Altered PIWI-LIKE 1 and PIWI-LIKE 2 mRNA expression in ejaculated spermatozoa of men with impaired sperm characteristics.

In about half the cases of involuntary childlessness, a male infertility factor is involved. The PIWI-LIKE genes, a subclade of the Argonaute protein family, are involved in RNA silencing and transposon control in the germline. Knockout of murine Piwi-like 1 and 2 homologs results in complete infertility in males. The aim of this study was to analyze whether the mRNA expression of human PIWI-LIKE 1-4 genes is altered in ejaculated spermatozoa of men with impaired sperm characteristics. Ninety male participants were included in the study, among which 47 were with normozoospermia, 36 with impaired semen characteristics according to the World Health Organization (WHO) manual, 5th edition, and 7 with azoospermia serving as negative control for the PIWI-LIKE 1-4 mRNA expression in somatic cells in the ejaculate. PIWI-LIKE 1-4 mRNA expression in the ejaculated spermatozoa of the participants was measured by quantitative real-time PCR. In nonazoospermic men, PIWI-LIKE 1-4 mRNA was measurable in ejaculated spermatozoa in different proportions. PIWI-LIKE 1 (100.0%) and PIWI-LIKE 2 (49.4%) were more frequently expressed than PIWI-LIKE 3 (9.6%) and PIWI-LIKE 4 (15.7%). Furthermore, a decreased PIWI-LIKE 2 mRNA expression showed a significant correlation with a decreased sperm count (P = 0.022) and an increased PIWI-LIKE 1 mRNA expression with a decreased progressive motility (P = 0.048). PIWI-LIKE 1 and PIWI-LIKE 2 mRNA expression exhibited a significant association with impaired sperm characteristics and may be a useful candidate for the evaluation of the impact of PIWI-LIKE 1-4 mRNA expression on male infertility.

INTRODUCTION

Infertility is generally defined as the inability to conceive spontaneously within a period of at least 12 months of unprotected intercourse. In more developed countries, a 12-month prevalence rate of up to 16.7% is reported, setting one of six couples at risk of unwilling childlessness.1 In approximately 40% of cases, male factors contribute to couple infertility. Causes for impaired sperm characteristics as the leading cause for male infertility are diverse, ranging from endocrine disturbances, immunological factors, increased scrotal temperature, urogenital tract infections as well as congenital or acquired urogenital abnormalities and genetic abnormalities (European Association of Urology [EAU] Male Infertility guidelines).2 However, cases of male idiopathic infertility are frequent and molecular data about their aetiology and predictors of impaired sperm characteristics are scarce.3

The members of the PIWI gene family, a subclade of the Argonaute proteins, are defined by their highly conserved P-element induced wimpy testis (PIWI) and Piwi-Argonaute-Zwille (PAZ) domains and their exclusive expression in the germline.45 Both domains function in the binding of small RNAs with the PAZ domain ensuring the binding of the 3'OH-overlap of small RNAs and the PIWI domain binding to the 5'-phosphate cap of small RNAs. This binding implicates slicer function and posttranscriptional silencing of target mRNAs.678 PIWI-LIKE proteins are shown to interact with a subclass of small RNAs called piwi-interacting RNAs (piRNAs), which are longer than other small RNA classes such as microRNAs (26–31 nt instead of 18–25 nt), share a preference for a 5'uridine, and are processed from a few distinct genetic loci.91011 At the cellular level, piRNA/PIWI complexes trigger the silencing of retrotransposons, selfish genetic elements activated for instance by demethylation processes during gametogenesis and able to reintegrate at random loci in the genome.121314 PIWI-LIKE genes also act at the chromatin level and exert epigenetic silencing of distinct genomic areas through regulation of histone K9 methylation.15161718 Furthermore, in male knockout mice, the loss of murine Piwi-like 1 (Miwi) results in a meiotic arrest at the round spermatid stage,19 while loss of murine Piwi-like 2 (Mili) results in a zygotene-pachytene spermatocytic block,20 resulting in male mice sterility. Taken together, PIWI-LIKE genes play an essential role in male gametogenesis by protecting the differentiating germ cells' genomic stability and the individual's male fertility.

A clear correlation between the dysregulated expression of human PIWI-LIKE 1 (HIWI) and tumorigenesis and prognosis can be observed in several tumor entities, including seminomas,21 breast cancer,2223 and soft-tissue sarcoma.24 On the other hand, the data on the impact of human PIWI-LIKE gene mutation or dysregulation on male fertility are scarce. Gu et al.25 identified single-nucleotide polymorphisms in PIWI-LIKE 3 (nonsynonymous) and in PIWI-LIKE 4 (in the 3' untranslated region) to be associated with the occurrence of oligozoospermia. Furthermore, PIWI-LIKE 2 hypermethylation was associated with spermatogenic failure due to germ cell maturation defects in a cohort of 32 patients with azoospermia or severe oligozoospermia.26 Recently, it was shown that single-nucleotide mutations in the D-box domain of PIWI-LIKE 1 prevent ubiquitination and protein decay, resulting in azoospermia probably due to failure in the histone-to-protamine exchange.27 However, no data exist about the predictive potential of human PIWI-LIKE gene expression in ejaculated spermatozoa, which are the cells classically used in andrological practice for the assessment of fertility status and therapy decision.

The aim of this study was to determine whether mRNA expression of the four human PIWI-LIKE orthologs (PIWI-LIKE 1/HWI, PIWI-LIKE 2/HILI, PIWI-LIKE 3/HIWI3, and PIWI-LIKE 4/HIWI2) were detectable in ejaculated spermatozoa of patients. Furthermore, we wanted to evaluate whether these mRNA expressions were associated with impaired clinical semen characteristics.

PATIENTS AND METHODS

Study subjects and ethical approval

Participants were recruited from the couple fertility clinic of the Center for Reproductive Medicine and Andrology, University Hospital Halle (Saale, Germany). Altogether, 90 participants exhibiting different diagnoses of normozoospermia, asthenozoospermia, teratozoospermia, oligozoospermia, and different combinations, as well as azoospermia were enrolled. All patients were examined by an experienced andrologist. A detailed overview of the studied cohort is given in . The study was approved by the Ethics committee of the Medical Faculty of the Martin-Luther University Halle-Wittenberg. All participants gave written informed consents.

Clinical data assessed for this study contained age, sperm concentration, motility and morphology, occurrence and concentration of round cells, viability, semen volume, and pH. Every ejaculate analysis was conducted in the andrology laboratory of our clinic according to the World Health Organization (WHO) manual, 5th edition.28 The categories for the subdivision of the cohort (oligozoospermia, asthenozoospermia, necrozoospermia, teratozoospermia, and normozoospermia) were applied according to the reference values described in the WHO manual.28 Samples from the seven azoospermic patients were included in this study for the validation of the spermatozoa-specific PIWI-LIKE mRNA detection.

Specimen preparation and RNA isolation

After collection and ejaculate analysis according to the WHO manual, 5th edition, the specimens were centrifuged (1000 g, 10 min) immediately to separate spermatozoa and seminal plasma. The cell pellet was snap frozen and cryopreserved at −80°C until analysis.

Total RNA isolation from ejaculated spermatozoa was performed by the phenol/chloroform method. Briefly, the sperm pellet was dissolved in 1 ml TRIzol reagent (Invitrogen, Carlsbad, CA, USA) and RNA separated by the addition of chloroform (AppliChem, Darmstadt, Germany) and centrifugation. To remove remaining traces of DNA, the solution was treated with DNase I (Qiagen, Hilden, Germany) for 15 min. RNA was precipitated by addition of isopropanol (AppliChem) and incubation at −20°C overnight. RNA pellet was washed twice with ice-cold ethanol and dissolved in 30 μl RNAse-free water. The RNA concentration and purity was assessed by absorption spectrometry in a spectrophotometer.

cDNA synthesis

The cDNA synthesis was carried out with RevertAid H Minus First strand cDNA synthesis kit (ThermoScientific, Waltham, MA, USA) and random primer according to manufacturer's protocol. Then, 1 μg of total RNA was applied for each reaction, yielding a sufficient amount of cDNA for double measurement of each gene.

Quantitative real-time-PCR

The measurement of the PIWI-LIKE 1–4 mRNA expression in the specimen was quantified by quantitative real-time-PCR. The reaction mix was set up with the HotStart Taq-Polymerase Kit (Qiagen) according to the manufacturer's protocol. Each human PIWI-LIKE gene (PIWI-LIKE 1–4) was amplified by using specific TaqMan primers (PIWI-LIKE 1: Hs_00380290; PIWI-LIKE 2: Hs_01032719; PIWI-LIKE 3: Hs_00908837; PIWI-LIKE 4:Hs_00895218; Applied Biosystems, Darmstadt, Germany). Quantitative real-time PCR measurements applying TaqMan assays were performed with the following program: (1) 95°C for 15 min, (2) 95°C for 20 s, and (3) 60°C for 1 min (repeat cycle 2–3 for 45 times).

The mRNA expression of hypoxanthine phosphoribosyl transferase (HPRT; forward primer: 5'-TTG CTG ACC TGC TGG ATT AC-3'; reverse primer: 5'-CTT GCG ACC TTG ACC ATC TT-3') was measured by using Maxima SYBR Green/ROX Master Mix (ThermoScientific) according to the manufacturer's protocol. HPRT mRNA is the most stable expressed according to free software package NormFinder (available from: http://www.mdl.dk/publicationsnormfinder.htm) and was chosen as the reference gene. The mRNA expression for each gene was calculated by the Δ CT method.29

Statistical analyses

Statistical analyses were performed with SPSS 19.0 (SPSS Inc., Chicago, IL, USA). Bivariate correlation analyses according to Spearman-Rho and nonparametric tests (Mann–Whitney U-Test) were conducted for the evaluation of the association between human PIWI-LIKE 1–4 mRNA expression and sperm characteristics. P < 0.05 was considered statistically significant.

RESULTS

Assessment of PIWI-LIKE mRNA expression in ejaculated spermatozoa

PIWI-LIKE 1–4 mRNA was measurable in ejaculated spermatozoa in different proportions. PIWI-LIKE 1 (100.0% of cases) and PIWI-LIKE 2 (49.4%) were more frequently detectable than PIWI-LIKE 3 (9.6%) and 4 (15.7%). HPRT was expressed stable in every sample and served as internal reference for the standardization of the comparison of the individual PIWI-LIKE 1–4 mRNA expressions. In connection with the use of 1 μg total RNA per se men sample for cDNA synthesis, the application of a reference gene in the calculation of the relative mRNA expression (given as Δ CT value) guarantees the independence of the PIWI-LIKE 1–4 transcript measurement from the number of spermatozoa in the ejaculate sample. On the other hand, in the cellular fraction of ejaculates from seven azoospermic patients analyzed as controls, PIWI-LIKE 1–4 mRNA was not detectable in any sample. Round cells (≥0.1 × 106 ml−1) could be detected in 13 ejaculate samples, while 70 specimens exhibited no accurate determinable round cell contamination (<0.1 × 106 ml−1). The mean concentration of round cells was 0.3 × 106 ml−1 (range: 0–3.8 × 106 ml−1), while the median was 0 ml−1. Furthermore, in nonparametric tests, no correlation between the occurrence of detectable concentrations of round cells in the ejaculate and the expression of either PIWI-LIKE 1–4 mRNA was observed (P ranging from 0.37 to 0.6; Mann–Whitney U-test). A nonparametric bivariate linear correlation analysis was performed to examine whether the mRNA expressions of individual human PIWI-LIKE genes were correlated with each other in human ejaculated spermatozoa. In the ejaculated sperm samples, only the mRNA expression of PIWI-LIKE 1 was correlated to PIWI-LIKE 2 expression (r = 0.543; P = 2.1 × 10−9).

PIWI-LIKE 1 mRNA is correlated with reduced sperm vitality

Further clinical and biographical data were studied in correlation with the individual PIWI-LIKE gene expression assessing bivariate linear regression analyses. The percentage of nonviable spermatozoa was inversely associated with the expression of the PIWI-LIKE 1 mRNA (r = −0.34; P = 0.018; n = 49). Additionally, there was a significant positive correlation between increased age of participant and increased PIWI-LIKE 1 (r = 0.29; P = 0.008; n = 83) or increased PIWI-LIKE 2 mRNA expression (r = 0.26; P = 0.016; n = 83). Furthermore, PIWI-LIKE 1 mRNA expression was inversely associated with the percentage of motile spermatozoa in the ejaculate (r = −0.23; P = 0.037), while PIWI-LIKE 2 mRNA expression was associated with sperm concentration (r = 0.27; P = 0.011) and especially the number of spermatozoa in the ejaculate (r = 0.31; P = 0.004).

Altered PIWI-LIKE 1 or PIWI-LIKE 2 mRNA expression was associated with sperm characteristics

The mRNA expression of the different human PIWI-LIKE genes in relation to the individual clinical sperm characteristics (concentration, total sperm count, motility, and morphology) was analyzed by applying nonparametric tests.

PIWI-LIKE 2 mRNA expression was decreased in samples exhibiting oligozoospermia (<39 × 106 cells per sample) in comparison with samples with normozoospermia (P = 0.022; Mann–Whitney U-test). The same was true when considering sperm concentration, decreased PIWI-LIKE 2 mRNA was also correlated with lower sperm concentration (<15 × 106 ml−1; P = 0.022; Mann–Whitney U-test). The relationship of the mRNA expression of PIWI-LIKE 1 and PIWI-LIKE 2 and total sperm count is shown in (Figure and ).

Box-plots comparing the relative mRNA expression of the PIWI-LIKE 1 and PIWI-LIKE 2 (expressed in Δ CT values) in ejaculated spermatozoa stratified to total sperm count and progressive motility. While (a) PIWI-LIKE 1 expression was not associated with sperm count, (b) samples from patients with oligozoospermia (<39 × 106 spermatozoa per ejaculate; n = 26) exhibited a decreased PIWI-LIKE 2 mRNA expression in comparison with normozoospermia (≥39 × 106 spermatozoa per ejaculate; n = 57, P = 0.022, Mann–Whitney U-test). (c) Samples from patients with asthenozoospermia (<32%; n = 23) exhibited an increased PIWI-LIKE 1 mRNA expression compared with normozoospermic patients (≥32%; n = 60; P = 0.048, Mann–Whitney U-test), while (d) PIWI-LIKE 2 mRNA expression was not associated with motility. Note the different y-axis scales indicating the higher expression of PIWI-LIKE 1 mRNA in comparison to PIWI-LIKE 2 mRNA. The boxes designate the distribution of 50% of the measured values; the lines designate the median value. The whiskers define the 2.5% and 97.5% quantiles.

Bearing in mind, that around half the samples had no detectable PIWI-LIKE 2 mRNA expression, we applied additional statistical analyses with the categories PIWI-LIKE 2 mRNA expression “detectable” and “nondetectable.” The results demonstrated that nondetectable PIWI-LIKE 2 mRNA expression was significantly associated with a lower sperm concentration (P = 0.017; Mann–Whitney U-test) and a lower sperm count (P = 0.004; Mann–Whitney U-test).

The association between progressive motility and PIWI-LIKE gene mRNA expression revealed that increased PIWI-LIKE 1 was associated with the occurrence of asthenozoospermia (<32% progressive motile spermatozoa; P = 0.048, Mann–Whitney U-test, (Figure and ). Neither PIWI-LIKE 1 nor PIWI-LIKE 2 mRNA expression was significantly associated with abnormal morphology (<4%; P = 0.24 and P = 0.19; Mann–Whitney U-test).

DISCUSSION

In this study, we evaluated the mRNA expression of the spermatogenesis-relevant genes PIWI-LIKE 1–4 in ejaculated spermatozoa of patients referred to our center for couple infertility. To the best of our knowledge, the analysis of mRNA expression of all four PIWI-LIKE genes in ejaculated spermatozoa has not been published. Heyn et al.26 analyzed PIWI-LIKE 2 transcript levels in spermatozoa of normozoospermic individuals in comparison to testicular samples with patients with spermatogenic arrest. In our study, we detected a significant correlation between the expression of PIWI-LIKE 1 and PIWI-LIKE 2 mRNA, suggesting an overlapping expression pattern of these genes in sperm development as it is described for the homologous Piwi-like 1 (Miwi) and Piwi-like 2 (Mili) genes in mice.30 Specifically, the expression of both murine gene homologs is detected in late stages of spermatogenesis, with Miwi expressed only in the mid-pachytene spermatocytes and later stages.1931 Most recently, Hempfling et al.32 reported an almost ubiquitous expression of PIWI-LIKE 1 and PIWI-LIKE 2 in testicular samples of patients with normozoospermia or hypospermatogenesis. In contrast, PIWI-LIKE 4 was only analyzed in samples of patients with spermatogenic arrest and showed a high expression in these samples, while PIWI-LIKE 3 was not detectable in any testicular sample.32

Intriguingly, increased PIWI-LIKE 1 and also PIWI-LIKE 2 mRNA expression was significantly correlated with elevated participants' age. An association between PIWI-LIKE gene expression and increasing age has not been described before and may be interesting in the context of aging males' protection of their genetic material against transposons activation, environmental stressors, and oxidative radicals. Furthermore, especially PIWI-LIKE 2 mRNA expression seems to play an essential role in these processes, leading to a worsened prognosis for soft-tissue sarcoma patients with a decreased mRNA expression of either PIWI-LIKE 2 or PIWI-LIKE 4.33 However, as age was only a secondary endpoint of this study and the age range was rather small, general conclusions on an age-dependence of PIWI-LIKE 1–4 mRNA expression in spermatozoa are not possible yet.

Analogous to the detrimental effect of a decreased PIWI-LIKE 2 mRNA expression in tumors on patients' outcome, a decreased expression of PIWI-LIKE 2 mRNA in our cohort was associated with a decreased total sperm count (P = 0.022). This result is in line with the results of Heyn et al.,26 showing a decreased PIWI-LIKE 2 transcript level due to PIWI-LIKE 2 promoter hypermethylation in patients' samples exhibiting different stages of spermatogenic failure. The application of the same amount of total RNA (1 μg) for cDNA synthesis of each patient's sample and the usage of a reference gene (HPRT) for the normalization of the gene expression both ruled out a bias in the measured PIWI-LIKE 2 mRNA expression, thus additionally strengthening this result. Furthermore, a significant association between an increased PIWI-LIKE 1 mRNA expression and a decreased percentage of progressive motile spermatozoa could be demonstrated in our study cohort, which may indicate the need for a PIWI-LIKE gene silencing in proper sperm development.

The protein expression of PIWI-LIKE 1 and 2 in human ejaculated spermatozoa is still rarely studied; however, recently Hempfling et al.32 demonstrated with immunofluorescence in testicular biopsies that PIWI-LIKE 1 protein is detectable in spermatocytes and spermatids, but not in elongated spermatids. This is a strong hint that the protein expression of the PIWI-LIKE genes in human spermatozoa resembles that detected in murine spermatozoa, where PIWI-LIKE 2 and 1 protein expression is diminished in late phases of spermatogenesis.30 Therefore, we would expect no detectable PIWI-LIKE 1 or 2 protein expression in normal sperm samples. From the recent literature, one could speculate that mammalian PIWI-LIKE 1 or 2 mRNA (or potentially protein expression) is a surrogate for proper sperm development, maturation, and better genomic integrity rather than a functional regulator of sperm activity.

In summary, we measured mRNA expression of the four human PIWI-LIKE genes in a cohort of 83 men with different semen status. PIWI-LIKE 1 and PIWI-LIKE 2 mRNA expressions were especially identified as putative new targets for further studies on the association between sperm-bound mRNA transcriptome and male fertility impairments. These interactions should be addressed in broader studies applying more holistic techniques such as microarrays and RNA-seq.

AUTHOR CONTRIBUTIONS

MG performed the data analysis and revised the manuscript; CM and LM carried out the clinical sample processing and the qPCR measurements; TG and HMB conceived the study design, prepared and revised the manuscript. All authors read and approved the final version of the manuscript.

COMPETING INTERESTS

All authors declare no competing interests.

Table 1

Participants’ clinical characteristics