Spermatozoa in mice lacking the nucleoporin NUP210L show defects in head shape and motility but not in nuclear compaction or histone replacement

Biallelic loss‐of‐function mutation of NUP210L, encoding a testis‐specific nucleoporin, has been reported in an infertile man whose spermatozoa show uncondensed heads and histone retention. Mice with a homozygous transgene intronic insertion in Nup210l were infertile but spermatozoa had condensed heads. Expression from this insertion allele is undefined, however, and residual NUP210L production could underlie the milder phenotype. To resolve this issue, we have created Nup210lem1Mjmm, a null allele of Nup210l, in the mouse. Nup210lem1Mjmm homozygotes show uniform mild anomalies of sperm head morphology and decreased motility, but nuclear compaction and histone removal appear unaffected. Thus, our mouse model does not support that NUP210L loss alone blocks spermatid nuclear compaction. Re‐analyzing the patient's exome data, we identified a rare, potentially pathogenic, heterozygous variant in nucleoporin gene NUP153 (p.Pro485Leu), and showed that, in mouse and human, NUP210L and NUP153 colocalize at the caudal nuclear pole in elongating spermatids and spermatozoa. Unexpectedly, in round spermatids, NUP210L and NUP153 localisation differs between mouse (nucleoplasm) and human (nuclear periphery). Our data suggest two explanations for the increased phenotypic severity associated with NUP210L loss in human compared to mouse: a genetic variant in human NUP153 (p.Pro485Leu), and inter‐species divergence in nuclear pore function in round spermatids.

haploid round spermatids that transform into mature spermatozoa during spermiogenesis.In this final phase, spermatids undergo extensive and unique structural modifications involving elongation and condensation of the nucleus, histone-to-protamine chromatin transition, acrosome formation, and flagellum biosynthesis to become highly specialized motile spermatozoa, 4 optimized for the faithful transmission of the paternal genetic and epigenetic information to the next generation.
6][7] The interpretation of genetic data can, however, be complicated by the genetic heterogeneity of male infertility which means that independent confirmatory cases are often not available, even in large studies of common phenotypes such as azoospermia. 8r this reason, candidate human gene mutations must often be transposed to an animal model to validate their pathogenicity, and elucidate the physiological role of the affected gene.
Recently, a homozygous loss-of-function mutation in the Nucleoporin 210 like gene (NUP210L) has been identified in an infertile man whose seminal analysis revealed spermatozoa with large uncondensed round heads that retain histones and have abnormal motility, indicating that the NUP210L protein contributes to chromatin compaction and flagellar biogenesis during human spermiogenesis. 9Such a phenotype is extremely rare and has only been described in this one case.
NUP210L is predicted to be a nuclear pore complex (NPC) protein and is predominantly expressed in the testis (https://gtexportal.org/ home/gene/NUP210L).Based on its primary sequence, its structure is similar to that described for its more widely expressed paralogue NUP210 (also known as gp210 or POM210): a peptide signal at the N-terminus, a large 1783 amino acid domain located in the luminal space between the two nuclear membranes, a transmembrane domain that enables NUP210 to be addressed to the nuclear pore and a short cytoplasmic C-terminal tail. 10 With NDC1 and POM121, NUP210, and NUP210L make up the known mammalian transmembrane nucleoporins (also known as pore membrane proteins-POMs). 11With its relatively large luminal domain, NUP210/NUP210L is probably the major component of the luminal ring that forms a cushion around the center of the NPC, possibly anchoring the NPC to the pore membrane and buffering its transport functions from forces in the nuclear envelope. 12 Human Protein Atlas, NUP210 and NUP210L appear to have complementary expression during human spermatogenesis, with NUP210 in spermatogonia and early spermatocytes, and NUP210L in late spermatocytes and spermatids (https://www.proteinatlas.org/),suggesting a specialized role for NUP210L in NPC function during spermiogenesis.
Male mice homozygous for a ROSA26-EGFP transgenic insertion in intron 17 of Nup210l have been reported to be infertile with elevated numbers of abnormal spermatozoa: immotile (86%) and mild sperm head defects (80%). 13In contrast to the human case, there was no evidence of nuclear compaction failure in the ROSA26-EGFP transgene insertion mouse.The effect of the ROSA26-EGFP insertion on NUP210L expression was not reported, however, and so residual expression of NUP210L or production of a functional truncated isoform might explain the milder phenotype in the mouse compared to the published human case, where the mutation is predicted to result in a complete loss of NUP210L function. 9 address this issue and determine whether NUP210L is essential for nuclear compaction, in mouse spermatids, we created a mouse mutant allele, Nup210l em1Mjmm .Here, we demonstrate that NUP210L is absent from germ cells in male mice homozygous for Nup210l em1Mjmm .
Their spermatozoa have a uniform moderately abnormal head morphology, frequent flagellar anomalies and decreased progressive motility.
Nevertheless, Nup210l em1Mjmm homozygous males are fertile, and do not show compromised histone removal or nuclear compaction failure, demonstrating that the Nup210l-KO mouse model does not validate the hypothesis that, in the human case, loss of NUP210L alone is sufficient to cause failure of germ cell chromatin condensation during spermiogenesis.We reveal that the human case is heterozygous for a rare missense variant with predicted pathogenicity in another nucleoporin gene encoding NUP153.We show that NUP153 and NUP210L colocalize to the nuclear periphery of elongating spermatids, in mouse and human, while in round spermatids they localize to the nuclear periphery in human but the nucleoplasm in mouse.We discuss how our findings might explain the exceptional nuclear compaction failure in the human case.

| Nup210l-knockout mice
Founder mice carrying the null allele Nup210l em1Mjmm were created using CRISPR-Cas9 methodology.Fertilized C57BL/6NCrl oocytes were injected with Cas9 nuclease, tracrRNA and the crRNA (guide RNA), at the transgenic animal service (SEAT) at Gustave Roussy.The targeting sequence of the crRNA was 5 0 -TACTAAGAGATATG-TATCGT-3 0 .For the creation and study of the Nup210l em1Mjmm model, the necessary ethical approval was obtained from the Animal experimentation Ethics committee of Marseille (Comité d' Ethique en Expérimentation Animale de Marseille-CEEA14).

| Human samples
Human testicular material (T0140007) came from a 50 year old man with brain death, within a protocol, approved by the French Research Minister, for the procurement of human tissue for use in research.We declared our protocol through the French Biomedicine Agency (Agence de la biomedicine) in January 2013.Testes were recovered during multi-organ retrieval for transplantation while the donor was under extracorporeal circulation and respiratory assistance.The donor had normal spermatogenesis based on histological analysis and presence of numerous epididymal and testicular spermatozoa.
The infertile man, 9 13-4587, was recruited at our public fertility center in Marseille and gave his informed consent that his DNA be used in research.DNA was obtained from the Centre of Biological Resources at La Timone Hospital, Marseille (CRB AP-HM).

| Mouse sperm analysis
The caudal epididymis was dissected into 2 mL of DMEM-HEPES, cut into small pieces and then incubated for 10 min at 37 C to liberate spermatozoa.Spermatozoa were counted and their motility assessed by direct microscopic observation.To assess morphology sperm smears were stained with SpermBlue, manufacturer's protocol (Microptic, Barcelona, Spain) and silver nitrate staining. 14

| Mouse histology
Testes and epididymides were collected from WT and Nup210l À/À mice, fixed in Bouin's solution overnight at room temperature and embedded in paraffin using standard protocols.Sections were stained with Mayer's hematoxylin, with periodic acid-Schiff (PAS) for testes or eosin for epididymides.Seminiferous tubule stages were determined according to Russell. 15

| Antibodies and lectin
The following primary antibodies were used for immunofluorescence (IF) and Western blot (WB) analyses at the given dilution: anti-NUP210L

| Statistical analyses
The significance of the difference between the means in study groups was calculated using Student's t-test.A p-value of <0.05 was considered significant.

| Additional methodology
Details are given in Supporting Information for: Mouse sperm analysis; Mouse Histology; Tissue preparation for Immunostaining; Immunostaining; Chromomycin A3 staining; Western blot analysis; and Transmission electron microscopy.

| NUP210L is not expressed in testis of Nup210l em1Mjmm/Mjmm KO mice
To study the biological function of NUP210L in vivo, we used CRISPR-Cas9 to create a mouse line carrying a null allele of Nup210l.
We selected an allele, Nup210l em1Mjmm (hereinafter referred to as Nup210l À ) with a single base insertion in exon 6 of the 40 exons that compose Nup210l (Figure 1A).To determine the transcripts produced by Nup210l À , we used RT-PCR to amplify across the targeted exon 6 (primers in exons 5 and 10).We obtained a single predominant band of the expected size for wild-type and Nup210l À heterozygotes but two bands were obtained for Nup210l À homozygotes (Figure 1B).Purifying and sequencing each RT-PCR product confirmed the presence of the insertion mutation in exon 6 and revealed the upper band to consist of all exons from 5 to 10, while the lower band did not include exon 6 (Figure 1C).Thus, the mutation reduces the efficiency of exon 6 splicing.Since exon 6 is 134 bp in length neither transcript produced from the Nup210l À allele preserves the Nup210l reading frame and both are predicted to be substrates for nonsense mediated decay.We conclude that Nup210l À is a null allele.
We confirmed that the NUP210L protein is not expressed in male mice homozygous for Nup210l À using immunofluorescence with an antibody raised against human NUP210L (HPA064245immunogen amino acids 530-613-exons 12-14; Figure 1D).In WT and heterozygous males, the antibody gave a diffuse granular nucleoplasmic signal in round spermatids that became concentrated at the posterior pole in late round and early elongating spermatids and was only detected at the caudal site in most elongating spermatids and spermatozoa.In testis from adult Nup210l À/À mice, no signal was observed on testicular germ cells, showing both that NUP210L is not produced from the Nup210l À allele and that the antibody HPA064245 detects NUP210L specifically by immunofluorescence.
Western blot analysis detected a weak band of approximately 210 kDa in WT and HTZ testis that is absent in Nup210l À/À testis.A strong band at 150 kDa, and a weaker band at 100 kDa were detected in all three genotypes showing that they are nonspecific (Figure 1E).

| Mice lacking NUP210L are fertile with normal testicular histology
Male and female Nup210l À/À mice were observed to develop normally and showed no signs of distress.To test fertility, male littermates Nup210l +/+ (n = 7) and Nup210l À/À (n = 8) were mated with two females for 10 weeks.No differences were seen in frequency of litters or litter size (Figure 2A).No age-dependent decrease in fertility was observed up to 6 months.The average weight of the testis from 6-month-old Nup210l À/À mice was comparable with that of Nup210l +/+ of the same age (Figure 2B).Nup210l À/À females also showed no signs of reduced fertility (normal litter size and frequency-unpublished data).
To determine the state of spermatogenesis in mice lacking NUP210L function, we performed histological analysis of testis sections stained with PAS-H.Histological analysis of Nup210l À/À testis revealed that all seminiferous tubules were rich in germ cells with abundant spermatozoa visible in the lumen (Figure 2D).The germ cell content and arrangement at each tubule stage was observed to be normal in Nup210l À/À adult males (Figure 3A).In addition, we quantified seminiferous tubule stages and did not observe a significant difference in their distribution between Nup210l À/À and Nup210l +/+ males (Figure 3B), indicating that no delay occurs during spermatogenesis in Nup210l À/À mice.

| Spermatozoa quality is diminished in mice lacking NUP210L
We next investigated spermatozoa quality in Nup210l À/À mice and during spermatogenesis (Figure 4A).In contrast, spermatozoon motility and morphology were affected in Nup210l À/À mice.We scored nonprogressive and progressive motility for six males of each genotype.All genotypes had spermatozoa with progressive motility.The Nup210l À/À males, however, consistently had a lower percentage of progressively motile spermatozoa compared to Nup210l +/+ littermates (Figure 4B).
The morphology of the sperm was examined by sperm blue and silver nitrate staining.Flagella of Nup210l À/À sperm frequently formed a hairpin loop (40%), bent in the distal mid piece, or were coiled (15%).These sperm tail abnormalities were significantly more common in Nup210l À/À mice than WT (Figure 4C,D).In addition, the sperm head was uniformly affected in homozygous mice with shortening of the acrosomal hook and narrowing of the base of the head (Figure 4D).
Transmission electron microscopy (TEM) was used to investigate the state of chromatin condensation as well as the organization and position of flagellar components.Epididymal sperm heads (n > 35) were examined for each of three Nup210l À/À males and one Nup210l +/À male.Sperm chromatin appears tightly compacted in both Nup210l +/À and Nup210l À/À mice, as evidenced by the high density of nuclear staining observed by TEM (Figure 5A).We observed the presence of straight, axoneme and the mitochondrial sheath appear normal in the straight and bent (double cross-section) flagella (Figure 5B).but aberrant in the coiled sections, as indicated by disordered axonemal microtubules (Figure 5C).
The head-tail junction appeared normal in spermatozoa with a straight or a coiled flagellum (Figure 5A,C).

| Normal chromatin compaction and histone removal in Nup210l À/À spermatozoa
To determine whether histone removal during spermatid elongation is affected in Nup210l À/À mice, we used immunofluorescence to track acetylated-histone H4 (acH4) (Figure 6A).We observed normal acH4 labeling in Nup210l +/+ and Nup210l À/À mice: the acH4 signal was first detected weakly throughout the nucleus of round spermatids at step 8, but became much stronger as spermatids began to elongate at step 9.As elongation progressed, the acH4 signal was gradually lost, receding from under the acrosome to the posterior nuclear pole in early condensed elongated spermatids, and was not detectable in later condensed spermatids or in spermatozoa.Our results indicate that histone removal from spermatid chromatin is unaffected in mice lacking NUP210L.
We then used chromomycin A3 (CMA3) staining to evaluate whether protamines were present in the chromatin of mature spermatozoa.CMA3 binds histonylated chromatin but not chromatin packaged with protamines.We observed that the frequency of CMA3-positive spermatozoa was low in both WT and Nup210l À/À spermatozoa (Figure 6B,C), indicating that the histone- to-protamine transition is essentially normal during spermiogenesis in Nup210l À/À mice.

| NUP210L-KO man carries potential pathogenic variant in nucleoporin NUP153
Our findings indicate that the absence of NUP210L in mice has no major effect on fertility, while it has been linked to a failure of nuclear remodeling and male infertility in humans. 9To search for a genetic basis for this discrepancy, we re-analyzed exome sequencing data for the infertile man (13-4587) looking for potentially pathogenic variants in known components of the NPC.We identified a heterozygous missense variant in the nucleoporin NUP153, p.Pro485Leu, that we verified by Sanger sequencing (Figure 7A,B).The affected proline is conserved in all mammals and the variant is predicted to be pathogenic by Polyphen2 (probably damaging) and SIFT (deleterious).In gnomAD, the frequency of NUP153-p.Pro485Leu is 0.0004 overall but, is highest (0.0022) in Bulgarian and southern European groups (n = 14 274 alleles).Based on a lower than expected number of LoF mutations (LOEUF = 0.174), gno-mAD estimates that there is a high probability of intolerance to loss of a single copy of NUP153 in human.Thus, if the p.Pro485Leu substitution has a negative effect on NUP153 function, a reduction in functional NUP153, coupled to complete loss of NUP210L, could explain the severity of the phenotype observed in the infertile man.

| NUP210L and NUP153 colocalize in mouse and human elongating spermatids
To evaluate if NUP210L and NUP153 could act together in human, we used IF analysis to determine their relative localization in spermatids.We observed that NUP210L localizes at the nuclear periphery of elongating spermatids, in human and mouse, and is gradually restricted to the posterior pole of the nucleus (Figure 7C).This caudal localization persists in both mouse and human spermatozoa.In round spermatids, however, NUP210L localization diverges between the two species: in human, NUP210L is restricted to the nuclear periphery except under the acrosome, while, in mouse, it is nucleoplasmic.
NUP153, like NUP210L, is nucleoplasmic in mouse round spermatids, but otherwise colocalizes perfectly with NUP210L, at the nuclear periphery in human round spermatids, and the caudal extremity in mouse and human elongating spermatids (Figure 7D).We also show that NUP153 localization in elongating spermatids is not altered in mice lacking NUP210L (Figure 7E).The colocalization of NUP153 and NUP210L shows that they could function together in elongating implies that species differences in NPC organization in round spermatids must be considered to explain the severe human phenotype.

| DISCUSSION
Loss of NUP210L function has recently been reported in an infertile man producing predominantly abnormal spermatozoa with large uncondensed heads and an elevated level of histone retention, indicating that NUP210L and the NPC play a central role in the remodeling of the spermatid nucleus during spermiogenesis. 9However, in a mouse model with a ROSA26-EGFP transgene insertion into intron 17 of Nup210l, homozygous males showed infertility but produced spermatozoa with condensed sperm heads. 13Since the effect of the ROSA26-EGFP intronic insertion on the expression of NUP210L was not determined, a possible explanation for this discrepancy was that the insertion did not completely inactivate Nup210l.Here, we have resolved this issue by creating a knockout mouse model in which we show that NUP210L is not expressed.Our results demonstrate that in the mouse the absence of NUP210L has no overt effect on histone removal or chromatin condensation, establishing that NUP210L loss in the mouse does not strengthen the case that loss of NUP210L alone is the cause of the striking failure of nuclear compaction observed in the human case. 9comparison with the published data for the ROSA26-EGFP 13 model indicates similarities between the sperm phenotypes of the two models: short acrosomal hook, tapering of the caudal extremity of the sperm head and a reduced percentage of progressively motile spermatozoa.Nonetheless, the homozygous ROSA26-EGFP male mice clearly show signs of a more severe phenotype: age-related degeneration of the seminiferous epithelium with reduced sperm production, acrosome anomalies and male infertility at all ages.The greater phenotypic severity in the ROSA26-EGFP model, compared to our model, may be due to differences in the genetic background or a dose-dependent negative effect of the ROSA26-EGFP allele, such as production of a truncated isoform of NUP210L or altered expression of a neighboring gene.Our knockout model presented here represents a reference for the phenotype associated with loss of NUP210L function in the mouse.
We identified that the infertile man 13-4587 9 is heterozygous for a potential pathogenic variant, p.Pro485Leu, in the nucleoporin NUP153, a critical component of the NPC nucleoplasmic basket.In somatic cells, NUP153 has diverse roles including regulating the export of mRNAs, 16 building chromatin architecture through interactions with CTCF and cohesins, 17  nuclear membrane. 18In spermatids, the NL is composed of B-type lamins and it is dismantled during spermiogenesis. 19,20NUP153 might therefore also facilitate B-type lamin extraction from the nucleus and contribute to the breakdown of the spermatid nuclear lamina.Of note in this regard, in Caenorhabditis elegans, the NUP210 orthologue is required for disassembly of the NL and NE breakdown during mitosis. 21Importantly, in human, we show here that NUP153 and Given the phenotype of large round uncondensed sperm heads, in the infertile man, it seems reasonable to assume that the primary defect occurs in his round spermatids as elongation begins.We show here that in round spermatids, NUP210L and NUP153 localize to the nuclear periphery in human, but to the nucleoplasm in the mouse.
Since this is observed for two nucleoporins, it favors the interpretation that NPC or NE organization in round spermatids is distinct between human and mouse.The localization of mouse NUP210L to the nucleoplasm is unexpected because NUP210 proteins carry a signal peptide to direct them to the NPC via the endoplasmic reticulum, 10 and mouse and human NUP210L proteins localize to the nuclear periphery when ectopically expressed in HeLa cells (unpublished data).This indicates that the signal peptide carried by mouse NUP210L may not be efficient in round spermatids, and indeed Signa-lIP 6.0 does not predict the peptide for mouse NUP210L, but does for human NUP210L (unpublished data). 22Our data show that in human, but not in mouse, NUP210L could play a role at the nuclear envelope in round spermatids, and this may underlie the greater phenotypic severity observed with the loss of NUP210L function in human compared to mouse.
In elongating spermatids, in both mouse and human, we show that NUP210L and NUP153 concentrate at the caudal pole of the nucleus.This is entirely consistent with electron microscopy studies in mouse and bovine showing that NPCs shift into a high-density array at the caudal pole when the manchette forms, and the spermatid nucleus begins to elongate. 23,24This NPC reorganization to the caudal pole indicates that this must represent the major site of exchange between the nucleoplasm and the cytoplasm during the elongation steps of spermatid differentiation.NUP210, has been shown to influence NPC spacing in HeLa cells, 25 and in Xenopus laevis there is recent evidence that NUP210 forms a buffering transluminal cushion around the NPC. 12 The absence of NUP210L could therefore conceivably reduce the capacity of traffic into and out of the elongating spermatid nucleus by negatively impacting the exceptional tight NPC packing, or the size and plasticity, of the pore channel.In support of the idea that a stressed histone-to-protamine transition could be one consequence of NUP210L loss, mice lacking the chromatin remodeling proteins TNP1 or PRM1, have spermatozoa with a short acrosomal hook and reduced motility as observed in our Nup210l-KO mice. 26Despite the absence of a severe phenotype in the NUP210L knockout mice presented here, our results show that NUP210L does play a role in morphogenesis of the sperm head and is required for efficient flagellum formation in the mouse.We conclude that although NUP210L is not required for spermatid nuclear condensation or male fertility in the mouse, it may optimize NPC array organization and functionality at the posterior pole in elongating spermatids.
Nup210l +/+ controls.Epididymal spermatozoa were recovered and we determined their number, motility and morphology.No difference was observed in the number of spermatozoa recovered F I G U R E 1 Generation of Nup210l À/À mice.(A) Schematic illustration of Nup210l, 40 exons, including the location of the C insertion mutation in exon 6. (B) RT-PCR analysis of adult mouse testis RNA, PCR products were separated on 3% agarose gel.Lane 1 (1 kb+) is the 1 kb plus size marker (ThermoFisher).The blue line indicates the wild-type band, while the red and green ones mark the two bands detected in Nup210l À/À mice.(C) Sanger sequencing of RT-PCR bands: top (blue)-wild-type (WT) RT-PCR amplicon, middle (red)-upper Nup210l À/À amplicon with the "C" insertion in exon 6, and bottom (green)-lower Nup210l À/À amplicon showing skipping of exon 6. (D) Immunofluorescence revealing the absence of NUP210L protein in Nup210l À/À mice testis-anti-NUP210L (green), lectin PNA labeling of the acrosome (red) and DAPI (blue).Scale bar is 5 μm.(E) Western blot analysis of testicular lysates from adult mice with anti-NUP210L.A band of approximately 210 kDa (arrowhead) was detected in wild-type (WT) and Nup210l +/À but not Nup210l À/À lysates.Sizes of Spectra size marker (ThermoFisher) are indication.[Colour figure can be viewed at wileyonlinelibrary.com] from Nup210l À/À and control mice showing that the rate of spermatozoa production is not affected by the absence of NUP210L coiled and bent (double cross-sections) flagella in homozygotes, consistent with our prior findings in bright field analysis.The structure of the F I G U R E 2 Fertility test and histological analysis of the testis from WT and NUP210L deficient mice.(A) No significant difference observed in litter size between Nup210l À/À and wild type males.p-Value = 0.1.(B) No difference observed in testes weight between Nup210l À/À and wild type males.Seven mice of each genotype were used.p-Value = 0.93.(C) Testes of Nup210l À/À mutant mice compared with their wild-type littermates.(D) PAS-hematoxylin staining of testis (i, iii) and H&E staining of cauda epididymis (ii, iv) from WT and Nup210l À/À mice.On the right are images with higher magnification showing compacted sperm heads in greater detail (ii 0 , iv 0 ).[Colour figure can be viewed at wileyonlinelibrary.com]

F I G U R E 3
Normal spermatid development and stage proportions in WT and Nup210l À/À mice males.(A) Nup210l À/À mice have a typical arrangement of germ cells with normal spermatid development.Stages of the seminiferous tubules were identified according to the morphology of spermatocytes and spermatids.Spermatids at different stages are illustrated.Pl, preleptotene; L, leptotene; Z, zygotene; P, pachytene; D, diplotene; RS, round spermatids; ES, elongating spermatids.(B) Quantification of the proportion of stages in Nup210l À/À and wild type males.Three mice of each genotype were analyzed.p-Value = 0.79 for the stage I, II, III, p-value = 0.41 for the stage IV, V, p-value = 0.61 for the stage VI, VII, VIII, p-value = 1 for the stage IX, and p-value = 0.84 for the stage X, XI, XII.p-Values were calculated by Student's t-test.[Colour figure can be viewed at wileyonlinelibrary.com]

4
Sperm quality analysis of Nup210l À/À mice.(A) Nup210l À/À males have normal sperm count.Sperm from cauda epididymis of 6-month-old wild type (n = 10) and Nup210l À/À mice (n = 10) were counted.p-Value = 0.07.(B) Progressive motility of sperm was reduced in Nup210l À/À males.Six mice of each genotype were analyzed.p-Value <0.0001 for progressive motility, p-value = 0.0001 for nonprogressive motility and p-value = 0.0008 for immotile sperm.(C) Quantification of coiled, bent and normal tail observed in the spermatozoa of Nup210l À/À and WT mice.p-Values <0.0001 for coiled tail, 0.006 for bent tail and 0.001 for normal tail.(D) Sperm morphology, as labeled by sperm blue stain (i-vi) and silver nitrate (vii-xii).Nup210l À/À sperm show bent and coiled flagellum, as well as changes in the morphology of head: shortened acrosomal hook (arrowheads) and narrowing of the base of the head (arrows).

F I G U R E 5
TEM analysis reveals normal chromatin condensation in Nup210l À/À (A) Sperm head of Nup210l À/À mice exhibit normal electron density under electron microscopy and normal head-tail coupling apparatus (HTCA).Magnified view of the sperm head in the right.(B) Ultrastructure of the Nup210l À/À sperm flagellum showing a normal microtubule structure in an uncoiled flagellum, but abnormal structures in a coiled one (C).spermatids, supporting the possibility that a partial loss of NUP153 function may explain the severe nuclear compaction failure associated with loss of NUP210L in human.Nevertheless, the distinct localization of NUP210L and NUP153 in human and mouse round spermatids and post-mitotic formation of the nuclear lamina through the targeting of B-type lamins to the inner F I G U R E 6 Normal chromatin compaction in Nup210l À/À mice.(A) Immunostaining images of testicular section from wild-type (WT) and Nup210l À/À mice are shown-anti-acetylated histone 4 acH4 (green), lectin PNA (red), and DAPI (blue).White arrows S8, S9, S12, and S16spermatids at step 8, 9, 12, and 16.Scale bar is 20 μm.(B) CMA3 were used to stain epididymal sperm (green) and DNA counterstained with Hoechst.Scale bar is 20 μm.(C) CMA3 positive and negative sperm were counted in Nup210l À/À and WT mice.Three mice of each genotype were analyzed, with 200 sperm counted for each mouse.p-Value = 0.37.[Colour figure can be viewed at wileyonlinelibrary.com] NUP210L colocalize to the nuclear periphery in round spermatids and at the caudal nuclear pole in elongating spermatids and mature spermatozoa, showing that they could have a functional interaction during human spermiogenesis.Furthermore, the underrepresentation of loss-of-function alleles for NUP153 in gnomAD indicates that there is intolerance for haploinsufficiency of NUP153 in human, implying that a partial reduction in NUP153 functionality could negatively impact the NPC and sensitize it to the absence of NUP210L in spermatids.Testing if this genetic combination underlies the severe human phenotype will require further similar human cases or a mouse model with the appropriate mutations in NUP210L and NUP153.F I G U R E 7 Patient 13-4587 is heterozygous for a potential pathogenic variant in the gene encoding nucleoporin NUP153 which colocalizes with NUP210L to the caudal nuclear pole in elongating spermatids and spermatozoa, in human and mouse.(A) Position of the p.Pro485Leu the NUP153 gene missense variant, p.Pro485Leu, heterozygous in patient 13-4587.(B) Sanger chromatogram displaying the sequences of both the control and patient (13-4587), with the serine codon in the blue rectangle.(C) Immunolocalization of NUP210L (green) during human and mouse spermatogenesis.The spermatocyte and subsequent stages of spermatid development are labeled.Different steps of spermatids are identified using lectin PNA (acrosome-red) and DAPI (DNA-blue).Scale bar is 5 μm.(D) Colocalization of NUP210L (red) and NUP153 (green) in human and mouse spermiogenesis.Scale bar is 5 μm.(E) No difference in NUP153 localization occurs in Nup210L À/À mice.Scale bar is 5 μm.[Colour figure can be viewed at wileyonlinelibrary.com]