An essential step in mammalian fertilisation is the sperm acrosome reaction (AR) – exocytosis of a single large vesicle (the acrosome) that surrounds the nucleus at the apical sperm head. The acrosomal and plasma membranes fuse, resulting in both the release of factors that might facilitate penetration of the zona pellucida (which invests the egg) and the externalisation of membrane components required for gamete fusion. Exocytosis in somatic cells is a rapid process – typically complete within milliseconds – yet acrosomal enzymes are required throughout zona penetration – a period of minutes. Here, we present the first studies of this crucial and complex event occurring in real-time in individual live sperm using time-lapse fluorescence microscopy. Simultaneous imaging of separate probes for acrosomal content and inner acrosomal membrane show that rapid membrane fusion, initiated at the cell apex, is followed by exceptionally slow dispersal of acrosomal content (up to 12 minutes). Cells that lose their acrosome prematurely (spontaneous AR), compromising their ability to penetrate the egg vestments, are those that are already subject to a loss of motility and viability. Cells undergoing stimulus-induced AR (progesterone or A23187) remain viable, with a proportion remaining motile (progesterone). These findings suggest that the AR is a highly adapted form of exocytosis.
Over 80 million people worldwide experience infertility and over one third of infertility cases are due to male factors, i.e. sperm dysfunction (http://www.who.int/reproductive-health/infertility/report.pdf), but 20-30% of these are classified `normal' by standard semen analyses (quantity, morphology and motility) (Liu and Baker, 2003). A crucial aspect of sperm function not amenable to such analysis is the acrosome reaction (AR), yet premature AR and/or AR failure are important causes of male infertility (Liu et al., 2006). The AR, the Ca2+-dependent exocytosis of the acrosome, requires multiple fusions between the outer acrosomal membrane (OAM) and the closely apposed plasma membrane. Although first described more than 50 years ago, detailed understanding of this process has remained elusive. The AR is considered to fulfil two key functions. First, proteolytic enzymes in the acrosome (including hyaluronidase and acrosin) are exposed, which might facilitate penetration of the zona pellucida, although acrosin knockouts in mice fail to affect fertilisation (Baba et al., 1994). Second, the acrosomal content disperses, revealing a new surface membrane – the inner acrosomal membrane (IAM) – which is a requirement for fertilisation. The AR is induced by binding to the zona pellucida (Kopf and Gerton, 1991), probably after `priming' by progesterone released from the oocyte and surrounding cells (Roldan et al., 1994; Schuffner et al., 2002; Harper et al., 2004). Exposure of acrosomal content therefore occurs when the sperm reaches the outer surface of the zona pellucida, yet for proteolytic activity of the enzymes to be of significance they must be present throughout zona pellucida penetration [a process that takes minutes (Yanagimachi, 1994)]. Previous methods for estimating the frequency of AR (% acrosome-reacted sperm) rely primarily on labelling acrosomal content or the OAM with fluorescent probes, but this approach requires sperm fixation and provides no dynamic information (e.g. Harper et al., 2003). To understand how the AR facilitates progression of the sperm through the zona pellucida by proteolytic activity and also exposes the IAM such that gamete fusion occurs, the details of the process must be observed in live cells. Attempts to visualise exocytosis by monitoring membrane loss in live human sperm have proved unsuccessful (Harper et al., 2006), but fluorescent tagging of acrosomal content in macaque sperm suggests a possible approach (Tollner et al., 2003). Here, we uncover the events of the AR in real time in live human sperm.
Results and Discussion
In this study, the progress of the AR (fenestration of plasma membrane, exposure and dispersal of acrosomal content and subsequent exposure of the IAM) was visualised in living human sperm using fluorescent labels for two different components of the acrosome. Soybean trypsin inhibitor (SBTI) binds to the acrosomal contents (specifically acrosin) (Arts et al., 1994) of mammalian sperm following membrane fenestration (Baibakov et al., 2007; Tollner et al., 2003). The complement regulatory protein CD46 (membrane cofactor protein) is localised solely on the IAM in human sperm (Anderson et al., 1989; Cummerson et al., 2006). As dispersal of the acrosomal content proceeds, binding sites for anti-CD46 antibodies are revealed, allowing exposure of the IAM to be monitored (Fig. 1A). Preliminary analyses on the kinetics of antibody binding to acrosome-reacted sperm (using 10 μM A23187 to pre-induce the AR) (Harper et al., 2006; Jamil and White, 1981; Tesarik, 1985) showed that anti-CD46 antibodies are capable of binding to exposed IAM within 10 seconds (saturation within 40 seconds; supplementary material Fig. S1), therefore providing temporal resolution within a fraction of a minute. Phycoerythrin (PE) and fluorescein isothiocyanate (FITC) anti-CD46 conjugates (referred to as fluorescent anti-CD46) were both required for these studies and had similar uptake characteristics (data not shown).
Temporal and spatial characteristics of the acrosome reaction
Following stimulation with 10 μM A23187, membrane fenestration and exposure of acrosomal content (accumulation of Alexa-Fluor-488-conjugated SBTI; Alexa-488–SBTI) occurred within 1 minute from initiation to completion (Fig. 1B,C). This was followed by a slower decline in fluorescence, indicating a gradual dispersal of content into the surrounding medium. Alexa-488–SBTI fluorescence stabilised at a raised plateau, consistent with previous suggestions that components of the acrosomal content remain attached to the sperm (Aitken and Brindle, 1993; Green and Hockaday, 1978; Kim et al., 2001; Tesarik et al., 1988). As the acrosomal contents dispersed, there was a strikingly slow accumulation of fluorescent anti-CD46 as the IAM became progressively exposed (Fig. 1B,C; supplementary material Movie 1). The initial rate of loss of acrosomal content (Alexa-488–SBTI staining) was significantly more rapid than the initial rate of exposure of the IAM (accumulation of fluorescent anti-CD46; P<0.01, one-tailed test; Fig. 1Di), indicating that some dispersal of acrosomal content generally precedes exposure of anti-CD46-binding sites. Heterogeneity was observed in the rates of dispersal (Fig. 1Dii,iii) and very occasionally complete failure of acrosomal-content dispersal prevented exposure of the IAM (Fig. 1Diii), a situation that might permit zona penetration but would preclude gamete fusion.
The proportion of cells in which AR occurred and the kinetics of the loss of acrosomal content (detected by exposure of the IAM) varied significantly between different AR-inducing conditions (Table 1, supplementary material Fig. S2A,B). A23187 (10 μM) induced AR in almost 50% of sperm, with a duration of 8.8±0.9 minutes (from initiation to completion of fluorescent anti-CD46 uptake; Fig. 1E). Stimulation with progesterone, a product of the cumulus oophorus (Hartshorne, 1989; Osman et al., 1989), at a dose of 3 μM [similar to concentrations within the cumulus and widely reported to induce AR in human sperm (Harper et al., 2006; Kirkman-Brown et al., 2000; Osman et al., 1989)], induced AR in 10% of sperm. The kinetics of fluorescent anti-CD46 uptake under these conditions were significantly slower than in A23187-treated sperm (P<0.05; Fig. 1E). Spontaneous AR occurred in 2% of sperm during a 60-minute incubation. In these cells, the rate of exposure of the IAM was similar to that occurring upon progesterone stimulation (Fig. 1E).
|.||AR (%) .||Latency of AR after stimulus (min) .||Time to 90% AR (min) .||Viability after AR (%) .||Time of death after AR (min) .||Motile after AR (%) .||Cells .||n .|
|.||AR (%) .||Latency of AR after stimulus (min) .||Time to 90% AR (min) .||Viability after AR (%) .||Time of death after AR (min) .||Motile after AR (%) .||Cells .||n .|
AR was induced by 10 μM A23187, 3 μM progesterone or occurred spontaneously. For each stimulus, the following variables were measured and the mean ± s.e.m. recorded: the percentage of sperm undergoing AR within the duration of the experiment, the latency from time of stimulation to initiation of AR (not recordable for spontaneous AR), the duration from initiation to 90% completion of AR (from normalised fluorescent plots of each cell), the percentage of sperm remaining viable following initiation of AR, the time between AR initiation and cell death, and the percentage of sperm remaining motile following initiation of AR. *Significant compared with percentage of spontaneously occurring AR, P<0.05, Mann-Whitney Test. †Significant compared to duration of A23187-induced AR, P<0.05, Mann-Whitney Test. n, number of experiments
These data show that exocytosis of the human sperm acrosome has unusual kinetics that reflect the requirements of fertilisation. The initial phase of the sperm exocytotic response is a compound fusion event of the OAM and overlying plasma membrane, requiring SNAREs (soluble NSF attachment protein receptors) (De Blas et al., 2005). It appears that this process, as in other cells employing this type of fusion machinery, is rapid. The time-course observed here (⩽1 minute) includes pore expansion, vesiculation and the time for Alexa-488–SBTI to accumulate on the acrosomal content upon fenestration. However, membrane fusion is followed by a uniquely slow dispersal phase, with acrosomal content being retained over the IAM during penetration of the egg vestments. Our observations are complemented by data from previous studies on fixed sperm that have suggested that the AR is a dual-stage process (Kligman et al., 1991) or possibly a continuous progression of many stages (Kim et al., 2001) taking several minutes (Jungnickel et al., 2007). The slow dispersal of acrosomal content described here might explain reports of `partial AR' in assays on sperm fixed in vitro (Jaiswal et al., 1999), which could simply reflect fixation prior to completion of dispersal. Whether the acrosomal contents are retained for such an extended period in vivo, when the sperm will be experiencing physical forces from penetration of the egg, is still to be confirmed. Occasionally, the acrosomal matrix failed to disperse (Fig. 1Diii). This might reflect a form of kiss-and-run or `kiss-and-hold' (Troyer and Wightman, 2002), in which exocytosis reversed or arrested after formation of a fusion pore. Alternatively the acrosomal matrix of these few cells might be effectively insoluble. By contrast, dispersal of a single chromaffin granule, detected by amperometry, is complete in <20 milliseconds (Pihel et al., 1996). A 700 nm diameter dense-matrix mast-cell granule disperses within 500 milliseconds (Pihel et al., 1996), yet this granule has a larger volume than the human acrosome and, even discounting the IAM, a smaller surface-area:volume ratio.
The AR has recently been proposed as a model for exocytotic membrane fusion (Tomes, 2007). If initiation and progression of membrane fusion is dependent upon specific proteins (such as SNAREs or Ca2+ channels), then fusion would spread through the acrosome (effectively a 2D organelle) from an initiation point. This may be reflected by the sites of initial Alexa-488–SBTI and fluorescent anti-CD46 binding. To investigate this we recorded the site of initial fluorescence uptake within three regions of the acrosome (anterior, mid-section and posterior; Fig. 2A). The site of initiation of the AR, observed as uptake of Alexa-488–SBTI, was not random (P<0.01, chi-squared test), occurring most frequently in the anterior acrosome (66.2±5.7%; Fig. 2B,C). This is consistent with previous reports in human (Yudin et al., 1988) and macaque (Tollner et al., 2003) sperm. By contrast, the location of initial binding of fluorescent anti-CD46 was random, indicating that, after the spread of fenestration from the anterior initiation point, subsequent dispersal of acrosomal contents was not spatially regulated (P>0.05, chi-squared test; Fig. 2B,D).
Kinetics of the acrosome reaction and relationship to cell death and motility
Because of the limitations of the current methods for assessment of the AR, little, if anything, is known about the relationship between the AR, and the viability and motility of sperm. We now describe this relationship in living human sperm detected by simultaneous recording of the AR (monitored by accumulation of FITC–anti-CD46) and propidium iodide (PI; a red nucleic-acid dye that is excluded from viable cells). Viability was maintained during A23187-induced AR in over 90% of sperm and was, on average, sustained for 30 minutes (Table 1, Fig. 3A,B, Fig. 2B and supplementary material Movie 2). By contrast, although only 14% of sperm accumulated PI during a 60-minute incubation, sperm that underwent spontaneous AR were from this subpopulation (94.7±4.3% of spontaneous AR cells; Table 1, Fig. 3C,D), with PI uptake occurring prior to or co-incident with initiation of the AR. However, over 40% of sperm undergoing progesterone-induced AR remained viable for several minutes following the initiation of AR (Table 1, Fig. 3E, supplementary material Fig. S2B). Analysis of motility, by simultaneous capture of bright-field images, revealed that a loss of motility either preceded or correlated with the initiation of both A23187-induced and spontaneous AR (Fig. 3A,C, arrows show the timing of motility loss). This suggests that, if these findings are mirrored in vivo, human sperm undergoing spontaneous AR in the female tract prior to reaching the egg would not reach the oocyte. Almost 10% of sperm undergoing progesterone-induced AR remained motile (Table 1, supplementary material Movie 3), although motility appeared to become sluggish (Fig. 3F). Importantly, within a field of sperm there was no evidence of non-random clumping in induction of AR (an acrosome-reacting sperm did not induce AR in `bystander' sperm).
The observed loss of viability in acrosome-reacting cells might be in part a reflection of the cell immobilisation that was necessary for monitoring the AR (although this might also resemble events occurring in zona-bound sperm). In our experiments, spontaneous AR was strongly associated with (probably caused by or a cause of) cell death. Whether spontaneous AR is a selective mechanism in humans (to remove poor-quality sperm) is yet to be established. We have suggested previously that progesterone-induced AR serves this function, removing cells that fail adequately to regulate Ca2+ (Publicover et al., 2007). Therefore, in humans, fertilisation will be restricted primarily to sperm that reach the egg with their acrosome intact and undergo induced AR on the surface of the zona.
In summary, we describe the first observations and quantitative information on the kinetics of the AR, a form of exocytosis adapted to its unique function and crucial role in mammalian (and therefore human) fertilisation. Initially, there is rapid fusion and vesiculation of the OAM with the overlying plasma membrane, with fusion apparently spreading from the apical part of the acrosome. The cell then retains the acrosomal contents attached to the IAM, for several minutes following their exposure, despite the high surface-area:volume ratio of the acrosome. After 8-12 minutes, the acrosomal contents are sufficiently dispersed to reveal sites on the IAM that are required for gamete fusion. Cells undergoing AR lose viability and motility, either immediately or after a period of minutes. Spontaneous or prematurely induced AR will therefore eliminate these cells from the `race'. Together, these data provide a completely new, real-time view of agonist-activated SNARE-dependent exocytosis in a cell in which this fusion is a single, irreversible step, providing a major advance in the understanding of this pivotal event. In a broader context, this information is important for future understanding of fertilisation and male infertility.
Materials and Methods
Antibodies and fluorescent probes
FITC- and PE-conjugated murine monoclonal antibodies to human CD46 (E4.3) were from BD Biosciences (UK). Fluorescently conjugated SBTI (Alexa-488–SBTI) was from Invitrogen (UK).
Preparation and capacitation of spermatozoa
Freshly ejaculated human semen samples were obtained from healthy volunteers with informed consent and local ethical-committee approval, and allowed to liquefy for 30 minutes. Highly motile sperm were recovered by direct swim-up in Sperm Wash Modified–Human Tubular Fluid medium (Conception Technologies, USA) (Cummerson et al., 2006). Sperm concentration was adjusted to 1.5×106 cells/ml and 1 ml aliquots were capacitated for at least 4 hours at 37°C, 5% CO2.
Real-time fluorescence imaging
Confocal imaging was carried out using a Zeiss LSM510 Meta with an XL incubator (maintained at 37°C, 5% CO2, in humid conditions) through a fluar 40× (1.3 NA) oil-immersion objective unless otherwise stated. FITC and Alexa Fluor 488 were excited at 488 nm with emitted light captured through a 505-550 nm bandpass filter. PE was excited at 488 nm with emitted light captured through a 560 nm longpass filter. Images were captured every 30 seconds unless stated otherwise using Time Series software (LSM510 or AutoTimeserie v.1.19 LSM32) (Rabut and Ellenberg, 2004) with autofocus.
Visualisation of the acrosome reaction using Alexa-488–SBTI and PE–anti-CD46
Capacitated sperm were plated onto poly-D-lysine (10%; Sigma, UK)-pre-coated 35 mm glass-based dishes (IWAKI, Japan) and incubated for 20 minutes (37°C) to enable the sperm to adhere. Non-adhered sperm were removed by media replacement. Sperm were incubated with 5% mouse serum followed by addition of Alexa-488–SBTI (2 μg/ml) and PE–anti-CD46 (0.05 μg/ml) prior to imaging. At a recorded time point during imaging, 10 μM A23187 was added to the dish to induce the AR. The initial site of probe uptake was observed using a fluar 100× (1.45 NA) oil-immersion objective.
Simultaneous recording of acrosome reaction, viability and flagellar activity
Cells were imaged in the presence of the green-fluorescent anti-CD46 conjugate (FITC–anti-CD46) to enable simultaneous recording with PI (5 μg/ml; Sigma, UK) for detection of the AR and viability. Where described, stimulants of the AR (10 μM A23187 or 3 μM progesterone; Sigma, UK) were added directly to the dish. Bright-field transmission images were captured simultaneously with fluorescence images, enabling visualisation of flagella movement throughout the duration of each experiment. Motility was assessed by movement during the raster scan inherent in laser scanning microscopy.
Single-cell data processing and analysis
Data were processed using AQM Advance 6 (Kinetic Imaging, UK). Raw intensity values from the whole head (for measurement of the AR and cell death) were taken for each sperm. Time of occurrence of the AR and cell death was recorded as the first point at which dye began to accumulate in the sperm head (assessed visually and by fluorescence uptake plots). Data from each sperm undergoing the AR/cell death were normalised as a percentage of maximum fluorescence increase for each individual sperm (minimum values immediately before the AR/cell death; maximum being maximum recorded fluorescence values following the AR/death). Although variability in fluorescence intensity values within experiments was low (see supplementary material Fig. S1B), values were normalised because of variation between experiments. Complete AR was recorded as the time taken to reach 90% increase in fluorescence per sperm.
All data are presented as mean values ± s.e.m. unless stated otherwise. All statistical tests are two-tailed Mann-Whitney tests for nonparametric data unless stated otherwise.
The authors thank Jane Rees, Chris Barratt, John Aitken, Chris Wood, Alan Morgan, Dave Spiller and Alan McNeilly for comments and suggestions on the manuscript. The Centre for Cell Imaging was supported by the BBSRC.