| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Male Reproductive Tract |
Division of Biological Sciences, Graduate School of Science, Hokkaido University, Sapporo, 060-0810 Japan
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
gene regulation, male reproductive tract, spermatid, spermatogenesis, testis
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Mammalian spermatogenesis encompasses three phases: 1) proliferation and renewal of spermatogonia by mitosis, 2) meiosis, and 3) morphogenesis of the products of meiosis, the spermatids, to maturity during spermiogenesis [4]. Interestingly, apoptosis occurs most frequently in the germ cells at the second phase [5–9]. In addition, the last phase (spermiogenesis) involves drastic morphological and functional differentiation of newly formed spermatids into mature testicular sperm. Considering the above, it would be useful to investigate the roles of proteinases in the phases of spermatogenesis. A recent demonstration of the presence of Gzmb in testicular Sertoli cells and germ cells, both of which are noncytotoxic, is of particular interest [10]. The authors postulate that in the testes, Gzmb may facilitate migration of developing germ cells. This study strongly suggests that in addition to its roles in the immune system, Gzmb plays roles in human reproductive function [10].
In our attempts to search for serine proteinases expressed in the germ cells of adult mouse testes, we isolated and identified cDNA encoding a novel type of enzyme. Based on the remarkable amino acid sequence homology to other serine proteinase family members, the current molecule was considered to belong to the granzyme family and was designated granzyme N (Gzmn). Here, we report the expression and characterization of this novel granzyme. The present results indicate that Gzmn may be involved in the spermatogenesis of mouse testes. To the best of our knowledge, this is the first report on a granzyme exclusively expressed in the germ cells of mammalian testes.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Male C57BL/6NCrj strain mice were used. The animals were kept under controlled conditions (25°C, 14L:10D) and allowed free access to food and water. The animals were killed by cervical dislocation, and the relevant tissues were rapidly removed, frozen in liquid N2, and stored at –80°C until use. The experimental procedures were approved by the Animal Care and Use Committee of the Center for Experimental Plants and Animals, Hokkaido University.
cDNA Cloning of Mouse Gzmn
Total RNAs were prepared from frozen mouse testes and used to select poly(A)+ RNAs. The first strand of cDNA was synthesized using a SuperScript First-Strand synthesis system (Invitrogen Corp., Carlsbad, CA) according to the manufacturer's protocol. Polymerase chain reaction (PCR) amplification was performed using the following two degenerate oligonucleotide primers based on the cDNA sequence for conserved regions in serine proteinases: a sense primer (mSPS; 5'-GT(G/T)(C/G)T(G/ T)(A/T)C(A/T)GCTGC(A/T/C)CACTG-3') corresponding to the amino acid sequence Val-Leu-Thr-Ala-Ala-His-Cys, and an antisense primer (mSPAS; 5'-(A/T)GGGCC(A/T)CC(A/T/G)GAGTC(A/T)CC-3') corresponding to the amino acid sequence Gly-Asp-Ser-Gly-Gly-Pro-Leu. The PCR conditions were 94°C for 5 min followed by 30 cycles of 94°C for 1 min, 53°C or 55°C for 1 min, and 72°C for 1 min, with a final extension at 72°C for 7 min. Fragments of 400–500 base pairs (bp) in size were recovered from the PCR products by agarose gel electrophoresis. The products were subcloned into pBluescript(II)KS+ vectors (Stratagene, La Jolla, CA) cut with EcoRV. Forty-six clones for annealing at 53°C, and 61 clones for annealing at 55°C, were obtained from the PCR products after transformation of the recombinant plasmids into Escherichia coli JM109 cells. Among these, one clone (448 bp) was a cDNA fragment of a new granzyme (Gzmn). This cDNA fragment was used as a probe for further experiments.
A mouse testes cDNA library was constructed using the 
gt10 vector. (Stratagene) with 5 µg of poly(A)+ RNA and was packaged using Gigapack III extract (Stratagene). Screening was conducted for approximately 1 x 106 plaques from the library using the 448-bp EcoRV fragment according to the procedure previously described [11, 12]. This resulted in a new full-length granzyme clone of 975 bp.
Northern Blot Analysis
Twenty-five micrograms of total RNA from each tissue were separated by formaldehyde agarose gel electrophoresis and blotted onto a Nytran membrane (Schleicher & Schuell, Inc., Dassel, Germany). The blots were hybridized with the 32P-labeled, 448-bp EcoRV fragment as a probe at 42°C for 16 h in 50% formaldehyde, 5x standard saline-sodium phosphate-ethylenediaminetetraacetic acid (SSPE) buffer/1% (SDS)/5x Denhardt solution, and denatured salmon sperm DNA (100 µg/ml). The membranes were washed with increasing stringency; 0.1x standard saline-sodium citrate (SSC) buffer/0.1% SDS was used for the final wash at 42°C. A cDNA fragment of glyceraldehyde 3-phosphate dehydrogenase (GAPDH), ß-actin (mßA), or cold shock domain protein A (Csda) [13] (also called mouse Y-box protein 3 [14] or MSY4 [15]) was used as a control. The cDNA fragments of mouse GAPDH (306 bp), mßA (293 bp), and Csda (397 bp) were amplified with a set of primers (Table 1) using total RNA isolated from the ovaries (for GAPDH) or testes (for mßA and Csda).
|
Isolation of Splenocytes and Interleukin-2 Activation
Resting spleen cells were isolated from mice before culture according to a procedure described previously [16]. In brief, spleen cells were cultured in RPMI 1640 medium with 10% heat-inactivated fetal bovine serum, 2 mM glutamine, 100 U/ml of penicillin, 100 µg/ml of streptomycin (Gibco BRL, Life Technologies, Tokyo, Japan), and 100 U/ml of interleukin (IL)-2 (Sigma Chemical, St. Louis, MO). After 7 days, the cells were harvested for reverse transcription (RT)-PCR analysis.
RT-PCR Analysis
Total RNAs were isolated from testes of 3-, 4-, and 8-wk-old mice. After DNase I treatment, the RNA (5 µg) was reverse-transcribed into cDNA at 37°C for 1 h using the SuperScript First-Strand synthesis system and a poly dT-oligonucleotide. A 1/100 aliquot of cDNA thus synthesized was then subjected to PCR with 0.5 U of Ex Taq polymerase (hot start version; TaKaRa, Tokyo, Japan) and specific oligonucleotide primers (Table 1). Amplifications were carried out for granzyme A (Gzma), Gzmb, granzyme C (Gzmc), granzyme D (Gamd), granzyme E (Gzme), granzyme F (Gzmf), granzyme G (Gzmg), granzyme K (Gzmk), granzyme M (Gzmm), and Gzmn, and the products were electrophoresed in agarose gel (2.5%; thickness, 0.5 cm) and stained with ethidium bromide (1 µg/ml, 10 min). A negative control experiment was performed without reverse transcriptase.
Total RNAs isolated from heat-stressed mouse testes were used for RT-PCR. A 1/25 aliquot of cDNA was subjected to PCR amplification under the following conditions: 25 cycles at 65°C for Gzmn, 20 cycles at 65°C for GAPDH, and 20 cycles at 55°C for heat shock protein (Hsp) 70-1. The primers used for Gzmn, GAPDH, and Hsp70-1 [17] are indicated in Table 1.
In some experiments, the signal intensities of RT-PCR products were quantified with an ATTO Light-Capture AE-6960 (ATTO Ltd., Tokyo, Japan) and were normalized against a corresponding relative intensity of GAPDH in each sample.
In Situ Hybridization Analysis
In situ hybridization studies were performed as described previously [11]. A full-length Gzmn cDNA (975 bp) was used to generate both antisense and sense digoxigenin (DIG)-labeled riboprobes with a DIG RNA Labeling Mix (Roche Diagnostics, Mannheim, Germany). Testes of 8-wk-old mice were frozen and sectioned (thickness, 10 µm). The sections were treated with 10 ng/ml of proteinase K (Roche), postfixed, and acetylated. Prehybridization was performed with hybridization buffer (50% formaldehyde, 6x SSPE, 1% SDS, 5x Denhardt solution, and 500 ng/ml of tRNA) for 1 h at room temperature. Hybridization was carried out in hybridization buffer containing 200 ng/ml of cRNA probes at 70°C for 18 h. The sections were first washed once for 20 min with 2x SSC and then twice for 20 min each time with 0.2x SSC at 70°C. The hybridized probes were detected with anti-DIG-alkaline phosphatase conjugate in a buffer containing nitro blue tetrazolium chloride and 5-bromo-4-chloro-3-indolyl phosphate in darkness.
Preparation of Specific Gzmn Antibody
The mouse Gzmn cDNA (amino acids Glu-19 to Leu-248) was ligated in-frame between the EcoRV and EcoRI sites of pET30a (Novagen, Madison, WI) after creating an EcoRV site at the 5'-ends and an EcoRI site at the 3'-ends by PCR with pfu polymerase (Novagen). The orientation and sequence of the cDNA in the pET plasmid were confirmed by DNA sequencing. The ligated vector was transformed to E. coli strain BL21(DE3) pLysS (Novagen). The conditions for the cell culture and the preparation of recombinant protein using an Ni2+-chelate column were as described previously [18]. The protein thus prepared was a fusion protein containing 48 extra amino acids (all of which originated from the plasmid sequence) at the N-terminus in addition to the Gzmn sequence. The purified recombinant protein was used as an antigen to raise specific antibody in rabbits. The antisera were obtained from blood collected after five booster injections. Rabbit antiserum was affinity-purified as follows: The recombinant Gzmn protein separated by SDS-PAGE was transferred to a polyvinylidene difluoride (PVDF) membrane (Millipore, Bedford, MA). The membrane was incubated with the antiserum overnight, washed extensively with PBS, and then eluted with 50 mM glycine-HCl (pH 2.5). The eluate was immediately neutralized and concentrated for Western blot analysis. Starting with 0.5 ml of antiserum, 50 µl of purified Gzmn antibody in PBS were prepared.
Preparation of Testes Extracts and Western Blot Analysis
Testes from 1-, 2-, 3-, 4-, and 8-wk-old untreated mice, cryptorchid testes, and heat-stressed testes were separately homogenized with five volumes of SDS-PAGE sample buffer. The homogenates were boiled for 20 min and centrifuged at 13 000 x g for 10 min. The resulting supernatants were used as samples.
Five-microliter aliquots of the extracts prepared as described above were subjected to SDS-PAGE in 12% polyacrylamide gel under reducing conditions. The proteins were then electroblotted onto PVDF membranes. The purified Gzmn-specific antibodies or preimmune sera were diluted 1: 1000 with PBS containing 0.05% Tween 20 (TPBS; pH 7.4). The secondary antibody, goat anti-rabbit immunoglobulin (Ig) G (Amersham Biosciences, Piscataway, NJ), was diluted 1:5000 with TPBS. Signals associated with testicular Gzmn were detected using an ECL Western blot detection kit (Amersham Biosciences) according to the protocol provided by the manufacturer.
Gzmn Construction and Expression in COS-7 Cells
To produce a recombinant pro-Gzmn protein identical to mature Gzmn but with a FLAG-tag and an HA-tag in its N-terminus and C-terminus, respectively, the expression vector pMH-Gzmn was constructed as follows: First, a 91-nucleotide sequence corresponding to the Gzmn sequence from Met-1 to Glu-20 with a FLAG-tag (DYKDDDDK) at the carboxyl terminus was amplified using Gzmn-primer-A (5'-GGAATTCATGCTACCAGTTCTGATTCT-3') and Gzmn-primer-B (5'-CTTATCGTCGTCATCCTTGTAATCCTCCTCTGCTCCATCTCCTA-3') with full-length Gzmn cDNA as a template. Next, a 717-bp nucleotide sequence, which corresponded to FLAG-tag followed by the sequence of Val-21 to Leu-248, was amplified using the primer set of Gzmn-primer-C (5'-GATTACAAGGATGACGACGATAAGGTCATCGGGGGCCATGAGGT-3') and Gzmn-primer-D (5'-CCGGATATCGAGCAGCTTCATGTTTGTGC-3'). The above two PCR products were purified from agarose gels. Using the PCR products as templates, an amplification was carried out with Gzmn-primer-A and Gzmn-primer-D. This sequence was ligated in-frame between the EcoRI and HindIII sites of pMH vector (Roche). The COS-7 cells were cultured in Dulbecco modified Eagle medium containing 10% fetal bovine serum and 1% penicillin-streptomycin-Glutamine (Invitrogen). Transfection of pMH-Gzmn and Mock vectors was performed with lipofectamine reagent (Invitrogen) in serum-free medium. Transfected cells and conditioned medium were obtained at 48 h after transfection. The expression was assessed by Western blot analysis with anti-FLAG antibodies (Sigma) or anti-HA antibodies (Roche).
N-Glycosylation Analysis of Gzmn
The COS-7 cells expressing Gzmn were collected, sonicated in PBS, and centrifuged [12]. The resulting supernatant was used as the cell lysate. The conditioned medium was used after being concentrated 20-fold. Testes from 8-wk-old mice were homogenized in 0.1 M Tris-HCl (pH 8.6) containing 0.5% SDS and 10 mM ß-mercaptoethanol, and homogenate was centrifuged at 14 000 x g for 15 min. The resulting supernatant was used as the testes extract. The samples were treated with glycopeptidase F under denaturing conditions using a Glycopeptidase F kit (TaKaRa). Negative-control experiments were performed without the enzyme.
Immunohistochemistry
Cryostat cross-sections (thickness, 7 µm) were prepared from frozen testes and mounted on gelatin-coated slides. Sections were dried overnight by cold air from a dryer. After washing in cold PBS, nonspecific binding sites were blocked with 10% normal goat serum in PBS for 1 h at room temperature. The purified Gzmn-specific antibodies or preimmune sera were diluted 1:10 with 10% normal goat serum in PBS and applied to the sections for 1 h at room temperature. After the sections were washed in cold PBS, they were fixed with 4% paraformaldehyde in PBS for 20 min at room temperature. The sections were washed in PBS, then incubated with biotinylated goat secondary antibody specific to rabbit IgG for 30 min at room temperature. After an additional wash in PBS, bound antibody was detected by alkaline phosphatase-conjugated streptavidin for 30 min and visualized using VECTOR Red (Vector Laboratories). Toluidine blue was used to counterstain the nuclei, followed by the addition of aqueous mounting media.
Experimental Cryptorchidism
Surgery was performed using 8-wk-old mice under light ether anesthesia as described previously [5]. Briefly, a midline abdominal incision was made, and the left testis was manipulated into the abdomen and then sutured to the abdominal wall. As a control, the right testis was manipulated into the abdomen and then returned to the scrotum. The testes were removed at Postoperative Days 0 (untreated), 2, 4, 6, 8, and 10 for total RNA extraction and for tissue extraction. Northern blot analysis was sequentially conducted using a blot membrane containing 10 µg of the above total RNAs with the 32P-labeled, 448-bp Gzmn cDNA; the 32P-labeled, 397-bp Csda cDNA; and the 32P-labeled, 30-bp 18S ribosomal RNA-specific cDNA fragment (5'-CGGCATGTATTAGCTCTAGAATTACCACAG-3'). The tissue extracts were subjected to Western blot analysis under reducing conditions. All experiments were performed in triplicate.
Heat Shock
Eight-week-old male mice were exposed to a single heat shock as previously described [19]. Mice were restrained, and the lower half of the torso of each animal was submerged in a warm water bath. Heat shock was performed at 43°C for 20 min. As a control experiment, mice were subjected to heat treatment at 33°C for 20 min. At 8 and 16 h after heat shock, mice were killed, and both testes were collected for RT-PCR and Western blot analysis of Gzmn. As controls, Hsp70-1 and GAPDH were chosen. Three animals were used for each group.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
As described in Materials and Methods, RT-PCR with two degenerated oligonucleotide primers (mSPS and mSPAS) was conducted using total RNAs prepared from adult mouse testes. Among the 107 PCR products analyzed, one with a length of 448 bp had a nucleotide sequence homologous to those of granzymes. This 448-bp cDNA fragment was then used as a probe for screening a mouse testes cDNA library, and a total of nine positive clones were obtained. Based on preliminary nucleotide sequence analyses, the clones could be grouped into two types. The two representative cDNA clones, Gzmn-1 and Gzmn-2, were further analyzed to determine their complete nucleotide sequences (Fig. 1). The Gzmn-1 was a clone of 975 bp and consisted of a 102-bp 5'-noncoding region, a 747-bp coding sequence, and a 126-bp 3'-noncoding region (the nucleotide sequence has been deposited in the DDBJ/ EMBL/GenBank Data Bank under accession no. AB049454) (Fig. 1A). The Gzmn-2 was a clone of 1006 bp and was made up of a 102-bp 5'-noncoding region, a 642-bp coding region, and a 262-bp 3'-noncoding region. The two clones differed from each other in that Gzmn-2 contained an additional 34-bp nucleotide segment in the Gzmn-1 sequence (Fig. 1B). The insertion of the 34-bp segment appears to occur after the first nucleotide of a codon for Gly (Gly-202). This alternative splicing event results in the generation of Gzmn-2. Because of a frame shift caused by the insertion of the segment, a premature stop codon is produced in Gzmn-2. As will be described below, Gzmn-2 does not appear to encode a functional proteinase.
|
The Gzmn-1 encodes 248 amino acid residues with a calculated molecular weight of 26 886. The sequence of the 18-amino acid hydrophobic signal peptide was predicted according to von Heijne [20]. The activation site is presumably the region from the COOH-terminal to Glu-20 (the Met at the initiation codon is assigned the number 1). The amino acid sequence contains one possible N-glycosylation site (Asn-154). Three essential residues (His-65, Asp-109, and Ser-204) required for proteolytic activity are conserved. On the other hand, the Gzmn-2 clone encodes a putative protein of 213 amino acid residues with a molecular weight of 23 161. The amino acid sequence of this protein is the same as that of the Gzmn-1 protein up to Lys-201, but its COOH-terminal 1/3 portion is totally different from the predicted structure of the counterpart Gzmn-1 protein. Because the Gzmn-2 protein lacks the essential serine residue, it should not function as a proteinase even if the corresponding mRNA were translated. (In the following section, we deal mainly with the mGzmn-1 protein.)
A homology search in the NCBI Protein BLAST database revealed that the current molecule was highly homologous to mouse Gzm subfamily members. The amino acid sequence identities of Gzmn were 71% for Gzmf [21], 70% for Gzme [21] and Gzmg [22], 68% for Gzmd [21], and 55% for Gzmb [23] and Gzmc [24]. On the other hand, Gzma [25], Gzmk [26], and Gzmm [16] exhibited identities as low as 33%. Interestingly, the amino acid sequence of Gzmn was highly homologous (72% sequence identity) to that of rat NK cell protease 7 [27].
During the course of the present study, a mouse genomic sequence corresponding to Gzmn appeared in the public database (GenBank accession no. AC091783). The alignment analysis between the current cDNA and its genomic sequence revealed that the mouse Gzmn gene spans approximately 9 kilobases (kb) of genomic DNA, with six exons and five introns. Both Gzmn-1 and Gzmn-2 are clearly shown to be two splicing products of the mouse Gzmn gene (Fig. 1C).
Expression of Gzmn mRNA
Northern blot analysis of Gzmn mRNA was conducted using total RNAs isolated from various tissues of the adult mouse. The Gzmn mRNA was detected only in the testes, and its size was approximately 1.1 kb (Fig. 2A).
|
The expression of Gzmn mRNA in the testes during postnatal development was also examined (Fig. 2B). The Gzmn transcript was detected at 3 wk after birth. The expression became more prominent as the animal reached sexual maturity.
The relative expression of Gzmn-1 and Gzmn-2 mRNA in the testes was examined. For this purpose, two primers (primer SS and primer AS) (see Table 1 and Fig. 1, A and B) were used for RT-PCR analysis of total testicular RNAs isolated from 3-, 4-, and 8-wk-old mice. The primers were designed to amplify both types of cDNAs simultaneously. After PCR reactions using different numbers of cycles for mice of different ages, the amplified products were analyzed by acrylamide gel electrophoresis (Fig. 2C). The PCR products were detected by 22 cycles of reaction with the total RNA from 8-wk-old mouse testes. On the other hand, at least 27 cycles were required to detect the same products when the total RNAs of 3- and 4-wk-old mice were used as templates. The results showed that the intensities of PCR products amplified at 27 cycles were greater in 4-wk-old mice than in 3-wk-old mice. Therefore, quantitative analysis of band intensities was conducted after 27 cycles, when the PCR reactions were linear. The results showed that the amounts of Gzmn product (Gzmn-1 + Gzmn-2) in 4- and 8-wk-old mice were 2.6- and 2.3-fold greater, respectively, than the amount in 3-wk-old mice; this finding was compatible with the data of Northern blot analysis (Fig. 2B). In addition, two products corresponding to Gzmn-1 and Gzmn-2 were detected at all time points examined. A similar densitometric analysis revealed that the amount of Gzmn-2 product was 1.5- to 1.9-fold greater than the amount of Gzmn-1.
We next examined the expression of the Gzmn gene in mouse spleen cells. When cells were treated with IL-2, a remarkably high level of Gzmb gene expression was seen (Fig. 2D). On the other hand, no Gzmn mRNA was detected in the spleen cells, even after treatment with IL-2. For comparison, total RNAs from the testes were used for PCR amplification under the same conditions. The two Gzmn fragments were detected, again confirming the presence of two splicing products in the mouse testes. A very faint band corresponding to the Gzmb product was visible.
We assessed the relative testicular expression levels of the 10 functional Gzm genes by RT-PCR analysis with the respective specific primers (Fig. 3). Total RNAs isolated from 8-wk-old mice were used as templates. When using 25 cycles of PCR amplification, only the products corresponding to Gzmn and Gzmk were detected, with the band intensity of the former being much greater than that of the latter (Fig. 3A). The PCR analysis using 30 cycles yielded an additional three granzyme species: Gzma, Gzmc, and Gzmm (Fig. 3B). The five other granzyme species (Gzmb, Gzmd, Gzme, Gzmf, and Gzmg) were not amplified at all under these conditions. As described above, we could detect a faint PCR product band corresponding to Gzmb (Fig. 2D). This discrepancy probably resulted from the difference in PCR analysis conditions between the two experiments; PCR experiments (Fig. 2D) were conducted using a fivefold greater amount of testes total RNA as a template than was used for the experiment shown in Figure 3. At any rate, the expression level of Gzmb is undoubtedly very low in the testes. In RT-PCR using total RNA from IL-2-treated splenocytes as a template, the products for most granzyme species (i.e., all species except Gzmf, Gzmg, and Gzmn) were amplified (data not shown). These results clearly indicate that Gzmn is the dominant species of this family of enzymes expressed in the mouse testes.
|
Distribution of Gzmn mRNA-Positive Cells in the Adult Mouse Testes
The localization of Gzmn mRNA in the mouse testes was examined by in situ hybridization. A DIG-labeled antisense probe for Gzmn was used to detect the mRNA in 8-wk-old mouse testes. The analysis revealed that some, but not all, seminiferous tubules gave specific, positive signals for Gzmn expression (Fig. 4), probably because of the cycles of spermatogenesis. No signal was detected in the interstitial tissues. To identify the Gzmn-positive cell types in the tubules, the sections were counterstained with methyl green. Pachytene spermatocytes, round spermatids, and elongated spermatids in the positive tubules were specifically stained, but spermatogonia and testicular sperm were not. No positive signals were detected using the sense probe.
|
Expression of Gzmn Protein in the Mouse Testes and COS-7 Cells
Testes extracts of mice at various postnatal developmental stages were examined by Western blot analysis using purified Gzmn antibody. The antibody detected two polypeptide bands (Mr = 30 and 31 kDa) when using the extract of 4-wk-old mouse testes and when using the extract of 8-wk-old mouse testes (Fig. 5A). No signal was detected with the preimmune serum (Fig. 5B). The above-mentioned two bands were also detectable with the 3-wk-old mouse testes extract when the membrane was stained by longer exposure (data not shown); they were not detected with the sperm extract (data not shown). These results demonstrate that Gzmn protein is, indeed, present in the mouse testes.
|
To examine whether Gzmn is a secretory protein, we constructed an expression vector, Gzmn-FLAG-HA, to produce a recombinant protein containing FLAG-tag and HA-tag at the N-terminus and C-terminus of the mature Gzmn (Fig. 5B, top). After COS-7 cells were transfected with the vector, the cell lysate and conditioned medium were prepared. Western blot analysis of the samples gave two closely migrating bands with Mr values of 35 and 36 kDa, both of which were detectable with FLAG-tag and HA-tag antibodies (Fig. 5B, bottom). These results clearly indicate that the sequence of hydrophobic amino acid residues at the N-terminus of Gzmn functions as a signal peptide and that Gzmn is a secretory protein.
The Gzmn was consistently detected in two forms with the extract of testes and COS-7 cells (Fig. 5, A and B). Therefore, the nature of this multiplicity was examined. The conditioned medium and cell lysate obtained from COS-7 cells transfected with Gzmn-FLAG-HA vector were treated with or without glycopeptidase F and were subjected to Western blot analysis using the FLAG-tag antibody. As shown in Figure 5C (top), treatment with the enzyme caused disappearance of the 36-kDa polypeptide band in the cell lysate and conditioned medium. Similarly, the glycopeptidase F treatment of testes extracts strengthened the signal intensity of the 30-kDa polypeptide band, with a concomitant disappearance of the 31-kDa band (Fig. 5, bottom). These results indicate that Gzmn undergoes N-glycosylation when it is expressed in COS-7 cells as well as in the testes. Therefore, the multiplicity of the protein is considered to result from the posttranslational attachment of a carbohydrate during synthesis.
Immunohistochemical Localization of Gzmn in the Adult Mouse Testes
Sections of the adult mouse testis were immunohistochemically analyzed with Gzmn-specific antibody (Fig. 6). The sections were counterstained with toluidine blue to identify cells expressing the antigen. Some, but not all, pachytene spermatocytes were positively stained. Some of the round spermatids (stages I–VII) and elongated spermatids (stages VIII–XII) also showed positive immunoreactivity for Gzmn [28]. Spermatogonia and testicular sperm showed no immunoreactivity. No specific signal was observed with the preimmune serum. Some positive staining for the antibody was observed in the myoid and interstitial areas. These areas also stained positive with the preimmune serum, however, so the staining was considered to be nonspecific.
|
Change in Expression of Gzmn mRNA in Experimental Cryptorchid Testes and Heat-Stressed Testes
The Gzmn mRNA expression was examined in the surgically induced cryptorchid testes by Northern blot analysis. The expression level of Gzmn mRNA was unchanged up to Postoperative Day 8 after the induction of cryptorchidism (Fig. 7A). On Day 10, the mRNA level in the cryptorchid testes was drastically decreased compared with that in the scrotal testes. We also examined the expression of Csda under the same conditions. The Csda gene expression was induced soon after the operation, and such an increased expression level was maintained even at Postoperative Day 10 in the treated testes. This expression pattern of the Csda gene was essentially the same as that reported in two recent studies [29, 30], confirming the validity of our cryptorchidism experiment. Expression of 18S rRNA was not affected by the treatment. The level of Gzmn protein in the extracts of the cryptorchid testes decreased on Postoperative Day 6 and continued to decline steadily thereafter (Fig. 7B, left). In the three control scrotum testes, no change in the protein level was detected (Fig. 7B, right).
|
Next, we investigated the effect of local heating of the testes on the levels of Gzmn mRNA. After bathing the lower abdomen of the male mouse in water at 43°C for 20 min, Gzmn mRNA levels in the testes were analyzed by RT-PCR using the total RNA. The local heating of the testes did not result in change in Gzmn mRNA levels at either 8 or 16 h after treatment (Fig. 7C). As a positive control, expression of the Hsp70-1 gene in the treated testes was examined. An elevated level of Hsp70-1 mRNA was detected at both 8 and 16 h after the local heating, which was consistent with a previous report that a 43°C heat shock induced expression of the Hsp70-1 gene [19]. The result of the present control experiment indicated that the results of the present heat shock experiment were reliable. Expression of a housekeeping gene (GAPDH) was not affected by the treatment. Western blot analysis revealed that the level of Gzmn protein in the testes extract was not altered by the local-heating treatment (Fig. 7D).
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
To provide insight regarding the substrate preference of Gzmn, we compared key amino acid residues of the primary structure of Gzmn with those of its related molecules. It is generally believed that the specificity of serine proteinases is mostly governed by four residues—189, 192, 216, and 226 (numbering based on the zymogen chymotrypsinogen)—among which residues 189 and 226 are most critical in the S1 pocket. Gzmn has Ala-198 (corresponding to chymotrypsinogen residue 189), Lys-201 (corresponding to chymotrypsinogen residue 192), Val-221 (corresponding to chymotrypsinogen residue 216), and Gly-229 (corresponding to chymotrypsinogen residue 226). Three of these residues are identical to those of rat NK cell protease 7 [27], which shows the highest homology (72% identity) to Gzmn. In analogy to rat NK cell protease 7, the current granzyme may have a preference for large hydrophobic amino acids in the P1 position because of its Ala-198 and Gly-201. We hypothesized that Gzmn is a proteinase with chymotrypsin-like specificity. To determine whether this prediction was valid, we tried to obtain an active recombinant Gzmn using a variety of expression systems, such as expressions in E. coli and mammalian cells, but this attempt was unsuccessful (data not shown). Thus, the above idea must be experimentally confirmed by future biochemical studies.
An interesting feature of Gzmn is its restricted expression. In our Northern blot analysis, the mRNA for this granzyme was detected exclusively in the testes. This expression pattern was surprising, because granzymes are generally thought to be specific for activated CTLs and NK cells [31, 32]. Therefore, we initially suspected that the pronounced expression of Gzmn in the testes might have resulted from lymphocytes residing in the interstitial tissues of this organ. Unexpectedly, in situ hybridization and immunohistochemical analyses revealed that the cells expressing Gzmn were spermatocytes and spermatids in the seminiferous tubules. The complete absence of Gzmn transcript in the interstitial tissues indicates that the testicular lymphocytes are not responsible for the expression. In this context, it is particularly interesting to note a recent finding on Gzmb expression in the human testes [10]: Those authors localized Gzmb mRNA and the protein to both Sertoli cells and germ cells, such as primary spermatocytes and round spermatids. The present study also confirmed the testicular expression of Gzmb in the mouse by RT-PCR, although the level was much lower than that of Gzmn.
At present, 11 Gzm genes have been identified for the mouse, including granzyme L (Gzml), which is presumed to be a pseudogene with early termination sites [3]. Our current data regarding RT-PCR analysis for the 10 functional Gzm species revealed that in addition to Gzmn and Gzmb, four other genes (Gzma, Gzmc, Gzmk, and Gzmm) are expressed in the mouse testes (Figs. 2D and 3). However, with the exception of Gzmn, the mRNA levels of these Gzm species clearly are very low and only detectable by PCR. A densitometric evaluation indicated that the expression of Gzmn was at least two orders of magnitude higher than the expressions of the other Gzm species. Gzmn is thus the Gzm family member predominantly expressed in the mouse testes.
We demonstrated that Gzmn expression was not detected in the mouse splenocytes treated with IL-2. This is in contrast with the remarkably high level of Gzmb mRNA induced by the same treatment. Together, these results indicate that Gzmn is specifically expressed in the testes. They also suggest that Gzmn expression is controlled by a mechanism different from those used to control Gzm species involved in the pathway for lymphocyte-mediated killing in both the innate and adaptive immune systems [1, 33–35].
The Gzmn mRNA is present in the testes of mice older than 2 wk. Two distinct transcripts, Gzmn-1 and Gzmn-2, are detectable at an approximate ratio of 1:2 throughout the periods of its expression. At present, the physiological meaning of the presence of these two distinct Gzmn mRNAs is not known. If they were both translated, then two forms of putative precursor proteins (without the 18-amino acid signal peptide but with the putative two-residue propeptide) would be detected in the testes extracts: a protein of 230 amino acid residues (expected molecular weight, 25 009) that includes the three catalytic triad residues, and a protein of 195 amino acid residues (expected molecular weight, 21 284) without the catalytic Ser residue. However, the molecular mass of the testicular Gzmn recognized with the specific antibody was approximately 30 kDa. These values are larger than those calculated from the amino acid sequences of these proteins. Such differences in molecular size between the theoretical and experimental values have previously been reported for other proteins, including serine proteinase family members [11, 12, 18, 36]. Interestingly, the testicular tissues in the present study contained Gzmn in two forms (Mr = 31 kDa and 30 kDa). We clearly demonstrated that this multiplicity was the result of N-glycosylation of the gene product. The presence or absence of an attached carbohydrate produces two forms of the protein. Indeed, Gzmn contains a single N-glycosylation site at Asn-154. In analogy to the activation of other granzymes, Gzmn may undergo proteolytic conversion by removal of two amino acid residues (Glu-Glu) from the NH2-terminus of the proenzyme. This proteolytic processing may occur in cytoplasmic granules of spermatocytes and spermatids, presumably because of the action of a coexisting cysteine proteinase dipeptidyl peptidase I [37–40]. The idea that the 31- and 30-kDa proteins may represent the inactive precursor and active mature Gzmn, respectively, therefore is ruled out completely. Although the present study demonstrates that the mouse germ cells express two distinct forms of Gzmn transcript (Gzmn-1 and Gzmn-2), it is not clear whether Gzmn-1, Gzmn-2, or both are translated in the mouse testes. Further studies are necessary to solve this problem.
The in vivo role of Gzmn is not known, but its testes-specific expression strongly suggests a biological role closely associated with testicular function. In the adult mouse, spermatogenesis is characterized by continuous germ cell maturation toward the center of the seminiferous tubules: mitotic proliferation of spermatogonia, meiotic division of spermatocytes, differentiation of spermatids, and finally, release of testicular sperm into the tubule lumen. The present study demonstrated that the Gzmn gene is expressed in the spermatocytes and round spermatids. Because apoptosis occurs most frequently among these germ cells [5–9] and the participation of Gzma [41], Gzmb [42], and Gzmc [43] in cytotoxic cell killing has been clearly established, we considered it useful to examine whether Gzmn is involved in apoptosis during spermatogenesis. In a previous study using an adult-mouse model of experimental unilateral cryptorchidism, DNA fragmentation consistent with apoptosis was observed on Day 6 in the cryptorchid testes, with subsequent loss of testicular weight, histologic evidence of germ cell loss, and histochemical staining of apoptotic germ cells [9]. In the present study, we showed that the Gzmn mRNA level declined on Day 10 in the cryptorchid testes but not in the scrotum testes. On the other hand, in the treated testes, Gzmn protein levels remained unchanged up to Day 4 but then gradually decreased thereafter. The nature of this differential change between Gzmn mRNA and protein in the cryptorchid testes remains to be explored. Nevertheless, we tentatively presume that Gzmn is not directly involved in the apoptotic process induced by experimental cryptorchidism, although further experiments are needed to validate this hypothesis. Interestingly, a remarkable drop in the Gzmn mRNA level occurred while the transcript of Csda, an mRNA repressor that is highly expressed in mouse spermatogenic cells [14, 29, 30], was still at an induced level. As demonstrated in the current study, Gzmn mRNA is localized in pachytene spermatocytes, round spermatids, and elongated spermatids.
Heat shock proteins, including Hsp70-1, are required for spermatogenesis, and they protect cells from environmental hazards, such as heat, radiation, and chemicals [44]. A recent study has shown that testicular expression of Hsp70-1 is induced by a single heat shock treatment of the testes [19]. That same study also showed that germ cell apoptosis was initiated in the heat-stressed testes simultaneously with the Hsp70-1 expression [19]. In the present study, we confirmed that Hsp70-1 expression was induced by heat shock treatment. However, the treatment had no effect on the expression level of Gzmn mRNA or Gzmn protein. These findings at least appear to exclude the possibility that Gzmn is involved in the early event of germ cell apoptosis induced by heat shock. Our current data clearly show that the Gzmn expression profile in the heat-stressed testes is distinct from the expressions of either Csda or Hsp70-1, indicating that expression of the Gzmn gene is under different regulation than those of the other two genes, despite all three being expressed in the germ cells. We tentatively assume that Gzmn plays a role in spermatogenesis in the mouse testes, but the precise role of this protein remains to be elucidated.
|
FOOTNOTES |
|---|
2 Correspondence: Takayuki Takahashi, Division of Biological Sciences, Graduate School of Science, Hokkaido University, Sapporo 060-0810, Japan. FAX: 81 11 706 4851; ttakaha{at}sci.hokudai.ac.jp
Received: 7 April 2004.
First decision: 24 April 2004.
Accepted: 15 July 2004.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  D. Kaiserman, C. H. Bird, J. Sun, A. Matthews, K. Ung, J. C. Whisstock, P. E. Thompson, J. A. Trapani, and P. I. Bird The major human and mouse granzymes are structurally and functionally divergent J. Cell Biol., November 20, 2006; 175(4): 619 - 630. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
