Abnormal fertility, acrosome formation, IFT20 expression and localization in conditional Gmap210 knockout mice

Abstract

GMAP210 (TRIP11) is a cis-Golgi network-associated protein and a Golgi membrane receptor for IFT20, an intraflagellar transport component essential for male fertility and spermiogenesis in mice. To investigate the role of GMAP210 in male fertility and spermatogenesis, floxed Gmap210 mice were bred with Stra8-iCre mice so that the Gmap210 gene is disrupted in spermatocytes and spermatids in this study. The Gmap210^flox/flox: Stra8-iCre mutant mice showed no gross abnormalities and survived to adulthood. In adult males, testis and body weights showed no difference between controls and mutant mice. Low-magnification histological examination of the testes revealed normal seminiferous tubule structure, but sperm counts and fertility were significantly reduced in mutant mice compared with controls. Higher resolution examination of the mutant seminiferous epithelium showed that nearly all sperm had more oblong, abnormally shaped heads, while the sperm tails appeared to have normal morphology. Electron microscopy also revealed abnormally shaped sperm heads but normal axoneme core structure; some sperm showed membrane defects in the midpiece. In mutant mice, expression levels of IFT20 and other selective acrosomal proteins were significantly reduced, and their localization was also affected. Peanut-lectin, an acrosome maker, was almost absent in the spermatids and epididymal sperm. Mitochondrion staining was highly concentrated in the heads of sperm, suggesting that the midpieces were coiling around or aggregating near the heads. Defects in acrosome biogenesis were further confirmed by electron microscopy. Collectively, our findings suggest that GMAP210 is essential for acrosome biogenesis, normal mitochondrial sheath formation, and male fertility, and it determines expression levels and acrosomal localization of IFT20 and other acrosomal proteins.

INTRODUCTION

Golgins, a group of coiled proteins, localize to the Golgi and Golgi-associated vesicles (3). Golgins are typically anchored to the cytosolic face of the Golgi membrane by their extreme C-termini. These features have led to the proposal that golgins extend into the cytoplasm to capture, or tether, transport vesicles to Golgi membranes, which can be followed by membrane fusion to facilitate both membrane traffic and assembly or maintenance of Golgi cisternae (48).

The golgin GMAP210 (Golgi-microtubule-associated protein of 210 kDa), also known as thyroid hormone receptor interactor 11 (TRIP11), is thought to act as a tether at the cis-Golgi as a result of dual recognition of the small GTPase ADP-ribosylation factor (ARF)1 by a carboxy-terminal GRIP-related Arf-binding (GRAB) domain and curved membranes by an amino-terminal lipid packing sensor (ALPS) motif (3, 7, 11, 47). Like other golgins, GMAP210 has been demonstrated to have an important role in maintaining a morphologically normal and functional Golgi apparatus (46). Mice lacking GMAP210 die at birth with a pleiotropic phenotype that includes growth restriction, ventricular septal defects of the heart, omphalocele, and lung hypoplasia (14). The physiological role of GMAP210 is indicated by the fact that mutations in human GMAP210 cause the neonatal lethal skeletal dysplasia, achondrogenesis type 1A; however, whether this arises from altered glycosylation or reduced secretion of extracellular matrix proteins remains to be ascertained (54).

It has been shown that GMAP210 is important for some aspects of cilium biology by identifying this Golgin as a Golgi membrane receptor for intraflagellar transport 20 (IFT20) (14). It was found that cells lacking GMAP210 have normal Golgi structure, but IFT20 was no longer localized to this organelle. GMAP210 is not absolutely necessary for ciliary assembly, but cilia on the GMAP210 mutant cells are shorter than normal and have reduced amounts of the membrane protein polycystin-2 localized to them. This evidence suggests that GMAP210 and IFT20 function together at the Golgi in the sorting or transport of proteins destined for the ciliary membrane.

Our laboratory discovered that IFT20 is essential for male fertility and spermatogenesis in mice, and its major function is to transport cargo proteins for sperm flagellum formation (66). GMAP210 is a Golgi membrane receptor for IFT20; therefore, we hypothesized that this Golgi-associated protein would have a role in spermatogenesis and be essential for male fertility. To test this hypothesis, we generated a male germ cell-specific Gmap210 knockout (KO) mouse model by crossing floxed Gmap210 mice to Stra8-iCre transgenic mice. Our studies demonstrated that GMAP210 is essential for normal spermiogenesis and male fertility. As expected, IFT20 expression and localization were abnormal in conditional Gmap210 mutant mice. However, the major phenotypes between the two mutants were different. GMAP210 has an essential role in acrosome formation and determines the localization of acrosomal proteins. It has little role in sperm core axoneme formation.

MATERIALS AND METHODS

Ethics statement.

All animal studies were approved by Wayne State University Institutional Animal Care and Use Program Advisory Committee (Protocol no. IACUC-18-02-0534).

Generation of male germ cell-specific Gmap210 knockout mice.

Stra8-iCre mice were purchased from Jackson Laboratory (stock no: 008208). Transgenic mouse line Stra8-cre expresses improved Cre recombinase under the control of a 1.4 Kb promoter region of the germ cell-specific stimulated by retinoic acid gene 8 (Stra8) (49). Gmap210^flox/flox mice were generated by Dr. Gregory J. Pazour of the University of Massachusetts Medical School (14) using cells generated by the European Conditional Mouse Mutagenesis Program project. To generate germ cell-specific Gmap210 knockout mice, we followed the same breeding strategy as the one used to generate germ cell-specific Ift140 knockout mice (65). Briefly, 3- to 4-mo-old Stra8-cre males were crossed with 3- to 4-mo-old Gmap210^flox/flox females to obtain Stra8-iCre; Gmap210^flox/+ mice. The 3- to 4-mo-old Stra8-iCre; Gmap210^flox/+ males were crossed back with 3- to 4-mo-old Gmap210^flox/flox females again, the Stra8-iCre; Gmap210^flox/flox were considered to be the homozygous knockout mice (KO), and Stra8-iCre; Gmap210^flox/+ mice were used as the controls. Mice were genotyped by PCR using multiplex PCR mix (Bioline, cat no. BIO25043). The presence of the Stra8-iCre allele was evaluated as described previously (49). To genotype the offspring, genomic DNA was isolated as described previously (66). The following primers were used for genotyping: Gmap210 forward: 5′-CATGGATTGCTTTGCATTGT-3′; Gmap210 reverse: 5′-AAGAGTGTTTAGAACCTGGACAACTT-3′, and Stra8-iCre genotypes were determined as described as previously (66).

Assessment of fertility and fecundity.

To assess fertility and fecundity, 6-wk-old or adult Gmap210 KO and control males were paired with 3- to 4-mo-old wild-type females for at least 2 mo. The number of pregnant mice and the number of offspring from each pregnancy were recorded.

Spermatozoa counting.

Spermatozoa were collected from cauda epididymides and fixed in 2% formaldehyde for 10 min at room temperature. After washing and resuspending in PBS buffer, the spermatozoa were counted using a hemocytometer chamber under an optical microscope. The sperm count was calculated as previously described (32).

Spermatozoa motility assay.

Assessment of sperm motility was carried out following a previously described procedure (65). Briefly, the sperm from the epididymis was squeezed and placed in warm PBS or the noncapacitating medium. Ten minutes after collection of sperm, sperm motility was observed on a prewarmed slide using the reverse microscope of the Nikon TE200E with SANYO color CCD, high-resolution camera (CCV-3972) and HD studio peak (version 14.0) software. Eight fields were analyzed for each sperm sample. Individual spermatozoa were tracked using ImageJ (NIH, Bethesda, MD) and the plug-in MTrackJ. Sperm motility was calculated according to the curve velocity (VCL), which is the curvilinear distance (DCL) of spermatozoa in seconds (VCL = DCL/t).

Histological staining on tissues.

Mouse testes and epididymides were prepared for histological analysis using previously described methods (65). Briefly, samples were collected and fixed in PBS buffer containing 4% formaldehyde, sectioned into 5 µm slides, and stained with hematoxylin and eosin using standard procedures.

Immunofluorescence analysis of isolation of spermatogenic cells and testis sections.

Testes and cauda epididymal sperm from Gmap210 control and KO adult mice were dissected in a petri dish with 5 mL DMEM containing 0.5 mg/mL collagenase IV and 1.0 mg/mL DNAse I (Sigma-Aldrich), which was incubated afterward for 30 min at 32°C with gentle stirring. Released spermatogenic cells were pelleted by centrifugation (5 min at 1,000 revolutions/min, 4°C). After washing with PBS, the cells were fixed with 5 mL of 4% paraformaldehyde (PFA) containing 0.1 M sucrose at room temperature. The dispersed, mixed testicular cells were washed three times with PBS. Afterward, the cells were resuspended in 2 mL of warm PBS, and 50 μL of cell suspension was loaded to the slide and allowed to air dry. Cells were permeabilized with 0.1% Triton X-100 (Sigma-Aldrich) for 5 min at 37°C, washed with PBS three times, and blocked with 10% goat serum in PBS for 1 h. Cells were then washed three times with PBS and incubated overnight with the indicated primary antibodies. After extensive washing with PBS, the cells were incubated with the same as used for Western blot analysis for 1 h, but the dilutions were 10 times higher. The slides were washed with PBS and mounted in VectaMount with DAPI (Vector Laboratories, Burlingame, CA) and sealed with nail polish. Images were captured by confocal laser-scanning microscopy (Zeiss LSM 700). Staining of cauda epididymal sperm mitochondria was carried out using MitoTracker probes (Molecular Probes). Isolated germ cells and epididymal sperm were also stained for acrosome using peanut-lectin.

To conduct immunofluorescence on testis sections, fresh testes were fixed with 4% paraformaldehyde in 0.1 M PBS (pH 7.4), and 5 µm paraffin sections were made. For the immunofluorescence, the sections were incubated with the indicated primary antibodies at 4°C overnight. The procedure then followed the one used for immunofluorescence using isolated testicular cells described above. Dr. Richard Oko’s laboratory at Queen’s University, Canada provided the following primary antibodies: anti-ZPBP1 (1:200, anti-rabbit); anti-MMP2 (1:200, anti-rabbit); and anti-SPACA1 (1:200, anti-rabbit).

Scanning electron microscopy.

Mouse testes and cauda epididymal sperm were fixed in 0.1 M sodium cacodylate containing 1% paraformaldehyde and 3% glutaraldehyde (pH 7.4) at 4°C overnight and processed for electron microscopy as reported (65).

Transmission electron microscopy.

Mouse testes and cauda epididymal sperm were fixed in the same buffer as scanning electron microscopy. Images were taken with a Hitachi HT7700 transmission electron microscope.

Western blot analysis.

Western blot analysis was executed as previously described (32). The membranes were immunoblotted with the following indicated antibodies at 4°C overnight: anti-GMAP210 (1:1,000; Dr. Pazour’s laboratory) (14), anti-IFT20 (1:2,000; Dr. Pazour’s laboratory) (43), anti-AKAP4 (1:4,000; Dr. George L. Gerton at University of Pennsylvania) (34), anti-ODF2 (1:1,000; cat. no. 12058-1-AP, Proteintech Group), anti-SPAG16 (1:2,000; Zhang’s laboratory) (67), and anti-β-actin (1:2,000; cat. no. 58169, Cell Signaling Technology). Anti-SPACA1 (1:4,000), anti-ZPBP1 (1:1,000), and anti-MMP2 (1:1,000) were from Dr. Oko’s laboratory. After being washed, the membrane was incubated with the secondary antibody conjugated with horseradish peroxidase, including donkey anti-rabbit IgG (1:2,000; cat. no. NA934, GE Healthcare) or sheep anti-mouse antibody (1:1,000, cat. no. NA931, Amersham Biosciences). Signals were detected with Super Signal West Pico Chemiluminescent Substrate and West Femto Maximum Sensitivity Substrate (Thermo Fisher). The gray value of the target bands was analyzed with ImageJ software, and each experiment was repeated three times to obtain the mean value.

Statistical analysis.

Measurement data are expressed as the mean ± SE. The two-tailed student’s t test was used for evaluation of significant differences between two normally distributed groups. Statistical significance is defined as P < 0.05.

RESULTS

GMAP210 is highly expressed in the testis and developmentally regulated during spermatogenesis.

GMAP210 protein expression was analyzed in various tissues, including the heart, brain, spleen, lung, liver, kidney, muscle, and testis. GMAP210 protein was highly expressed in the testis and liver using Western blotting (Fig. 1A). During testicular development, germ cell expression levels of GMAP210 protein were present only after day 16 postbirth, and the expression level appeared to increase with age (Fig. 1B).

Fig. 1.

Golgi-microtubule-associated protein of 210 kDa (GMAP210) expression in the wild-type mice. A: examination of GMAP210 protein expression in the indicated tissues from wild-type mice. Note that GMAP210 protein is highly expressed in liver and testis. B: analysis of testicular GMAP210 expression during the first wave of spermatogenesis of wild-type mice. Note that the protein is only present after day 16 after birth and that the expression level appears to be increased with time.

Impaired fertility associated with reduced sperm number and motility of conditional Gmap210 knockout mice.

To study the role of GMAP210 in mouse spermatogenesis, male germ cell-specific conditional knockout mice were generated (Supplemental Fig. S1; Supplemental Material is available at https://doi.org/10.6084/m9.figshare.8651573). GMAP210 protein expression in the mutant mice was examined with Western blotting. Control mice showed robust GMAP210 signals in the testis; however, its signal was barely detectable in conditional Gmap210 mutant mice (Fig. 2A).

Fig. 2.

Reduced fertility and sperm number of conditional Golgi-microtubule-associated protein of 210 kDa (Gmap210) knockout (KO) mice. A: examination of testicular GMAP210 protein expression of control and conditional Gmap210 KO mice. Note that GMAP210 protein was significantly decreased in the KO mice. B: sperm count in the control and conditional Gmap210 KO mice. The KO mice have a significantly reduced sperm number. C: percentage of motile sperm. D: sperm motility. All experiments were repeated three times. Statistical data between two groups were analyzed using a t test; n = 3. Values are means ± SE. In B–D, **P < 0.01, compared with control.

Fertility of the homozygous conditional Gmap210 mutant mice was also examined. All control mice were fertile, but more than 50% of the conditional Gmap210 knockout mice were infertile, and the fertile knockout mice sired fewer numbers of pups. The body weight, testis weight, and testis/body weight were not statistically different between the control and homozygous mutant mice (Table 1). Epididymal sperm numbers in mutant mice was significantly lower than those in controls (Fig. 2B). Sperm motility was also examined. Most control sperm were motile and displayed vigorous flagellar activity and progressive long-track forward movement (Supplemental Video SA; see https://doi.org/10.6084/m9.figshare.8664311). Few sperm are motile in conditional Gmap210 mutant mice (Supplemental Video SB; see https://doi.org/10.6084/m9.figshare.8664311). The percentage of motile sperm (Fig. 2C) and motility itself (Fig. 2D) were significantly reduced in the conditional knockout mice.

Table 1.

Fertility of control and conditional Gmap210 knockout mice

Genotype	Fertility	Litter Size	Testis Weight, mg	Body Weight, g	Testis/Body Weight, mg/g
Control	6/6	7 ± 2.2 (n = 6)	206.9 ± 10.8 (n = 6)	26.2 ± 1.8 (n = 6)	7.9 ± 1.5 (n = 6)
Gmap210 KO	4/9	2 ± 1.3 (n = 4)	207.4 ± 11.2 (n = 9)	25.6 ± 1.6 (n = 9)	8.1 ± 2.0 (n = 9)

Values are means ± SE; n = no. of mice. All control mice were fertile, but more than 50% of conditional Golgi-microtubule-associated protein of 210 (Gmap210) knockout (KO) mice were infertile, and the fertile KO mice sired fewer pups. The differences in testis weight, body weight, and testis/body weight were not statistically significant between the control and homozygous mutant mice.

The conditional Gmap210 knockout mice developed more oblong heads.

Morphology of epididymal sperm was originally examined by light microscopy. Sperm from the control mice had well-developed, hook-shaped heads and long and smooth tails (Supplemental Fig. S2; see https://doi.org/10.6084/m9.figshare.8657816). In contrast, sperm from conditional Gmap210 knockout mice had more oblong heads, even though the tails looked normal with light microscopy (Supplemental Fig. S2). Sperm morphology was further examined by scanning electron microscopy. As observed under light microscopy, sperm from control mice were observed to have long, smooth tails with normally condensed heads (Fig. 3A). However, multiple abnormalities were observed in sperm of mutant mice, including more oblong heads (Fig. 3, B, C, and D) and some sperm with disrupted middle pieces, although most tails appeared to be normal (Fig. 3D).

Fig. 3.

Examination of epididymal sperm by scanning electron microscope. Sperm from the control mice showed normal morphology (A). In the knockout (KO) mice, nearly all sperm showed more oblong heads (arrows in B, C, D). Most tails appeared to be normal; however, some sperm had a disrupted midpiece/flagellum junction (arrowhead in D). Gmap210, Golgi-microtubule-associated protein of 210 kDa.

Testis histology of control and conditional Gmap210 knockout mice.

Histology of adult testis of control and conditional Gmap210 mutant mice was also examined, and three stages of spermatogenesis are illustrated. Control seminiferous tubules in Stage VII show normal step 7 round spermatid above pachytene spermatocytes (P), with step 16 elongating spermatids lining the lumen and their tails extending toward the center. In Stage VIII, step 8 round spermatids were seen with adjacent residual bodies (Rb) that remained after spermiation of mature sperm. In Stage X, step 10 elongating spermatids face the lumen (Fig. 4A).

Fig. 4.

Testis histology from control and conditional Golgi-microtubule-associated protein of 210 kDa (Gmap210) knockout (KO) mice. Images showing cross sections of seminiferous tubules. Bar = 20 μm. A: control seminiferous tubules are seen at three stages. In Stage VII, step 7 round spermatid are found above pachytene spermatocytes (P) and step 16 elongating spermatids line the lumen with their tails extending toward the center. In Stage VIII, step 8 round spermatids are seen with adjacent residual bodies (Rb) that remain after spermiation of the mature sperm. T, tail of sperm being released. In Stage X, step 10 elongating spermatids face the lumen. B: Gmap210 KO seminiferous tubule in Stage VII has normal step 7 round spermatids and pachytene spermatocytes (P). Abnormal step 16 elongating spermatids (Ab16) are seen lining the lumen with their tails (T) forming a swirl. The heads of step 16 spermatids have abnormal shapes and lie among components of the cytoplasmic lobes before spermiation. C: Gmap210 KO Stage IX with normal step 9 spermatids and the resorption of some residual bodies (Rb). However, numerous abnormal step 16 spermatids (Ab16) are seen with more oblong heads surrounded by cytoplasm. D: Gmap210 KO Stage XI showing abnormal heads of step 11 elongating spermatids (Ab11). P, pachytene spermatocyte.

In Gmap210 mutant mice, the Stage VII seminiferous tubule had normal step 7 round spermatids and pachytene spermatocytes (P). However, abnormal step 16 elongating spermatids (Ab16) are seen lining the lumen with their tails (T) forming a swirl. The heads of step 16 spermatids had abnormal, more oblong shapes suspended among components of cytoplasmic lobes before spermiation (Fig. 4B). Stage IX in Gmap210 mutant mice had normal step 9 spermatids and showed some resorption of residual bodies (Rb). However, numerous abnormal step 16 spermatids (Ab16) were observed with more oblong heads, which were surrounded by excess cytoplasm (Fig. 4C). Stage XI in the Gmap210 mutant mice showed abnormal heads of step 11 elongating spermatids (Ab11) (Fig. 4D).

Epididymis histology of conditional Gmap210 knockout mice.

In controls, the cauda epididymis lumen was filled with highly compacted sperm having normal morphology and showing an alignment of heads and tails (Fig. 5A). In Gmap210 mutant mice, the epididymal lumen was filled with abnormal, globozoospermia-like heads and disorganized tails that appear to be short in length. Large residual bodies of cytoplasm were also scattered in the lumen (Fig. 5B).

Fig. 5.

Histology of cauda epididymis from control and conditional Golgi-microtubule-associated protein of 210 kDa (Gmap210) knockout (KO) mice. Bar = 20 μm. A: control epididymis showing an epithelium (Ep) lining the lumens that are filled with normal sperm aligned with their heads (Hd) and tails (T). B: Gmap210 KO epididymis showing a lumen filled with abnormal sperm, having more oblong heads (Ab) and disorganized tails (T). Residual bodies of cytoplasm (Rb) are also scattered in the lumen and sometimes associated with the more oblong sperm heads. Ep, epithelium; TRIP11, thyroid hormone receptor interactor 11 (GMAP210).

Testicular ultrastructure of conditional Gmap210 knockout mice.

In control mice, normal spermatid ultrastructure was observed in all stages of spermatogenesis (Figs. 6–8). A concave Golgi complex formed over the round spermatid nucleus and released granules that fused to form an acrosomal vesicle with acrosomal granule that flattened over the area of the nuclear envelope underlined by the nuclear lamina (Fig. 6). Elongating spermatids formed normal flagellar axonemes and acrosomal vesicles and granules, and the elongated nuclei were highly condensed (Figs. 7 and 8).

Fig. 6.

Transmission electron microscopy of wild-type (WT) and Golgi-microtubule-associated protein of 210 kDa (Gmap210) knockout (KO) mice round spermatids at specific stages of the seminiferous epithelial cycle. A: Stage II-III. WT step 2 spermatid has a prominent Golgi apparatus (G) with Golgi membrane stacks (St) forming an arch over the area adjacent to the nucleus (N). Proacrosome vesicles (Pa) with granules are seen between the trans-Golgi face and the nuclear membrane. B: Stage II-III. Gmap210 KO step 2–3 spermatid has a more diffuse Golgi apparatus (G) that does not arch over the nucleus (N). However, proacrosome vesicles (Pa) with granules are found adjacent to the nucleus membrane. C: Stage V. WT step 5 spermatid has a prominent acrosomic granule (Ag) in the elongated acrosomic vesicle (Av) that is applied to the nuclear lamina (NL) of the nuclear (N) membrane. A prominent Golgi apparatus (G) arches over the acrosomic vesicle with proacrosome vesicles (Pa) and granules found at the membrane of the acrosomic vesicle. D: Stage V. Gmap210 KO step 5 spermatid Golgi apparatus (G) continues to appear more diffuse compared with the WT. However, proacrosome vesicles (Pa) are abundant, with some vesicles adhering to the area of the forming nuclear lamina (NL), which also has premature indentations of the nuclear (N) membrane, possibly the future marginal ring. E: Stage VI-VII. WT step 6–7 spermatid shows a well-developed acrosomic cap over one-third of the nucleus (N). The forming acrosome has a flattened granule (Ag) and vesicle (Av) applied over the well-formed nuclear lamina (NL). Golgi apparatus (G) remains arched over the acrosomic vesicle. F: Stage VI-VII. Gmap210 KO step 7 spermatid has a prominent nuclear lamina (NL), with attached proacrosome vesicles (Pa) and granules. Vesicles sometimes rest in heavy indentations of the nuclear (N) membrane, at premature sites where the marginal ring will form (*) and appear to form a subacrosomal layer (SAL). Golgi apparatus (G) remains diffuse and does not show the typical arching of Golgi saccule stacks. G: Stage VIII. WT step 8 spermatid has a flattened acrosomic granule (Ag) and vesicle (Av) applied to the nuclear lamina (NL) of the nucleus (N) and forms slightly indented regions marking the marginal ring (Mr). Nucleus and edge of the acrosome have moved to the plasmalemma, where the Sertoli cell has formed an apical ectoplasmic specialization (Es). Implantation fossa (If) holds the proximal centriole (Pc) that forms the flagellum. Cytoplasmic lobes (Cl) of excess cytoplasm and organelles have developed from the elongated step 16 spermatids (not shown). H: Stage VIII. Gmap210 KO step 8 spermatid has a prominent nuclear lamina (NL) at the nuclear membrane, with formation of the marginal ring (Mr) at the outer edge of the acroplaxome. There is no evidence of a forming acrosome (*), although strands of endoplasmic reticulum are seen running parallel to the acroplaxome and may contribute to the formation of a pseudoacrosome in some spermatids. Implantation fossa appears abnormal with the proximal centriole (Pc) attached to a protrusion of the nucleus (N). Bar = 1 μm (A, C–E, G, H); bar = 600 nm (B); bar = 800 nm (F).

Fig. 8.

Transmission electron microscopy of wild-type (WT) and Golgi-microtubule-associated protein of 210 kDa (Gmap210) knockout (KO) mice germ cells of the seminiferous epithelium showing misalignment of the flagellum and mitochondria and abnormal nuclear shapes. A: in the WT, step 15 spermatids show normal mitochondria (M) surrounding the axonemal complex (Ax) in midpieces of the elongating flagellum. Surrounding cytoplasm is limited by the spermatid plasmalemma (Pl). B and C: Gmap210 KO step 15–16 spermatids showing normal axonemal complex (Ax) formation with surrounding outer dense fibers (Odf) in the midpiece. However, mitochondria (M) are missing in some locations (*). D: Gmap210 KO step 15 spermatid showing numerous abnormalities. Mitochondria (M) are clustered some distance away from the forming axonemal complex (Ax), which is also not aligned with the nucleus (N). There is also an excessive amount of cytoplasm surrounding the nucleus and therefore the nucleus does not approach the Sertoli cell membrane. Nucleus is distorted with small protrusions of nucleopodes (Np) and does not have an acrosomal cap. E: Gmap210 KO step 15 spermatid in Stage VII showing an abnormal nucleus (N) with nucleopode projections (Np). Nuclear lamina/pseudoacrosome (NL/P) appears to have grown away from the nucleus, forming a pocket for cytoplasm (Ct). The forming flagellum with a long axonemal complex (Ax) is capped by the head-to-tail coupling apparatus (Htc), but is incorrectly placed at the cephalic region of the nucleus. An excessive amount of cytoplasm, with mitochondria (M) misaligned, extends around the flagellum and nucleus, with no formation of an apical ectoplasmic specialization. Bar = 1 μm (A, B, D, E); bar = 200 nm (C).

Fig. 7.

Transmission electron microscopy of wild-type (WT) and Golgi-microtubule-associated protein of 210 kDa (Gmap210) knockout (KO) mice elongating spermatids at specific stages of the seminiferous epithelial cycle. A: Stage IX. WT step 9 spermatid has a prominent, flattened acrosome (Ac) applied to the nuclear lamina (NL) that covers one pole of the nucleus (N). Sertoli cell (Sc) apical ectoplasmic specialization (Es) has a dotted appearance along the edge where the two cells meet. A marginal ring (Mr) is formed where the acrosome terminates and the manchette (Ma) begins. An implantation fossa (If) with the proximal centriole is found opposite the acrosome. B: Stage IX. Gmap210 KO step 9 spermatid nucleus is distorted in shape. Nuclear lamina (NL) is present, but a true acrosome is missing, although a single, large acrosomal granule (Ag) is observed opposite the acroplaxome. A pseudoacrosome (P) or flattened saccule is formed in some cases over the acroplaxome, as shown here, similar to that formed in the Hrb-null mutation (19). A large manchette (Ma) is formed, with an attachment to the perinuclear ring (Pr) that is forming abnormally over a protrusion of the nucleus. Sertoli cell (Sc) cytoplasm surrounds the spermatid, but an apical ectoplasmic specialization does not always form adjacent to the nucleus in the mutant spermatids, probably due to the lack of a true acrosome. C: Stage XI. WT step 11 spermatid has a well-developed acrosome (Ac) along the Sertoli (Sc)-germ cell border, where the nucleus (N), nuclear lamina (NL), and acrosome meet the plasmalemma. The manchette has bundles of microtubules extending from the nucleus. D: Stage XI. Most Gmap210 KO step 11 spermatids show nuclear (N) distortion with abnormal extensions of the nuclear lamina (NL) and pseudoacrosome (P) into the Sertoli cell (Sc) cytoplasm, even forming pockets of Sertoli cell cytoplasm within nuclear indentations of the spermatids. Sertoli cell has formed apical ectoplasmic specialization junctions (Es) on the acroplaxome/pseudoacrosome extensions. E: Stage VII-VIII. WT step 16 spermatids show well-developed heads with condensed nuclei (N), capped with distinct acrosomes (Ac). These spermatids will be released into the lumen after the Sertoli cell (Sc) has phagocytized the forming cytoplasmic lobes (Cl) of excess germ cell cytoplasm and organelles. F and G: Stage VII-VIII. Gmap210 KO step 15–16 spermatids have highly condensed nuclei (N) that display massive distortion. This abnormality includes the formation of nucleopodes (Np) that are small protrusions of nuclear material covered by the nuclear lamina (NL), similar to those formed in the Hrb-null mutation (19) and the TMF null mouse (29). The excess cytoplasm found in the elongating spermatids does not form proper cytoplasmic lobes and thus appears to cause the flagellum to curl around the nucleus, as illustrated in (F), where the nucleus with a protruding acroplaxome is seen adjacent to a cross-section of the developing flagellum, with a normal axoneme but missing two mitochondria (*). Bar = 1 μm (A–G).

In conditional Gmap210 knockout mice, round spermatids displayed a more diffuse Golgi apparatus that did not have the typical arch over the nucleus (Fig. 6). Although the Golgi complex was atypical during the Golgi phase, proacrosomic vesicles with granules were formed on the Trans face adjacent to an apparently normal nuclear lamina along one aspect of the nucleus of round spermatids (Fig. 6). Some proacrosomal vesicles attached to the nuclear membrane but did not fuse to form an acrosomal vesicle and thus did not form an acrosome.

Elongation of the spermatid nucleus, beginning in Stage IX, was also abnormal (Fig. 7). The nuclear chromatin became condensed as the spermatids attempted to elongate, but the nucleus was misshapen, forming nucleopods, which are often associated with abnormal extensions of the nuclear lamina (Figs. 7 and 8). The flagellar axoneme appeared to be normal with an intact ‘9+2’ microtubule array, outer dense fibers, and fibrous sheath (Fig. 8); however, the mitochondrial sheath often had missing mitochondria. Mitochondria were also seen clustered some distance from the forming axonemal complex, which did not always align with the nucleus. There was also an excessive amount of cytoplasm surrounding the nucleus and midpiece in the elongated spermatids, which prevented the nucleus from approaching the Sertoli cell membrane at the ectoplasmic specialization. Mitochondrial clusters were observed in this excess cytoplasm, suggesting a failure to form cytoplasmic lobes and residual bodies. Cross sections of the axoneme were also present in this region, providing ultrastructural evidence of partial coiling of the flagellum within the excessive accumulation of spermatid cytoplasm (Fig. 8, Supplemental Fig. S3; see https://doi.org/10.6084/m9.figshare.8659469).

Reduced expression and abnormal localization of IFT20 in conditional Gmap210 knockout mice.

IFT20 is a binding partner of GMAP210. To examine its expression level and localization in the conditional Gmap210 mutant mice, Western blot analysis and immunofluorescence staining were performed. Testicular IFT20 expression levels were significantly reduced in the Gmap210 mutant mice (Fig. 9A). In control mice, immunofluorescence staining demonstrated IFT20 presence in Golgi bodies of spermatocytes (Fig. 9Ba) and localized in the Golgi/acrosome region of round spermatids (Fig. 9, Bb and Bc) and the manchette of elongating spermatids (Fig. 9Bd).

Fig. 9.

Significantly reduced levels and disrupted localization of intraflagellar transport 20 (IFT20) in the conditional Golgi-microtubule-associated protein of 210 (Gmap210) knockout (KO) mice. A: IFT20 expression was significantly reduced in the Gmap210 KO mice, as examined by Western blot analysis. Three independent experiments were performed, and representative data of Western blot results are shown. *P < 0.05, compared with control (Student’s t test). B: examination of IFT20 localization in the control (a–d) and conditional Gmap210 KO mice (e–h). Immunofluorescence staining was conducted on germ cells collected from the control and conditional Gmap210 KO mice. In controls, IFT20 was present in Golgi bodies of spermatocytes (arrows in a) and localized in the acrosome region of round spermatids (arrowheads in b and c) and manchette of elongating spermatids (dashed arrow in d). In the conditional Gmap210 KO mice, IFT20 was no longer present in these structures (e–h). Instead, the protein was dispersed as individual vesicles (arrows in e–g). In the elongating spermatids of conditional Gmap210 KO mice, the manchette appears to be normal (dashed arrow in h), although IFT20 is absent. α-tubulin staining is a marker of the manchette microtubules.

However, in the Gmap210 knockout mice, IFT20 was no longer present in these structures (Fig. 9B, e–h). Instead, the protein was dispersed as individual vesicles in the cytoplasm (Fig. 9B, e–g). The manchette appeared to be normal in the conditional Gmap210 knockout mice; however, IFT20 was missing in this structure (Fig. 9Bh).

Abnormally developed acrosome and mitochondrial sheath in the conditional Gmap210 knockout mice.

Immunofluorescence staining was conducted on the testicular germ cells of control and the Gmap210 knockout mice. In control mice, normal peanut-lectin signal was observed in the acrosome region of round and elongating spermatids (Fig. 10Aa). However, the peanut-lectin staining was either absent or disrupted in the Gmap210 knockout mice (Fig. 10, Ab and Ac).

Fig. 10.

Abnormal acrosome development and mitochondrial sheath formation in the conditional Golgi-microtubule-associated protein of 210 (Gmap210) knockout (KO) mice. A: examination of the acrosome in isolated testicular germ cells by immunofluorescence staining. In control mice, normal lectin signal was observed in round and elongating spermatids (arrows in a). However, in conditional Gmap210 KO mice, lectin labeling was either absent (b) or disrupted (dashed arrow in c). B: examination of the acrosome in epididymal sperm by immunofluorescence staining. In control mice, normal acrosome staining with the lectin was observed (arrows in a). However, in conditional Gmap210 KO mice, most sperm had no acrosome staining (b). Occasionally, acrosome staining was observed, but the staining was either present over the entire head (dashed arrow in c), or highly concentrated in certain areas (arrow in d). C: examination of the mitochondrial sheath in epididymal sperm by immunofluorescence staining with MitoTracker Red. In control mice, specific staining was present only in the midpiece (arrow in a). In Gmap210 KO mice, staining of the midpiece was barely observable. Instead, a strong signal was seen over the head region of the abnormal sperm (arrowheads in b–d). In some sperm, some staining was observed in the midpiece but only near the head (dashed arrows in c and d). DIC, differential interference contrast.

The acrosome was also examined in epididymal sperm. In control mice, normal peanut-lectin labeling of the acrosome was observed (Fig. 10Ba). However, in conditional Gmap210 knockout mice, most sperm had no acrosome staining (Fig. 10Bb); however, an occasional staining with peanut-lectin was observed in some sperm, but the signal was either present over the entire head (Fig. 10Bc) or highly concentrated in certain areas, without the normal formation of a cap over the nucleus (Fig. 10Bd).

To stain mitochondria, epididymal sperm were collected from the control and conditional Gmap210 knockout mice. In control mice, specific staining for mitochondria was present only in the midpiece of sperm (Fig. 10Ca). In the Gmap210 knockout mice, the midpiece staining pattern was barely observable. Instead, a strong signal was seen surrounding the heads of sperm (Fig. 10C, b–d), or in some cases, the staining was observed in what appeared to be a shortened midpiece near the head and in midpieces coiled around the head (Fig. 10, Cc and Cd).

Testicular expression levels of selected acrosomal proteins, including zona pellucida binding protein 1 (ZPBP1), sperm acrosome associated 1 (SPACA1), and matrix metallopeptidase 2 (MMP2) were examined by Western blotting. Compared with the control, ZPBP1 and SPACA1 expression levels were significantly reduced in the conditional Gmap210 knockout mice; however, MMP2 level was not changed (Fig. 11A).

Fig. 11.

Abnormal expression pattern of acrosomal proteins in the conditional Golgi-microtubule-associated protein of 210 (Gmap210) knockout (KO) mice. A: examination of expression levels of the selected acrosomal proteins by Western blotting. Notice that the expression levels of selective acrosomal proteins, including sperm acrosome associated 1 (SPACA1) and zona pellucida binding protein 1 (ZPBP1), were dramatically reduced in the knockout mice; however, matrix metallopeptidase 2 (MMP2) level was not changed. B: examination of localization of SPACA1 by immunofluorescence. Notice that strong signal of SPACA1 was detected in the acrosome of spermatids from the control mice; the signal was much weaker in the conditional Gmap210 knockout mice.

Localization of these acrosomal proteins was also examined by immunofluorescence. SPACA1, ZPBP1, and MMP2 were closely associated with the acrosome in control mice, and their signals in the acrosome were dramatically reduced in the conditional Gmap210 knockout mice (Figs. 11B, 12, and 13).

Fig. 12.

Abnormal expression pattern of zona pellucida binding protein 1 (ZPBP1) in the conditional Golgi-microtubule-associated protein of 210 (Gmap210) knockout mice. Examination of localization of ZPBP1 by immunofluorescence. Notice that strong signal of ZPBP1 was detected in the acrosome of spermatids from the control mice; the signals were much weaker in the conditional Gmap210 knockout mice.

Fig. 13.

Abnormal expression pattern of matrix metallopeptidase 2 (MMP2) in the conditional Golgi-microtubule-associated protein of 210 (Gmap210) knockout mice. Examination of localization of matrix metallopeptidase 2 (MMP2) by immunofluorescence. Notice that strong signal of MMP2 was detected in the acrosome of spermatids from the control mice; the signals were much weaker in the conditional Gmap210 knockout mice.

Ultrastructural changes in epididymal sperm of conditional Gmap210 knockout mice.

Ultrastructural changes in cauda epididymal sperm of the conditional Gmap210 knockout mice were examined by transmission electron microscopy. Control mice had highly condensed nuclei with normal morphology and well-developed flagella (Fig. 14, A and B). In conditional Gmap210 knockout mice, most sperm had more oblong or misshapen heads (Fig. 14, C, E, and G). The core axonemal structure of the midpiece appeared to be normal (Fig. 14D). Excess cytoplasm or residual bodies, which were not resorbed by Sertoli cells in the testis, were often observed in the preparations and were seen standing alone or aggregated near the nucleus (Fig. 14, C, E, and H).

Fig. 14.

Ultrastructural changes of epididymal sperm of the conditional Golgi-microtubule-associated protein of 210 (Gmap210) knockout (KO) mice. The control mice had nicely condensed nuclei (A) and well-developed tails (B). M, midpiece; P, principle piece. Most sperm from conditional Gmap210 KO mice had more oblong heads or distorted nuclei (black arrows in C, E, and G). The midpiece appeared to be normal in some sperm (D) but appeared to be disorganized with too many mitochondria in others (E), possibly due to the failure of cytoplasmic lobe formation and removal of residual bodies in the testis. Mitochondria components not assembled into sperm flagella were present (arrowheads in F); some more oblong heads were surrounded by excess cytoplasm (dashed arrow in G). Sloughed residual bodies were also observed in the sperm preparations (dashed arrows in H).

DISCUSSION

Cilia are cell organelles that extend from the surface of nearly all mammalian cells (6). Cilia are formed by a process that requires intraflagellar transport (IFT), a bidirectional motility along axonemal microtubules that is essential for the formation (ciliogenesis) and maintenance of most eukaryotic cilia and flagella (44, 50). The process of IFT involves movement of large protein complexes called IFT particles or trains from the cell body to the ciliary tip, followed by their return to the cell body (5). Among the known IFT particle proteins, IFT20 is the only one present in the Golgi complex. It has been shown that GMAP210 is a Golgi membrane receptor for IFT20. Our laboratory discovered that the IFT20 protein is highly expressed in the testis, and it is essential for male fertility and spermiogenesis in mice (66). Therefore, the present study of GMAP210 was initiated to extend our understanding of its relationship to IFT20 in the regulation of spermatogenesis and to examine the cause of male infertility in the conditional germ cell mutant.

This study demonstrated that GMAP210 is essential for acrosome formation and male fertility. In conditional Gmap210 mutant mice, fewer germ cells were able to complete spermatogenesis, which led to a significant reduction in sperm counts compared with controls. Although testis histology showed normal mitosis and meiosis in the mutant mice, spermiogenesis was abnormal and consistent with the predicted role of GMAP210 in spermatogenesis and male fertility.

A striking phenotype observed in the Gmap210 mutant mouse was the formation of more oblong sperm heads that lacked acrosomes, which are also referred to as globozoospermia-like. Globozoospermia is a common human infertility syndrome caused by spermatogenetic malfunction (26, 27, 52), which results in abnormal shapes of the sperm heads and malformation of the acrosome, and, in the most serious cases, total absence of acrosomes. The acrosome is a Golgi-derived vesicular organelle that flattens over the anterior half of the sperm nucleus. It is linked to the nuclear membrane through a thin perinuclear theca layer (39, 40) that contains the acroplaxome cytoskeletal components of the subacrosomal layer (22, 24). During the Golgi and cap phases of spermiogenesis, the acrosome is formed by fusion of proacrosomal vesicles derived from the trans-Golgi network (1, 30, 35, 39) and is gradually modified until it reaches its final shape. The size and inner structure of the mature acrosome show considerable variation between species (55).

In the Gmap210 mutant spermatids, a nuclear lamina formed at the nuclear membrane and small proacrosomal vesicles, with and without dense content, were seen attached to the area forming a partial subacrosomal-perinuclear theca layer, with what appeared to be ultrastructurally an acroplaxome. However, the vesicles did not fuse and therefore a cap-like acrosomal vesicle could not form. Failure of proacrosomal vesicle fusion has been observed in other knockout mice (4, 17–19, 21, 25, 63), along with treatment with the chemical N-butyldeoxynojirimycin (39). In each case, there appears to be a disruption in the spermatid Golgi transport system or the trafficking of proteins to the acrosomal membrane or acrosome-associated proteins in the perinuclear theca layer. Specific effects on these proteins in Gmap210 mutant spermatids remain to be determined.

Numerous mouse knockout models have been found to produce globozoospermic-like phenotypes, including Hrb, Gopc, Pick1, Hsp90b1, Dpy19l2, Tmf1, Atg7, Mfsd14a, Spaca1, Smap2, Zpbp1/Zpbp2, GM130, Brd7, Slc9a8, and AU040320 (2, 10, 15–18, 21, 29, 31, 37, 45, 59, 60, 62, 63). Another example is homozygous mutation (L967Q) in the Vps54 gene (41), as well as siRNA knockdown of lamin A/C (51). In all of these globozoospermia genetic models, the common factors are disruption of acrosome biogenesis and an association with the Golgi complex involving a range of complex elements, from protein processing in the endoplasmic reticulum and fusion of proacrosomal vesicles to the adhesion of the acrosome with the acroplaxome (8, 9). It is noteworthy that in nearly every knockout animal model that results in abnormal formation of the acrosome or failure to form an acrosome, there is also abnormal nuclear shaping and the formation of more oblong heads, as well as nucleopodes, similar to those described for the Hrb mutant mouse (25).

Our intensive transmission electron microscopy studies demonstrated that Golgi bodies were more diffuse and did not arch over the nucleus in the absence of GMAP210. Proacrosome vesicles were present but not fused into large acrosome granules. Thus, acrosome biogenesis was affected in the absence of GMAP210. Round heads and abnormal manchette morphology might be the secondary defects of disrupted acrosome biogenesis. Consistently, localization and/or expression levels of some acrosomal proteins, including ZPBP1 (31, 36, 64), MMP2 (12), and SPACA1 (13), were also changed in the absence of GMAP210, which also contributed the defect in acrosome formation. Interestingly, even though testicular ZPBP1 and SPACA1 levels were reduced in the conditional Gmap210 knockout mice, MMP2 level was not changed, which indicates that different acrosomal proteins have different fate in germ cells in the absence of GMAP210.

It is noteworthy to consider that the formation of an acrosome is very important for normal elongation of the sperm head, possibly through the Sertoli cell F-actin hoops/acroplaxome/manchette interaction (23, 24). More oblong sperm heads with failure to form the acrosome are key features of globozoospermia and result from a disruption in this complex morphogenetic programing that is taking place during late spermiogenesis. Thus, the observed head abnormalities in Gmap210-deficient spermatozoa are likely to be secondary effects resulting from the failure to form an acrosome.

Interestingly, abnormal mitochondria assembly was discovered in the developing spermatids and cauda sperm in the absence of GMAP210. Besides being involved in acrosome biogenesis, GMAP210 might also play a role in sperm mitochondria sheath formation. Abnormal sperm mitochondria sheath formation might also be caused by unknown mechanisms when sperm pass through epididymis. This was also observed in other knockout mouse models in which acrosome biogenesis was affected (33, 37).

Even though GMAP210 and IFT20 are binding partners, the two proteins play very different roles in the regulation of male germ cell development. We previously demonstrated that IFT20 is also essential for male fertility and spermiogenesis in mice (66). However, acrosome formation was not affected in the conditional Ift20 knockout mice (66), and the phenotype was different from the Gmap210 knockout mice as described here. In the Ift20 knockout mice, epididymal sperm of the mutant Ift20 mice exhibited abnormal 9+2 axonemal structures, indicating improper flagella assembly, which is consistent with defective function of an IFT protein; however, axonemal structures appeared to be normal in the Gmap210 knockout mice, indicating that the protein does not have a role in axoneme formation. In the Ift20 knockout mice, cytoplasmic lobe material remained in the midpiece region, which resulted in bent sperm, but in the Gmap210 knockout mice numerous residual bodies failed to be resorbed and were released into the lumen with the abnormal sperm. Sperm tails also showed a difference in the two models. Ift20 knockout sperm tails were kinked or shortened, whereas those of the Gmap210 knockout mice were smooth with normal lengths. The difference in phenotypes might lie in the fact that IFT20 conducts dual functions in germ cell development: 1) directing sperm flagella formation by IFT mechanisms (61) and 2) helping to regulate normal autophagy, as is required to clear redundant cytoplasmic components through its association with the autophagy core protein ATG16L (42), which would be involved in the removal of excess cytoplasm by the residual body and subsequently phagocytized by Sertoli cells (56–58). In the Gmap210 knockout mice, some residual bodies were not resorbed, which might have been caused by the disrupted localization of IFT20 protein in the absence of GMAP210 and a potential interference in the autophagy role of IFT20.

Given that IFT20 localization was disrupted in Golgi bodies of the Gmap210 mutant spermatids, it appears that one of the key mechanisms of GMAP210’s role in spermatogenesis is to maintain normal localization of IFT20, as well as its expression level. Compared with controls, testicular IFT20 protein expression in the conditional Gmap210 knockout mice was significantly decreased, and IFT20 localization was abnormal. Usually IFT proteins function as adaptors because they collaborate with other proteins to form complexes to transport cargo (28). Other IFT20 binding partners were identified, including sperm flagellar protein 2 (SPEF2), coiled-coil domain containing 41 (CCDC41), and the exocyst and biogenesis of lysosome-related organelles complex-1 (BLOC-1) (20, 53). Our recent studies showed that IFT20 is present on the acrosome surface, and the localization seems to be dependent on other proteins (66). IFT20 is considered to be a cargo of other proteins, including GMAP210, and thus is not thought to be involved in acrosome biogenesis. However, GMAP210 is responsible for the transport of vesicles from the Golgi for acrosome formation, including those that would contain IFT20. In the absence of GMAP210, acrosome biogenesis is disrupted and the IFT20 docking site on the acrosome surface is no longer present. Even though the IFT20 level was reduced in the conditional Gmap210 knockout mice, and its localization was also affected, the remaining IFT20 was sufficient for the germ cells to develop normal flagella through IFT mechanism. Thus, IFT20 transport into the developing flagellum was not disturbed.

In summary, a role for GMAP210 in spermatogenesis was discovered. GMAP210 is required for the fusion of proacrosomal vesicles in the normal process of acrosome formation as well as normal expression level and localization of IFT20 and some acrosomal proteins. It is essential for male fertility. Mutation of Gmap210 resulted in the formation of abnormal spermatozoa with more oblong heads and may represent another genetic factor associated with globozoospermia and reduced fertility in men.

GRANTS

This research was supported by NIH National Institute of Child Health and Human Development Grants HD076257 and HD090306 and by Wayne State University Start-up Fund and Research Fund (to Z. Zhang), National Institute of General Medical Sciences Grant GM060992 (to G. J. Pazour), National Natural Science Foundation of China (81671514 and 81502792), the China Scholarship Council Fund (201708420223), Excellent Youth Foundation (2018CFA040) and Youth Foundation (2018CFB114) of Hubei Science and Technology Office, and Special Fund of Wuhan University of Science and Technology for Master Student’s short-term studying abroad.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

Z.W., Y.S., W.L., S.A.K., G.J.P., R.A.H., and Z.Z. conceived and designed research; Z.W., S.M., Q.H., Y.T.Y., L.S., S.Z., T.Z., W.L., B.H., and L.Z. performed experiments; S.M., Q.H., L.S., S.Z., T.Z., B.H., L.Z., S.A.K., and G.J.P. analyzed data; W.L., S.A.K., R.A.H., and Z.Z. interpreted results of experiments; Y.T.Y. prepared figures; Y.S. drafted manuscript; Z.Z. edited and revised manuscript; Z.W., Y.S., S.M., Q.H., Y.T.Y., L.S., S.Z., T.Z., W.L., B.H., L.Z., S.A.K., G.J.P., R.A.H., and Z.Z. approved final version of manuscript.