| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Department of Anatomy and Developmental Biology,3 Graduate School of Medicine, Chiba University, Chiba 260-8670, Japan
Department of Cell Biology,4 Cancer Institute, Japanese Foundation for Cancer Research, Tokyo 170-8455, Japan
Laboratory of Zoology,5 Graduate School of Agriculture, Kyushu University, Fukuoka 812-8581, Japan
ABSTRACT
Gopc (Golgi-associated PDZ- and coiled-coil motif-containing protein)–/– mice are infertile, showing globozoospermia, coiled tails, and a stratified mitochondrial sheath. Transmission electron microscope (TEM) images of the spermatozoa were studied quantitatively to analyze disorganization processes during epididymal passage. Factors maintaining straight tail and normal mitochondrial sheath were also studied by TEM and immunofluorescent microscopy. Sperm tails retained a normal appearance in the proximal caput epididymidis. Tail disorganization started between the proximal and the middle caput epididymidis, and the latter is the major site for it. The tail moved up through the defective posterior ring and coiled around the nucleus to various degrees. Tail coiling occurred in the caput epididymidis suggesting it was triggered by cytoplasmic droplet migration. SPATA19/spergen-1, a candidate mitochondrial adhesion protein, remained on the stratified mitochondria, while GPX4/PHGPx, a major element of the mitochondrial capsule, was unevenly distributed on them. From these findings, we speculate GPX4 is necessary to maintain normal sheath structure, and SPATA19 prevents dispersal of mitochondria, resulting in a stratified mitochondrial sheath formation in Gopc–/– spermatozoa. The epididymal epithelium was normal in structure and LRP8/apoER2 expression suggesting that tail abnormality is due to intrinsic sperm factors. Three cell structures are discussed as requisite factors for maintaining a straight tail during epididymal maturation: 1) a complete posterior ring to prevent invasion of the tail into the head compartment, 2) stable attachment of the connecting piece to the implantation fossa, and 3) a normal mitochondrial sheath supported by SPATA19 and supplied with sufficient and normally distributed GPX4.
epididymis, sperm maturation, teratozoospermia
During spermiogenesis, the spermatozoa eliminate most of their cytoplasm, and the organelles become closely packed. Attachments between structural components must be significant to maintain cell integrity in the uniformly designed and vigorously motile spermatozoa. The final form of mature spermatozoa is established during passage though the epididymis, where they undergo functional and morphological modifications, called "epididymal maturation" (for review [1]). The most common morphological change during maturation is the migration of the cytoplasmic droplet from the neck region to the distal end of the midpiece [2]. This can be considered as the final elimination of excess cytoplasm. Although the epididymal region for droplet migration varies from species to species, it generally occurs in the caput epididymidis [3].
It is said that approximately one in 10 couples seek medical help for the problem of subfertility. About 60% of subfertility in men is attributed to genetic factors [4], and the elucidation of the underlying gene defects with mouse models is an area of intensive research [5–7]. Among them, Gopc deficient mice are expected as a good model to scrutinize globozoospermia [6, 8, 9], since similar round-headed spermatozoa have been reported in infertile patients [10–13].
GOPC predominantly localizes at the trans-Golgi region in round spermatids [8], and contains one PDZ domain, two coiled coil motifs, and two evolutionarily conserved domains. The PDZ domain is required for binding to the membrane protein, Frizzled, while the coiled-coil motifs and conserved domains are required for its localization to the Golgi [14]. More recently, a specific interaction of GOPC with GOLGA3/golgin-160, a Golgi-localized protein, has been reported [15]. In addition, FIG (fused in glioblastoma), a human homologue of mouse GOPC with 92.3% identity, has been shown to interact with a fusion protein, STX6/syntaxin-6 [16]. These evidences support the idea that GOPC plays a role in vesicle transport from the Golgi apparatus.
Our previous study showed that the primary morphological defect in Gopc–/– spermatozoa was a failure of Golgi-derived proacrosomal vesicles to fuse at the round spermatid stage [17]. In Gopc–/– spermatozoa, the absence of the acrosome triggered malformation of head elements such as the nucleus, the manchette, the postacrosomal sheath, and the posterior ring, which normally develop sequentially [17]. Despite the various malformations of the head components, tail elements, such as the axoneme, the outer dense fibers (ODFs), the fibrous sheath, and the mitochondrial sheath appeared to be organized normally during spermiogenesis [17, 18]. A straight tail was maintained until the spermatozoa reached the proximal caput epididymidis. However, most of the spermatozoa taken from the cauda epididymidis had a tail that was coiled around the nucleus. The mitochondria of such spermatozoa generally gathered at the proximal part of the midpiece, and formed a "stratified mitochondrial sheath." As a consequence, the ODFs of the distal midpiece were denuded from the mitochondrial sheath [18]. In the present study, we were interested in examining four factors in this unique material: 1) determining the epididymal regions, where tail disorganization occurs, 2) the mechanism of tail coiling around the nucleus, 3) the process of mitochondrial sheath disorganization, and 4) the molecules maintaining a normal mitochondrial sheath. From our results, we deduced the factors involved in maintaining a straight tail and a single layered mitochondrial sheath during epididymal passage.
In the literature, tail abnormalities have been reported to start either during spermiogenesis or during epididymal maturation. In another mouse model of globozoospermia, Hrb–/– spermatozoa showed a similar phenotype to that of Gopc–/– spermatozoa including coiling of the tail around the nucleus [19, 20]. However, in Hrb–/– spermatids, the mitochondria did not organize a normal sheath and tail coiling started during spermiogenesis [20]. Another tail abnormality originating in the testis, axonemal dysgenesis, has been reported in mice with a t complex mutation [21] or with insertional mutation of the Theg (testicular haploid expressed gene) [22]. A similar phenotype has also been reported in a human infertile patient [23]. In addition, epididymis originated tail abnormality, angulation, or hairpin curve formation at the cytoplasmic droplet has been reported in a boar [24] and in transgenic mice [25, 26]. In Ros1–/– mice, the initial segment of the epididymis failed to develop [25], and the sperm tail angulated during passage through the epididymis [26]. In this case, tail abnormality was attributed to epididymal dysfunction rather than sperm themselves.
Three molecules, SPATA19 (spermatogenesis associated 19), GPX4 (glutathione peroxidase 4), and LRP8 (low density lipoprotein receptor-related protein 8), were of interest in relation to the maintenance of the normal mitochondrial sheath. SPATA19 is a strong candidate to act as a binding molecule between adjacent mitochondria in the sheath [27] because it contains a mitochondria-targeting signal and aggregates mitochondria when artificially expressed in cells [28]. A selenoprotein, GPX4, is known to be a multi-functional protein [29], which is believed to act as a peroxidase in early-stages spermatids. In spermatozoa, GPX4 is the structural protein that makes up to more than 50% of the mitochondrial capsule, and stabilizes the mitochondrial sheath [30]. Spatiotemporal changes in the distribution of GPX4 in subcellular compartments of the spermatogenic cells have been reported in the rat [31]. Within mitochondria, GPX4 moved from the matrix to the outermost region of the membranes in step 19 spermatids [31]. In relation to GPX4, LRP8, a receptor of RELN/reelin and LDL (low-density lipoprotein), is also of interest. It is highly expressed in the epididymal principal cells [32], and Lrp8–/– male mice are infertile because they have reduced GPX4 levels resulting in angular tails [33]. From TEM observations, the mitochondria of Lrp8–/– spermatozoa were reported to be highly irregular in size and shape [33]. In the present work, immunocytochemistry was used to show the presence of SPATA19 and GPX4 on the disorganized mitochondrial sheath of Gopc–/– spermatozoa, and this result is discussed in relation to the maintenance of a normal sheath structure. Expression of LRP8 was also examined to determine whether the tail abnormality was caused by epididymal disfunction or by an intrinsic sperm factor.
Gopc mutant mice were supplied by the Department of Cell Biology, Japanese Foundation for Cancer Research (JFCR), Cancer Institute [8]. Animal handling was performed in accordance with the guideline for the care and use of laboratory animals of Chiba University and with the approval of the Animal Research Committees of the JFCR Cancer Institute and Chiba University.
Four Gopc–/– and 2 Gopc+/+ mice were used 10 weeks after birth. They were anesthetized with Nembutal (Abbott Laboratories, Abbott Park, IL) or diethyl ether, and the epididymides were initially fixed with 2.5% Hepes buffered glutaraldehyde by perfusion through the left ventricle. The organs were removed, and small blocks of tissue were collected from five different regions of the epididymis (Fig. 1, a–e): proximal caput (a), middle caput (b), distal caput (c), corpus (d), and cauda epididymidis (e). The initial segment (Fig. 1I) of the caput epididymidis was omitted from sampling because of the paucity of luminal spermatozoa. Each tissue block was cut into small pieces and immersed in the same fixative for 2 h, rinsed in buffer, and then post-fixed with OsO4. Thereafter, the samples were dehydrated through a graded ethanol series and embedded in Epon 812. Ultrathin sections were cut on an ultramicrotome (Ultracut FC 4E; Reichert-Jung, Wien, Austria), picked up on #300 copper mesh, and stained with uranyl acetate and lead citrate. Observation was performed with a TEM (1200EX; JEOL Inc., Tokyo, Japan).
|
To scrutinize the tail disorganization of Gopc–/– spermatozoa in relation to the different epididymal regions, we tried to quantify the following three items from sectioned TEM images—I: the degree of tail coiling, II: the rate of appearance of the most distal segment of the tail, III: the rate of appearance of the stratified mitochondrial sheath and bare ODFs. Since sectioning images were used, the data obtained were indexed to show a tendency. In order to obtain reliable data, more than 100 cells were examined for each epididymal region of each mouse.
Degree of tail coiling: Counting method of the sectioned tail components in the perinuclear cytoplasm. TEM images of Gopc–/– spermatozoa were recorded on film at a direct magnification of 3000x using a montage technique whenever possible. These images at a final magnification of 4500x were used as the sample to generate the basic data. Four practical counting conditions were followed. First, only spermatozoa cut through the central portion of the nucleus were used. Second, the number of tail-components was only counted in the perinuclear cytoplasm. Third, the number of innermost tail components that appeared on the cut surface was counted, with the exception of the axoneme, which was too slim to count precisely at the magnification used. Thus the following categories of tail structures were counted (Fig. 2): 1) the mitochondrial sheath, only when ODFs did not appeared on the cut surface; 2) a set, or a part of a set, of ODFs of the midpiece (if the mitochondrial sheath and ODFs appeared together, only the number of ODFs was counted); and 3) the fibrous sheath of the principal piece. Fourth, if the tail components were cut longitudinally, the innermost structure was counted as one as long as it appeared continuously (Fig. 2). The number of tail sections inevitably varied with the direction of sectioning as shown in Figure 2.
|
Rate of appearance of the most distal piece of the tail. As another indicator of the degree of coiling, the most distal piece of the tail that was rolled up into the perinuclear cytoplasm was recorded as midpiece, principal piece, or absent.
The rate of appearance of the stratified mitochondrial sheath and bare ODFs. The presence or absence of the stratified mitochondrial sheath and bare ODFs irrespective of their position in or out of the perinuclear cytoplasm, was recorded separately in each spermatozoon.
Details of the method of calculation for each data set are explained in the figure legend for each graph. To indicate individual variations in each corresponding region, the mean values and standard deviations (mean ± SD) for measurements from four animals were calculated. The statistical significance of differences for these values in different regions was assessed using Welch t-test, which is a test for two parametric and unpaired groups with different variance.
Antibodies and Fluorescent Probes
The following primary and secondary antibodies and fluorescent probes were used: Anti-SPATA19: a rabbit antibody was raised against an 11 amino acid sequence at the hydrophilic region of SPATA19 (NKTEEEASRSI) [28]; Anti-GPX4: rabbit antibody to the 15 amino acid sequence at the C-terminus of human GPX4 was given by Dr. S. Yokota, Biology Laboratory, University of Yamanashi, Faculty of Medicine, Japan [31]; and Anti-LRP8: Goat anti-human LRP8 antibody (catalog number: SC-10113, Santa Cruz Biotechnology Inc, CA). The secondary antibodies: Alexa Fluor 546 goat anti-rabbit IgG (Molecular Probes Inc, OR) and biotin-conjugated rabbit anti-goat immunoglobulins (Dako Japan Inc., Tokyo). Probe for mitochondria: MitoTracker Green (Molecular Probes Inc, OR). Probe for nucleus: Hoechst 33258 (Sigma Inc., MO).
Immunofluorescent Labeling of Spermatozoa
Spermatozoa were collected from the cauda epididymidis of adult Gopc–/– or Gopc+/+ mice, and stored in CZB [34] containing 10% of glycerin at –20°C until use. Thawed sperm were mounted on glass slides, dried, and incubated with anti-SPATA19 (1:100 dilution) or anti-GPX4 (1:250 dilution) rabbit polyclonal antibodies in PBS containing 5% of normal goat serum and 3% of BSA at 4°C overnight and were rinsed in PBS. They were then incubated with the secondary antibody, Alexa Fluor 546 conjugated goat anti-rabbit IgG (0.5 µg/ml), for 1 h and counterstained with Hoechst 33258 (10 µg/ml) and MitoTracker Green (100 nM) for nucleus and mitochondria, respectively. After washing with PBS, they were evaluated with Olympus BX50 (Olympus Ltd., Tokyo, Japan) equipped with an imaging system composing of appropriate filters for fluorescence and a Charge Coupled Device (CCD) camera RETIGA Exi FAST 1394 (Qimaging Alliance, BC, Canada). Acquisition and storage of the data were controlled by a Macintosh computer (G5) running SlideBook 4 (Nippon Roper, Tokyo, Japan). All figures were prepared offline from the images using an Adobe Photoshop (version 7.0) in Macintosh computers.
Immunohistochemistry Using Streptavidin-Biotin Complex (S-ABC)
Adult Gopc+/– and Gopc–/– mice were anesthetized with Nembutal (Abbott Laboratories) or diethyl ether and fixed with 4% paraformaldehyde by perfusion through the left ventricle. Testes and epididymides were removed and immersed in the same fixative for 18 h. The organs were then processed for paraffin embedding and cut at a thickness of 3 µm. The sections were autoclaved in a 0.01 M citrate buffer at 120°C for 5 min to activate the antigen. After treatment with 0.1% Triton X-100 for 30 min, endogenous peroxidase activity was suppressed with 0.3% H2O2 in methanol for 30 min. Nonspecific binding of the antibody was suppressed by incubation in PBS containing 5% fetal bovine serum for 30 min at room temperature. The sections were then incubated with goat anti-LRP8 antibody at a 1:40 dilution at 4°C overnight and rinsed in PBS. The samples were then treated with biotin-conjugated rabbit anti-goat immunoglobulins (1:100 dilution) for 1 h followed by incubation in a streptavidin-HRP (horse radish peroxidase) solution (DAKO LSAB 2 System, HRP DAKO, Dako Japan Inc., Tokyo) for 30 min at room temperature. Immunohistochemical reactions were visualized using 3,3'-diaminobenzidine (DAB) and H2O2.
Immature spermatozoa from the proximal caput epididymidis. Gopc–/– spermatozoa had a round nucleus, and the acrosome was absent in most cases. The posterior ring was either absent or defectively formed and the nucleus was often located in the cytoplasmic droplet (Fig. 3A). In this epididymal region, the tail was straight and the mitochondrial sheath was monolayered. In higher magnification images of the mitochondrial sheath, structural differences were noted between Gopc+/+ and Gopc–/– spermatozoa (Fig. 3, B and C). In Gopc+/+ spermatozoa, the mitochondria characteristically had a uniform size and an electron-dense matrix with narrow intracristal space. The mitochondria and ODFs were regularly spaced, and the outer mitochondrial membrane was clearly visible (Fig. 3B, arrows). In contrast, Gopc–/– mitochondria were swollen to varying degrees, and the matrix appeared paler and the number of cristae was reduced. The mitochondria and ODFs were irregularly spaced, and the outer mitochondrial membrane could be recognized only in limited areas (Fig. 3C, arrows). The degree of deviation from the normal structure varied in each mitochondrion.
|
Tail coiling. Gopc–/– spermatozoa from the middle caput through to the cauda epididymidis showed abnormal morphology of both head and tail regions (Fig. 4, A–F), the most common abnormality being coiling of the tail around the nucleus. The coiling started as bending of the nucleus toward the midpiece and dislocation of the connecting piece from the implantation fossa (Figs. 2 and 4B). In the next step, the nucleus appeared to roll down the midpiece (Fig. 4C). The plasma membrane covering the tail progressively expanded and joined to the membrane covering the perinuclear cytoplasm (Fig. 4C). Separation of the connecting piece from the nucleus was observed in later stages. The degree of coiling tended to be low in the middle caput epididymidis (Fig. 4, A and B), and increased progressively from the distal caput through the cauda epididymidis (Fig. 4, D and F). In the cauda epididymidis, the final degree of tail coiling varied for each spermatozoon. Coiling often stopped at the annulus (Fig. 4D). In such spermatozoa, only the midpiece was seen in the perinuclear cytoplasm (Fig. 4, D and E). In spermatozoa showing the greatest degree of coiling, the whole length of the tail including the principal piece was rolled up (Fig. 4F). This image corresponds to the light microscopic image of a whole mounted spermatozoon shown in Figure 4F (inset). Only 2% of cauda epidiymal spermatozoa showed this image in Gopc–/– mouse.
|
Appearance of a stratified mitochondrial sheath and of bare ODFs. The other common abnormality was the formation of a stratified mitochondrial sheath (Figs. 2 and 4A–D) which resulted in baring of the ODFs at the midpiece (Fig. 4, B and D). This was first observed in the middle region of the caput epididymidis (Fig. 4, A and B). In such spermatozoa, many mitochondria detached from the ODFs but kept their contact with each other. It brought about stratification and shortening of the mitochondrial sheath. As a consequence, the mitochondrial sheath shifted proximally and the distal ODFs lost their mitochondrial sheath. Such ODFs are referred to as "bare ODFs" throughout this paper. Bare ODFs are either in direct contact with the plasma membrane where they project from the perinuclear cytoplasm (Fig. 4B inset) or are exposed to cytoplasm around the nucleus (Fig. 4D, double arrows). Such spermatozoa had either straight (Fig. 4A) or coiled tail (Fig. 4, C, D and F). In the lower left spermatozoon of Figure 2, the mitochondria retained a single layered arrangement, but were widely separated from the sharply curved ODFs (short arrows). These observations suggested that tail coiling facilitated separation of the mitochondria from the ODFs.
A summary of the tail coiling processes and the formation of a stratified mitochondrial sheath in Gopc–/– spermatozoa during epididymal passage is provided in diagram form in Figure 5.
|
Details of data collection are given in Materials and Methods and in each figure legend. In each subject, individual variation is indicated by a standard deviation (SD), which tended to be small in the proximal and the middle caput epididymidis, and large in the corpus and cauda regions. It indicated that individual variation became large in the distal epididymis. By Welch's t-test, highly significant (P < 0.01) or significant (P < 0.05) differences were found only when the data from all regions were compared with those of the proximal and/or the middle caput epididymidis. Differences from the proximal caput groups that were detected are indicated with asterisk(s) (**P < 0.01; or *P < 0.05) above the columns in each graph (Figs. 6–8). Similarly, significant differences from the middle caput groups are indicated with plus signs (++P < 0.01; or +P < 0.05) (Figs. 6–8). The statistical results indicated that the tail deformation starts between the proximal and the middle caput segments, and the middle caput epididymidis was the major region of tail disorganization irrespective of individuals.
|
|
Three Numerical Indexes of Tail Coiling
Mean number of tail sections. As the simplest indicator of tail coiling, the mean number of tail sections per sperm section with standard deviation (M ± SD) is shown in Figure 6 (Top). The result indicated that the number progressively increased until distal caput epididymidis irrespective of individuals.
Percentage of more than two tail sections as an indicator of tail coiling. If the tail is straight, as in spermatozoa from the proximal caput epididymidis (Fig. 3A), the number of tail sections in the perinuclear cytoplasm is 0 or 1. If two or more tail sections are present, this indicates that the tail is not straight but coils around the nucleus.
Thus, the percentage of spermatozoa showing more than two tail sections was plotted on Figure 6, Middle. The number was low (0.7 ± 0.6%) in the proximal caput, increased significantly in the middle caput, and continued to increase significantly in the distal caput epididymidis. This result indicated that tail coiling started between the proximal and the middle caput epididymidis, and continued until the distal caput epididymidis irrespective of individuals. Thus, the major site of tail coiling was the middle caput epididymidis.
Percentage of more than three tail sections as an indicator of tail encircling. If there are more than three tail sections per sperm section, this indicates the tail coils around the nucleus more than once in a progressive form of tail coiling (Fig. 4, E and F). Such spermatozoa were absent from the proximal caput epididymidis (Fig. 6, Bottom). A statistically significant increase in the number of such highly coiled tails was found only when data from the proximal caput were compared with those from the middle and distal caput groups. These results indicated that encirclement of the nucleus also started between the proximal and the middle caput as a progression of tail coiling, and that the final degree of encirclement varies in each mouse.
Rate of appearance of the most distal piece of tail in the perinuclear cytoplasm. The most distal piece of tail sections was classified into three groups, absent, midpiece, and principal piece, to show the extent to which the tail pieces rolled up into the perinuclear cytoplasm. Results were plotted separately in Figure 7. It was noteworthy that group 2, midpiece tail sections, appeared with the highest frequency in all areas except the proximal caput epididymidis, where group 1, "absent", predominated (Fig. 7, Top). There was a significant increase in the number of midpiece tail sections appearing in the middle and the distal caput epididymidis in comparison with the adjacent proximal epididymal regions (Fig. 7, Middle). In contrast, sections containing the principal piece did not show a significant increase until the distal caput epididymis (Fig. 7, Bottom). These results provide a quantitative support for the presence of a barrier to tail coiling between the middle and the principal pieces.
|
Rate of appearance of stratified mitochondrial sheath and bare ODFs. Both stratified mitochondrial sheaths and bare ODFs were virtually absent from the proximal caput (Fig. 8). The number of both structures increased significantly between the proximal and the middle caput epididymidis. Spermatozoa with a stratified mitochondrial sheath were seen in approximately half of the Gopc–/– spermatozoa in the distal caput through to the cauda groups. Although the percentage of bare ODFs also increased in the epididymis, it was always lower than that of the stratified mitochondrial sheath. These differences could be explained by the relative sizes of the two structures.
Immunocytochemistry on whole mounted spermatozoa. In our previous paper, the characteristics of mature Gopc–/– spermatozoa from the cauda epididymidis were fully reported at light microscopic level [18]. Here, we only describe new findings using specific antigens to SPATA19 and GPX4, which are interesting because of their roles in relation to the maintenance of the mitochondrial sheath.
SPATA19. In both Gopc+/+ and Gopc–/– spermatozoa, positive staining was localized only in the midpiece. Merged images of SPATA19 labeled with Alexa fluor 546 (red) and mitochondria visualized with MitoTracker Green (green) appeared yellow in both Gopc+/+ (Fig. 9A, Top to Bottom) and Gopc–/– spermatozoa (Fig. 9, B and C, Top to Bottom). Results show that the distribution of SPATA19 corresponded well with that of the mitochondria even after disorganization of the mitochondrial sheath into the stratified form.
|
GPX4. GPX4 immunoreactivity (red) was detected along the midpiece (midochondrial region; green) and on the head (blue) in Gopc+/+ spermatozoa (control; Fig. 10A, Top to Bottom), but only on the mitochondrial region around the nucleus in Gopc–/– spermatozoa (Fig. 10, B and C, Top to Bottom). In addition, in Gopc–/– spermatozoa, GPX4 immunoreactivity on the mitochondria showed various intensity (negative to strongly positive), showing a mixture of green, yellow, and orange colors in the merged images (Fig. 10, B and C, Top to Bottom). The round nucleus was negative in the reaction. This indicates that GPX4 was expressed in Gopc–/– spermatozoa, but was unevenly distributed in the disorganized mitochondrial sheath.
Immunohistochemical localization of LRP8 in the epididymis. Immunohistochemistry localized LRP8 to the Golgi apparatus of the epididymal principal cells both in Gopc+/– and Gopc–/– mice. The initial segment was negative, the proximal caput was positive, and the strongest positive reaction was observed at the middle caput epididymidis (Fig. 11). The strength of the positive reaction gradually decreased toward the corpus epididymidis, and the cauda epididymidis was virtually negative. Based on the immunocytochemical results, Figure 12 shows a schematic drawing of the elements and molecules thought to maintain the normal mitochondrial sheath structure.
|
|
Our present results clearly demonstrate that the tail deformation starts between the proximal and the middle caput epididymidis, and the latter is the major site of tail disorganization irrespective of individuals. Here, we discuss the mechanisms of tail coiling and formation of a stratified mitochondrial sheath in Gopc–/– spermatozoa, and consider the sperm elements and molecules involved in maintaining a normal tail structure.
Morphological Basis and Mechanisms of Tail Coiling
As a morphological basis of tail coiling, two structural abnormalities were noted in the pre-coiling spermatozoa in our previous study [17, 18]: either an absent or a defectively formed posterior ring, and the presence of residual perinuclear cytoplasm (Fig. 3A). The present study adds two more factors—a weak adhesion between the connecting piece and the implantation fossa and an insufficiently stiff mitochondrial sheath, which will be discussed later in relation to sheath disorganization.
The normal posterior ring separates the sperm into head and tail compartments. Formation of a defective posterior ring and the presence of excessive perinuclear cytoplasm were intimately related [17]. In normal spermatids, the cytoplasm of the apical head is compressed by close contact between the plasma membrane and the outer acrosomal membrane [35], while the bulk of cytoplasm moves distally along the manchette [36]. In Gopc–/– spermatids, the absence of the acrosome leads the presence of excessive perinuclear cytoplasm, malformation of the manchette, the absence of the postacrosomal sheath, and a defective posterior ring [17]. As a consequence, the globular nucleus is often present in the cytoplasmic droplet (Fig. 3A). Thus, a deficiency in the posterior ring allows tail penetration into the perinuclear cytoplasm which provides space for the invading tail in the head compartment [18].
Circumstantial evidence suggests that tail coiling is related to the migration of the cytoplasmic droplet from the neck to the distal end of the midpiece, since both phenomena occur in the caput epididymidis. However, the precise mechanism and the motive force of the droplet migration are not known yet. In Gopc–/– spermatozoa, the coiling started as a bending of the nucleus toward the midpiece within the residual cytoplasm (Figs. 2 and 4B). As a result, the nucleus occupied the central portion of the residual cytoplasm and the tail was pushed to the periphery. When the nucleus moves along the midpiece together with the cytoplasm, the tail coiled around the nucleus because the dislocated connecting piece maintained partial contact with the nucleus at least in the initial stage of coiling (Figs. 2 and 4B). In effect, the tail rolls up and winds around the nucleus (Figs. 2 and 4B–D). Further evidence to support this idea is that the rolling often stops at the end of the midpiece (Fig. 4D), where there is the annulus that stops cytoplasmic migration in normal spermatozoa. The statistical analysis on the midpiece (Fig. 7, Middle) and the principal piece (Fig. 7, Bottom) in the perinuclear cytoplasm also supports that the annulus prevents further tail coiling and only when the annulus is broken, the principal piece can wind around the nucleus. There is no doubt that dislocation of the connecting piece from the nucleus accelerates tail coiling, because the sharp bending at the neck increases the curvature of the tail, thus it facilitating tail coiling within the limited volume of the perinuclear cytoplasm. The present study has shown that the dislocation started in the early phase of tail coiling (Figs. 2 and 4B), indicating a weakness in the adhesion between the connecting piece and the implantation fossa of the nucleus and suggesting that the dislocation is one of causes of tail coiling, rather than a result of it.
One question that arises is what is the driving force of tail coiling? Motility might be a candidate, but only about 7% of Gopc–/– spermatozoa are motile, and most of the motile spermatozoa displayed only sluggish movement [8]. Moreover, sperm motility is known to be dormant in the epididymis, and the spermatozoa were fixed in situ by perfusion in the present TEM study. Therefore, tail motility could be denied as the driving force in Gopc–/– spermatozoa, although agonal movement at fixation cannot be ruled out. In general, all liquids tend to keep spherical shape by surface tension, if there is no surrounding force. This notion is also applicable to cell shape. Spermatids are supported by the surrounding Sertoli cells in the testis. While in the epididymis, spermatozoa are in the luminal fluid. Under this condition, retraction force acting on the protruded tail could be a driving force for the tail coiling in the Gopc–/–spermatozoa, since they have the spherical mass of residual cytoplasm around the nucleus, the defective posterior ring, and the weak neck.
Mochida et al. [37] studied tail-coiling process in the Hook1/Azh–/– spermatozoa, whose tail showed fundamentally similar final form to those of Gopc–/– spermatozoa (Figs. 4F and 5G). However, the coiling process appeared to be different between Hook1–/– and Gopc–/– spermatozoa. In Hook1–/– spermatozoa, tail coiling started as a bending or a looping at the midpiece of late spermatids, and progressed in the epididymis. The authors suggested that the plasma membrane fusion was involved in coiling processes, since the spermatozoa appeared to have normal posterior ring. On the other hand, in Gopc–/– spermatozoa, plasma membrane fusion within a spermatozoon had never been observed during coiling processes.
Stratification of the Mitochondrial Sheath and Formation of Bare ODFs
In Gopc–/– spermatozoa, the mitochondrial sheath frequently stratified and shifted proximally. Consequently, ODFs became denuded of the mitochondrial sheath in the distal midpiece, as suggested in a previous paper [18]. Similar midpiece disorganization has been reported in Pvrl2/Nectin-2–/– mouse spermatozoa [38]. Here we consider the molecular mechanism underlying stratification of the mitochondrial sheath. SPATA19 has been suggested to act as an adhesive molecule between adjacent mitochondria [27, 28]. It may correspond with the cross-filaments linking adjacent outer mitochondrial membranes [39]. The present study showed that SPATA19 localizes on the mitochondrial sheath both in Gopc+/+ and Gopc–/– spermatozoa, even after disorganization of the mitochondrial sheath into the stratified form in the latter (Fig. 9). This result supports the idea that SPATA19 works as an adhesive molecule between the adjacent mitochondria of the sheath, and explains why the mitochondria retain an aggregated form even after stratification. Moreover, mitochondria freed from the ODFs have a wider area for attachment with each other. This is likely to support formation of a stratified mitochondrial sheath.
GPX4 is the only intracellular antioxidant enzyme known to directly reduce lipid hydroperoxide in membrane [40]. Recently, however, evidence is accumulating that GPX4 is a multi-functional protein [29], and while acting as a peroxidase in spermatids, it is a structural protein that stabilizes the mitochondrial sheath in spermatozoa [30]. Interestingly, infertile male patients with reduced sperm-GPX4 expression have been reported [41, 42]. TEM images of mitochondria in GPX4-defective spermatozoa show they are characterized by an irregularity in size and shape, by swelling, and by the paler appearance of the matrix, and a reduction of the cristae [33, 41]. In Gopc–/– spermatozoa, mitochondria showed similar ultrastructural characteristics to those of GPX4-defective spermatozoa, even before disorganization (Fig. 3C). In addition, GPX4 was unevenly expressed in the disorganized mitochondrial sheath (Fig. 10, B and C) suggesting local deficiency of GPX4. It is clear that the stability of the mitochondrial sheath in Gopc–/– spermatozoa is insufficient to maintain a single layered structure, and it is possible that a local deficiency in GPX4 may explain the fragility of the mitochondrial sheath. Tail coiling in Gopc–/– spermatozoa may, therefore, be partially explained by the local absence of GPX4, since in normal spermatozoa, the mitochondrial sheath is arranged around the flagellum, and must give some rigidity to the midpiece.
It is known that some tail abnormalities are due to epididymal disfunction. In relation to GPX4, LRP8 is interesting, because it is expressed in the epididymal epithelium. Lrp8–/– spermatozoa showed similar symptoms to GPX4-defective spermatozoa [33], and it has been suggested that LRP8 controlled the functional expression of sperm GPX4. In the present study, the principal cells of the caput epididymidis did not show a significant difference in expression of LRP8 between Gopc+/– and Gopc–/– mice. The immunohistochemical localization, however, showed minor differences between our data and previously reported results. A strong positive reaction has been reported at the luminal surface of the epididymal epithelium [32], while we found it associated with the Golgi apparatus. The reason for these differences is unclear, although it may be due to differences in tissue preparation such as fixation or in the antibodies used. As commented in the Introduction, sperm tail angulation was induced in Ros1/c-Ros–/– spermatozoa because of the undifferentiated nature of the initial segment of the epididymis [25]. In Gopc–/– mice, the initial segment developed normally in the fine structural characteristics (not shown). Therefore, the tail disorganization of Gopc–/– spermatozoa is presumed to be caused by factors intrinsic to the sperm themselves rather than to epididymal disfunctions.
One question that arises is whether tail coiling and mitochondrial detachment from ODFs are related or independent. Tail coiling is not essential for stratification, because it can be seen in sperm with straight tail (Fig. 4A). However, the mitochondria were often widely separated from the sharply curved ODFs (Fig. 2). These observations suggest that tail coiling facilitates the separation of the mitochondria from ODFs, which is likely to be beneficial for stratified mitochondrial sheath formation.
Role of GOPC During Spermiogenesis and Sperm Maturation
The function of GOPC molecules is still not clearly understood. However, from its molecular characterization and localization, it has been suggested that GOPC plays a role in vesicle transport from the Golgi apparatus [14], in addition to the role in membrane fusion. Axonemal proteins of the mammalian spermatozoa are thought to emerge from the distal centriolar region independent of the formation of head components. In ameba, it has been observed that flagellar proteins were synthesized on mRNAs localized close to the basal bodies [43]. However, in the flagella of mammalian spermatozoa, additional components, such as ODFs, a fibrous sheath, and a mitochondrial sheath are added around the axoneme in the later stage of spermiogenesis. It has been reported that KRT5/SAK57, an acidic keratin, was temporarily stored in the manchette, and then became a component of ODFs and the longitudinal columns of the fibrous sheath after the dispersion of the manchette [44]. In Gopc–/– spermatozoa, it is possible that there may be insufficient storage and transport of materials in the flagella due to the defectively formed manchette [17]. It may also cause tail disorganization during passage through the epididymis.
The present results suggest that tail disorganization of Gopc–/– spermatozoa is largely due to weak adhesion between the following structures: 1) the plasma membrane and nuclear envelope at the posterior ring, which results in coiling of the tail around the round nucleus, 2) the connecting piece of the flagellum and the implantation fossa of the nucleus, which causes dislocation of the tail from the nucleus and facilitates tail coiling, and 3) the mitochondria and ODFs, which result in detachment of the mitochondria from the ODFs. However, SPATA19 maintains adhesion between the individual detached mitochondria leading to formation of a stratified mitochondrial sheath. During passage through the epididymis, the migration of the cytoplasmic droplet containing the nucleus triggers tail disorganization in Gopc–/– spermatozoa, which have weak adhesion points. Overall the events observed in Gopc–/– spermatozoa provides a good model for analyzing malformation of both spermatozoa heads and flagella, and may prove useful in explaining clinical symptoms in human infertile patients.
ACKNOWLEDGMENTS
The authors are grateful to Mr. Tohru Mutoh and Mrs. Kyoko Kamimura for their technical assistance, and to Drs. Roy Jones and Elizabeth A. Howes for correcting English usage. The GPX4 antibody was a gift from Dr. S. Yokota.
FOOTNOTES
1Supported by a Grant-in-Aid for Scientific Research from the Japan Society for Promotion of Science to F.S.-T. (17590150) and, in part to K.T. (16390046). 
Correspondence: 2FAX: 81 43 226 2021; e-mail: fstoyota{at}faculty.chiba-u.jp
Received: 8 November 2006.
First decision: 8 January 2007.
Accepted: 8 March 2007.
REFERENCES
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
