|Home | About | Journals | Submit | Contact Us | Français|
Infertility in men and women is frequently associated with genital contamination by various commensal or uropathogenic microbes. Since many microorganisms are known to release quorum-sensing signals in substantial amounts, we raised the question whether such molecules can directly affect human spermatozoa. Here we show that farnesol and 3-oxododecanoyl-l-homoserine lactone, employed by the opportunistic pathogenic yeast Candida albicans and the gram-negative bacterium Pseudomonas aeruginosa, respectively, induce multiple damage in spermatozoa. A reduction in the motility of spermatozoa coincided in a dose-dependent manner with apoptosis and necrosis at concentrations which were nondeleterious for dendritic cell-like immune cells. Moreover, sublethal doses of both signaling molecules induced premature loss of the acrosome, a cap-like structure of the sperm head which is essential for fertilization. Addressing their mechanism of action, we found that the bacterial molecule, but not the fungal molecule, actively induced the acrosome reaction via a calcium-dependent mechanism. This work uncovers a new facet in the interaction of microorganisms with human gametes and suggests a putative link between microbial communication systems and host infertility.
The phenomenon of quorum sensing (QS) has gained intensive attention not only in the study of microbial communication within defined bacterial populations but also in the study of interkingdom signaling and pathogenicity. QS is defined as a means for microorganisms to sense their population density via the release of signaling molecules, to which they in turn respond. Reaching a threshold concentration in a bacterial population, these molecules can coordinately regulate a multitude of different effects, such as bioluminescence, biofilm formation, and virulence gene expression. Different QS molecules, such as the autoinducing peptides or N-acylhomoserine lactones, have been described in the past for many bacterial species (17, 40).
A eukaryotic QS system was first evidenced in Candida albicans, a major human fungal pathogen and frequent commensal of the gastrointestinal and genitourinary tracts (10, 23). The isoprenoid alcohol farnesol was shown to inhibit the transition of C. albicans yeast cells to filamentous growth forms, and an opposite effect was recorded for tyrosol, another QS molecule in this fungus (5). Both these molecules are detectable in the supernatants of dense C. albicans yeast cell cultures in micromolar amounts. Recent observations indicate that the effects of farnesol are much more complex, since physiological farnesol concentrations of approximately 35 μM favor stress resistance in C. albicans (38) and even support antagonistic properties on other microbes. For example, farnesol was shown to induce apoptosis in the filamentous fungus Aspergillus nidulans (29) and to inhibit biofilm formation in the bacterial pathogen Staphylococcus aureus (13). There has also been particular interest in the impact of QS molecules on human cells, because evidence suggests implications in immunomodulation and the proliferation of distinct immune and malignant cells (11, 22, 26).
Genital infections caused by various microbial pathogens are frequently associated with infertility in men and women worldwide (24). Little attention has been paid, however, to potential direct influences of commensal or pathogenic microorganisms on human gametes, and therefore their interaction with human spermatozoa remains largely elusive. Nevertheless, adverse effects of microbes on sperm could be observed during in vitro coincubation experiments with uropathogenic bacteria and yeasts (12), even in the presence of cell-free supernatants of C. albicans cultures (37). Since QS molecules are expected to be released by microorganisms in substantial amounts in vitro as well as in the human host, we raised for the first time the question whether such molecules can directly affect human spermatozoa. To address this possibility, we monitored the impact of selected QS molecules on sperm parameters which are crucial for fertilization. Here we studied not only the impact of C. albicans farnesol but also that of 3-oxododecanoyl-l-homoserine lactone (3-oxo-C12-HSL), which is employed by the gram-negative, opportunistic pathogenic bacterium Pseudomonas aeruginosa, a frequent inducer of urinary tract infections (34). These studies revealed that distinct microbial QS molecules elicit multiple detrimental effects on human spermatozoa. They not only can impair sperm motility but also can induce spermatozoal cell death and, most notably, at sublethal doses can cause premature acrosome loss, a phenotype which is known to prevent the penetration of the oocyte by the sperm (19).
Human semen from healthy donors, provided with informed consent and local ethics committee approval, were analyzed according to WHO guidelines (39). In some of the experiments the isolated spermatozoa were pooled, and in other cases they were derived from individual donors. The observed effects were consistent despite the identity of the donor or whether they were derived from a single donor or from a pool. Sperm motility was assessed microscopically by scoring the percentage of progressive motile (A+B), nonprogressive (C), and immotile (D) spermatozoa (39). Progressive motile sperm were recovered by a two-step pure sperm gradient (Nidacon) and adjusted to 1 × 108/ml sperm preparation medium (MediCult). To monitor sperm motility in the presence of QS molecules, 4 × 106 sperm were incubated with different concentrations of farnesol (Fluka), tyrosol (Fluka), or 3-oxo-C12-HSL (Cayman) in 100 μl of SynVitroFlush medium (MediCult). Control samples were incubated with the solvent as recommended by the manufacturers. Sterile filtered supernatant aliquots of the C. albicans wild-type-strain SC5314 (8) and the Δtup1 mutant Bca2-10 (2) were obtained from cultures grown in synthetically defined medium (6.7 g yeast nitrogen base with ammonium sulfate [MP Biomedicals] and 20 g glucose per liter) for 48 h at 30°C. Thirty-microliter C. albicans supernatant aliquots were mixed with 4 × 106 sperm in a volume of 100 μl SynVitroFlush medium.
The hypoosmotic swelling test (14) was used to classify spermatozoa as osmotically intact by tail swelling. Briefly, the swelling test was performed by diluting 20 μl of sperm suspension (2 × 106 sperm) with 200 μl of hypoosmotic solution (7.35 g sodium citrate and 13.51 g fructose in 1 liter of distilled water). After incubation for 60 min at 37.5°C with 5.7% CO2, spermatozoa were centrifuged for 5 min at 900 × g. The pellet was resuspended in 20 μl hypoosmotic solution and smeared on a glass slide, and subsequently 200 spermatozoa were scored. Spermatozoa were classified as osmotically intact if tail swelling was observed. Spermatozoa were classified as osmotically incompetent if a straight tail was observed (14). Acrosome staining with fluorescein isothiocyanate (FITC)-Pisum sativum agglutinin (PSA) (Sigma) was used to identify intact acrosomes as follows. An aliquot of 20 μl of sperm suspension (2 × 106 sperm) was centrifuged for 5 min at 900 × g, resuspended in 2-μg/ml Hoechst solution, and incubated for 10 min in the dark. The sperm suspension was centrifuged again to remove excess stain, and the sperm pellet was resuspended in 20 μl SynVitroFlush medium. Twenty-microliter droplets of the sperm suspension were smeared on glass slides and allowed to dry. The slides were fixed in ice-cold methanol for 30 s. After drying, the fixed sperm cells were incubated for 30 min with 30 μl of 50 μg/ml FITC-PSA (Sigma) in water. After washing in water and mounting with Moviol, acrosome staining was classified as follows: intact acrosomes displayed a uniform green fluorescence in the acrosomal region of the sperm head, whereas acrosome loss was indicated by absent fluorescence or equatorial segment staining. A total of 200 spermatozoa were scored per sample. To monitor the effect of farnesol and 3-oxo-C12-HSL on acrosome loss under conditions of low calcium levels, SynVitroFlush medium with 5 mM EDTA was used. A 10 μM concentration of the calcium ionophor calcimycin (A23187) (Sigma), dissolved in dimethyl sulfoxide, was used as a positive control to induce the acrosomal reaction in spermatozoa. The Halosperm kit (Halotech DNA), a sperm chromatin dispersion test, was used to detect DNA fragmentation in sperm as follows. Aliquots of 25 μl of sperm suspension (2.5 × 106 cells) were mixed with 50 μl of low-melting-point agarose at 37°C. Fifty microliters of the mixture was pipetted onto a glass slide provided with the Halosperm kit, covered with a glass coverslip, and left to dry at 4°C for 5 min. Glass coverslips were removed carefully, and slides were immersed horizontally in a tray with freshly prepared denaturant solution for 7 min at room temperature and then transferred in a tray with lysis solution for 25 min at room temperature. After washing with water, slides were dehydrated in 70%, 90%, and 100% ethanol (2 min for each step) and allowed to dry. Sperm cells were stained with Wright's stain and classified by bright-field microscopy. Five hundred spermatozoa per sample were scored. Spermatozoa without DNA fragmentation show a large halo, which is absent or very small in spermatozoa with fragmented DNA (6).
Human monocyte-derived dendritic cells (DCs) were obtained from peripheral blood mononuclear cells by a standard protocol (27). In brief, after dilution with 50 ml phosphate-buffered saline-0.1 M citrate, blood was distributed over a density gradient (leukocyte separation medium; PAA Laboratories) and centrifuged at 400 × g for 30 min at room temperature. Monocytes were collected from the interface; washed with phosphate-buffered saline; resuspended in RPMI 1640 medium (PAA Laboratories) supplemented with 10% fetal bovine serum, 2 mM glutamine, and 50 μg/ml gentamicin; and incubated for 1 h at 37°C with 5.7% CO2 on plastic cell culture dishes (Greiner). The nonadherent cells were removed after 1 h. The adherent fractions (monocytes) were cultured for 2 to 3 days in RPMI 1640 medium supplemented with 10% fetal bovine serum, 2 mM glutamine, 50 μg/ml gentamicin, 20 U/ml granulocyte-macrophage colony-stimulating factor, and 16 U/ml interleukin-4. Cytokines and medium were replaced every second day. The purity of immature DCs (iDCs) was indicated by fluorescence-activated cell sorter (FACS) analysis using HLA-DR antibody.
FACS analyses detected apoptotic and necrotic sperm and iDCs by use of the annexin V-FITC kit (Bender Medsystems). In brief, an aliquot of 2 × 106 cells was incubated with 195 μl of prediluted binding buffer and 5 μl of annexin V-FITC. The cells were incubated in the dark at room temperature for 10 min. The suspension was centrifuged at 900 × g for 5 min and resuspended in 190 μl prediluted binding buffer with 10 μl propidium iodide. The cells were detected using log forward- and log side-scatter dot plots and density plots. FACS analysis was performed by use of a FACScan 2.0 cytometry system equipped with an argon laser emitting at 488 nm (Becton Dickinson). The fluorescence was measured on the FL1 fluorescence channel equipped with a 530-nm band-pass filter. Ten thousand cells were analyzed per sample and counted at low flow rate. Fluorescence data were collected by using logarithmic amplifiers, and forward scatter data were collected using linear amplifiers. The mean fluorescence values were determined with CellQuest 33 (Becton Dickinson) software.
Fluorescence microscopy was performed with a Zeiss LSM 510 inverted confocal laser scanning microscope equipped with a Zeiss Axiovert 100 microscope. Imaging scans were acquired with an argon laser with a wavelength of 488 nm and corresponding filter settings for FITC and parallel transmission images. The cells were observed with a ×63 immersion oil objective. Light microscopy was performed with an Olympus IX 51 microscope equipped with an Octax Eye USB2 camera. The cells were observed with a ×40 achromat objective. The imaging software was Octax EyeWare Mx.
All data were expressed as the mean ± standard deviation (SD). Differences were analyzed by the two-tailed unpaired Student t test. In all analyses, a P value of <0.05 was considered statistically significant.
Using an in vitro model of freshly isolated, highly motile human spermatozoa, we first investigated whether sperm motility is affected by soluble factors produced by C. albicans yeast cell cultures. For this purpose, the motility of spermatozoa was monitored in the presence of supernatant aliquots of the widely used C. albicans strain SC5314 grown in synthetically defined medium for 48 h. Motility analysis indicated a time-dependent loss of progressive motility for spermatozoa incubated with C. albicans supernatant (Fig. (Fig.1A).1A). As a control, spermatozoa were incubated with equal amounts of the growth medium. From these results we hypothesized that C. albicans QS molecules such as farnesol or tyrosol could be potential candidates which might add to the inhibitory effect of C. albicans cultures on sperm motility. Farnesol appeared to be the more likely candidate, since this molecule is known to induce apoptosis in different carcinoma cells (15, 28). We also tested the supernatant of C. albicans strain Bca2-10, a Δtup1 deletion mutant which was recently shown to produce larger amounts of farnesol in the culture supernatant than wild-type cells (16). A stronger inhibition of progressive sperm motility was detected for the supernatant of the C. albicans Δtup1 mutant than for the wild-type control filtrate (Fig. (Fig.1A).1A). However, the culture supernatants are probably rather complex and might contain factors other than QS molecules which could inhibit sperm motility. Therefore, we next studied a possible impact of pure farnesol and tyrosol on this phenotype.
Increasing concentrations of farnesol were analyzed for a possible effect on sperm motility at different time points, i.e., 1, 3, and 24 h. The degree of motility inhibition positively correlated with the concentration of farnesol, with a dramatic effect at a 50 μM concentration of the molecule already after 1 h of incubation (Fig. (Fig.1B).1B). In contrast, no reduction of progressive sperm motility was seen in the presence of tyrosol under the tested conditions (Fig. (Fig.1C1C).
The observed adverse impact of farnesol on spermatozoal motility did not provide evidence about sperm viability. In order to detect viable spermatozoa among nonmotile semen samples, we first employed the hypoosmotic swelling test, which is used in the in vitro fertilization laboratory as an indicator of intact sperm plasma membranes. The assay detected a loss of membrane integrity in the majority of spermatozoa after exposure to 50 μM farnesol for 1 h (Fig. (Fig.2A).2A). This result supported the view that the farnesol-induced reduction in motility coincides with spermatozoal membrane damage and killing. Consistently, light microscopic inspection revealed a complete membrane rupture of several sperm heads under these conditions, which was already apparent after 1 hour of treatment with only 25 μM farnesol (Fig. (Fig.2B).2B). Control spermatozoa did not reveal significant morphological alterations compared to farnesol-treated samples.
The observed cytotoxic effect on sperm could be a consequence of an apoptosis-like cell death. To investigate this possibility, apoptotic spermatozoa were identified via binding of annexin V-FITC to externalized phosphatidylserine, whereas necrotic cells were visualized with propidium iodide. In the presence of 25 μM farnesol for 3 h, a considerable proportion of necrotic spermatozoa was detected by flow cytometry (32.7%), in contrast to the control (12.0%) (Fig. (Fig.3A).3A). Under these conditions, a small yet considerable proportion of spermatozoa (6.3%) was identified to be apoptotic. Almost the entire population of the sperm sample incubated with 50 μM farnesol for 3 h was identified to be necrotic. Addressing the question whether this fast-acting deleterious effect of farnesol is specific for spermatozoa, we also tested human iDCs, which are involved in the defense against cervix-invading pathogens (7). As indicated in the lower panel of Fig. Fig.3A,3A, a strong induction of apoptosis and/or necrosis could not be detected in iDCs under the tested conditions. Because apoptosis and necrosis are assumed to coincide with other cellular alterations (33), we also analyzed whether farnesol induces spermatozoal DNA fragmentation. The assay used identifies spermatozoa with intact DNA by a halo of dispersed DNA loops, which is not visible in sperm with fragmented DNA. After 1 h of treatment with 25 μM farnesol, only 50% of the spermatozoal heads showed an extended halo of DNA (Fig. (Fig.3B),3B), indicating a strong impact of farnesol on DNA fragmentation and supporting its role as a trigger of apoptosis. Notably, spermatozoa with fragmented DNA were already detected after treatment with 5 μM farnesol.
Sperm viability and motility are necessary but not sufficient parameters for successful fertilization. Premature acrosome loss and/or acrosome reaction failure are also important causes of male infertility. In the next experiment, the proportion of intact sperm acrosomes was measured after 1 h of incubation in the presence of farnesol. Here, only those acrosomes which showed a complete positive FITC-conjugated PSA staining were considered to be intact. Acrosome loss was already apparent in a considerable proportion of sperm treated with sublethal doses of 5 μM farnesol (Fig. (Fig.4),4), indicating an enhanced sensitivity of the sperm acrosome to farnesol.
Given that microbial signaling is also assumed to play a key role in host adaptation processes in bacteria, we also investigated whether bacterial QS molecules affect sperm function. As an example, we monitored a potential effect of the widely studied 3-oxo-C12-HSL secreted by P. aeruginosa (26). In the motility assay used, increasing concentrations of 3-oxo-C12-HSL did not significantly affect the progressive motility of spermatozoa after 1 h and 3 h; however, after 24 h a strong reduction of sperm motility was observed for concentrations of ≥25 μM (Fig. (Fig.5A).5A). At this time point, a strong increase of apoptotic and necrotic cells in the presence of a 50 μM concentration of the molecule was revealed by flow cytometry (Fig. (Fig.5B).5B). Similar to our results obtained with farnesol, the deleterious effect of 3-oxo-C12-HSL was more pronounced on spermatozoa than on iDCs. Nevertheless, treatment with 50 and 100 μM 3-oxo-C12-HSL for 24 h resulted in a detectable proportion of necrotic iDCs. In comparison to the impact on sperm motility and viability, an enhanced adverse effect of 3-oxo-C12-HSL on sperm acrosomes was detected, which was seen already after 1 h of incubation at low concentrations of 1 to 5 μM (Fig. (Fig.5C5C).
We hypothesized that the acrosome loss detected in the presence of farnesol and 3-oxo-C12-HSL could be either a result of unspecific membrane damage in the acrosomal region of the sperm head or a consequence of an actively induced acrosome reaction. To investigate the latter possibility, the effect of sublethal doses of the QS molecules on sperm acrosomes was monitored in the presence of a calcium chelator. This experiment is based on the knowledge that the oocyte-induced acrosomal reaction is mediated by increased levels of free intracellular calcium in the sperm cytoplasm and is specifically prohibited in the absence of extracellular calcium. The proportion of intact sperm acrosomes was measured after 1 h of incubation in the presence of 5 mM EDTA at increasing concentrations of farnesol or 3-oxo-C12-HSL (Fig. (Fig.6).6). As a positive control, spermatozoa were treated under the same conditions with calcimycin, a known inducer of the acrosome reaction. As expected, the presence of calcimycin reduced the proportion of spermatozoa with intact acrosomes, an effect which was abrogated in the presence of EDTA. Notably, the effect of 3-oxo-C12-HSL on sperm acrosomes was also abolished in the presence of the calcium chelator, suggesting that this molecule also actively induces the acrosome reaction in human spermatozoa via a calcium-dependent mechanism. In contrast, the detrimental effect of farnesol on sperm acrosomes was observed irrespective of the presence of the calcium chelator (Fig. (Fig.66).
Altered sperm phenotypes such as reduced motility, acrosome loss, or DNA fragmentation can promptly lead to fertilization failure; e.g., the proportion of sperm with fragmented DNA is negatively correlated with pregnancy, which was determined only for semen samples with less than 30% of fragmentation-positive sperm (18). A first hint that secreted factors of C. albicans might be harmful for spermatozoa was described by Tian and coworkers (37), who demonstrated spermatozoal impairment by C. albicans culture supernatants. Our present findings support this observation and attribute a specific antispermatozoal role to one of the known secreted C. albicans QS molecules, i.e., farnesol, a nonsterol isoprenoid, which is also a catabolite of the cholesterol biosynthetic pathway. We found that a loss of progressive sperm motility in the presence of farnesol coincided with apoptosis and necrosis at molecule concentrations of ≥25 μM, whereas DNA fragmentation and acrosome loss were detected at even lower concentrations. A farnesol concentration of 50 μM induced necrosis in almost the entire sperm population tested. Our results therefore indicate a considerable sensitivity of spermatozoa to the molecule, a finding which is supported by the observation that similar deleterious effects of farnesol were not detected in immature dendritic control cells. However, farnesol was previously shown to induce apoptosis in other fungal organisms and, interestingly, also specifically in distinct human carcinoma cell lines at concentrations of approximately 60 to 250 μM (15, 28). The antitumor activity of farnesol has even led to consideration of its potential use as a therapeutic agent, a proposition which, however, should consider the effects of farnesol on spermatozoa.
A crucial step in the process of fertilization is the acrosome reaction, which is irreversible and takes place in the acrosome of the sperm head as it contacts the zona pellucida of the oocyte (25). Our finding that comparably low and sublethal concentrations of farnesol induce acrosome loss in spermatozoa therefore appears to be highly important and might explain reports which attribute fertilization failures to semen that was contaminated by C. albicans but showed normal sperm parameters in terms of viability and motility (3). This enhanced sensitivity of sperm acrosomes was also observed in coincubation experiments using the P. aeruginosa QS molecule 3-oxo-C12-HSL. Interestingly, 3-oxo-C12-HSL has structural similarity to farnesol, and cross talk between C. albicans and P. aeruginosa via their QS molecules even has been described previously (9). In our studies, significant acrosome loss was visible in spermatozoa treated with comparably low concentrations of 3-oxo-C12-HSL for only 1 hour, whereas motility loss, apoptosis, and necrosis were revealed after treatment for 24 h at higher concentrations. This finding let us raise the question whether acrosome loss caused by sublethal doses of the two QS molecules was a result of unspecific membrane disorder in the acrosomal region of the sperm head or a consequence of an actively induced acrosome reaction. Although components of the oocyte zona pellucida have been suggested to be natural inducers of the acrosome reaction, other external signals, such as progesterone, have also been reported to promote this process, which is triggered by altered levels of free intracellular calcium (21). Interestingly, both the QS molecules 3-oxo-C12-HSL and farnesol have been linked before with a potential impact on intracellular calcium levels. Higher concentrations (100 μM) of 3-oxo-C12-HSL were revealed to induce apoptosis in a detectable proportion of murine fibroblasts via an increase of intracellular calcium (31), and farnesol has been identified as a potent blocker of smooth muscle L-type calcium channels (20). In our experiments, concentrations of 1 to 5 μM 3-oxo-C12-HSL were sufficient to induce the calcium-mediated acrosome reaction in a considerable proportion of sperm, an observation which supports the view that 3-oxo-C12-HSL likely acts as a true signal on human spermatozoa. In contrast, the farnesol-induced acrosome loss was also measured in the presence of the calcium chelator and is hence assumed to be a result of enhanced membrane damage within the region of the sperm acrosome. The cytotoxic effects of 3-oxo-C12-HSL were strongly pronounced in spermatozoa; however, at higher concentration (100 μM), detectable deleterious effects were also observed in iDCs. This finding supports recent data which demonstrated immune modulatory and, at higher concentrations (50 to 100 μM), specific cytotoxic effects for 3-oxo-C12-HSL on macrophages, neutrophils, DCs, and CD4+ T cells but not on CD8+ and different epithelial cells (1, 35, 36).
Addressing the question whether microbial signaling molecules are secreted during host colonization in considerable amounts, efforts were successfully undertaken to detect bacterial QS molecules even in the sputa of cystic fibrosis patients infected with P. aeruginosa (32). The highest concentrations of 3-oxo-C12-HSL up to 600 μM have been measured in P. aeruginosa biofilm cultures grown in vitro, which were thus assumed to be potentially significant (4, 30). However, one has to consider that many host niches are known to be encountered by mixed microbial populations, which likely employ multiple communication systems and consequently produce complex mixtures of signaling molecules. From our data we conclude that soluble QS molecules of different microbial origins may elicit diverse detrimental effects on human spermatozoa. This finding appears to be pathologically important, since many host surfaces of the genitourinary tract are colonized by dense communities of various microbial species, some of which might even synergistically contribute to spermatozoal impairment. Our results provide first insights into a not-yet-understood aspect of host-microbe interaction and present potential consequences thereof.
We thank Joachim Morschhäuser and Bernhard Hube for critical reading of the manuscript and Alexander D. Johnson for kindly providing C. albicans strain Bca2-10.
This work was supported by the HKI and the DFG-funded excellence graduate school Jena School for Microbial Communication.
Editor: A. Casadevall
Published ahead of print on 17 August 2009.