The Apoptotic Function Analysis of p53, Apaf1, Caspase3 and Caspase7 during the Spermatogenesis of the Chinese Fire-Bellied Newt Cynops orientalis

Background Spontaneous and stress-induced germ cell apoptosis during spermatogenesis of multicellular organisms have been investigated broadly in mammals. Spermatogenetic process in urodele amphibians was essentially like that in mammals in spite of morphological differences; however, the mechanism of germ cell apoptosis in urodele amphibians remains unknown. The Chinese fire-belly newt, Cynops orientalis, was an excellent organism for studying germ cell apoptosis due to its sensitiveness to temperature, strong endurance of starvation, and sensitive skin to heavy metal exposure. Methodology/Principal Findings TUNEL result showed that spontaneous germ cell apoptosis took place in normal newt, and severe stress-induced apoptosis occurred to spermatids and sperm in response to heat shock (40°C 2 h), cold exposure(4°C 12 h), cadmium exposure(Cd 36 h), and starvation stress. Quantitative reverse transcription polymerase chain reactions (qRT-PCR) showed that gene expression of Caspase3 or Caspase7 was obviously elevated after stress treatment. Apaf1 was not altered at its gene expression level, and p53 was significantly decreased after various stress treatment. Caspase assay demonstrated that Caspase-3, -8,-9 enzyme activities in newt testis were significantly elevated after heat shock (40°C 2 h), cold exposure(4°C 12 h), and cadmium exposure(Cd 36 h), while Caspase3 and Caspase8 activities were increased with Caspase9 significantly decreased after starvation treatment. Conclusions/Significance Severe germ cell apoptosis triggered by heat shock, cold exposure, and cadmium exposure was Caspase3 dependent, which probably involved both extrinsic and intrinsic pathways. Apaf1 may be involved in this process without elevating its gene expression. But starvation-induced germ cell apoptosis was likely mainly through extrinsic pathway. p53 was probably not responsible for stress-induced germ cell apoptosis in newt testis. The intriguing high occurrence of spermatid and sperm apoptosis probably resulted from the sperm morphology and unique reproduction policy of Chinese fire-belly newt, Cynops orientalis.


Introduction
Apoptosis is evolutionarily conserved, genetically driven and energy-dependent programmed cell death, which plays pivotal roles in development and homeostasis of multicellular organisms [1,2]. A large body of evidence proved the great importance of apoptosis for spermatogenesis in vertebrates and invertebrates [3,4,5,6,7,8]. Spermatogenesis is a well-organized complex process, in which diploid spermatogonia proliferate and differentiate into terminally haploid mature functional sperm. However, during normal spermatogenesis not all spermatogonia in the testis can undergo spermiogenesis and become mature sperm, for most of them are eliminated through spontaneous germ cell apoptosis [7,8,9]. Notably, germ cell apoptosis can be induced by various environmental or physiological stress, such as heat shock, UV irradiation, toxicant exposure, hormonal withdrawal and oxidative stress [3,10,11,12,13,14,15,16]. Both spontaneous apoptosis and stress-induced apoptosis are indispensible for healthy germ cell development and functional sperm maturation [17]. Above all, apoptosis maintains cellular homeostasis and a suitable ratio of Sertoli cells to germ cell. Furthermore, apoptosis serves to eliminate potentially defective germ cells and ensure high-quality gametes production [18,19,20]. Finally, apoptosis keeps germline genetic integrity by clearing off mutant cells and prevents transmission of mutations to progeny and offspring [21,22,23]. Two different mechanisms have been elucidated for cell apoptosis: the extrinsic receptor-mediated pathway and the intrinsic mitochondria-dependent pathway. The extrinsic pathway was initiated with the extracellular death ligands, like Fas ligand (FasL), binding to their corresponding cell-surface receptors such as Fas. Then a death-inducible signaling complex (DISC) formed, leading to the activation of caspase8. Eventually, the activated caspase8 triggered activation of downstream effector caspases: caspase3, 6 and 7 that were directly responsible for cell demise [24,25,26]. By contrast, the intrinsic pathway was mainly regulated by Bcl-2 family members including Bax, Bak, Bcl-2, Bcl-xl and so on [21], which could positively or negatively regulate mitochondrial outer membrane permeabilization to promote the release of cytC (cytochrome C) and other apoptotic molecules. In cytosol, cytC, Apaf1 and procaspase9 were assembled into apoptosome to activate downstream effector caspase3, 6 and 7, leading to cellular components degradation and cell death [25]. Both the extrinsic and intrinsic pathways proved to participate in germ cell apoptosis during spermatogenesis. On the one hand, Baxmediated apoptosis was indispensable for spermatogenesis [21]. Bax was elucidated to be essential for the progression of the first wave of spermatogenesis in mice [27], and it was found highly expressed in testis of young mice but only expressed at low level in spermatogonia of adult mice [28,29]. The Bax-null male mice were infertile as a result of disordered germ cell apoptosis, which caused excessive numbers of spermatogonia and preleptotene spermatocytes in spermatogenesis [13,21,30]. On the other hand, Fas and FasL were also evidenced essential for germ cell apoptosis during spermatogenesis. Fas was expressed in germ cells and FasL detected in Sertoli cells [31,32,33,34]. Testicular Fas participated in germ cell apoptosis triggered by testosterone withdrawal [35], UV radiation [36] and toxicants exposure [32,36].
Both the two distinct caspase-mediated apoptosis pathways could be regulated by p53. p53 was the most prominent tumor suppressor, and many types of cancers resulted from functionally impaired p53 [37,38] that lost its capability in DNA repairing and cell apoptosis induction. It was demonstrated that p53 played critical functions in germ cell apoptosis both through extrinsic pathway and intrinsic pathway, since it could regulate not only Bax, Bid, Bcl-2, Bcl-xL, Puma, Noxa, Apaf1 and IAP, but also death receptors like CD95, Fas, Apo-1 and DR5 [39,40,41,42,43,44]. p53-mediated apoptosis was found responsible for the initial phase of germ cell apoptosis during spermatogenesis in cryptorchid mice [31,45,46]. Moreover, under genotoxic stress, p53 was involved in apoptosis of spermatogonia, spermatocytes and spermatids [39,40,41]. A higher percentage of morphologically abnormal spermatozoa in the ejaculate and reduced fertility occurred to p532/2 mice, which was probably due to failed germ cell apoptosis [45].
Previous researches on germ cell apoptosis mainly focused on mammals for an elaborate and comprehensive knowledge of mammalian spermatogenesis. Actually, the spermatogenetic process in urodele amphibians is essentially like that in mammals in spite of morphological differences. During the spermatogenesis of Cynops orientalis, spermatogonia underwent several cycles of cell division to become spermatid, and later in spermiogenesis, the spermatid eventually matured into sperms. Mature sperm commonly contained typical acrosome, long and narrow nucleus and long flagellum, and a large number of mitochondria were arranged around the midpiece of the sperm flagellum. Like other newt species, the testis of Cynops orientalis consists of lobules in successive zones, and spermatogenesis proceeds synchronously in these zones. Therefore, it is easy to isolate and identify germ cells of particular spermatogenic stage, which is very suitable for germ cell apoptosis analysis (unpublished data). As an important urodele amphibian, the Chinese fire-belly newts, Cynops orientalis, is an excellent organism for studying germ cell apoptosis due to its advantageous physiological features: sensitiveness to temperature, strong endurance of starvation and sensitive skin to environmental heavy metal. Coincidentally, temperature, starvation and heavy metals are often utilized to investigate germ cell apoptosis. Besides, germ cell apoptosis have been reported in Japanese fire-belly newts and marbled newts, and some unidentified caspase was reported involved in germ cell apoptosis in Japanese fire-belly newts [47,48,49,50,51,52,53,54]. However, the detailed mechanism of germ cell apoptosis was not elucidated. In the present study, we investigated the germ cell apoptosis induced by heavy metal exposure, starvation and temperature stress in Chinese fire-belly newts, Cynops orientalis. We provided evidence that stress-induced germ cell apoptosis occurred by activating Caspase3 both through extrinsic pathway and intrinsic pathway. In addition, Apaf1 probably was involved in the stress-induced germ cell apoptosis, whereas p53 was not likely responsible for this process, which needs further investigation. Adult Chinese fire-belly newts, Cynops orientalis, were collected from Fuyang Mountainous Area (Hangzhou, China). Animals were kept in big aquatic tank in laboratory (12 h light:12 h darkness at 2262uC. All the animals were anesthetized by placing them for 20 min in a solution of 0.1% MS222 (Sigma, St. Louis). Fresh samples of testis, liver, muscle, intestine and heart were dissected and stored at -80uC for total RNA extraction.

Animals and Sampling
Full-length cDNA Cloning of p53, Apaf1, Caspase3 and Caspase7 Total RNA from the testis of adult Chinese fire-belly newt, Cynops orientalis, was extracted using Trizol (Invitrogen, USA) according to the instruction book, and reverse transcription was performed using BioRT cDNA First Strand Synthesis Kit (Bioer, Japan). Degenerate primers were designed using CODEHOP program supplied by Blocks WWW Server (http://blocks.fhcrc. org/) based on conserved protein sequence, and gene specific primers designed by Primer 5.0 and Oligo 6. All degenerate primers and gene specific primers used for cloning p53, Apaf1, Caspase3 and Caspase7 were presented in Table 1 and Table 2. Polymerase chain reactions (PCR) were run in Mygene Series Peltier Thermal Cyclers (Hangzhou, China), and the PCR program for amplifying cDNA fragment was as follows: 94uC for 5 min, 15 cycles in a touch down program (94uC for 30 s, 55uC for 30 s and 72uC for 30 s, followed by a 0.5uC decrease of the annealing temperature per cycle), and then 32 cycles were performed (30 s at 94uC, 30 s at 48uC and 30 s at 72uC) with 10 min at 72uC for the final extension. The RACE (Rapid amplification of cDNA ends) PCR were run according to Smart RACE cDNA Amplification Kit (ClonTech) and 39Full RACE Amplification Kit (Takara). The target PCR products were cloned into pMD18-T vector (TaKaRa) and sequenced by Sangon Company (Shanghai, China).
Protein Alignment and Structure Prediction of p53, Apaf1, Caspase3 and Caspase7 The protein alignment and structure prediction were performed using Vector NTI10 (Invitrogen, USA) and online softwares (http://www.expasy.org/tools/scanprosite/) (http:// smart.embl -heidelberg.de/)(http://zhanglab.ccmb.med.umich. edu/I-TASSER/) [55,56,57]. The GenBank accession numbers of Apaf1 proteins in protein alignment are as follows: Cynops orientalis (JQ320086); Homo sapiens (O14727. Analysis of tissue distribution of p53, Apaf1, Caspase3 and Caspase7 mRNA Total RNA was extracted from muscle, heart, testis, intestine and liver using Trizol (Invitrogen, USA), and reverse transcription was performed using BioRT cDNA First Strand Synthesis Kit (Bioer, Japan). RT-PCR was used to analyze the gene expression level of p53, Apaf1, Caspase3 and Caspase7 in various tissues. The primers used for gene expression analysis were listed in Table 1 and Table 2, and the program was run as follows: 94uC for 5 min; 30 s at 94uC, 30 s at 60uC, and 30 s at 72uC (32 cycles); 10 min at 72uC for final extension. The b-actin served as the control, and the density of amplified bands were analyzed using Tanon GIS system (Tanon 2500).The gene expression levels of p53, Apaf1, Caspase7 and Caspase3 were presented as the ratio of their relative amount to b-actin transcripts.    Stress challenging the Testis of Chinese Fire-belly Newt,

Cynops orientalis
Chinese fire-belly newts were initially raised in their best conditions and then divided into several groups for stress treatments that included cadmium exposure, heat shock, cold exposure and starvation. For cadmium exposure, newts were injected subcutaneously with cadmium chloride (CdCl 2 ) dissolved in saline (0.65% NaCl) according to the dosage of 50 mg/kg body weight. The newts in control group were injected only with 0.65% NaCl. Then these newts were raised for 12 h, 24 h and 36 h respectively at room temperature (2262uC). The newts for cold exposure were kept in the incubator at 4uC (4 h and 12 h) while the newts for heat shock were exposed to 38uC (2 h, 4 h and 12 h) and 40uC (2 h and 4 h). The newts raised at room temperature served as the control. For starvation treatment, the newts were starved for one month, and those with daily feed served as the control. In each stress group, all newts were recovered in normal conditions for 4 h before their testes dissection for RNA extraction and TUNEL analysis.

Analysis of Gene Expression of p53, Apaf1, Caspase3 and Caspase7 after Stress Treatment
Total RNA was extracted from the testes of stress-treated newts and control newts with Trizol (Invitrogen, USA). Reverse transcription was conducted with Prime ScriptTM RT reagent kit (TakaRa, Japan) according to the instruction book. SYBR Green PCR mastermix (Takara, Japan) was used for quantitative RT-PCR, and 10 ml reaction mixture was added into Axygen 384well reaction plates sealed with optical adhesive covers (Axygen). The cDNA levels for p53, Apaf1, Caspase3 and Caspase7 were normalized to b-actin. The primers used for realtime PCR analysis were listed in Table 1 and Table 2. The ABI PRISM 7900HT instrument (Applied Biosystems, USA) was applied for quantitative RT-PCR with each sample prepared in triplicates. The program for quantitative PCR was run as follows: 95uC for 30 s, then 40 cycles were performed (5 s at 95uC and 31 s at 60uC), and then 95uC for 15 s, 60uC for 1 min, 95uC for 15 s.

TUNEL Assay
The testes of stress-treated newts and normal newts were dissected, and then wrapped in OCT compound (SAKURA, Japan) for section preparations with Micron HM525. TUNEL staining was performed on tissue sections using in situ cell death detection kit according to the instruction book (Beyotime, China). Briefly, the testis sections were firstly fixed using 4% paraformaldehyde (PFA) for one hour, and later washed three times with PBS. Then the sections were submerged in PBS with 0.1% Triton-X-100 in ice bath for 2 min. After further washes with PBS, the reaction was performed in the terminal deoxynucleotidyl transferase (TdT) buffer with fluorescein labeled dUTP, and the sample were incubated with reagent for one hour at 37uC with plastic membrane mounted to avoid reagent evaporation. Washed with PBS, the sections were incubated with mixture of DAPI and antifade reagent before coverglass mounting. The positive control was prepared by Positive TUNEL preparation kit (Beyotime, China), in which the sections were treated with DNase I for 30 min at 25uC and PBS wash before TUNEL labeling reaction. The TUNEL results were analyzed with a Zeiss laser scanning confocal microscope (LSM-710) (Carl Zeiss, USA).

Caspase Activity Assay
Caspase activities in testes of newts were analyzed using caspase activity assay kits, which were based on the capabilities of caspase-3, -8, -9 to cleave acetyl-Asp-Glu-Val-Asp p-nitroanilide (Ac-DEVD-pNA), acetyl-Ile-Glu-Thr-Asp p-nitroanilide (Ac-IETD-pNA) and acetyl-Leu-Glu -His-Asp p-nitroanilide (Ac-LEHD-pNA) respectively producing yellow formazan product p-nitroaniline (pNA). After heat shock (40uC,2 h), cold exposure (4uC, 12 h), cadmium exposure (36 h) and starvation (one month), newts testes were dissected and rinsed with cold PBS, and then homogenized in ice bath using lysis buffer, with 5 min further extraction in ice bath. The tissue lysates were centrifuged at 18,0006g for 10 min at 4uC. Caspase-3, -8, -9 activities in the supernatant was assayed using the kit (Beyotime, China), and the absorbance at 405 nm was determined to quantify the caspase activities. All the experiments were performed in triplicates, and the caspase activities in the testes of stress-treated newts were presented as percentage of enzyme activity to that in control newts.

Transmission Electronic Microscopy (TEM)
The newt testis was initially fixed with 2.5% glutaraldehyde and then post-fixed in 1% osmium tetroxide solution at room temperature. After pre-stained with uranyl acetate, the sample was dehydrated with acetone and ethanol before their bedding in Spurr's resin. The sections were prepared with a Leica Ultracut UC6 Ultramicrotome (Leica Microsystems AG, Wetzlar, Germany). After staining with uranyl acetate and lead citrate, the sections were observed under transmission electronic microscopy (PHILIPS TECNAI 10).

Statistical Analysis
All data obtained in real-time RT-PCR and caspase activity assay were analyzed using EXCEL and SPSS 9.0 software and presented as mean6standard error. Student's T-test was used to evaluate the significant difference of gene expression levels and caspase activities between experimental group and control group. It was considered to be statistically significant when P value ,0.05.
Caspase3 and caspase7. Multiple protein alignment of Caspase3 and Caspase7 in Cynops orientalis with their homolog showed that Caspase3 in Cynops orientalis have 56.2% identity with Danio rerio homolog, 54.4% with Xenopus laevis, 57.6% with Homo sapiens and 56.3% with Mus musculus. Caspase7 had 67.6% identity with Danio rerio homolog, 64.6% with Xenopus laevis, 62.6% with Homo sapiens and 62.3% with Mus musculus. His134 and Cys176 in Caspase3, and His149 and Cys191 in Caspase7 were predicted to be critical for catalytic activity and they were conserved compared with their homolog ( Figure 5). The putative cleavage sites at Asp39 and Asp188 in Caspase3, and Asp24 and Asp30 in Caspase7 were also highly conserved, at which Caspase3 and Caspase7 can be separated into functional large subunit and small subunit respectively ( Figure 5). The predicted subdomain structures of putative Caspase3 and Caspase7 in Cynops orientalis were very similar. They shared 50.6% identity, and both of them had conserved Asp cleavage sites and four loops that contributed to catalytic groove conformation. Additionally, the predicted 3-D Figure 1. Multiple alignment of Apaf1 in Cynops orientalis with its homolog from several vertebrates. The protein alignment was created using Vector NTI10 (Invitrogen) with the following sequences: Cynops orientalis(JQ320086); Homo sapiens (O14727.2); Mus musculus(NP_033814.2); Xenopus laevis;(NP_001085834.1); Danio rerio(NP_571683.1). The identities of Apaf1 in Cynops orientalis with its homolog were 61.4% (Homo sapiens), 55.7% (Xenopus laevis), 51.2% (Danio rerio) and 59.9% (Mus musculus) respectively. doi:10.1371/journal.pone.0039920.g001 structures of Caspase3 and Caspase7 also had high similarity ( Figure 6).
Tissue specific Gene Expression of p53, Apaf1, Caspase3 and Caspase7 RT-PCR result showed that p53, Apaf1, Caspase3 and Caspase7 were expressed in various examined tissues. In the testis, Caspase3 and Apaf1 were expressed at slightly higher level than p53 and Caspase7 (Figure 7).

Gene Expression of p53, Apaf1, Caspase3 and Caspase7 in the Testis after Stress Treatment
In the testes of starved newts, gene expression of p53 was slightly but significantly decreased while Apaf1 didn't showed significant alteration. Notably Caspase3 showed a more than 20fold significant increase in its gene expression whereas Caspase7 show no significant change (Figure 8 A and B). After cadmium exposure for 12 h, 24 h and 36 h, the expression of p53 showed slight but significant decrease while Apaf1 was slightly increased though without significance. Gene expression of Caspase3 and Caspase7 were decreased after cadmium exposure for 12 h and 24 h, but only the change of Caspase3 mRNA level was significant. After 36 h exposure to cadmium, Caspase3 gene expression was increased by about 8 fold though without significance, whereas no significant alteration of Caspase7 was detected ( Figure 8C and D).
After 4 h exposure to cold, gene expression of p53 in newt testis was significantly decreased whereas its expression was elevated after 12 h though without significance. Differently the gene expression of Apaf1 were decreased but only showed significant difference after cold exposure for 12 h. Gene expression level of Caspase3 was greatly increased after cold exposure though without statistical significance whereas Caspase7 didn't showed significant change after cold exposure ( Figure 9A and B). In heat shock group, expression of p53 showed significant decrease after exposure to 40uC, and no significant change was found after exposure to 38uC. Apaf1 was only significantly increased after exposure to 38uC for 2 h, without any significant changes in other groups. Caspase3 were significantly increased after heat shock at 40uC for 2 h and 38uC for 24 h without no significant alterations in other groups. Caspase7 was only significantly increased in the testes of newts exposed to 38uC for 24 h (Figure 9C and D).

TUNEL Analysis of Germ Cell Apoptosis in Newt Testis
TUNEL assay showed spontaneous germ cell apoptosis occurred in the testis of normal newts, and apoptotic early spermatids ( Figure 10A1-3) and mature sperms ( Figure 11A1-3) were detected. Compared with normal newts, TUNEL results showed severer germ cell apoptosis occurred to early spermatids and mature sperm in the testis of newts treated by starvation ( Figure 10B1-3, Figure 11B1

Ultrastructure of Newt Mature Sperm
The TEM results showed that the nucleus of newt mature sperm was long and narrow shaped (Figure 14 A). A large number of mitochondria were observed arranged surrounding the sperm nucleus (Figure 14 B).

p53, Apaf1, Caspase3 and Caspase7 were Conserved in Chinese Fire-belly Newt, Cynops Orientalis
We have cloned the full-length cDNA of p53, Apaf1, Caspase3 and Caspase7 from the testis of Chinese fire-belly newt, Cynops orientalis. p53 and Apaf1 protein share about 50% identity with their homolog, and they have conserved functional subdomains. The DNA binding domain of p53 is highly conserved and six key amino acids essential for binding capability are totally the same among the examined homolog [58]. Besides, the transactivation domain and profine-rich domain conferred p53 possible capability in gene expression regulation while tetramerization domain indicated p53 probably became functional in a tetramer form [58,59]. As for Apaf1, CARD is an essential functional domain because it is responsible for recruiting procaspase9 for apoptosome formation [24,25,60]. The NOD domain is involved in Apaf1 oligmerization into apoptosome [61], and the C-terminal 12 WD40 repeats contributed to cytC binding to initiate downstream cascade of caspases [24,25,61,62,63]. The 3-D structure of p53 and Apaf1 was quite similar to their counterparts in mammals [58,64]. Base on the highly conserved subdomain structures and  Caspase3 and Caspase7 showed 54%-57% and 62%-67% identity with their homolog respectively. As critical cleavage sites, Asp24/Asp203 (Caspase7) and Asp39/Asp188 (Caspase3) are conserved, revealing a similar cleavage mechanism. Besides, being key catalytic residues, the highly conserved His134/Cys176 (Caspase3) and His149/Cys191 (Caspase7) indicated an similar catalytic performance [65,66]. As effector caspases, Caspase3 and Caspase7 in Cynops orientalis have highly similar subdomain structure and 3D structures, which further demonstrated that their similar performance in cell apoptosis resulted from resembled structures [67].
RT-PCR results showed that p53, Apaf1, Caspase3 and Caspase7 were expressed in various examined tissues. In the testis, Caspase3 and Apaf1 were expressed at higher level than p53 and Caspase7, which indicated more active Caspase3 and Apaf1 in newt testis. Caspase3 and Caspase7 have been found expressed during normal spermatogenesis, and play important functions in spontaneous germ cell apoptosis [23]. p53 and Apaf1 were also proved expressed in testis, and they are critical for sperm development and male fertility [23,68,69]. Knockout or overexpression of p53 would result in abnormal spermatogenesis [22,23] while knockout of Apaf1 could induce male infertility [70].

Spontaneous Germ Cell Apoptosis Occurred in Newt Testis
During normal spermatogenesis, spontaneous germ cell apoptosis commonly occurred to spermatogonia and spermatocytes in many mammalian species including rats, hamsters, mice and humans [2,3,4,5,6,7,8,9,45,71,72,]. Since the spermatogonia and spermatocytes were active in DNA replication, which made them vulnerable to damage [21,45,73,74,75]. In some cases, apoptosis even occurred to spermatids or sperm, though the frequency is low [2,76]. In the present study, we only observed apoptotic spermatids and sperm without apoptotic spermatogonia and spermatocytes detected, whereas severe apoptotic spermatogonia were observed in positive control sections. The result indicated the intriguing germ cell apoptosis in newts was probably resulted from specificity of this species. The spontaneous germ cell apoptosis in newt testis may be responsible for clearing off abnormal spermatogenic cells and ensuring production of high quality sperm [27,28,30,77]. Figure 7. p53, apaf1, caspase3 and caspase7 gene expression in various tissues of Cynops orientalis. Total RNA was extracted from fresh testis, muscle, liver, intestine and heart, and then transcribed into cDNA for RT-PCR analysis. b-actin served as an internal control. p53, apaf1, caspase3 and caspase7 are expressed in all the examined tissues. In the newt testis, Apaf1 and Caspase3 were expressed at higher level than p53 and Caspase7. The intensity of the signal was analyzed by Tanon-Gis system and the data were processed with SPSS 9.0. The gene expression level in testis were presented as their percentage of b-actin. Significance of difference was assessed by SPSS9.0. P,0.05. n = 3. doi:10.1371/journal.pone.0039920.g007 Caspase3 and Caspase7 are Involved in Stress-induced Germ Cell Apoptosis When spermatogenesis was threatened by environmental or physiological stress, stress-induced germ cell apoptosis was often initiated to remove abnormal germ cells [78,79,80]. For instance, hormone deprivation, heat shock, low temperature, starvation and heavy mental exposure could aggravate germ cell apoptosis in testes [50,77,81,82]. In our research, negative effects were exerted by cadmium exposure (36 h), heat shock (40uC 2h), cold exposure (4uC 12h) and starvation (one month) on germ cell maturation in newt, leading to severer spermatid and sperm apoptosis. Obviously, the germ cell development in Cynops orientalis was vulnerable to stress, and germ cell apoptosis had been triggered by temperature stress, starvation and heavy metal exposure.
Caspase3, Caspase6 and Caspase7 were three important effector caspases that directly determined cell death in cell apoptosis signaling pathway [66,83]. Previous researches on newts have demonstrated that high caspase activity in toxicant-treated testes was responsible for chromatin condensation and apoptotic body formation [49]. In our study, after stress treatment, not only gene expressions of Caspase3 or Caspase7 in newt testis were obviously elevated, but also the Caspase3 enzyme activity in the testis was significantly increased. Therefore, the stress-induced germ cell apoptosis was dependent on both gene expression and enzyme activation of Caspase3.

Possible Mechanism for Starvation-induced Germ Cell Apoptosis
Nutrients are essential for male reproduction performance, for it can supply energy like glucose and growth factors that tightly correlates with cell metabolism, growth and proliferation [84,85]. Without sufficient growth factor and glucose, cells firstly reduced its metabolism and ceased proliferation, and further nutrients deprivation would initiate cell apoptosis [86,87,88,89,90,91,92,93]. In our study, after one-month food deprivation, we found severer germ cell apoptosis occurred in the testis of Cynops orientalis compared with control newt, which probably resulted from serious shortage of growth factor and glucose. It was demonstrated that cell apoptosis triggered by nutrients deprivation was caspase-dependent, and executed mainly by intrinsic mitochondria-dependent pathway [85,94,95,96]. In detail, nutrients deprivation could change cellular metabolism, and disrupted the overall balance between proapoptotic and antiapoptotic Bcl-2 family members. Then mitochondrial physiology would be altered to facilitate cytC release into the cytosol. Together with Apaf1 and procaspase9, cytC was assembled into the apoptosome to activate caspase9 that further activated effector caspases to induce cell death [93,97,98]. In addition, p53 was elucidated to be involved in starvation-induced cell apoptosis [99], and Fas-dependent extrinsic pathway proved critical for germ cell apoptosis after growth factor withdrawal [100]. However, in our study, after one month starvation, either the gene expression of Apaf1 and p53 or Caspase9 enzyme activity in the testis of Cynops orientalis was significantly decreased. Therefore intrinsic pathway was probably not involved in starvation-induced germ cell apoptosis. In contrast, Caspase8 enzyme activity in the testis was slightly elevated after starvation treatment, which indicated that the extrinsic pathway might participate in the starvation-induced germ cell apoptosis in Cynops orientalis.

Possible Mechanism for Cadmium-induced Germ Cell Apoptosis
Cadmium is one of the most toxicants in terrestrial and aquatic environment, and both cadmium pollution and cadmium injection into testis can induce testicular impairment, germ cell death, poor semen quality and even male infertility [80,101,102,103,104,105,106,107]. Cadmium triggered germ cell apoptosis by disrupting endocrine secretion and enhancing reactive oxygen species (ROS) production. On the one hand, cadmium is an endocrine disruptor that can affect functions of Leydig cell and inhibit testicular testosterone secretion [4,35,108,109,110,111,112,113], both of which were essential for spermatogenesis [2,114,115]. Once the testosterone was inhibited by cadmium, massive testicular germ cell apoptosis would occur [101,106]. Fas and FasL have been elucidated to play pivotal roles in germ cell apoptosis after testosterone  [32,35,36,77,80,103,116,117]. In this study, the enzyme activity of Caspase8 in the testis was elevated by 100%, and it demonstrated that the cadmium exposure probably disrupted the hormone secretion and the extrinsic cell death pathway was initiated. On the other hand, cadmium could induce germ cell apoptosis by enhancing generation of ROS like superoxide ion, hydroxyl radicals and hydrogen peroxide, and these oxidative stress could result in peroxidation, mitochondrial dysfunction or DNA damage in germ cell and oxidative damage in Leydig cells [80,118,119,120,121,122,123,124]. Oxidative stress could induce p53 mediated cell apoptosis in the intrinsic mitochondriadependent pathway [125,126,127,128]. The Caspase9 enzyme activity in the testis was raised by 80%, and probably the mitochondria-dependent intrinsic pathway was also involved in the cadmium-induced germ cell apoptosis. As the key participant of intrinsic cell death pathway, Apaf1 probably participate in the cell apoptosis in spite of no elevation in its gene expression.
Under oxidative stress, p53 could be activated to promote reduced glutathione generation and sestrins expression, thus protecting DNA from damage by decreasing ROS [129]. But we found that gene expression of p53 was reduced after cadmium exposure, and previous study also showed cadmium suppressed p53 gene expression in the testes and liver [101,130]. So we deduced that ROS induced by cadmium may result in devastating DNA damage, which directly lead to cell apoptosis in a p53independent pathway.

Possible Mechanism for Heat-and Cold-induced Germ Cell Apoptosis
Suitable environmental temperature is important for male reproduction. Heat shock can induce germ cell death and even male infertility while cold temperature leads to abnormal sperm production [53]. In adult rat testis, heat shock (43uC for 15 min) caused germinal epithelium damage, abnormal Sertoli cell and germ cell death [11,131,132] that mainly occurred to meiotic spermatocytes and early spermatids [11]. The intrinsic mitochondria-dependent pathway proved critical for heat-induced germ cell apoptosis, in which Bcl-2 family members played important functions [77,81,133,134,135]. Bax induced the cyt C release from the mitochondria to initiate the activation of Caspase9, which eventually activated downstream effector caspases (caspase3, 6 and 7) leading to cell death [81,136]. Moreover, Fas-dependent extrinsic pathway was also involved in heat-induced testicular germ cell apoptosis in cryptorchid testis [31]. We detected that enzyme activities of Caspase8 and Caspase9 in newt testis were significantly enhanced by 139% and 96% respectively after heat shock, indicating the involvement of extrinsic and intrinsic pathway in the germ cell apoptosis. Though without significant alteration, Apaf1 also should be involved in heat-induced germ cell apoptosis regarding its key status in intrinsic pathway. Though p53-dependent apoptosis was reported to be responsible for the initial phase of germ cell loss in experimental cryptorchid testis [31], p53 gene expression was significantly decreased in our study, indicating it probably didn't participate in this process.  Low temperature could trigger germ cell apoptosis mainly by reducing FSH secretion, which was critical for male reproduction especially in seasonally breeding animals [53]. FSH not only supported proliferation and maturation of Sertoli cells [137], but also maintained spermatogonia viability and proliferation [52,138]. Additionally, FSH could inhibit PRL induced spermatogonia cell death both in vivo and in vitro [51,53]. In newts, spermatogonia apoptosis occurred after cold exposure for one week (12u and 8uC) whereas no degeneration was found in newts kept at 22uC. Notably, cold triggered germ cell death was significantly suppressed by FSH at 12uC and 22uC but not at 8uC [50,53], indicating the irremediable germ cell apoptosis induced by lower temperature. Therefore severe germ cell apoptosis triggered by cold exposure (4uC) in our research to a big extent resulted from significant decrease in FSH secretion. It was elucidated that FSH can determine cell destiny by regulating several important proteins, including Survivin, PDCD6 and DR5, all of which were key members in intrinsic pathway and extrinsic pathway in cell apoptosis. High level of FSH could upregulate the expression of Survivin and suppress the expression of PDCD6 and DR5 [139], thus to promote cancer development and inhibit cell apoptosis. When FSH level in the testis was decreased after cold exposure, the expression of Survivin was probably reduced while the inhibition of PDCD6 and DR5 was terminated. In this way, both extrinsic and intrinsic pathway could participate in coldinduced germ cell apoptosis, which was evidenced by the significantly elevated enzyme activities of Caspase8 (150%) and Caspase9 (50%) in our research. Moreover increased gene Figure 13. Caspase-3, -8, -9 activities in the testis of newts treated with various stress. The Caspase activities in the testis of newts were analyzed after stress treatment including starvation (one month), cadmium injection (36 h), cold exposure (4uC, 12 h) and heat exposure (40uC, 2 h). Caspase-3, -8, -9 activity was measured by colorimetric assays based on caspase-3, -8, -9 to change Ac-DEVD-pNA, Ac-IETD-pNA and Ac-LEHD-pNA into a yellow formazan product (pNA), respectively. The caspase activities in stress exposure groups were presented as the percentage of enzyme activity in the control group. Value represents the Mean6SD, and T-test in SPSS9.0 applied to analyze the significance of differences between stress treated group and control group. * P,0.05 when compared with control. n = 3. doi:10.1371/journal.pone.0039920.g013

Possible Mechanism for Intriguing Sperm Apoptosis
The most intriguing phenomenon in our study is that germ cell apoptosis commonly occurred to spermatids and mature sperm, but not to spermatogonia and spermatocytes. Differently, other researches on newts only found apoptotic spermatogonia, spermtocytes and only abnormal mature sperm [47,48,49]. In previous researches, apoptotic mitotic diploid spermatogonia and meiotic primary spermatocytes were most common, and then was the apoptotic spermatids, either in spontaneous germ cell death or stress-induced cell apoptosis [3,45,140,141,142,143]. Only in recent years, more and more researches reported the occurrence of mature sperm apoptosis [76,144,145]. Several key apoptotic proteins were found in mature spermatozoa. For instance, in ejaculated mature sperm, functional p53 was localized in 43% of semen samples, and PARP in 75-100% of the sperm [144,146]. Moreover, active caspase1, caspase3 and caspase8 were found predominantly in the postacrosomal region of live sperm, and active caspase9 was particularly localized in the midpiece [76]. In human sperm, procaspases 8 and 9 and their active forms, together with caspase3, were also observed [146]. Furthermore, sperm fractions with low motility exhibit higher levels of active caspase3 compared with high-motility fractions [147]. Active caspase3 found in the sperm midpiece were closely associated with DNA fragmentation in low motility semen specimen, suggesting that caspase-dependent apoptotic mechanisms could originate in the cytoplasmic droplet or within mitochondria and function in the nucleus [148]. The activation of caspases in low-motility sperm samples may be attributed to the role played by cytC/Apaf-1 complex, which can activate caspase9 that was followed by activation of downstream caspase3, caspase6 and caspase7 [145].
We hypothesized that several physiological features of Chinese fire-belly newts might account for why mature sperm were easily involved in cell apoptosis. Above all, newt mature sperm had a long and narrow shaped nucleus which may be quite vulnerable to environmental stress (Figure 14 A). Moreover, similar to the mitochondria in midpiece of mammalian sperm, a large number of mitochondria were arranged surrounding the nucleus of newt sperm (Figure 14 B), which provided a possible platform for mitochondria-dependent cell death. Finally, it probably resulted from the specificity of this species. Without stringent controlling of early stage spermatogenic cell proliferation, the newts can firstly ensure high production of spermatids and sperm, and at the late stage of spermatogenesis, the quality of spermatids and sperm could be strictly controlled by removing abnormal spermatids and sperm via germ cell apoptosis.

Conclusion
From the results we can get several conclusions 1) p53, Apaf1, Caspase3 and Caspase7 are conserved in the testis of Cynops orientalis, and Caspase3, Caspase7 and Apaf1 were probably involved in the stress-triggered germ cell death while the roles played by p53 needs further investigation; 2) Spontaneous germ cell apoptosis occurred during spermatogenesis; 3) cadmium exposure, cold exposure and heat shock triggered severer apoptosis of spermatids and sperm probably both via extrinsic pathway and intrinsic pathway while starvation induced germ cell apoptosis possibly mainly via the intrinsic pathway ( Figure 15). Hypothesized sperm features and unique reproduction policy responsible for high occurrence of spermatid and sperm apoptosis. The nuclear morphology and mitochondria arrangement may be the physiological basis for observed high frequency of spermatid and sperm apoptosis, and a unique reproduction policy of newts may also account for this phenomenon. doi:10.1371/journal.pone.0039920.g015