During meiosis I, ring-shaped cohesin complexes play important roles in aiding the proper segregation of homologous chromosomes. RAD21L is a meiosis-specific vertebrate cohesin that is required for spermatogenesis in mice but is dispensable for oogenesis in young animals. The role of this cohesin in other vertebrate models has not been explored. Here, we tested if the zebrafish homolog Rad21l1 is required for meiotic chromosome dynamics during spermatogenesis and oogenesis. We found that Rad21l1 localizes to unsynapsed chromosome axes. It is also found between the axes of the mature tripartite synaptonemal complex (SC) in both sexes. We knocked out rad21l1 and found that nearly all rad21l1^-/- mutants develop as fertile males, suggesting that the mutation causes a defect in juvenile oogenesis, since insufficient oocyte production triggers female to male sex reversal in zebrafish. Sex reversal was partially suppressed by mutation of the checkpoint gene tp53, suggesting that the rad21l1 mutation activates Tp53-mediated apoptosis or arrest in females. This response, however, is not linked to a defect in repairing Spo11-induced double-strand breaks since deletion of spo11 does not suppress the sex reversal phenotype. Compared to tp53 single mutant controls, rad21l1^-/- tp53^-/- double mutant females produce poor quality eggs that often die or develop into malformed embryos. Overall, these results indicate that the absence of rad21l1^-/- females is due to a checkpoint-mediated response and highlight a role for a meiotic-specific cohesin subunit in oogenesis but not spermatogenesis. 

Materials and Methods

Ethics statement

            The UC Davis Institutional Animal Care and Use Committee (IACUC) has approved of this work under the protocol #20199; For noninvasive procedures (e.g. fin clips for genotyping), zebrafish were anesthetized using tricaine. Invasive surgical methods were performed on fish euthanized by submerging fish in ice water.

Data availability

            All data and related metadata underlying the findings have been deposited and are available at DRYAD (doi:10.25338/B8V91Q).

Zebrafish strains

            Zebrafish husbandry was performed as previously described [78]. The wild-type NHGRI strain was used in the production of the rad21l1^uc89 mutants. Fish used in experiments were outcrossed to the AB strain background 3-4 times. The spo11^-/- strain is in the AB background and described in Blokhina 2019. The tp53^-/- mutant is described in [63]. All test crosses were performed with wild-type AB strain fish. 

rad21l1^-/- mutant generation

            The rad21l1^uc89 mutants were generated using transcription activator-like effector nucleases (TALENs) to target exon 2 and genotyped using high resolution melt analysis (HRMA). TALEN target sequences: NG-NI-NG-NH-HD-HD-HD-NI-NI-HD-NG-HD-NG-NG-HD-NI-HD-HD-half repeat NG and NH-HD-NH-NI-NH-HD-HD-NI-NH-NI-NG-NG-NG-NG-NH-NH-HD-NH-half repeat NI. Injected founder fish were raised to adulthood and outcrossed to wild-type fish. The resulting offspring were screened for mutations in rad21l1 via HRMA and subsequent sequencing. HRMA primer sequences are: Fwd: 5’-CGCCGAGACATGTTTTATGCCC-3’, Rev: 5’-TCAAACACGTGGGCTTTGGT-3’. HRMA was performed with 20X Eva Green dye (VWR, Radnor, PA, Catalog #89138-982) using a CFX-96 real time PCR machine and Precision Melt Analysis software (BioRad, Hercules, CA). Mutants were backcrossed to either AB or NHGRI strain. The sex reversal phenotype was specific to populations genotyped as rad21l1^-/- indicating that it is unlikely due to off-target effects. The phenotype correlation remained consistent through 5-6 crosses. 

Genotyping

            Mutant identification: Genomic DNA was extracted and samples were analyzed with HRMA [42]. Primers for Rad21l1 genotyping were the same as described in the rad21l1 mutant generation. Primers for Spo11 were Fwd: 5’-TCACAGCCAGGATGTTTTGA -3’ and Rev: 5’-CACCTGACATTGCAGCA-3’ with an annealing temperature of 61° C. Primers for Tp53 were Fwd: 5’-CTCCTGAGTCTCCAGAGTGATGA-3’ and Rev: 5’-ACTACATGTGCAATAGCAGCTGC-3’. Genomic DNA was extracted and samples were analyzed as described for rad21l1 mutants except that the reaction was done in 2 mM MgCl₂ with an annealing temperature of 65° C. Two HRMA runs were required to confirm the three genotypes resulting from a tp53^+/-  incross; the first run distinguished heterozygous from homozygous (wild-type and mutant) samples. Homozygous samples were run again under the same conditions but spiked with wild-type DNA to differentiate wild-type and mutant samples.

Antibody generation

            Guinea pig anti-zebrafish Rad21l1 polyclonal antibody production: An N-terminal fragment of Rad21l1 cDNA was amplified with Phusion DNA polymerase (Thermo Fisher Scientific, Catalog #: M0530L) using the following primers: Fwd: 5’-aactttaagaaggagatataccatgTCAAGCTTTTGCCTTCCTGT-3’ and Rev: 5’-tctcagtggtggtggtggtggtgctcAAGCATGCAGAAAAATAAGGCT-3’. The Rad21l1 PCR product was then cloned into pET28b using NEBuilder HiFi DNA Assembly Master Mix (NEB, Catalog #: E5520S). BL21 (DE3) cells containing pRARE and Rad21l1 overexpression construct were grown in 2.6 L of LB with kanamycin and chloramphenicol until an OD600 = 1 and induced with a final concentration of 1 mM IPTG at room temperature for six hours. The Rad21l1 peptide was purified under denaturing conditions using Novagen NiNTA purification resins (Sigma, Catalog #: 70666) according to the manufacturer’s instructions. The Rad21l1 peptide was concentrated to a final concentration of 1 mg/ml in PBS using a 10 kDa centrifugal filter (Sigma, Catalog # UFC901008). The Rad21l1-derived peptide was injected into three guinea pigs by Pocono Rabbit Farm and Laboratory following the 91-day polyclonal antibody production protocol. 

Chromosome spreads and staining

            All chromosome spreads and staining were performed as previously described [42,79]. Antibodies and dilutions described in S1 Table. 

Adult testis section and whole mount preparation and staining

            Protocols including “whole mount testes staining “ and “testes section preparation and staining” were performed as previously described [42]. Antibodies and dilutions described in S1 Table. 

 Whole mount juvenile gonad staining 

            Juvenile gonad staining was performed similarly to the adult protocol with some modifications:

Dissection and fixation: Euthanized fish were decapitated and cut open along the ventral midline to expose the body cavity. Alternatively, an additional cut was made at the anal fin to expose body cavity if fish was too small to make a ventral cut. The fish were fixed in 4% PFA in PBT at 4° C for 16-18 hours with gentle rocking. The fish were placed into fresh tubes and washed in 0.2% PBT 3 times for a minimum of 5 minutes each. Gonads were dissected out in PBT and placed into a ceramic 12-well plate, 1 gonad per well.

Primary antibody staining: Gonads were washed in an antibody block composed of 5% goat serum and 5% BSA in 0.2% PBT for 1 hour minimum on a 2D rocker at room temperature. Primary antibody chicken anti-Ddx4 [42] was added at 1:500 final dilution. Plate was left rocking gently overnight at 4° C.

Secondary antibody staining: The gonads were washed 3 times for a minimum of 30 minutes in PBT, then washed in antibody block as described above. Secondary antibody anti-chicken Alexa Fluor 488 was added at 1:300 final dilution. Plate was left rocking gently overnight at 4° C.

Glycerol dehydration and mounting: The gonads were washed 2 times for a minimum of 10 minutes each and dehydrated in a series of glycerol (Sigma-Aldrich, Catalog #: G5516-1L) washes for 1 hour minimum each: 30% glycerol with DAPI at 1:5000 dilution in PBT, 50% glycerol with DAPI at 1:5000 dilution in PBT, and 70% glycerol in PBT without DAPI. The gonads were mounted in 70% glycerol without DAPI on slides with vacuum grease applied to the four corners to hold the coverslip in place. Slides were stored at 4° C until imaging.

 Histology: The mutant and control females were crossed with wild-type males the morning of fixation to reduce the overall numbers of eggs in the ovaries. Animals were euthanized and decapitated posterior to the pectoral fins and cut open the ventral side to the genital pore using scissors and placed in Bouin's fixative overnight at room temperature with gentle shaking (10ML/fish). H&E staining of paraffin-embedded gonads was done according to [80]. Images were taken at 5X magnification on a Zeiss Axiophot microscope using a Leica DFC500 camera.

Imaging

           Images of chromosome spreads and whole mount gonads stained with Ddx4 (Vasa) and DAPI were collected at the Department of Molecular and Cellular Biology Light Microscopy Imaging Facility at UC Davis. Chromosomes spreads were imaged using the Nikon N-SIM Super-Resolution microscope in 3D-SIM imaging mode with APO TIRF 100X oil lens. The images were collected and reconstructed using the NIS-Elements Imaging Software. Sections and fluorescent whole mounts were imaged using the Olympus FV1000 laser scanning confocal microscope. Images were processed using Fiji ImageJ software. Only linear modifications to brightness and contrast of whole images were applied. Images of eggs and embryos were acquired using a dissecting microscope.

 Test crosses

            To analyze fertility, individual mutant fish were placed in a divided mating tank overnight with a single AB strain wild-type fish of the opposite sex. The divider was removed soon after onset of light, and any eggs produced were collected with a strainer, rinsed thoroughly with system water, and placed in a petri dish at 30° C. At 6 hours post fertilization (hpf), embryos were transferred to embryo medium (1X E3 media has final concentrations of 5 mM NaCl, 0.17 mM KCl, 0.3 mM CaCl₂ dihydrate, 0.33 mM MgSO₄ heptahydrate, 6 µM methylene blue) and categorized. Fertilized eggs were kept at 30° C and monitored at 24 and 48 hpf for morbidity and mortality. All data related to graphs in Figs 2-4, and 5 are reported in S1 File.

RNA extraction

            Testes were dissected from as described above and transferred to tubes containing 200 µL Tri Reagent Solution (ThermoFisher, Catalog #AM9738). Individual testes were homogenized with a plastic pestle (USA Scientific, Catalog #1415-5390) and every 2 minutes over 10 minutes and kept at room temperature. 20 µL of 1-Bromo-3-chloropropane (BCP) was added to the samples then vortexed and incubated at room temperature for 5 min before a 15 minute centrifugation at 14,000 RPM. 80 µL of the top clear layer was transferred to a fresh RNase-free 1.5 mL tube. 0.8 µL of 20 µg/mL glycogen (ThermoFisher, Catalog #R0551) and 80 µL of 100% isopropanol were added to each tube and the samples were placed at -20º C overnight. The tubes were centrifuged at 14,000 RPM for 30 min and the supernatant was removed. The pellets were washed with 300 µL of 75% ethanol and allowed to air dry. RNA pellets were resuspended in 7 µL of nuclease-free water (ThermoFisher, Catalog #4387937). RNA was quantified using a NanoDrop 1000 Spectrometer (ThermoFisher). RNA was treated with DNase I to remove genomic DNA (ThermoFisher, Catalog #18068-015). First-strand cDNA was synthesized using SuperScript First-Strand Synthesis System for RT-PCR (ThermoFisher Catalog #11904-018).

RT-qPCR

qPCR reactions were prepared with SsoAdvanced Universal SYBR Green Supermix (Bio-Rad, Catalog #1725271) using cDNA and the following primers (final concentration, 0.05 µM): eF1a (Housekeeping) Fwd: 5’-CTACCTACCCTCCTCTTGGTCG-3’, Rev: 5’-CCTTAAGTAGAGTGCCCAGGT-3’; Rad21a Fwd: 5’-TACCTGCATAGTGAGATGTTCTGT-3’, Rev: 5’- ACAGAACAATGGAGGAAAAACAAC-3’; Rad21b Fwd: 5’-TATCCGTGCATGTGCATTTT-3’, Rev: 5’-CCTCTGGCTACATGATTTGC-3’; Rec8a Fwd: 5’-TGGTGAAGCCTATCCCTCCA-3’, Rev: 5’-CTTCTGGCTCTGGTGGTTGT-3’; Rec8b Fwd: 5’-AGATTCCCCCAAGCAAGTTCA-3’, Rev: 5’-ACAAACTGCATTTAAACTGACCTCT-3’.  Cq values were determined for each primer pair and normalized to eF1a control.  The fold change between mutant and wild-type was analyzed using the 2^-ΔΔCt method. Raw Cq values and calculations are in S2 File.

Fig 1. Rad21l1 expression and loading. (A) Rad21l1 loading during prophase I of meiosis in spermatocyte nuclear surface spreads. Rad21l1 (magenta) loads onto chromosome axes simultaneously with Sycp3 (green) and is also dispersed as foci throughout the spread in leptotene. In early zygotene, Sycp1 (cyan) lines start near the telomeres and synapsis extends inward through late zygotene and are present end-to-end along axes. Note: there is some asynapsis in the pachytene which may indicate that the cell was either in very early or very late pachytene. The merged images are Rad21l1 and Sycp3 channels only. Mag images are magnifications from the Merge panels; the regions magnified are indicated by white boxes. Panel series a-p scale bar = 5 µm. Mag panel series q-t scale bar = 2 µm. (B) Rad21l1 loading during prophase I of meiosis in oocyte nuclear surface spreads. Panels a-e are arranged similarly to the corresponding panels of part (A). 

Fig 2. rad21l1^-/- mutants are predominantly male due to late sex reversal. (A) TALEN generated 17-bp deletion leads to a frameshift mutation resulting in a truncated 27 amino acid (aa) Rad21l1 protein with the conserved Rec8/Rad21-like family domains (1-100 aa and 495-543 aa) disrupted or deleted. Rec8/Rad21-like domains (purple boxes); altered amino acid sequence (red box). The ATG translational start site is located at the 4th-6th nucleotide from the end. (B) Spermatocyte nuclear spreads stained for telomeres (cyan), Sycp3 (green), and Rad21l1 (magenta). Rad21l1 forms lines of foci along the Sycp3 axis in rad21l1^+/- spermatocytes. In the rad21l1 mutant, no lines of Rad21l1 foci are seen. The rad21l1 mutant spermatocytes can form axes and pair homologs albeit with some asynapsed regions. Scale bar = 5 µm. (C) Sexed offspring of a rad21l1^+/- incross show a depletion of females in rad21l1^-/- fish. Data pooled from multiple crosses. (D) Sections of gonads prepared from 35-36 dpf rad21l1^+/+ and rad21l1^+/- (labelled rad21l1⁺) and rad21l1^-/- fish and stained for DNA (gray) and Ddx4 (also known as Vasa; green). At 35-36 dpf, oocytes are present in 10/12 rad21l1^+/- and 6/11 rad21l1^-/- samples. Scale bar = 50 µm. Arrows represent different cell types: (pink- diplotene oocytes, yellow- spermatocytes / sperm; red- premeiotic germ cells). (E) Whole mounts of gonads from 40 and 45 dpf are stained for DNA (gray) and Ddx4 (green). At 40 dpf, oocytes are present in 11/21 rad21l1^+/+ and 1/19 rad21l1^-/- samples. At 45 dpf, oocytes are present in 7/9 rad21l1^+/+ and 1/10 rad21l1^-/- samples. Scale bar = 50 µm. Fisher’s exact test used for all statistical analysis. ns = p>0.05, ** = p<0.01, **** = p<0.0001. 

Fig 3. Rad21l1 is dispensable for male fertility. (A/B) Data resulting from test crosses between rad21l1^-/- males and wild-type females to assess fertility and reproductive phenotype. rad21l1^+/+ male tank mates were used as controls. No significant difference in the number of eggs the males caused the females to release, the composition of the resulting clutch at 6 hpf, or the survival of the embryos through 48 hpf. Data pooled from 14 crosses over 5 weeks using the same pool of 14 rad21l1^-/- males, 12/14 of which crossed successfully at least once. Unpaired, two-tailed student t-test used for statistical analysis, ns = p>0.05. (C) Whole mount adult testes stained for DNA (gray) and Ddx4 (green) showing a phenotypic range of gonad morphology in rad21l1^-/- males. All samples except #5 displayed large clusters of mature sperm. Images marked as A and B were taken from the same sample to show variation within a single gonad. Wild-type tank mates used as controls. Scale bar = 50 µm. (D) Testes sections stained with a PNA telomere probe (Tel; magenta), an antibody to ?H2AX (green), and DAPI (blue), showing that telomere clustering and DSB localization (?H2AX) are normal in the rad21l1 mutant. Scale bar = 5 µm. (E) Fold change of relative mRNA expression for rad21l1 paralogs, rec8b, rec8a, and rad21a in wild-type versus the rad21l1^-/- mutant (Mut). Fold change was determined using the 2^-ΔΔCq method. eF1a (housekeeping) mRNA levels were used as a reference. 

Fig 4. tp53 knockout restores females to rad21l1 mutant population, but rad21l1;tp53 double mutant females produce poor quality eggs and malformed embryos.

(A) Sex ratios of all genotypes resulting from a rad21l1^+/-;tp53^+/- incross. Data pooled from 3 crosses. Fisher’s exact test used for statistical analysis. 0, 1, and 2 on the x-axis refer to the number of wild-type copies of tp53 and rad21l1. (B) Data resulting from test crosses between rad21l1^-/-;tp53^-/- females and wild-type males to assess fertility and reproductive phenotype. rad21l1^+/+;tp53^-/- female tank mates used as controls. No significant difference in the number of eggs the females released. rad21l1^-/-;tp53^-/- double mutant females release a significantly greater percentage of eggs that fail to be fertilized or display premature decomposition. (C) Representative images of clutches from double mutant and control females at 6 hpf showing lower overall quality of eggs released from double mutant females. (D) Images i-iii show examples of eggs described in the text at 6 hpf. Panel i shows a normal egg (left) and a tiny egg (right). Panel ii shows prematurely decomposing eggs and panel iii shows opaque eggs. All images are the same magnification. (E) Of normal embryos at 6 hpf, 32.4% are dead or malformed at 24 hpf and 38.7% by 48 hpf. Unpaired, two-tailed student t-test used for statistical analysis. (F) Representative images showing the range of malformations seen in developing embryos from rad21l1^-/-;tp53^-/- females at 24 and 48 hpf. ns = p>0.05, * = p<0.05, ** = p<0.01, *** = p<0.001.

Fig 5. spo11;rad21l1 double mutants are infertile males. (A) Sex ratios of all genotypes resulting from a rad21l1^+/-;spo11^+/- incross. Data pooled from 6 crosses. ns = p>0.05. 0, 1, and 2 on the x-axis refer to the number of wild-type copies of spo11 and rad21l1. (B) Whole mount testes stained for DNA (gray) and Ddx4 (green). WT and rad21l1^-/-;spo11^+/+ samples display clusters of mature sperm, while rad21l1^+/+;spo11^-/- and rad21l1^-/-;spo11^-/- samples do not. Scale bar = 30 µm.

S1 Fig. Alignment of zebrafish Rad21l1, Rec8a, and Rec8b proteins. Alignment of zebrafish Rad21l1 (ENSDARP00000074083), Rec8a (ENSDARP00000116796), and Rec8b (ENSDARP00000091417) using the Snapgene (v 5.1.4.1) Clustal Omega tool. Yellow shading indicates amino acids of Rec8a and Rec8b that match the Rad21l1 references sequences. The consensus sequence threshold was set at > 50%. Amino acids 329-516 (highlighted) were expressed to create the Rad21l1 antibody in Guinea pigs.

S2 Fig. Rad21l1 loading during prophase I of meiosis in spermatocyte nuclear surface spreads. The images are blown up images shown in Fig 1 panel t.  A. Rad21l1 (magenta); Sycp3 (green). B. Rad21l1 (gray); Sycp1(blue). Scale bar = 5 µm. 

S3 Fig. Schematic of the stages of oocyte and gonad differentiation during larval stages of development in zebrafish based on [56,81]. The gonad is considered bipotential starting at 5 dpf and by 30 days animals are differentiated as either female or male. Starting around 13 days, Stage IA oocytes appear, representing the leptotene (L) through pachytene (P) stages of meiotic prophase. The transition to form a testis begins ~20 dpf, after the animals have already started producing Stage IA oocytes. During testis transitioning (~20-30 dpf), the Stage IA oocytes undergo apoptosis and spermatogenesis ensues. In animals that go on to become females, the Stage IA oocytes develop further to Stage IB representing cells in diplotene (D). Note that differentiated females continue to form Stage IA oocytes through adulthood and do not rely solely on the pool generated when the gonad is the bipotential phase. 

S4 Fig. Oocytes in the rad21l1^-/- mutant reach the dictyate stage. Images are blown up from Fig 2E. The DAPI channel is enhanced to show the lampbrush chromosomes at the follicle stage in wild-type and mutant. The magenta arrows point to the lampbrush chromosomes at the diplotene stage.  

S5 Fig. H & E staining of gonad from a rare rad21l1^-/- female compared to wild-type. Stages of oocytes indicated (IB, II, III). (A) Ovary section from rad21l1^+/+ female; (B) Ovary section from rad21l1^-/-female. Scale bar = 500µm. 

S1 Table. Antibodies used in this study

S1 File. Master data sheet

S2 File. Final_qPCR_MasterSheet