Apoptotic sperm biomarkers and the correlation between conventional sperm parameters and clinical characteristics

The principal aim of this retrospective study was to examine the relationship between sperm apoptotic biomarkers and the patient's biclinical characteristics, the conventional sperm parameters and the results of assisted reproductive technology. Sperm analysis, activated caspases, annexin V staining for phosphatidylserine (PS) externalisation and labelling assay for DNA fragmentation were assessed in 122 males of infertile couples. Fifty‐seven couples were allocated to the natural conception group, and 65 couples underwent IVF or ICSI. Semen of IVF/ICSI patients showed a higher proportion of apoptotic spermatozoa in their spermatozoa when compared with a natural conception group (p < .05). Sperm apoptotic biomarkers correlated with age, FSH, and conventional sperm parameters. DNA fragmentation correlated positively with the percentage of semen having externalised PS (r = .78, p = 0) and activated caspases (r = .71, p = 0). Patients without clinical pregnancy had higher frequency of DNA fragmentation, externalised PS and activated caspases compared to patients with clinical pregnancy (p < .001). The best specificity and greater sensitivity were obtained with the test of the DNA fragmentation compared to the other biomarkers. Among the apoptotic biomarkers, only DNA fragmentation was found to predict natural or assisted pregnancy better than conventional sperm parameters.


Summary
The principal aim of this retrospective study was to examine the relationship between sperm apoptotic biomarkers and the patient's biclinical characteristics, the conventional sperm parameters and the results of assisted reproductive technology. Sperm analysis, activated caspases, annexin V staining for phosphatidylserine (PS) externalisation and labelling assay for DNA fragmentation were assessed in 122 males of infertile couples. Fifty-seven couples were allocated to the natural conception group, and 65 couples underwent IVF or ICSI. Semen of IVF/ICSI patients showed a higher proportion of apoptotic spermatozoa in their spermatozoa when compared with a natural conception group (p < .05). Sperm apoptotic biomarkers correlated with age, FSH, and conventional sperm parameters. DNA fragmentation correlated positively with the percentage of semen having externalised PS (r = .78, p = 0) and activated caspases (r = .71, p = 0). Patients without clinical pregnancy had higher frequency of DNA fragmentation, externalised PS and activated caspases compared to patients with clinical pregnancy (p < .001). The best specificity and greater sensitivity were obtained with the test of the DNA fragmentation compared to the other biomarkers. Among the apoptotic biomarkers, only DNA fragmentation was found to predict natural or assisted pregnancy better than conventional sperm parameters.

K E Y W O R D S
apoptotic sperm biomarkers, conventional sperm parameters, IUI, patient clinical parameters, spontaneous pregnancy

| INTRODUCTION
Apoptosis is a complex process in which cell death signalling and regulatory pathways have been described exhaustively in somatic cells (Horvitz, 1999). Ultrastructural changes and expressions of typical markers of apoptosis have been described in human ejaculated spermatozoa (Baccetti, Collodel, & Piomboni, 1996). Although the germ cell apoptosis is necessary for the establishment and maintenance of normal spermatogenesis in the testes, the existence of apoptosis, which is an "active" cell death in spermatozoa, remains controversial (Lachaud, Tesarik, Canadas, & Mendoza, 2004). In effect, spermatozoa are highly differentiated cells which exhibit little transcriptional activity. However, many studies have shown a higher proportion of spermatozoa expressing apoptosis markers in the ejaculation of infertile patients, compared to fertile men (Said, Paasch, Glander, & Agarwal, 2004;Weng et al., 2002). This result is a testicular apoptosis initiated and aborted (Sakkas, Mariethoz, & St John, 1999), an apoptosis initiated and/or continued in the male genital tract (Grunewald et al., 2005). The meaning of the expression of markers of This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited. apoptosis in these cells is not clearly defined today. In addition, methods for measuring markers of apoptosis in somatic cells, particularly of activated caspases, were transposed human spermatozoa without prior approval. However, for sperm samples, the realisation of these measures shows some challenges such as the low amount of spermatozoa available, particularly in cases of oligozoospermia, and heterogeneous cell populations. To evaluate the functional quality of the sperm, we sought the relationship between the expression of these markers, sperm parameters and the results in assisted reproductive technology. The processes of apoptosis and stigma were assessed by measuring the spermatozoa of several biochemical markers of apoptosis (activated caspases, externalisation of PS and DNA fragmentation in order to define the best marker for apoptosis and to optimise the management of infertile patients especially in assisted reproductive technology.

| Patients
We studied 122 male subjects who underwent seminal fluid evaluation at the Laboratory of cytogenetics, molecular genetic and human reproductive biology CHU FarhatHached Sousse Tunisia. All subjects were the partners of women who had failed to conceive after 1 year of unprotected regular sexual intercourse. Patients' information remained confidential and within the institution, and a written consent for the treatment was obtained from all patients. This study was conducted according to guidelines established for research on human subjects (Ethical committee, CHU FarhatHached Sousse).

| Study selection and inclusion criteria
The subjects were all nonsmokers, not using any medication and abstained from alcohol. The patients were ascertained to be in good health by means of their medical histories and a clinical examination including routine laboratory test and screening. Men with infection, varicocele upon physical examination or azoospermia were excluded.
Women with an indication for IVF, such as tubal infertility, severe male factor infertility, severe endometriosis, unexplained infertility, immune infertility and anovulatory infertility, were also excluded. The inclusion criteria for the female partner were as follows: female age <40 years and female body mass index (BMI) <30 kg/m 2 .
Individuals were divided into two groups: • The normal semen group in which females were able to conceive naturally or after IUI constituted the "Natural conception group" (n = 57).
• The abnormal semen group (IVF/ICSI group) consisted of 65 couples: the cause of infertility was oligoasthenoteratozoospermia.

| Collection of semen samples
The samples from 122 subjects were collected by masturbation into sterile plastic jars, after 3 days of sexual abstinence. The basic semen parameters were evaluated according to the World Health Organization guidelines, a sperm concentration of >15 × 10 6 /ml, >4% morphologically normal cells, >40% motile spermatozoa (categories "a", "b", and "c"), and >58% alive spermatozoa (World Health Organisation, 2010).
As specified in the kit, the FAM-VAD-FMK inhibitor was dissolved in dimethyl sulphoxide (DMSO) to obtain a 150× stock solution. Aliquots of this solution were stored at ≤−20°C in the dark for 6 months protected from light. Before use, a 30× FLICA working solution was prepared by diluting the stock solution 1:5 in 200 μl PBS. For labelling 3 × 10 5 spermatozoa (dilute it to a total volume of 300 μl of PBS), it was incubated with 10 μl of 30× working solution for 1 hr at 37°C under 5% CO2 with tubes protected from light. Cells were washed twice with 1× wash buffer (PBS containing 0.5% bovine serum albumin and 0.05% sodium azide). Cell pellet was resuspended in 400 μl 1× wash buffer and incubated in 2 μl of propidium iodide (PI) for 15 min, protected from light at RT in order to assess viability.
Finally, the nuclei of spermatozoa were stained with 2 μl of Hoechst 33,342 (Sigma) for 5 min, followed by washing and resuspension of cells in wash buffer.
Cells were observed under a fluorescence microscope using a band-pass filter (excitation 490 nm, emission >610 nm) to view green fluorescence.
A minimum of four hundred spermatozoa were randomly assessed per slide in at least five fields. Four patterns of fluorescence were measured: 1. Viable spermatozoa stained only with Hoechst, without staining of activated polycaspases: Casp−/PI−) (blue).

| Annexin V binding assay
Translocation of PS to the outer leaflet of the plasma membrane was detected by annexin V-fluorescein isothiocyanate (FITC; Sigma, France).
PS is normally found on the cytoplasmic face of the plasma membrane, and translocation to the extracellular leaflet involves a plasma membrane alteration. As the nuclear and plasma membranes of dead cells are also damaged, the assay includes staining with the DNA stain propidium iodide (PI; Sigma) to assess viability. As cryopreservation is deleterious to the plasma membrane, this technique was only applied on fresh sperm. By fluorescence microscopy, annexin V-FITC was observed as green fluorescence and PI as red. For this assay, 1 μl of annexin V-FITC (Sigma, France) was freshly mixed with 0.5 μl PI in 100 μl buffer solution (10 mmol/L HEPES, 150 mmol/L NaCl, 5 mmol/L KCl, 1 mmol/L MgCl 2 , 1.8 mmol/L CaCl 2 ) according to the instructions provided by the manufacturer (Sigma). An equal volume of this preparation was gently added to the sperm aliquot, and the mixture was incubated in the dark, at room temperature, for 15 min. Analysis of the sample was performed by conventional epifluorescent microscopy (emission: 525-617 nm). At least 200 spermatozoa were randomly assessed per slide in a minimum of five fields. As mentioned in the literature, the intra-observer and interobserver variabilities were <6% for this technique (Barroso, Morshedi, & Oehninger, 2000). Four patterns of fluorescence were observed: (i) viable spermatozoa with externalised PS (green) that stained only with annexin V-FITC (AN+PI−) = apoptosis-live; (ii) dead spermatozoa without externalised PS (red) that stained only with PI (AN−PI+) = intact-dead; (iii) dead spermatozoa with externalised PS (red and green) that stained both with annexin V-FITC and PI (AN+PI+) = apoptosis-dead; (iv) intact spermatozoa without any staining (AN−PI−) = intact-live.

| TUNEL reaction
DNA fragmentation in spermatozoa was detected by the "in-situ cell death detection kit" (Roche, Belgium) according to the instructions of the manufacturer, as described by Stouffs et al. (2004). Ejaculated sperm cells were washed twice in 5 ml of phosphate-buffered saline (PBS) and once in 1 ml PBS, followed by at least one night and a maximum of 48-hr incubation at −20°C in acetic acid:methanol (1:3). The samples were spread on polylysine slides, and the cells were permeabilised in 0.1 mol/L sodium citrate for 30 min at 70°C, incubated in the TUNEL reaction mixture for 1 hr at 37°C after which the detection with converter-alkaline phosphatase and nitroblue tetrazolium-5-bromo-4chloro-3-idolyl phosphate (NBT-BCIP) (Roche) was performed. Slides were counterstained by methyl green and analysed by light microscopy (data analysis performed on at least 400 spermatozoa). For each test, positive (DNase: 2 UI/ml) and negative controls were included.
The degree of DNA fragmentation is measured with DNA fragmentation index (DFI).

| Ovarian stimulation and IUI or ICSI procedure
Women in IUI group received r-FSH and monitoring was performed with urine LH surge along with TVS until hCG administration.
Ovulatory trigger with urinary hCG 5,000 IU was administered intramuscularly when there was at least one follicle of 18 mm or more.
Semen was prepared for insemination by double-density gradient technique. A single IUI was performed between 44 and 48 hr after hCG injection.
A maximum of three cycles of IUI were performed.
Women in IVF/ICSI group underwent superovulation using a gonadotropin-releasing hormone analogue suppression protocol and recombinant gonadotropins (FSH per stimulated cycle varied between 1,800 and 3,500 UI depending on the individual response). As soon as at least three follicles of R18 mm were detected, ovulation was induced by 10,000 IU of human chorionic gonadotropin (hCG). Oocyte-cumulus complexes were recovered 36 hr after administration of hCG. After cumulus cell removal by treatment with 25 IU hyaluronidase (Origio, Berlin, Germany), followed by pipetting to remove the innermost layer of cells, the oocytes were then rinsed three times in flushing medium (Origio,10,845,060) and the morphology of the metaphase II oocytes was evaluated at the time of ICSI (from 2 to 4 hr after retrieval) under inverted microscope at ×400 magnification. Only metaphase II oocytes with normal and clear zona pellucida, small size of perivitelline space, without vacuoles or cytoplasmic granularity and normal polar body were injected with motile spermatozoa into the ooplasm according to previously described procedures (Bungum et al., 2004).
Around 18-20 hr post-ICSI, fertilisation was assessed by the presence of pro-nuclei. The fertilisation rate was calculated from the ratio of fertilised oocytes to the total number of viable injected metaphase II oocytes. Segmentation rate was calculated from the ratio of cleaved embryos to the number of fertilised oocytes. Embryo quality was assessed at day 3 post-oocyte retrieval. The nonfragmented embryos with eight regular blastomeres were considered the best quality and named type I embryos. On day 3, type I embryos were transferred after a nonselective quarter laser-assisted hatching technique, using a solid-state diode laser (ZILOS-tK; Hamilton Thorne Biosciences, Beverly, MA, USA) in G-2. Pregnancy rate was defined from positive blood HCG and ultrasonography findings showing at least one embryo with a foetal heart beat 5 weeks after transfer.

| Statistical analysis
Statistical analysis was performed using SPSS 19.0 software. All variables were initially tested to determine variance homogeneity and data normality. The values are expressed as mean ± mean ± SD (range) or percentage (%). The Student's t test was used for the comparison of percentages between IVF/ICSI group and natural conception group.
Pearson's correlation was performed to examine the relationship between sperm apoptotic markers and, respectively, male age, BMI (body mass index), FSH (follicle-stimulating hormone), conventional sperm parameters and the achievement of the natural outcome of the first IVF/ICSI attempt (fertilisation rate, segmentation rate, of type I embryos and pregnancy). A receiver-operating characteristic (ROC) curve was used to calculate the area under the ROC curve (AUC) to determine the optimal threshold for the prediction of pregnancy with the best sensitivity and specificity. Statistical significance was set at p < .05.

| Clinical characteristics and conventional sperm parameters
Bioclinical characteristics and conventional sperm parameters of the natural conception group (n = 57) and IVF/ICSI group (n = 65) are presented in Table 1. We found statistically significant differences between the natural conception group and the IVF/ICSI group with regard to both conventional sperm parameters (concentration, motility [a + b] and atypical forms) and FSH. However, clinical characteristics values did not appear to be significantly different between the two treatment groups.

| Comparison of apoptotic sperm biomarkers between natural conception group and IVF/ICSI group
Results of apoptotic marker analysis in semen of IVF/ICSI patients and natural conception group (control) are presented in Table 2  with each other. In fact, DNA fragmentation was significantly positively associated with, respectively, the percentage of dead and viable spermatozoa with externalised PS (r = .78, p = 0) and dead and viable spermatozoa with activated caspases (r = .71, p = 0) ( Table 4).

| Relationship between apoptotic biomarkers and the outcome of IUI and IVF/ICSI
Overall, the 57 couples underwent IUI; there were 12 pregnancies Correlation between conventional sperm parameters, apoptotic markers according to fertilisation, segmentation, type I embryos rate and pregnancy in the IVF/ICSI group is given in  with 100% sensitivity, 74.7% specificity (above the curve of CASP+ and annexin +), CASP+ >80% (CASP+; AUC = 0.72 with 87% sensitivity, 81% specificity) and AN+ >42% (AN+; AUC = 0.7 with 70% sensitivity and 60% specificity) were reliable predictors of pregnancy (p < .001). In addition, DNA fragmentation had the best predictive value in pregnancy when compared with the other apoptotic markers.

| DISCUSSION
As far as we know, our study is the first research that combines and compares three different apoptotic tests to analyse fresh semen of Our results confirm that infertile patients had lower sperm density, motile spermatozoa and normal forms, compared with the control group. Additionally, semen of patients with lowered fertilising ability contained higher percentages of spermatozoa with apoptosis markers. These findings are in accordance with the results of others studies (Grunewald, Said, Paasch, Glander, & Agarwal, 2008;Marchetti, Gallego, Defossez, Formstecher, & Marchetti, 2004).
This study shows that sperm apoptotic markers are correlated positively with age, and that apoptosis at least in part plays an important role in the ageing process and age-related infertility in men. Previous reports demonstrated the association between advancing age and the expression of early apoptotic markers as evidenced by the significant increase in plasma membrane translocation of PS, as well as with a more subtle proportion of spermatozoa carrying DNA fragmentation (Moskovtsev et al., 2010;Wyrobek et al., 2006). Moreover, it has been shown that apoptosis increases with age, producing an accelerated germ cell loss (Kimura et al., 2003), related to the falling androgen levels and/or an increase in oxidative stress in the tissue (Samanta, Sahoo, & Chainy, 1999).
In the current study, we have not shown a correlation between BMI and the percentage of apoptotic cells. This is not in accordance with the results obtained by others who found that men with BMI ≥30 kg/m2 had more apoptotic, fewer normal viable spermatozoa and lower percentage of spermatozoa with normal mitochondrial  (Almeida et al., 2005;Weng et al., 2002) and MMP integrity (Barroso et al., 2006;Benchaib et al., 2007;Grunewald et al., 2008;Marchetti, Ballot, Jouy, Thomas, & Marchetti, 2012).
In the current study, we found that FSH, DNA fragmentation, membrane asymmetry loss and activated caspases gave a high predictive value regarding the outcome of IUI treatments independently of the results of conventional spermatozoa analysis. In addition, in IVF/ICSI conception, the rate of fertilisation, segmentation and type I embryos is more altered in women whose partners had a high percentage of spermatozoa with caspases activated, DNA fragmentation and externalised PS. However, in some other studies, sperm apoptotic markers have no impact on the clinical results of IVF and ICSI cycles (Zorn et al., 2012).
Moreover, different associations of apoptotic markers have been proposed to predict pregnancy or miscarriage, and sperm DNA fragmentation was usually known as the best marker to predict the outcome of assisted reproduction (Bungum et al., 2004).
Using ROC analysis, we showed that in addition to DFI >15.5, the percentage of spermatozoa with activated caspases CASP+ >80% and the percentage of spermatozoa having externalised PS AN+ >42 were reliable predictors of pregnancy. But DNA fragmentation had always the best predictive value in pregnancy when compared with the other apoptotic markers.
In our study, caspases activation and PS externalisation appear to have a negative association with fertilisation and embryo development. Previous studies have shown that successful fertilisation requires also sperm plasma membrane with normal integrity (Cincik et al., 2007;Flesch & Gadella, 2000) and the numerous functions of the membrane are related to cell metabolism, for maintaining sperm motility, capacitation, acrosome reaction and sperm-oocyte interaction (Cross & Hanks, 1991).
It is known that assisted reproductive technologies (ART) have provided a treatment choice for many couples with unexplained infertility; however, the current success rates of these procedures remain suboptimal. One potential reason for these low rates may be the inclusion of apoptotic spermatozoa during intracytoplasmic sperm injection (ICSI) (Seli et al., 2004).
Elimination of the apoptotic spermatozoa before ART may be a safe method resulting in higher fertilisation rate and embryo quality in patients diagnosed with infertility.

| CONCLUSION
The positive correlation between respective semen with activated caspases and externalised PS and DNA fragmentation suggested that cytoplasm and plasma membrane integrity should be considered as a direct indicator of DNA integrity. In the light of these findings, apoptotic markers may replace conventional sperm parameters and should be taken into account when assessing male fertility potential and decision on the type of assisted reproduction technology.
F I G U R E 1 Receiver operating characteristic (ROC) curve of the ability of apoptotic markers to predict pregnancy in all studied couple (n = 122). Blue: ROC) curve of DFI, the area under the ROC curve was 0.79, the optimal threshold for prediction of pregnancy was DFI >15, 5%, with 100% sensitivity and 74.7% specificity; red: ROC curve of the percentage of spermatozoa with activated caspases (CASP+), the area under the ROC curve was 0.72. The optimal threshold for prediction of pregnancy was 80%, with 87% sensitivity and 81% specificity; Green: ROC curve of the percentage of spermatozoa with externalized PS (AN+), the area under the ROC curve was 0.7 and the optimal threshold for the prediction of pregnancy was 42%, with 70% sensitivity and 60% specificity