Apoptosis-related proteins in the testes of infertile men with varicocele


M. Fujisawa, Division of Urology, Department of Organs Therapeutics, Faculty of Medicine, Kobe University Graduate School of Medicine, 7-5-2 Kusunoki-cho, Chuo-ku, Kobe 650–0017, Japan.
e-mail: masato@med.kobe-u.ac.jp


Objectives To assess immunohistochemically the expression of apoptosis-related proteins in the testes of infertile men, to determine which of these proteins were related to hypospermatogenesis, as a previous report suggested that apoptosis was suppressed in infertile men with varicocele.

Materials and methods The expression of Bcl-2, Bax, caspase-1 (ICE) and caspase-3 (CPP32) were examined in bilateral testicular specimens from 26 infertile men with varicocele and six normal testicular specimens, using the avidin-biotin-peroxidase complex method. Clinical variables were also assessed.

Results Bax, ICE, and CPP32 were expressed in germ cells, while Bcl-2 was not. Differences in staining in left or right testes were not significant. In both testes of infertile patients with varicocele, significantly fewer germ cells stained for CPP32 than in controls (P < 0.001). For Bax and ICE, total germ cell staining was similar between these groups. Staining was less frequent in infertile patients for both CPP32 and ICE when the analysis was restricted to spermatogonia. Serum luteinizing hormone levels correlated positively with CPP32 staining (P = 0.0457).

Conclusions The reduced expression of CPP32 participates in regulating apoptosis in the testes of infertile men with varicocele.


Apoptosis is a form of cell death resulting from the activation of a genetically determined cell-suicide programme [1]. Germ cells undergo spontaneous degeneration during normal spermatogenesis, with 25–75% of germ cells dying in the course of differentiation [2–4]. This cell death occurs by apoptosis, which is regulated by intratesticular androgens and gonadotrophins [5–7].

Recently, several apoptosis-related genes have been reported [8–15]. The Bcl-2 gene was discovered through its involvement in the t(14;18) chromosomal translocations found in most non-Hodgkin's lymphoma [8]. Increased concentrations of Bcl-2 protein have been shown to delay or block apoptosis in a large variety of cells under various conditions [9,11]. Bcl-2 is expressed in a variety of cell types, particularly long-lived cells such as memory lymphocytes and some types of neurones, as well as in regenerative stem cells located along the basement membrane in several epithelia, including the skin, colon and prostate [10]. The Bax protein is homologous with Bcl-2 and functions as its principal inhibitor [12]. In some tissues, patterns of Bax and Bcl-2 expression vary reciprocally, suggesting active antagonism between these proteins. On the other hand, Bax is expressed more widely than Bcl-2, being present in some organs containing little or none of the latter [14]. Additionally, caspase-1 (ICE) and caspase-3 (CPP32), proteases belonging to the caspase family, often induce apoptosis when over-expressed in mammalian cells [13,15].

Previously we reported that less apoptosis was detectable in the testes of infertile men with varicocele than in those of normal men [16]. In the present study, to determine which apoptosis-related genes regulated apoptosis in hypospermatogenic testis affected by varicocele, we immunohistochemically compared the expression of Bcl-2, Bax, ICE, and CPP32 in the testes of infertile men and normal men. We also investigated relationships between apoptosis-related proteins and clinical findings.

Materials and methods

Bilateral testicular biopsy specimens (52) from 26 infertile men (mean age 34.5 years, sd 3.8, range 28–46) with left varicocele were assessed. After evaluation the varicocele was classified into three grades according to size; nine were grade I, seven grade II and 10 grade III. Semen samples were analysed three times after 5 days of abstinence from sexual intercourse, all using the WHO guidelines [17]. Serum concentrations of FSH, LH, testosterone and oestradiol were also evaluated. Normal control testicular specimens were obtained from six fertile volunteers at the time of vasectomy (mean age 38.7 years, sd 5.6, range 32–46). All biopsies were taken through a small incision after obtaining informed consent. Testicular specimens were fixed in 10% neutral buffered formalin or Bouin's solution for 3–6 h and embedded in paraffin wax. Samples fixed in Bouin's solution were stained with haematoxylin and eosin, and were used to determine Johnsen's mean score.

Formalin-fixed specimens were immunostained for Bcl-2, Bax, ICE and CPP32 as described previously [14,18,19]. Deparaffinized, rehydrated sections (4 µm) were heated in a microwave oven or pressure cooker in 10 mmol/L citrate buffer (pH 6) for antigen retrieval. Sections were immersed in 0.3% H2O2 in methanol for 30 min at room temperature to block endogenous peroxidase and were pre-incubated with normal mouse or rabbit serum for 30 min at room temperature to block nonspecific binding of antibodies. The sections were then incubated with primary antibody for 1 h at room temperature. The avidin-biotin-peroxidase-complex method (Vectastain Elite ABC kit, Vector Labs., CA, USA) was used for detection, with 3,3′-diaminobenzidine tetrachloride as the chromogen. Nuclei were counterstained with methyl green. The primary antibodies used were mouse antihuman Bcl-2 antiserum (1 : 100; Calbiochem, Massachusetts, USA), rabbit antihuman Bax antiserum (1 : 5000; Chemicon International, USA), rabbit antihuman ICE antiserum (1 : 40; Calbiochem), and mouse antihuman CPP32 antiserum (pre-diluted; Immunotech, Marseille, France). As a negative control for immunostaining, the second antibody was used without the first antibody for each protein.

Immunostained cells were identified by light microscopy at a total magnification of ×400. For one index of staining the number of stained cells was expressed relative to the total number of germ cells in the seminiferous tubules (IALL). For another index, the staining of basal cell (spermatogonia) was expressed relative to the total number of spermatogonia (ISP); both indices were expressed as percentages and all samples analysed while unaware of origin.

The results are expressed as the mean (sd), with the significance of differences between groups determined by anova and P < 0.05 considered to indicate statistical significance.


Figure 1 shows representative immunohistochemically stained sections from a patient with varicocele and from a normal control. No Bcl-2 staining was present in germ cells either infertile or normal men. Weak immunostaining for Bax was detected in the cytoplasm of germ cell at all stages of maturation. ICE was expressed mainly in spermatogonia and only rarely in spermatids, and showed a perinuclear distribution. CPP32 was expressed mainly in spermatogonia and only rarely in spermatids; expression was both nuclear and cytoplasmic. Sertoli cells expressed Bax while interstitial cells, including Leydig cells, expressed Bcl-2, Bax, ICE, and CPP32 (Fig. 1). In negative controls for immunostaining, there was no detectable expression for Bcl-2, Bax, ICE, and CPP32.

Figure 1.

Immunostaining of Bax (a,b), ICE (c,d) and CPP32 (e,f), with the normal and varicocele testes on the left and right, respectively. All ×200.

The staining indices for controls and patients with varicocele are shown in Table 1. Both testes of patients with varicocele showed a significantly lower IALL for CPP32 than that in normal subjects (P < 0.001), while the difference was not significant for Bax and ICE. As the expression of ICE and CPP32 was mainly restricted to spermatogonia, the ISP was calculated for ICE and CPP32 but not for Bax (Table 1). The ISP for both ICE and CPP32 was significantly lower in patients with varicocele than in normal subjects (P = 0.0012 and < 0.001, respectively). Neither IALL nor ISP differed between right and left testes of varicocele patients. Additionally, the IALL and ISP of ICE was significantly positively correlated with those for CPP32 (P = 0.0116 and 0.0058, respectively).

Table 1.  Staining indices, as IALL and ISP for Bax, ICE and CPP32 in controls and in infertile men with varicocele
Mean (sd)
index, %
 Bax15.5 (4.3)13.8 (7.1)12.9 (5.6)
 ICE27.4 (5.1)20.2 (8.2)18.7 (11)
 CPP3234.9 (6.1)25.2 (7.1)24.8 (5.8)
 ICE40.3 (6.7)23.9 (10.9)23.0 (14.6)
 CPP3264.2 (9.4)43.0 (12.2)45.4 (10.8)

The relationship between IALL or ISP and clinical grade of varicocele was also assessed. For CPP32, the two indices tended to be higher in grade I than in grade II or III varicocele but there was no significant difference between the grade of varicocele and either index for Bax or ICE. When analysed for Bax, ICE or CPP32 and serum concentrations of LH, FSH, testosterone or oestradiol, there was a significant positive correlation only between IALL for CPP32 and LH (P = 0.0457) (Fig. 2).

Figure 2.

The correlation between CPP32 IALL and LH; *P< 0.05.


Germ cell deletion during normal spermatogenesis has been estimated to result in the loss of 25–75% of potential mature sperm cells in the adult testis [2–4]. Recent studies have characterized this germ cell death as apoptosis [5,6]. We previously reported that apoptosis was decreased in testes of men affected by varicocele [16]. The present results show that ICE and CPP32 are expressed in germ cells of human testes, especially in basal cells such as spermatogonia, with expression only rarely in spermatids. In both testes of patients with varicocele, both indices were significantly lower for CPP32 and the ISP significantly lower for ICE than in specimens from normal subjects. No gene expression was different in the right or left testes. Both indices for ICE were positively correlated those for CPP32. Caspase families participate in a cascade; upstream caspases cleave and activate downstream caspases [13]. The present results suggest that there is a connection between ICE and CPP32 in the testes of infertile men with varicocele. Furthermore, apoptosis was positively correlated with both indices for CPP32; thus CPP32 may be important in regulating apoptosis of testicular germ cells. In testes of infertile men with varicocele, apoptosis may be suppressed as a result of decreased CPP32 expression. Krajewski et al.[20] considered CPP32 to have the best correlation with apoptosis, among identified caspases. CPP32 is activated when cells receive signals to undergo apoptosis in response to serum withdrawal [21], activation of Fas [22], treatment with ionizing radiation [23], and a variety of pharmacological agents [21]. Recently, Bcl-2 has been shown to prevent the processing and activation of CPP32 [24].

Bcl-2 was not expressed in the present human testicular germ cells, while these cells expressed Bax at low levels, with no significant difference between the testes of patients with varicocele and specimens from fertile men. Lin et al.[25] also reported that Bax and Bcl-2 were not detected immunohistochemically in the seminiferous tubules. Therefore, the mitochondrial pathway involving Bcl-2 and Bax may not be involved in the regulation of apoptosis in germ cells; other systems, e.g. the Fas-Fas ligand system, may induce CPP32 expression which has been described in the testis [26,27]. Fas is a transmembrane receptor protein capable of initiating apoptosis when stimulated by receptor cross-linking or binding to its ligand (FasL) [26]. In rat testis, Lee et al.[27] localized FasL to Sertoli cells and Fas to germ cells, suggesting that germ cell apoptosis could be mediated by Fas expression. As CPP32 is activated in Fas-mediated apoptosis [22], the Fas system may regulate activation of CPP32 and apoptosis in the testes. However, while Fas expression was detected in germ cells, we could not detect FasL in the human testis (data not shown).

Optimal spermatogenesis requires the stimulatory action of both FSH and LH [28]. LH-mediated testosterone secretion from Leydig cells is considered the primary stimulus for spermatogenesis [29,30]. Some recent studies have described the hormonal regulation of apoptosis during spermatogenesis. Billig et al.[6] reported that increases in germ cell apoptosis after treatment of male rats aged 16–32 days with a GnRH antagonist correlated with decreases in serum FSH, but not LH concentrations. Therefore FSH was important in controlling apoptosis in rats at this developmental stage. Tapanainen et al.[5] found that when gonadotrophin removal by hypophysectomy in immature rats induced apoptotic cell death in the testis, the effect could be suppressed by treatment with FSH, LH/hCG, or testosterone. Troiano et al.[7] reported that testosterone withdrawal in the adult rat induces apoptotic cell death during spermatogenesis. Therefore, gonadotrophins and testosterone are essential for preventing testicular cell death in both seminiferous tubules and interstitial cells [5]. In the present study there was a positive correlation between CPP32 staining in germ cells and LH, suggesting that LH is likely to regulate CPP32 expression.

In conclusion, CPP32 may be important in regulating apoptosis in the testes of infertile men with varicocele; in turn, CPP32 expression may be controlled by factors such as LH.


H. Tanaka, Instructor.

M. Fujisawa, Assistant Professor.

H. Tanaka, Instructor.

H. Okada, Assistant Professor.

S. Kamidono, Professor and Chairman.