roig et al., 2010).

Two previous studies utilizing two Trip13 hypomorphic (gene trap) alleles, one moderate and one severe, revealed a role for TRIP13 in meiotic recombination (Li and Schimenti, 2007; Roig et al., 2010). However, while the Trip13 moderate allele showed mostly intact chromosomal synapsis, the severe allele displayed chromosomal unsynapsis preferentially at chromosome ends (Li and Schimenti, 2007; Roig et al., 2010). Upon close examination, we also found unsynapsed ends in spermatocytes with the moderate Trip13 mutant (Guan et al., 2020). To rigorously ascertain the role of TRIP13 in meiosis, we generated Trip13-null mutants using frozen sperm from Trip13^+/- males with a knockout allele from the International Mouse Phenotyping Consortium (IMPC). The mouse Trip13 gene consists of 13 exons. This new Trip13 mutant allele harbors a deletion of a 24-kb region including all 13 exons and thus is expected to be null. We verified this deletion by PCR and sequencing. Interbreeding of Trip13^+/- mice produced fewer Trip13^-/- offspring than expected: Trip13^+/+, 80; Trip13^+/-, 220; Trip13^-/-, 51 (χ2=27, p=0.0001). The Trip13^-/- testis was much smaller than Trip13^+/- testis (Figure 2A). Western blot analysis showed that TRIP13 was present in reduced abundance in Trip13^+/- testis and not detected in Trip13^-/- testis, demonstrating that this mutant allele is null (Figure 2B). The abundance of SYCP3, a component of the synaptonemal complex, was also reduced in Trip13^-/- testis, suggesting a partial depletion of meiotic cells (Figure 2B). The testis weight of adult Trip13^-/- males was reduced by 74% in comparison with the wild type males (Figure 2C). Adult Trip13^-/- males lacked sperm in the epididymis (Figure 2D). Histological analysis showed that while spermatocytes in all stages of meiosis were present in adult Trip13^+/+ and Trip13^+/- testes, Trip13^-/- testis displayed complete meiotic arrest, evidenced by the presence of early spermatocytes and a lack of secondary spermatocytes, round spermatids, and mature sperm (Figure 2E). Trip13^-/- females were also sterile. The adult Trip13^-/- ovary was very small and showed a complete loss of oocytes (Figure 2 - figure supplement 1). These observations are similar to the meiotic arrest phenotype observed in the hypomorphic Trip13 mouse mutants (Li and Schimenti, 2007; Roig et al., 2010).

We investigated the effect of Trip13 deficiency on meiotic progression. The transverse filament/central element protein SYCP1 and the lateral element protein SYCP3 were used to determine the stage of prophase I spermatocytes. While the P20 wild type testis contained spermatocytes from leptotene through diplotene stages, about half of the Trip13^-/- spermatocytes were in the early pachytene stage and no cells were found at later stages of prophase I, showing meiotic blockade at the early pachytene stage in Trip13^-/- testes (Figure 3A). The early pachytene-like spermatocytes from Trip13^-/- testes contained unsynapsed chromosomal ends (Figure 3B). Mouse centromeres are telocentric. Co-staining with the centromere marker CREST showed that 94% of unsynapsed ends were centromeric ends (Figure 4A). While XY chromosomes were synapsed at the pseudoautosomal regions and thus were connected in wild type and Trip13^+/- pachytene spermatocytes, they were separate in Trip13^-/- pachytene-like spermatocytes (Figure 3B and Figure 3C). We found that SYCE1, a component of the SC central element, localized to the synapsed SC but not to unsynapsed regions in Trip13^-/- pachytene-like cells (Figure 3C). Confocal microscopy with super-resolution deconvolution confirmed that many homologous chromosomes in Trip13^-/- spermatocytes had split ends and some had regions of interstitial asynapsis (Figure 3D). We quantified these meiotic defects in spermatocytes: chromosomal asynapsis (Figure 3E), the number of asynapsed ends (Figure 3F), and the extent of XY asynapsis (Figure 3G). Previous studies also reported synaptic defects in spermatocytes from Trip13 hypomorph mutants (Li and Schimenti, 2007; Roig et al., 2010). Here we found that global loss of Trip13 caused similar defects in autosomal synapsis but a more severe defect in sex chromosome synapsis.

Although Trip13^+/- males were fertile, their testis weight and sperm count were significantly reduced in comparison with wild type (Figure 2C-D). The percentage of spermatocytes with defects in synapsis was higher in Trip13^+/- males than wild type (Figure 3E). In addition, the percentage of spermatocytes with asynapsed XY chromosomes was significantly higher in Trip13^+/- males than wild type (Figure 3G). As defects in synapsis can activate the meiotic checkpoint and thus apoptosis of affected spermatocytes, the increased synaptic defects likely caused the decrease in sperm output in Trip13^+/- males. These results demonstrate that TRIP13 regulates meiosis in a dosage-dependent manner.

TRIP13 localizes to telomeres and loss of TRIP13 causes peri-centromeric/telomeric asynapsis. Therefore, we asked whether telomere or centromere proteins were affected in Trip13^-/- spermatocytes. Chromosome spreads were immunostained for two centromere markers, CREST and CENPC (Figure 4A-B). CENPC is essential for recruiting kinetochore proteins to the centromere (Klare et al., 2015; Kwon et al., 2007). CREST and CENPC still localized to centromeres in Trip13^-/- spermatocytes. Because many Trip13^-/- spermatocytes had split ends, two CREST or two CENPC foci were observed at the split ends. To assess the telomere, spermatocytes were immunostained for TRF1 and MAJIN (Figure 4C-D). In meiotic cells, telomeres contain canonical telomere proteins such as TRF1/TRF2 and a meiosis-specific complex (MAJIN, TERB1, TERB2, and SUN1) (Shibuya et al., 2015). TRF1 is a telomere protein expressed in both somatic and germ cells (Karlseder et al., 2003; Long et al., 2017; Shibuya et al., 2014). MAJIN is a component of the meiosis-specific telomere complex that is important for telomere attachment to the inner nuclear membrane (Shibuya et al., 2015). Both TRF1 and MAJIN localized to telomeres in Trip13^-/- spermatocytes (Figure 4C-D). Taken together, these results suggest that TRIP13 is not required for recruitment of these centromere or telomere proteins.

In wild type meiotic cells, HORMAD1 and HORMAD2 localize to unsynapsed and desynapsed chromosomes (Fukuda et al., 2010; Kogo et al., 2012a; Shin et al., 2010; Wojtasz et al., 2012). Thus, HORMAD1 and HORMAD2 localized to the largely unsynapsed XY but not to synapsed autosomal SCs in wild type cells (Figure 5A-B). TRIP13 is essential for removing meiotic HORMADs from the chromosome axes, a function that is conserved in yeast, worms, and mammals (Wojtasz et al., 2009). We confirmed that HORMAD1 and HORMAD2 remained on the synapsed autosomes in Trip13^-/- pachytene cells (Figure 5A-B). HORMAD1 and HORMAD2 localize to the interior of the lateral elements in the SC (Xu et al., 2019). Confocal microscopy with super-resolution deconvolution revealed that HORMAD1 and HORMAD2 localized to the lateral elements of the synapsed autosomes in Trip13^-/- spermatocytes (Figure 5C-D). These results were consistent with accumulation of HORMAD1/2 in Trip13 hypomorphic mutant spermatocytes (Roig et al., 2010; Wojtasz et al., 2009). Our analysis of Trip13-null mutant confirmed that TRIP13 is essential for HORMAD1/2 removal from the synapsed SCs.

TRRIP13 localizes to the SC in early pachytene spermatocytes. We sought to address what proteins might recruit TRIP13 to the SC. We examined the role of HORMAD1, REC8, SYCP2, and SKP1 in TRIP13 localization in spermatocytes using respective knockout mice (Figure 6). We first examined the localization of TRIP13 in Hormad1^-/- spermatocytes. Hormad1^-/- spermatocytes exhibit limited synapsis (Daniel et al., 2011; Kogo et al., 2012b; Shin et al., 2010). In both wild type and Hormad1^-/- cells, TRIP13 localized to the synapsed regions and to the telomeres (Figure 6A). This result shows that HORMAD1 is not required for localization of TRIP13 to the SC.

REC8, a meiosis-specific cohesin, promotes synapsis between homologs and inhibits synapsis between sister chromatids (Xu et al., 2005). In the absence of REC8, synapsis occurs between sister chromatids rather than homologous chromosomes. In Rec8^-/- spermatocytes, TRIP13 localized to synapsed sister chromatids (Figure 6B). To further probe TRIP13 recruitment to the SC, a mouse line expressing a truncated SYCP2 was used (Yang et al., 2006). SYCP2 and SYCP3 are essential components of the axial/lateral elements of the SC and interact with each other. The mutant used in this study (referred to as Sycp2^-/-) expresses a truncated SYCP2 protein that lacks the C-terminal coiled-coil domain necessary for binding to SYCP3. Thus, SYCP3 failed to localize to the axial elements and formed aggregates in Sycp2^-/- spermatocytes (Figure 6C). Nevertheless, Sycp2^-/- spermatocytes formed short stretches of synapsis (Figure 6C). TRIP13 localized as filaments in Sycp2^-/- spermatocytes, strongly suggesting that its localization is independent of SYCP3 and the C-terminus of SYCP2 (Figure 6C).

Previous work has shown that SKP1, a key component of the SKP1, Cullin, F-box (SCF) complex E3 ligase, is important for HORMAD removal and chromosomal synapsis (Guan et al., 2020; Guan et al., 2022). SKP1 localizes to synapsed regions in meiotic germ cells and specifically to the lateral elements in the SC (Guan et al., 2020). We found that TRIP13 still localized to the synapsed regions in Skp1^cKO (conditional knockout) spermatocytes (Figure 6D). Taken together, these results show that the SC localization of TRIP13 is independent of HORMAD1, REC8, SYCP2, SYCP3, and SKP1, which all localize to the SC lateral elements. Such a finding is consistent with the localization of TRIP13 to the SC central region (Figure 1D).

In order to investigate the mechanism of TRIP13 recruitment to the SC, we generated 3×FLAG-TRIP13 (N-terminal tag) and TRIP13-3×FLAG (C-terminal tag) mice through the CRISPR/Cas9-medidated genome editing approach (Figure 7 - figure supplement 1). Two different types of tagged mice were generated in case that one fusion protein is not functional. Both alleles were transmitted through the germline of the founder mice. Western blot showed that TRIP13 existed in two isoforms in wild type (non-tagged) testis (Figure 7A). The major TRIP13 isoform was 50 kDa. The minor isoform was slightly larger than 50 kDa. Western blot analysis showed that the FLAG tagged TRIP13 fusion proteins were present in both FLAG-tagged testes. Both FLAG-tagged TRIP13 fusion proteins also existed in two isoforms in the homozygous testes, suggesting that they corresponded to the two wild type isoforms. The 3×FLAG-TRIP13 proteins were apparently slightly larger, possibly due to the linker in the N-terminally tagged proteins. The nature and physiological significance of these two isoforms were not clear, but could be due to post-translational modification or alternative splicing. Immunofluorescence analysis using anti-FLAG antibody showed that the FLAG tagged TRIP13 proteins, like wild type TRIP13, localized to both telomeres and synaptonemal complex in pachytene spermatocytes from both FLAG/FLAG homozygous testes (Figure 7B). Importantly, both N- and C-terminally tagged homozygous mice were fertile. These results demonstrate that both N- and C-terminal tagged TRIP13 proteins localize normally and are functional.

To identify TRIP13-associated proteins in testis, we performed immunoprecipitation (IP) using 3×FLAG-TRIP13, TRIP13-3×FLAG, and wild type (no tag) testicular protein extracts with anti-FLAG monoclonal antibody. The immunoprecipitated proteins were eluted with FLAG peptides and subjected to mass spectrometry for protein identification. As expected, TRIP13 had more peptides in tagged TRIP13 IP than wild type (Table 1). HORMAD2, a known TRIP13 substrate protein, was also enriched in tagged TRIP13 IP (Table 1). Intriguingly, a large number of RNA-binding proteins involved in RNA splicing were highly enriched in tagged TRIP13 IP: DDX46, PUF60, RBM25, RBM39, U2AF1, U2AF2, and SRSF11 (Table 1). The biological relevance of these RNA-binding proteins to TRIP13 function remains unknown. We cannot exclude the possibility that identification of these RNA splicing factors could be immunoprecipitation artifacts.

Mice were maintained and used for experimentation according to the protocol approved by the Institutional Care and Use Committee of the University of Pennsylvania.

The cryopreserved sperm from Trip13^+/- males were obtained from the Mutant Mouse Resource and Research Center (MMRCC) at UC Davis (Trip13^{tm1.1(KOMP)Vlcg/JMmucd}, MMRRC_050223-UCD). Genotyping of wild type and knockout Trip13 alleles was performed by separate PCR reactions using tail genomic DNA. Other mutant mouse lines used in this study were previously generated: Hormad1, Sycp2, and Rec8 (Guan et al., 2020; Shin et al., 2010; Yang et al., 2006). Genotyping PCR primer sequences are listed in Table 2.

Generation of 3×FLAG-Trip13 and Trip13-3×FLAG knockin mouse strains

The 3×FLAG-knockin mouse strains were generated by the CRISPR-Cas9 mediated genome editing approach. To generate N-terminally tagged knockin mice, one single-guide RNA (sgRNA) was designed to target the first exon of the mouse Trip13 (Figure 7 – figure supplement 1A). A ssDNA repair template was created using the sequence surrounding the start codon as 5′ and 3′ homology arms. The knockin template itself included the sequence for the 3×FLAG epitope and a three-residue linker sequence. The C- terminally tagged strain was generated similarly (Figure 7; figure supplement 1B). The guide RNA targeted exon 13 containing the stop codon. The ssDNA template contained the FLAG-encoding sequences and homology sequences flanking the insertion site. sgRNA sequences and ssDNA template sequences were shown in Table 2.

For the sgRNA, the oligo was phosphorylated, annealed, and cloned to PX330 plasmid (Addgene, Waterton, MA). After in vitro transcription with the MEGAshortscript T7 Kit (AM1354, Invitrogen) and purification with the MEGAclear Transcription Clean-Up Kit (AM1908, Invitrogen), a mixture of Cas9 mRNA [100ng/µl; Trilink, catalog number L-7206) + 0.5 µl of the sgRNA (35 ng/µl) + 100ng/ul of ssDNA template] was prepared and injected into zygotes. The injected zygotes were cultured in KSOM medium at 37°C in a 5% CO2 incubator until the two-cell stage. The two-cell embryos were transferred into oviducts of 0.5-day post-coitum pseudopregnant ICR foster mothers. Founder mice were bred to wild type mice to obtain germline transmission. The N-terminal 3×FLAG allele and the C-terminal 3×FLAG allele were PCR amplified and sequenced to confirm the insertion. PCR genotyping primers are listed in Table 2.

The short isoform of mouse MAJIN (amino acids 1-124; XM_036161650.1 and XP_036017543.1) was expressed as a 6xHis-MAJIN recombinant protein in E. coli using the pQE-30 vector. The recombinant protein was affinity purified with the Ni-NTA agarose. Two guinea pigs were immunized at Cocalico

Biologicals Inc. (Reamstown, PA), resulting in two antisera (UP-GP140 and UP-GP141). Both anti-sera were used for immunofluorescence of nuclear spreads of spermatocytes.

For histology, testes or ovaries were fixed in Bouin’s solution at room temperature overnight, embedded with paraffin, and sectioned at 8 μm. Sections were stained with hematoxylin and eosin. For immunofluorescence analysis, testes were fixed in 4% paraformaldehyde (in 1×PBS) overnight at 4°C, dehydrated in 30% sucrose (in 1×PBS) overnight, and sectioned at 8 μm in a cryostat. Surface nuclear spread analysis was described before (Peters et al., 1997). Briefly, testicular tubules or ovarian tissues were soaked in hypotonic treatment buffer (30 mM Tris, 50 mM sucrose, 17 mM trisodium citrate dihydrate, 5 mM EDTA, 0.5 mM DTT, 1 mM PMSF). Then the cells were suspended in 100 mM sucrose and spread by physical disruption on “PTFE” printed slides that were previously soaked with paraformaldehyde solution containing Triton X-100 and sodium borate. Antibodies used for immunostaining are listed in Table 3.

Histological images were captured on the Leica DM5500B microscope with a DFC450 digital camera (Leica Microsystems, Wetzlar, Germany). Most immunolabeled chromosome spread images were taken on the Leica DM5500B microscope with an ORCA Flash4.0 digital monochrome camera (Hamamatsu Photonics, Bridgewater, NJ). Confocal microscopy of immunolabeled chromosome spreads was performed on a Leica SP5 II confocal (Leica Microsystems, Wetzlar, Germany) with an 100× (1.46 NA) oil immersion objective lens. Images were deconvolved with Huygens Essential deconvolution software (Scientific Volume Imaging B.V., Hilversum, Netherlands).