Oligo-astheno-teratozoospermia in mice lacking ORP4, a sterol-binding protein in the OSBP-related protein family

Authors


  • Communicated by: Masayuki Miura

Abstract

Oligo-astheno-teratozoospermia (OAT), a condition that includes low sperm number, low sperm motility and abnormal sperm morphology, is the commonest cause of male infertility. Because genetic analysis is frequently impeded by the infertility phenotype, the genetic basis of many of OAT conditions has been hard to verify. Here, we show that deficiency of ORP4, a sterol-binding protein in the oxysterol-binding protein (OSBP)-related protein family, causes male infertility due to severe OAT in mice. In ORP4-deficient mice, spermatogonia proliferation and subsequent meiosis occurred normally, but the morphology of elongating and elongated spermatids was severely distorted, with round-shaped head, curled back head or symplast. Spermatozoa derived from ORP4-deficient mice had little or no motility and no fertilizing ability in vitro. In ORP4-deficient testis, postmeiotic spermatids underwent extensive apoptosis, leading to a severely reduced number of spermatozoa. At the ultrastructural level, nascent acrosomes appeared to normally develop in round spermatids, but acrosomes were detached from the nucleus in elongating spermatids. These results suggest that ORP4 is essential for the postmeiotic differentiation of germ cells.

Introduction

Spermatogenesis is the process of male germ cell proliferation and differentiation. The former process starts with divisions of the spermatogonial stem cells and continues with sequential cell divisions of spermatogonia, and meiosis of spermatocytes to form round spermatids. In the final stage of spermatogenesis, called spermiogenesis, round spermatids differentiate into elongated spermatids and eventually into spermatozoa. During spermiogenesis, postmeiotic round spermatids undergo dynamic morphologic changes, which include nuclear condensation and elongation, formation of the flagellum and acrosome, reorganization of organelles and elimination of cytoplasm upon spermiation (Yan 2009). Disruption of germ cell proliferation or induction of germ cell apoptosis results in low sperm concentration (oligozoospermia), which is major symptom of the male infertility. In contrast, defects in spermiogenesis cause poor sperm motility (asthenozoospermia) and abnormal sperm morphology (teratozoospermia). These spermiogenic defects often induce germ cell apoptosis, thus leading to a combination of the syndromes called oligo-astheno-teratozoospermia (OAT syndrome). Although some cases of OAT syndrome can be clinically associated with varicocoele or other pathologies, approximately 60% of OAT individuals are diagnosed as idiopathic, which is assumed to include genetic abnormalities (Lessard et al. 2007). However, the genetic factors underlying OAT are largely unknown.

Oxysterol-binding protein (OSBP) was first identified as a high-affinity cytosolic receptor for oxysterols, such as 25-hydroxycholesterol (Taylor et al. 1984). Thereafter, it was shown to interact with cholesterol and to have the ability to transfer cholesterol from donor to acceptor liposomes (Ngo & Ridgway 2009). OSBP possesses a pleckstrin homology (PH) domain at the amino-terminus, an FFAT (diphenylalanine in an acidic tract) motif at the middle region, and a 400-residue sterol-binding domain at the carboxy-terminus (Fig. S1 in Supporting Information). OSBP transiently localizes to the ER and Golgi apparatus via FFAT motif interaction with VAMP-associated protein (VAP) and phosphatidylinositol-4-phosphate (PI4P) binding by the PH domain, respectively (Lagace et al. 1997; Levine & Munro 1998; Wyles et al. 2002). Most eukaryotes have proteins homologous to OSBP, named ORP (OSBP-related protein). Mammals have 12 ORPs (OSBP and ORP1-11) that can be subdivided into six subfamilies based on gene organization and amino acid homology (Fig. S2 in Supporting Information). The defining feature of the ORP family is a C-terminal sterol-binding domain that harbors an ORP signature motif ‘EQVSHHPP’ that is conserved in orthologues from other species, such as Caenorhabditis elegans (four ORPs, Kobuna et al. 2010), Drosophila melanogaster (four ORPs, Alphey et al. 1998) and Saccharomyces cerevisiae (seven ORPs, Beh et al. 2001; Fig. S2 in Supporting Information).

ORPs have recently been characterized as putative cholesterol carriers. Mammalian OSBP, ORP5, ORP9 and yeast ORP, Osh4p, transfer sterols between liposomes in vitro (Raychaudhuri et al. 2006; Ngo & Ridgway 2009; Schulz et al. 2009). A crystal structure analysis of Osh4p suggested that sterol and membrane binding facilitates reciprocal changes in ORP conformation, which accelerates the sterol transport cycle between donor and acceptor membranes (Im et al. 2005). At the cellular level, the transport of newly synthesized cholesterol from the ER to the cell surface is enhanced by expression of mammalian ORP2 (Laitinen et al. 2002). In yeast, ORPs were shown to transport sterol from the plasma membrane to the ER, and disruption of ORPs was shown to cause abnormal cholesterol distribution in the cell (Raychaudhuri et al. 2006). In addition to intracellular sterol transport, ORPs have been implicated in a variety of cellular processes including cell signaling, vesicular trafficking, microtubule-dependent endosomal positioning and lipid metabolism (Wang et al. 2005; Yan et al. 2008; Ngo & Ridgway 2009; Rocha et al. 2009; Banerji et al. 2010; Ngo et al. 2010; Raychaudhuri & Prinz 2010). However, the in vivo function of ORPs in mammals remains elusive due to the lack of a genetic deficiency state in human subjects and the lack of appropriate animal models deficient in ORP proteins. This is the first report that shows the physiological function of ORP family proteins in mice.

Results

ORP4 is specifically expressed in postmeiotic germ cells in testis

We first analyzed the expression pattern of ORP4 in mice. Western blot analysis using an anti-ORP4 monoclonal antibody showed selective expression of ORP4 in brain and testis (Fig. 1A). In contrast, OSBP, the closest homologue of ORP4, was expressed ubiquitously in various mouse tissues (Figs S1–S3 in Supporting Information; Fig. 1A). In the brain, ORP4 is highly expressed in the hippocampus, particularly in the apical dendrite of hippocampal CA1 neuron. We found no overall morphological changes in the brain but slight abnormality in the apical dendrites of the hippocampal CA1 pyramidal cells in ORP4 knockout mice. In this study, we focused our further analysis on the testis. In situ hybridization analysis showed the luminal localization of ORP4 mRNA in seminiferous tubules in the testis (Fig. 1B), indicating that ORP4 was expressed in spermatids. No signals were detected in interstitial Leydig cells (Fig. 1B: arrows). Consistently, an increase in ORP4 expression was observed during the second to fourth postnatal week (Fig. 1C), which correlates with the initiation of postmeiotic spermiogenesis. We also carried out fractionation of the testicular cells and found that ORP4 mRNA was enriched in the germ cell fraction, but not in the Sertoli cell fraction (Fig. 1D; Fig. S4 in Supporting Information). Taken together, these data indicate that ORP4 is specifically expressed in postmeiotic germ cells, but not in somatic cell types, in the testis.

Figure 1.

ORP4 is specifically expressed in postmeiotic germ cells in testis. (A) Western blot analysis of various organs using antibodies against ORP4 and oxysterol-binding protein. (B) in situ hybridization of ORP4 mRNA in elongating spermatids in seminiferous tubules. Blue signal represents ORP4 mRNA enriched at the luminal region of seminiferous tubules. Lu, lumen of seminiferous tubules. Arrows indicate interstitial Leydig cells. Scale bars, 50 μm. (C) Western blot analysis of testes at different postnatal stages (0.5–15 weeks) using an anti-ORP4 antibody. (D) Quantitative RT-PCR analysis of fractionated testicular cells. ORP4 is selectively expressed in the germ cell-rich fraction (closed bars), not in the Sertoli fraction (open bars). Data are presented as mean ± SD (n = 3).

ORP4 knockout male mice were infertile

To determine the physiological role of ORP4 in mammals, we generated ORP4-deficient mice by gene targeting. An ORP4 targeting vector substituted a neomycin-resistant gene for exons 4–8 (Fig. 2A), which contain the FFAT motif and the anterior half of the sterol-binding domain (Fig. S1 in Supporting Information). In ORP4−/− mice, expression of ORP4 was not detected either in tissues (Fig. 2B) or in sections of seminiferous tubules (Fig. 2C). As described above, ORP4 was expressed in postmeiotic germ cells, such as spermatids and spermatozoa, but not in early germ cells (spermatogonia and spermatocytes) in ORP4+/+ mice (Fig. 2C). Crosses between heterozygous mice yielded offspring that segregated in a Mendelian fashion (Table 1). ORP4−/− mice were healthy and showed no overt developmental abnormalities. However, ORP4−/− males were consistently infertile despite normal mating behavior and ejaculation with normal vaginal plug formations (Table 2). However, ORP4−/− females showed normal fecundity.

Table 1. Genotypes of litters from ORP4+/− intercrosses
ORP4 genotype+/++/−−/−Total
  1. Yielded offsprings are segregated in a Mendelian fashion.

Number of live-born mice366732135
%26.749.623.7100
Table 2. ORP4−/− mice present male infertility
 Reproductive function n
Genotype♂Plugs/females tested (%)Pregnancies/females mated (%)
  1. A male was mated with two females and housed together for a month. Females were scored for pregnancy rates. See the text for the detailed assessment procedure.

+/++/+8/14 (57.1)14/14 (100)12
+/+−/−7/13 (53.8)12/12 (100)11
−/−+/+6/11 (54.5)0/11 (0)11
Figure 2.

Targeted disruption of ORP4. (A) Schematic representation of the ORP4 gene-targeting strategy. Homologous recombination of the targeting vector with the wild-type allele results in the replacement of 3.9 kilobases of the ORP4 genomic sequence containing exons 4–8 with the neo cassette. A thymidine kinase gene (TK) was used for negative selection of clones with random integration of the targeting vector. (B) Western blot analysis of testis extracts from ORP4+/+, ORP4+/− and ORP4−/− mice using an anti-ORP4 antibody. (C) Cross-sections through ORP4+/+ and ORP4−/− seminiferous tubules, stained with an anti-ORP4 antibody (green). The outline of the seminiferous tubule is indicated by a white dotted line. No signal was detected in the ORP4−/− seminiferous tubule. The bright blue staining represents DAPI-stained nuclei. Indicated are positions of spermatogonium (G), pachytene spermatocytes (P), round spermatids (Rs) and elongated spermatids (Es) as judged by nuclear morphology. Scale bars, 50 μm.

Deficiency of ORP4 results in oligo-astheno-teratozoospermia

To determine the reason for the infertility of ORP4−/− mice, we analyzed spermatozoa from the cauda epididymides. The number of spermatozoa recovered from the cauda epididymides of ORP4−/− mice (0.75 ± 0.45 × 106) was much smaller than the number recovered from ORP4+/+ mice (15.2 ± 2.0 × 106; Fig. 3A). Hematoxylin and eosin (H&E) staining also showed a remarkably reduced number of spermatozoa in the lumen of the cauda epididymides in ORP4−/− mice (Fig. 3B). In ORP4+/+ cauda epididymides, more than 80% of the spermatozoa had a typical falciform head, although a portion of the sperm showed morphological defects under the C57/BL6 background (Fig. 3C, +/+), as reported previously (Albert & Roussel 1984). In contrast, ORP4−/− mice showed a striking increase in the number of spermatozoa with abnormal morphology (Fig. 3C, −/−). More than half of ORP4−/− spermatozoa had irregularly shaped heads such as round-shaped heads and curled back heads. Symplasts, which are formed by undivided elongating spermatids, were also frequently formed in ORP4−/− mice. Spermatozoa derived from ORP4−/− mice showed little or no motility (Movies in Supporting information) and had no fertilizing ability in vitro (Fig. S5 in Supporting Information). To reveal the reason for the reduced content of spermatozoa, we carried out TUNEL staining to visualize apoptotic cells. TUNEL staining revealed a significant increase in apoptotic germ cells, especially of elongated spermatids with flagellum, in the seminiferous epithelium in ORP4−/− testis (Fig. 3D). This result suggests that ORP4−/− germ cells during spermiogenesis undergo apoptosis in the seminiferous epithelium, leading to the reduced spermatozoa in the cauda epididymides of ORP4−/− mice. Taken together, these data indicate that deficiency of ORP4 resulted in severe oligo-astheno-teratozoospermia (low sperm number, low sperm motility and abnormal sperm morphology).

Figure 3.

Deficiency of ORP4 results in oligo-astheno-teratozoospermia. (A) Numbers of spermatozoa from a single cauda epididymidis of ORP4+/+ and ORP4−/− mice. Data are presented as mean ± SEM (n = 3). (B) Hematoxylin/eosin-stained section of cauda epididymides from ORP4+/+ and ORP4−/− mice. Scale bars, 50 μm. (C) Pie charts showing multiple sperm types and shapes in ORP4−/− mice. Two hundred sperm cells were analyzed for each genotype. Representative pictures of typical sperm shapes are shown on the right. (a) Noncoiled symplast, (b) headless, (c) normal shape, (d) fully coiled intracellularly, (e) partly coiled intracellularly and (f) noncoiled irregularly shaped monoflagellated Scale bars, 2 μm. (D) Detection of germ cell apoptosis. TUNEL staining (green) in the seminiferous tubules of ORP4+/+ and ORP4−/− mice. The outline of the seminiferous tubule is indicated by a white dotted line. Note that most TUNEL-positive cells have a flagellum (arrows). Scale bars, 50 μm.

ORP4 knockout mice showed abnormal morphology of postmeiotic germ cells

To analyze the morphology of male germ cells before spermiation, we analyzed the morphology of spermatogonia, spermatocytes and spermatids in seminiferous tubules by toluidine blue staining and transmission electron microscopy. In ORP4−/− seminiferous tubules, spermatogenesis appeared to progress rather normally, but the morphology of elongating and elongated spermatids was severely affected (Fig. 4). Nascent acrosomes appeared to normally develop in a round spermatid (Fig. 5A, arrow & Fig. S6A in Supporting Information), although some unusual features, such as bleb-like structures, were formed (Fig. 5A, arrowheads). Many elongated spermatids showed abnormal morphology and were degraded, showing unusual symplasts (Figs 4C,E and 5B,C, asterisks) and deformed cytoplasmic organelles (Fig. 5B, arrowheads). The axonemal apparatus and the nine outer dense fibers were not affected in ORP4−/− spermatozoa (Fig. S6B in Supporting Information). Consequently, Sertoli cells showed increased phagocytic activity for the degenerated elongated cells and contained phagocytosed spermatids in the basal cytoplasm near the basement membrane (Figs 4F and 5D, asterisks). Remarkably, nuclear condensation was not affected in ORP4−/− spermatids (Fig. 5B, open arrows); in fact, offsprings after intracytoplasmic sperm injection (ICSI) with ORP4−/− sperm into oocyte grew normally (Fig. S7 in Supporting Information), and no significant difference was found between ORP4−/− sperm and ORP4+/− sperm in the cleavage rate, implantation rate and birth rate after ICSI (Table 3). Thus, these data suggest that early spermatogenesis is not affected in ORP4−/− mice. Consistent with this idea, a bromodeoxyuridine (BrdU)-labeling assay showed that BrdU was efficiently incorporated into proliferating spermatogonia in ORP4−/− seminiferous tubules in a manner similar to that in ORP4+/+ tubules (Fig. 6A). Furthermore, flow cytometry of germ cells stained with propidium iodide showed that the distribution of DNA content in ORP4−/− germ cells was not affected. These observations indicate that spermatogonia proliferation and subsequent meiosis occur normally in ORP4−/− mice (Fig. 6B). Together, these data indicate that ORP4 is required for the postmeiotic differentiation of germ cells, but not for the proliferation and meiosis of early germ cells.

Table 3. ICSI with ORP4 heterozygotic (+/−) and homozygotic (−/−) elongated spermatozoa
Genotype of malesNo. oocytes culturedNo. (%) 2-cellNo. embryos transferredNo. (%) implantedNo. (%) offspringNo. (%) weaned
  1. No statistical significance is found in the cleavage rate [No. (%) of 2-cell], implantation rate [No. (%) of implanted] and birth rate [number (%) of offspring and number (%) of weaned]. Spermatids were retrieved from ORP4+/− or ORP4 −/− testes. See the text for the detailed techniques. Data were examined by Fisher's exact test for the duplicates. Elongated spermatids had almost the same ability to produce normal pups, suggesting the intactness of the paternal genome.

+/−5242 (81)4228 (67) 21 (50)21 (100)
−/−8765 (75) 6538 (58) 26 (40) 26 (100)
Figure 4.

Abnormal morphology of postmeiotic germ cells in ORP4 knockout mice. Light microscopy images of ORP+/+ (A and B) and ORP4−/− (C, D, E and F) spermatogenesis. Toluidine blue staining. (A) Stage V with step 5 (S5) and 15 (S15) spermatids. (B) Stage VII with step 7 (S7) and 16 (S16) spermatids. (C) Stage III with step 3 (S3) and 14 (S14) spermatids. (D) Stage V with step 5 (S5) and 15 (S15) spermatids. (E) Stage IX with step 9 (S9) spermatids. (F) Stage XII with step 12 (S12) spermatids. (C, D and F) Germ cells at various stages are exfoliated into the lumen (**). (C and E) Unusual symplasts (*) are frequently formed. (F) Sertoli cell (S) increases the phagocytic activity (*). P, pachytene spermatocytes; BM, basement membrane; G, spermatogonia; S, Sertoli cells; Sec, secondary spermatocytes. Scale bars, 10 μm. (G) Schematic drawing to show the occurrence of abnormal morphology in postmeiotic germ cells of ORP4 knockout mice. Original data are shown in (A–F). Red, light blue and green for the abnormality occurred in the nucleus, acrosome and cytoplasmic organelles. Spermatogenesis proceeded normally in ORP4−/− testes until approximately round spermatid production. In spermiogenesis, early round spermatids appeared to normally develop, but bleb-like structures in nuclei (red) were observed in mid- to late round spermatids. In elongating and elongated spermatids, degeneration was found severe especially in acrosome (light blue) and cytoplasmic organelles (green). Sertoli cells with phagocytosed cells were often observed.

Figure 5.

Abnormal spermatogenesis of ORP4−/− testes. Transmission electron micrographs. (A) Stage VII. (B) Approximately stages VIII-IX. (A) Nascent acrosomes appear to normally develop (arrow) in a step 7 (S7) round spermatid, although unusual bleb-like structures are formed at the peripheral region of the nuclei (arrowheads). (B) Although the acrosomes of S8–9 spermatids appear to be normally formed, the acrosomes of S16 elongated spermatids are detached from the nuclei (arrows). Nuclear condensation occurred normally (open arrows). Arrowheads indicate degenerated organelles. Note that a symplast is frequently observed in ORP4−/− testes (*). (C) A huge degenerated symplast (*) in the cauda epididymal lumen. Inset; magnified sperm head region from Figure 5C (square area). Note that the tail (arrow) is unusually bent toward the head region. (D) Accelerated phagocytosis of Sertoli cells. Degenerated elongated spermatids (*) are found in the basal cytoplasm of Sertoli cell (S) near the basement membrane (BM). Stage XII. (E) Spermatogenesis in wild-type (control). Scale bars, 2 μm (A–E) and 0.5 μm (C, inset).

Figure 6.

ORP4 is not required for proliferation and meiosis of early germ cells. (A) Immunohistochemical detection of bromodeoxyuridine (green) in ORP4+/+ and ORP4−/− seminiferous tubules 24 h after its administration. Scale bars, 500 μm. (B) Flow cytometry histogram of a suspension of testicular cells stained with propidium iodide. Percentage of ORP4+/+or ORP4−/− cells in each gate is indicated. Data are presented as mean ± SEM (n = 6 for each group).

Removal of cytoplasm from the sperm head is impaired in ORP4 knockout mice

During spermiogenesis, wild-type spermatids undergo a complex reconstruction program including nuclear condensation and elongation, formation of the flagellum and acrosome, rearrangement of mitochondria and elimination of cytoplasm from the sperm heads (Russell et al. 1990; Toshimori 2009). To further analyze the postmeiotic differentiation of ORP4−/− germ cells, we stained spermatozoa using MitoTracker and a cytoplasm marker. In spermatozoa from ORP4+/+ cauda epididymides, mitochondria localized to the middle piece of sperm tails (Fig. 7A, +/+, brackets). In contrast, mitochondrial sheath was coiling around irregularly round heads and lacking from the midpiece of tails of most ORP4−/− spermatozoa remaining in cauda epididymides (Fig. 7A, −/−, arrows), which did not show any abnormalities in flagella formation. To stain the cytoplasm, we used an antibody to ubiquitin, which is highly expressed in the cytoplasm of late spermatids (Zheng et al. 2007). In ORP4+/+ spermatozoa, ubiquitin was detected in the cytoplasmic droplet, which is located between the neck and the middle piece of the sperm tail (Fig. 7B, +/+, arrowheads). The cytoplasmic droplet is a remnant of the cytoplasm that is shed off in the late phase of spermiogenesis (Hermo et al. 1988). In contrast, ubiquitin was often observed in the cytoplasm at the head region of ORP4−/− spermatozoa (Fig. 7B, −/−, arrows), indicating that detaching and removal of cytoplasm from the sperm head are impaired during late spermiogenesis in ORP4−/− mice.

Figure 7.

Lack of ORP4 causes aberrant cytoplasm removal from the head region and mislocalization of MN13 molecule in mature sperm. (A) Mitochondrial sheath configurations in ORP4+/+ and ORP4−/− spermatozoa visualized by MitoTracker red. Arrows indicate the head of spermatozoa. A bracket indicates the middle piece of the sperm tail where mitochondria are commonly found. Scale bars, 10 μm. (B) Immunofluorescent detection of ubiquitin in ORP4+/+ and ORP4−/− spermatozoa (green). Arrows and arrowheads indicate the head and the cytoplasmic droplet of spermatozoa, respectively. Ubiquitin is detectable in the cytoplasmic droplet of the ORP4+/+ sperm, whereas it is in the head of the ORP4−/− sperm. Note that the cytoplasmic droplet is not observed at the corresponding region of the ORP4−/− sperm. Scale bars, 10 μm. (C) MN9 and MN13-immunofluorescence in ORP4+/+ and ORP4−/− mature sperm. The MN9 signal is found in both ORP4+/+ and ORP4−/− sperm. Note that the MN13 signal is normally found in ORP4+/+ sperm, but only a trace amount is found in ORP4−/− sperm. Also note that the tail is bent toward the head region in many ORP4−/− sperm as shown in Figure 5C (arrow; direction of sperm tail). MN9 (Equatorin, an acrosome indicator); red. MN13 (a perinuclear theca indicator at the postacrosomal region); green. Nucleus; blue. Scale bar, 1 mm.

MN13 antigen, an oocyte activation-related protein, is mislocalized in ORP4−/− sperm

To reveal the reason for acrosome detachment from nucleus, we stained spermatozoa with MN9 and MN13, monoclonal antibodies raised against mouse spermatozoa (Toshimori et al. 1991, 1992). MN9 antigen is a sperm-specific type 1 transmembrane protein, named equatorin, which is localized at the equatorial region of the acrosome (Toshimori et al. 1992, 1998; Manandhar & Toshimori 2001; Yoshinaga et al. 2001; Yamatoya et al. 2009; Yoshida et al. 2010). MN13 antigen is localized on the postacrosomal region of the sperm head and is involved in oocyte activation (Toshimori et al. 1992; Manandhar & Toshimori 2003; Ito et al. 2009, 2010). In ORP4−/− spermatozoa, equatorin was normally found at the equatorial region of the acrosome (Fig. 7C, MN9). Equatorin is transported to the equatorial region of the acrosome in early round spermatids (C. Ito and K. Toshimori, unpublished data). Thus, this result suggests that transportation of equatorin to the acrosome is achieved in early round spermatids in ORP4−/− mice. In contrast, the MN13 signal was faint in ORP4−/− spermatozoa, whereas it was normally expressed at the posterior head region in the wild-type spermatozoa (Fig. 7C, MN13). MN13 antigen has been reported to be translocated from the peri-acrosomal region in round spermatids to the postacrosomal region in elongating spermatids (Ito et al. 2010). Therefore, the strikingly reduced MN13 signal on the postacrosomal region suggests that stability and/or transport of MN13 antigen is obstructed in ORP4−/− spermatids during spermiogenesis. Finally, to examine the localization of ORP4 in spermatids when cytoplasmic removal has just been executed, we enzymatically prepared cells from seminiferous tubules and found that ORP4 was largely detectable at the equatorial segment, the region adjacent to the postacrosomal region (Fig. 8A,B).

Figure 8.

ORP4 is largely detectable at the equatorial region. (A) Schematic drawing to show sperm regions. Left, Lateral view. Right, Frontal view. AA, anterior acrosome. ES (PA), equatorial segment (posterior acrosome). Postacrosomal region. (B) Immunofluorescent detection of ORP4 within enzymatically dispersed cells at different stages from seminiferous tubules (green). ORP4 is largely detectable at the equatorial region (ES, arrow).

Discussion

Deficiency of ORP4 causes oligo-astheno-teratozoospermia

In this study, we showed that ORP4 is expressed in postmeiotic germ cells, and lack of ORP4 causes male infertility due to oligo-astheno-teratozoospermia. The cause of severe reduction in sperm count is likely to be extensive apoptosis of postmeiotic spermatids, because we found a significant increase in TUNEL-positive spermatids without any defects in mitosis and meiosis. In contrast, abnormal sperm morphology appears to reflect defects in postmeiotic differentiation of germ cells, because deformed postmeiotic spermatids were often observed in the seminiferous tubules of ORP4−/− mice. Moreover, low sperm motility seems to be the result of abnormal mitochondrial distribution of ORP4−/− sperms.

ORP4 is required for attachment of acrosome to nucleus during spermiogenesis

In this study, we found that MN13 antigen is not localized on the postacrosomal region of the head of ORP4−/− spermatozoa. MN13 antigen has been reported to translocate from the peri-acrosomal region in round spermatids to postacrosomal region (PAR) in elongating spermatids (Ito et al. 2010). To achieve the MN13 translocation, the acrosome should be attached to the nucleus through contact with the perinuclear theca (PT; Oko & Maravei 1994; Ito et al. 2009; Toshimori 2009). At the ultrastructural level, the acrosome was frequently detached from the nucleus in ORP4−/− elongated spermatids (Fig. 5B, arrows). These observations suggest that the formation of the PT is affected in ORP4−/− spermatids. Why is the compartment around PT more vulnerable than other intracellular compartments of spermatids to the deficiency of ORP4? Comprehensive examination on sterol levels of each compartment in murine spermatids has been conducted (Toshimori et al. 1985). According to the study, although PAR (the basal part of PT and the destination of MN13) is almost free from sterol, the equatorial segment (the apical part of PT and the corridor of MN13 for its translocation) is the richest in sterol. It is noteworthy that ORP4 is basically cytosolic but largely detectable at the equatorial segment after cytoplasmic removal has been executed (Fig. 8B). Among several compartments, the equatorial segment is specifically rich in vimentin intermediate filament whose role in sterol transport is well described (Friend 1989; Evans 1994). ORP4 has a cholesterol binding activity and is considered to have an ability to use this highly dynamic intermediate filament as a scaffold or track for transport of cholesterol or regulatory sterols between organelle (Wyles et al. 2007). As vimentin filaments contact the Golgi apparatus and endosomal/lysosomes via associated proteins (Gao & Sztul 2001; Styers et al. 2004), and possibly the ER via VAP–ORP4 interaction, lack of ORP4 may further provoke broad abnormalities.

Possible function of ORP4, a member of the ORP subfamily I, during spermiogenesis

The mammalian ORP family is classified into six subfamilies based on gene organization and amino acid homology (Fig. S2 in Supporting Information). ORP4 and its closest homologue, OSBP, fall into the ORP subfamily I. The ORP subfamily I is evolutionarily conserved in various species including human (ORP4 and OSBP), xenopus (ORP4 and OSBP), zebrafish (ORP4 and OSBP), D. melanogaster (CG6708) and C. elegans (obr-1). In addition to an ORP signature motif ‘EQVSHHPP’, the members in the ORP subfamily I possess their unique sequences in the sterol-binding domain (Fig. S3 in Supporting Information). So far, OSBP has been shown to bind cholesterol through the sterol-binding domain. OSBP also interacts with PI4P in the Golgi through the PH domain (Lagace et al. 1997; Levine & Munro 2002) and with VAP in the ER through the FFAT motif (Wyles et al. 2002; Loewen et al. 2003). Furthermore, OSBP has been reported to transfer cholesterol in vitro and maintain the Golgi cholesterol level at the cellular level (Banerji et al. 2010), suggesting that OSBP functions to carry cholesterol to the Golgi. Very recently, we showed that depletion of OSBP by siRNA caused the mislocalization of Golgi v-SNAREs GS28 and GS15 throughout the cytoplasm without affecting the perinuclear localization of Golgi t-SNARE syntaxin5 and reduced the level of a Golgi enzyme, mannosidase II (Nishimura et al. 2013). These abnormalities are rescued by the expression of siRNA-resistant OSBP, but not by ORP4, suggesting no redundant roles between OSBP and ORP4. As OSBP, ORP4 binds cholesterol (Moreira et al. 2001; Wyles et al. 2007) and interacts with VAP through the FFAT motif (Wyles & Ridgway 2004). We measured the cholesterol level in the testis and did not see the significant difference between wild-type and knockout mice. We could not measure the cholesterol level in purified sperm because of low abundance in ORP4 knockout mice. Given that ORP4 plays a role in transporting cholesterol in intracellular organelles such as OSBP, ORP4 knockdown may change the subcellular localization of cholesterol, but may not affect overall cholesterol level in cells.

Recently, mutants of CG6708, the sole homologue of ORP4 and OSBP in D. melanogaster, were reported to show male sterility due to defects in ‘germ cell individualization’ (Ma et al. 2010). In germ cell individualization, a process corresponding to the cytoplasmic removal of spermatids in mammals, 64 spermatids in a cyst are separated from each other as they discard their cytoplasm and organelles. In mutants of CG6708, most spermatids are enclosed within one membrane as a consequence of the defect in the cytoplasmic removal. Furthermore, CG6708 mutants show abnormal sterol distribution in a cyst. Notably, the sterility of CG6708 mutants is rescued by supplementation with excess cholesterol or its metabolic intermediate 7-dehydrocholesterol. Given the similarity of the structure and phenotype between mouse ORP4 and D. melanogaster CG6708, aberrant cytoplasmic removal in ORP4−/− mice may be caused by abnormal sterol distribution in spermatids.

Experimental procedures

Establishment of monoclonal antibodies against mouse ORP4

A recombinant mouse ORP4 protein (1–180 amino acids) that was expressed and purified by an Escherichia coli pET expression system (Novagen, Madison, WI, USA) was injected into the hind foot pads of WKY rats (SLC, Japan) using Freund's complete adjuvant. The enlarged medial iliac lymph nodes were used for cell fusion with mouse myeloma PAI cells. In the present investigation, PU17 and PU18 were used for Western blotting and immunohistochemistry at 1 : 500 and 1 : 50 dilutions, respectively.

Western blotting

Murine tissues were homogenized in quadruple volumes (w/v) of SET buffer (containing 0.25 m sucrose, 1 mm EDTA, 10 mm Tris–HCl pH 7.4) with protease inhibitors (0.5 mm phenylmethylsulfonyl fluoride, 2 μg/mL pepstatin, 2 μg/mL leupeptin, 2 μg/mL aprotinin). After centrifugation at 1000 g at 4 °C, the supernatants were used as the total protein extracts. The protein concentrations of samples were determined by the BCA assay (Pierce, Rockford, IL, USA). Each total protein extract was separated by SDS-PAGE and transferred to PVDF membranes. The membranes were blocked with 5% (w/v) skim milk in TTBS buffer (10 mm Tris–HCl, pH 7.4, 150 mm NaCl, 0.05% (w/v) Tween-20) and incubated with the specific monoclonal anti-mouse ORP4 antibody PU17. After incubation with horseradish peroxidase-conjugated anti-rat IgG antibody (GE Healthcare, Waukesha, Wisconsin, USA), ORP4 was detected by enhanced chemiluminescence.

In situ hybridization

A 444-bp DNA fragment corresponding to the nucleotide positions 109–552 of mouse ORP4 (GenBank accession number NM_152818) was subcloned into pGEMT-Easy vector (Promega) and was used for generation of sense or antisense RNA probes. Paraffin-embedded testicular sections (6 μm) of mouse (C57BL/6 mouse, male, 8 weeks; Sankyo Labo Service) were obtained from Genostaff Co., Ltd. For in situ hybridization, the sections were hybridized with digoxigenin-labeled RNA probes at 60 °C for 16 h. The bound label was detected using NBT-BCIP, an alkaline phosphate color substrate. The sections were counterstained with Kernechtrot (Muto Pure Chemicals Co., Ltd).

Fractionation of the testicular cells

Isolation of Sertoli cells and germ cells from the decapsulated testes was carried out essentially as described previously (Scarpino et al. 1998), except that we used RPMI1640 medium containing 1 mg/mL collagenase IV and 0.5 mg/mL Dispase (Roche) to remove interstitial Leydig cells. To digest the seminiferous tubules into Sertoli cells and germ cells, we used 2.5 mg/mL Trypsin solution (Invitrogen) containing 1 μg/mL DNase.

Quantitative real-time PCR

Total RNA was isolated using ISOGEN (Nippongene, Toyama, Japan) according to the manufacturer's protocol. Complementary DNA was prepared from total RNA using the High Capacity cDNA Reverse Transcription Kit (Applied Biosystems, Foster City, CA, USA). Quantitative PCR containing SYBR-Green fluorescent dye (ABI) was carried out in 35 cycles (LightCycler, Roche). Relative mRNA levels were enumerated by the delta-delta Ct method with beta-actin as a housekeeping gene. Analysis of melting curves confirmed product quality after each PCR. The primers used were as follows: AAT GCC TGT GGC TTG TAT C (gata-1 Forward), AGA GTG TTG TAG TGG TCG TTT G (gata-1 Reverse); TGC CTG GTA TTC CCA AGA TTG (crem Forward), CGG CTG ATT GTG CTG CAT A (crem Reverse); GCT TCT TTG CAG CTC CTT CGT (beta-actin Forward), ATG CCG GAG CCG TTG TC (beta-actin Reverse).

Generation of ORP4 knockout mice

Mouse ORP4 genomic DNA sequences were cloned from a C57BL/6-derived genomic DNA and subcloned into the pPNT vector. The targeting vector contains a PGK-neo cassette with 5.3 kb of ORP4 homologous regions upstream of exon 4 and with 5.7 kb homologous regions downstream of exon 8. The targeting vector was linearized and electroporated into RENKA ES cell lines (TransGenic Inc., Kumamoto, Japan) derived from the C57BL/6 mouse substrain. Stable clones were selected for G418 resistance. Of 282 G418-resistant clones, four clones were correctly targeted, as confirmed homologous recombination by Southern blot analysis with 5′, 3′ and neo probes. Blastocysts were obtained by aggregation methods of ICR morula embryos and three different ES cells, to recipient uterus as after 3 days of pseudopregnancy. Chimeric males were mated to C57BL/6 females, and offspring animals (F1) were genotyped by Southern blot analysis. F2 homozygous mutant mice were then generated by intercrossing F1 heterozygous males and females. Mice were genotyped by PCR analyses of genomic DNA isolated from tail biopsies.

Fertility assessment

A 12-week-old male was mated with two 12-week-old females and housed together for 30 days (N = 8–9/genotype). Females were scored for the presence of copulation plugs and pregnancy.

Testes analyses

Hematoxylin and eosin staining: Mice under anesthesia were perfused with phosphate-buffered saline. Testes and epididymides were removed and left in Bouin's fixative at 4 °C for 48 h. Paraffin sections (5 μm) were prepared and stained with hematoxylin and eosin.

Conventional transmission electron microscopy: Adult testes and cauda epididymides were fixed with 2.5% glutaraldehyde in a 30 mm HEPES (2−[4 + (2-hydroxyethyl)−1piperazinyl] ethanesulfonic acid; Nacalai Tesque, Inc., Kyoto, Japan) buffer by perfusion through the left ventricle and then immersed in the same fixative overnight at 4 °C. The tissues were cut into small pieces, fixed with 2% osmium tetroxide, dehydrated through a graded ethanol series and embedded in Epon 812 (TAAB Laboratories Equipment, Berks, UK). One-micrometer-thick sections were made for 1% toluidine blue (Wajdeck GmbH & Co, Munster, Germany) staining using an ultramicrotome (Ultracut E; Reichert-Jung, Vienna, Austria). Then, ultrathin sections were made and stained with uranyl acetate and lead citrate. The sections stained with toluidine blue were observed with an Olympus BX50 light microscope and analyzed using images captured by a CCD camera CoolSNAP (PhotoMetrics, Huntington Beach, CA, USA). The ultrathin sections were observed using a JEOL JEM-1200 EX TEM (JEOL, Tokyo, Japan).

Immunohistochemistry: Fixed testes were cryoprotected in a series of sucrose solutions (20–30%) in PBS at 4 °C and then embedded in Tissue-Tek OCT compound. Cryo-sections (5 μm) were blocked with 3% bovine serum albumin in PBS with 0.05% Tween-20 for 30 min, incubated with an ORP4 antibody PU18 (see above) overnight at 4 °C, washed four times with PBS with 0.05% Tween-20 and incubated with Alexa-488 secondary antibody (Molecular Probes) for 1 h at room temperature. DAPI was used for nuclear staining. Detection of proliferating cells: BrdU (Sigma) dissolved in PBS was injected intraperitoneally at 50 mg/kg of body weight. After 24 h, males were dissected and testes were fixed in Bouin's fluid for 24 h. Subsequent immunostaining was carried out as described above. An anti-BrdU monoclonal antibody (Abcam) was used at 1 : 200 dilution as a primary antibody. FACS analysis: The ploidy determination of gem cells was estimated by flow cytometry DNA analysis as previously described (Imai et al. 2009).

Sperm analyses

Motility and fertilization capacity of sperm was evaluated as previously described (Sato et al. 2010). As for mature sperm staining, epididymides were dissected and put into RPMI1640 containing fetal bovine serum. An aliquot of 6 mL sperm suspension was attached to poly-L-lysine-coated slides and incubated at 37 °C for 2 h. Sperms were fixed in 3.7% formaldehyde and permeabilized in PBS containing 0.5% Triton X-100. After 10-min permeabilization at room temperature, coverslips were blocked with 3% bovine serum albumin in PBS for 30 min and then incubated overnight at 4 °C with antiubiquitin monoclonal antibody (Santa Cruz) at 1 : 200 dilution. Mitochondrial sheaths were visualized using the mitochondrion-specific vital dye MitoTracker Red (Molecular Probes) as previously described (Kang-Decker et al. 2001). As for indirect immunofluorescence with MN13 or MN9 antibodies, mature sperm were treated with primary antibodies MN9 (1/20 000) and/or MN13 (1/10 000 dilution) of the stock solution, at a final concentration 0.02 μg/mL and 0.1 μg/mL, respectively. Then, the samples were treated with goat fluorescence-labeled secondary anti-mouse IgG or IgM (Invitrogen). The samples were visualized with an Olympus BX50 epifluorescence microscope (Olympus Corporation, Tokyo, Japan) in which a UPlanApo 100× NA 1.35 oil objective lens was equipped with an imaging system composed of the appropriate filters for fluorescence and a Charge Coupled Devices (CCD) camera RETIGA Exi FAST 1394 (Qimaging, Surrey, BC, Canada). Acquisition and storage of the data were controlled by SlideBook 4 software (Intelligent Imaging Innovations, Denver, CO, USA).

Intracytoplasmic sperm injection (ICSI)

Microinsemination: Elongated spermatids or testicular spermatozoa were collected from ORP4−/− testis and injected into the artificially ovulated oocytes from B6D2F1 wild (ORP4+/+) ovaries using a Piezo-driven micromanipulator as described before (Ogonuki et al. 2010). Fertilized oocytes were cultured for 24 h, and two-cell embryos were transferred into the oviducts of pseudopregnant ICR females. Pups were produced naturally or by caesarian procedure on day 19.5 and raised by lactating foster ICR dams. Data were examined by Fisher's exact test for the duplicates.

Acknowledgements

We thank Ms K. Kamimura for the help for electron microscopy. We also thank Dr H. Imai (Kitasato University) for technical advice in flow cytometric analysis. We are grateful to Ms H. Fukuda-Hirata for technical assistance and Dr T. Taguchi for critical reading of the manuscript and providing helpful comments.

Ancillary