Astaxanthin Prevents Human Papillomavirus L1 Protein Binding in Human Sperm Membranes

Astaxanthin (Asta), red pigment of the carotenoid family, is known for its anti-oxidant, anti-cancer, anti-diabetic, and anti-inflammatory properties. In this study, we evaluated the effects of Asta on isolated human sperm in the presence of human papillomavirus (HPV) 16 capsid protein, L1. Sperm, purified by gradient separation, were treated with HPV16-L1 in both a dose and time-dependent manner in the absence or presence of 30 min-Asta pre-incubation. Effects of HPV16-L1 alone after Asta pre-incubation were evaluated by rafts (CTB) and Lyn dislocation, Tyr-phosphorylation (Tyr-P) of the head, percentages of acrosome-reacted cells (ARC) and endogenous reactive oxygen species (ROS) generation. Sperm membranes were also analyzed for the HPV16-L1 content. Results show that HPV16-L1 drastically reduced membrane rearrangement with percentage of sperm showing head CTB and Lyn displacement decreasing from 72% to 15.8%, and from 63.1% to 13.9%, respectively. Accordingly, both Tyr-P of the head and ARC decreased from 68.4% to 10.2%, and from 65.7% to 14.6%, respectively. Asta pre-incubation prevented this drop and restored values of the percentage of ARC up to 40.8%. No alteration was found in either the ROS generation curve or sperm motility. In conclusion, Asta is able to preserve sperm by reducing the amount of HPV16-L1 bound onto membranes.


Introduction
HPV (Human Papilloma Virus) is responsible for the 5.2% (3% in women and 2% in men) of cancers in the world, with prevalence in cervical, ano-genital, head, and neck cancers [1]. Oncogenic strains of HPV DNA were demonstrated in almost all cervical malignancies in women [2], and, on a global scale, also in 5% of men [3], in the form of penile cancer (45%) with the HPV 16 member mostly responsible (60%) followed by the HPV 18 accounting for 13% of the cases [4].
Besides sexual inter-infections between sexual partners leading to the wider virus propagation, it has been recently shown that HPV DNA can cause a detrimental effect on early embryo development and clinical reproductive outcomes, since HPV DNA can be harbored inside the blastocyst stage

L1 Treatment: Effect on Sperm L1 Location
Aliquots of sperm, isolated and pooled as described in Methods, were analyzed for L1 protein binding and distribution with immunocytochemistry, immediately (T 0 ), or after incubation in capacitating conditions, in the absence (C), or presence of 1 (L1 1), 10 (L1 10) or 13 (L1 13) µg/mL of L1 protein (Figure 1a). The expected absence of anti-L1 fluorescence in T 0 and C confirmed that samples were free from any HPV pre-existing infection. Conversely, L1-capacitated sperm fluorescence indicated that L1 bound to cells in a concentration-dependent way, with a percentage of marked cells at 1, 10 and 13 µg/mL reaching 55.3% ± 3.6%, 98.1% ± 1.6% and 99.1% ± 0.7% of cells compared to control 0%, respectively (p < 0.0001) (Figure 1b). Due to its ability to bind to all sperm, a concentration of 10 µg/mL was chosen as suitable concentration to further investigate the time-dependent effect.

L1 Treatment: Effect on Sperm L1 Location
Aliquots of sperm, isolated and pooled as described in Methods, were analyzed for L1 protein binding and distribution with immunocytochemistry, immediately (T0), or after incubation in capacitating conditions, in the absence (C), or presence of 1 (L1 1), 10 (L1 10) or 13 (L1 13) µg/mL of L1 protein (Figure 1a). The expected absence of anti-L1 fluorescence in T0 and C confirmed that samples were free from any HPV pre-existing infection. Conversely, L1-capacitated sperm fluorescence indicated that L1 bound to cells in a concentration-dependent way, with a percentage of marked cells at 1, 10 and 13 µg/mL reaching 55.3% ± 3.6%, 98.1% ± 1.6% and 99.1% ± 0.7% of cells compared to control 0%, respectively (p < 0.0001) (Figure 1b). Due to its ability to bind to all sperm, a concentration of 10 µg/mL was chosen as suitable concentration to further investigate the timedependent effect.  . Before (T 0 ) or after (T 1 ) 180 min of incubation in capacitating conditions in absence (C) or presence of L1 (1, 10 or 13 µg/mL), sperm were analyzed for L1 presence by immunofluorescence cytochemistry as described in Methods. Similarly, time-dependent evaluation was carried out with L1 (10 µg/mL) for 30, 60, 90, 180 min. (a) Immunofluorescence analysis was performed with anti-L1 antibody (green) and Hoechst (blue) was used to visualize nuclei. Corresponding phase-contrast images for each condition were shown. (b) Dose dependent evaluation of protein L1 localization in C, L1 1 µg/mL, L1 10 µg/mL, and L1 13 µg/mL samples. The graph in the upper side shows the percentage of cell marked by L1 protein, whereas the histogram at the bottom shows L1 distribution between the mid-piece (filled blue) and the head (striped). (c) Time dependent evaluation of protein L1 localization in sample L1 10 µg/mL at different incubation times (30, 60, 90, 180 min). The number of sperm stained with anti-L1 antibody in any part of the cells (stained) is expressed as % of the total number of cells analyzed, whereas the number of cells showing L1 in the head or mid-piece is expressed as means ± SD% of marked cells. The figure is representative of 12 separate experiments conducted in triplicate ( a : p < 0.0001 comparing each sample against C values, by using Dunnett's test, following a significant one-way ANOVA).
Experiments with L1 10 µg/mL were carried out for 30, 60, 90, 180 min of incubation and sperm fluorescence evaluated. Interestingly, already after 30 min of L1 incubation, 91.8% ± 3.7% of sperm presented L1 related fluorescence (p < 0.0001 compared to control), which was progressively more marked by prolonging incubation time up to 3 h.
A qualitative analysis was also performed to evaluate the main site of protein binding (bottom panels of b and c). In the dose-dependent experiments, almost 100% of the marked cells showed fluorescence at the mid-piece, eligible as the principal binding point right after the first 30 min of incubation with 10 µg/mL, HPV L1 ( Figure 1c). Interestingly, with 1 µg/mL only 20% of the marked cells showed L1 binding also in the head, whereas the maximal effect was observed with 10 µg/mL (60%). The L1 distribution between mid-piece and head did not change with incubation time (Figure 1, panels b and c).

L1 Treatment: Effect on Sperm Motility
Aliquots were then assessed for motility by computer-assisted sperm analysis (CASA) ( Table 1). Samples were incubated in the presence of increasing L1 concentration (1, 10, and 13 µg/mL as representing the concentrations able to infect 100% of cells), after a pre-incubation of 30 min in the presence or absence of 2 µM Asta [20,21]. The increase of values of all parameters (except motility) is evident with all treatments compared to T 0 , meaning that capacitation had occurred leading to hyper-motility. However, when values from treated cells were compared to C (capacitated without Asta and/or L1), there was no significant difference, also for samples pre-incubated alone or with Asta and L1. Table 1. Sperm motility and kinematic parameters observed in different samples. Motility and kinematic parameters of spermatozoon were evaluated with computer-assisted sperm analysis (CASA) at T 0 (before starting incubation) and after 180 min of incubation in capacitating conditions in absence (C) or presence of L1, Asta or both. Motility = progressive and non-progressive motility (%); VSL = straight-line velocity (µm/s); VAP = average path velocity (µm/s); ALH = amplitude of lateral head displacement (µm).

Effect of L1 and Asta Alone or in Association on Sperm Capacitation
We previously showed that capacitation of sperm is linked to membrane rearrangement with lipid rafts relocation to the apical part of the head [20]. This rearrangement allows the gathering of the src kinase family member, Lyn, which activates and increases the Tyr-phosphorylation of the sperm head proteins [21]. To evidence the effects of L1 on sperm capacitation parameters, aliquots of sperm treated with L1 (10 µg/mL), after a pre-incubation of 30 min in the presence or absence of Asta, as described above, were analyzed for their shifted rafts (CTB), Lyn location, Tyr-P, and acrosome reaction (ACR).
L1 incubation affected rafts relocation ( Figure 2, CTB), with the sample showing only 20% of sperm with rafts relocated on the head (sample L1 compared to C). The corresponding quantification is in Table 2 (21.3% ± 4.2% compared to 72.0% ± 3.6% for L1 10 µg/mL and C, respectively, p < 0.001). These results suggest that capacitated-related hyper-motility was not affected by Asta, by L1, or by both Asta and L1.

Effect of L1 and Asta Alone or in Association on Sperm Capacitation
We previously showed that capacitation of sperm is linked to membrane rearrangement with lipid rafts relocation to the apical part of the head [20]. This rearrangement allows the gathering of the src kinase family member, Lyn, which activates and increases the Tyr-phosphorylation of the sperm head proteins [21]. To evidence the effects of L1 on sperm capacitation parameters, aliquots of sperm treated with L1 (10 µg/mL), after a pre-incubation of 30 min in the presence or absence of Asta, as described above, were analyzed for their shifted rafts (CTB), Lyn location, Tyr-P, and acrosome reaction (ACR).

Figure 2.
Localization of L1 protein, membrane rafts, Lyn and Tyr-P protein in human sperm during capacitation in absence or presence of L1 and/or Asta. Sperm cells, at T0 or incubated in capacitating conditions for 180 min in absence (C) or presence of L1 10 µg/mL, Asta 2 µM or both, were analyzed for L1 localization (green), CTB labelling (green), Lyn (red) and Tyr-P (green) localization by immunofluorescence cytochemistry as described in Methods. Hoechst (blue) was used to visualize nuclei and images were merged. The figure is representative of 12 separate experiments. Figure 2. Localization of L1 protein, membrane rafts, Lyn and Tyr-P protein in human sperm during capacitation in absence or presence of L1 and/or Asta. Sperm cells, at T 0 or incubated in capacitating conditions for 180 min in absence (C) or presence of L1 10 µg/mL, Asta 2 µM or both, were analyzed for L1 localization (green), CTB labelling (green), Lyn (red) and Tyr-P (green) localization by immunofluorescence cytochemistry as described in Methods. Hoechst (blue) was used to visualize nuclei and images were merged. The figure is representative of 12 separate experiments. Table 2. Sperm biochemical parameters observed in different samples. Sperm biochemical parameters were evaluated with fluorescence microscopy at T 0 (before starting incubation) and after 180 min of incubation in capacitating conditions in absence (C) or presence of L1 (L1 1, 10 or 13 µg/mL), Asta or both (Asta+L1 10 µg/mL) as described in Methods.
The lack of translocation causes the consequent decrease of Lyn gathered on the apical part of the head (Figure 2, Anti-Lyn) ( Table 2, 16.3% ± 3.5% compared to 63.1% ± 2.5% for L1 10 µg/mL and C, respectively, p < 0.001), followed by the net reduction of the percentage of cells showing Tyr-P of the head (Figure 2, Anti-P-Tyr) ( Table 2).
In these conditions only a reduced percentage of cells underwent the acrosome reaction compared to the C sample (19.4% ± 2.8% of acrosome-reacted cells (ARC) in the L1 10 µg/mL sample compared to 65.7% ± 4.2% in C sample, p < 0.001) ( Table 2).
When Asta was added to the capacitating medium, the values of capacitation only slightly improved, as previously shown [21], since sperm were already in optimal conditions and did not need membrane raft translocation (CTB), Lyn gathering, and Tyr-P. Asta also prevented L1 binding to the membranes of sperm ( Figure 2 Anti-L1, compare L1 with Asta+L1 panels) with a net reduction of protein of about 52% (98.1% ± 1.6% of sperm infected by L1 compared to 45.5 ± 4.2% in the presence of Asta, p < 0.001, Table 2). In the same way, also Lyn relocation and the following Tyr-P of the head were partially restored. Our results show that Asta can recover about 50% of the L1-induced alterations, with cell showing rafts relocation switching from 21.3% ± 4.2% to 50.2% ± 2.7% (p < 0.001), Lyn relocation from 16.3% ± 3.5% to 43.2% ± 3.0% (p < 0.001) and Tyr-P of the head from 11.3% ± 1.7% to 43.6% ± 3.9% (p < 0.001) ( Table 2, comparing line L1 10 µg/mL to Asta+L1 10). Consequently, also the L1-induced reduction of the ARC percentage was greatly restored by the presence of Asta (Table 2), shifting from 19.4% ± 2.8% to 40.8% ± 2.6% (p < 0.001), which, also if still far from the 65.7% ± 4.2% of the C sample, accounts for twice the value of the L1-treated sample (L1 10 µg/mL: 19.4% ± 2.8%).

Effect of Asta on L1 Binding to Sperm Membranes
Samples were also analyzed for the L1 protein binding to the subcellular fractions. After incubation in the absence or presence of Asta, L1, or both, aliquots from each sample were sonicated, sub-cellular fractions were separated on gradient and further centrifuged, as described in Methods. Membranes, cytosol, heads, and flagella were subjected to Western blotting and revealed with ant-L1 antibodies.
L1 was detected on membranes from L1-treated sperm, (Figure 3, lanes L1 and Asta+L1), but not in T 0 , C and Asta samples did not show any response, as expected. The sample treated with Asta+L1 showed a reduction in the L1 content, with Asta reducing the amount of L1 by about 30%, compared to L1 sample (206 ± 17 vs. 145 ± 14 ng of L1 protein/30 × 10 6 cells p < 0.001).
When analyzed in other compartments, such as heads and tails, L1 detection was practically negligible ( Figure S1), thus indicating that the main site of L1 binding was on the membrane ( Figure S2) Compared to data in Figure 1 and Table 2, showing that Asta decreased the percentage of marked cells from 98.1% ± 1.6% to 45.5% ± 4.2% (p < 0.001) ( Table 2) seems not to account for the higher reduction found with immunofluorescence. We must consider that, when evaluated with immunocytochemistry, we account for the number of cells presenting L1-related fluorescence, without considering the amount of the protein effectively linked. By observing the L1 distribution between mid-piece and head, it is reasonable that the binding of L1 starts from the mid-piece and gradually spreads over the whole cell, thus increasing the concentration of L1 for each cell.
Compared to data in Figure 1 and Table 2, showing that Asta decreased the percentage of marked cells from 98.1% ± 1.6% to 45.5% ± 4.2% (p < 0.001) ( Table 2), data from the Western blotting of membranes confirmed the net reduction of the amount of L1 in the presence of Asta. The decrease seems not to account for the higher reduction found with immunofluorescence. We must consider that, when evaluated with immunocytochemistry, we account for the number of cells presenting L1related fluorescence, without considering the amount of the protein effectively linked. By observing the L1 distribution between mid-piece and head, it is reasonable that the binding of L1 starts from the mid-piece and gradually spreads over the whole cell, thus increasing the concentration of L1 for each cell.

Effect of Asta and L1 on ROS Production
When the effects of the different treatments were assayed on the ROS generation curve ( Figure  4), samples were incubated in the absence (C), or presence of Asta, or L1, or both (Asta+L1), and analyzed for ROS generation in a luminometer for 180 min with luminol as luminescent source.

Effect of Asta and L1 on ROS Production
When the effects of the different treatments were assayed on the ROS generation curve (Figure 4), samples were incubated in the absence (C), or presence of Asta, or L1, or both (Asta+L1), and analyzed for ROS generation in a luminometer for 180 min with luminol as luminescent source.
All samples showed similar curves of ROS generation compared to the control (C), thus suggesting that none of the treatments affected H 2 O 2 formation in sperm.
All samples showed similar curves of ROS generation compared to the control (C), thus suggesting that none of the treatments affected H2O2 formation in sperm. To initiate a successful infection, HPV must bind to heparan sulfate proteoglycan (HSPG) by exposing L1 from its capside structure and through L1 form an endocytic complex [25], that is also responsible for virus internalization [26].
It has been recently confirmed that also in human sperm, GAGs mediate the viral binding to the cell surface, in particular the GAG syndecan-1 (Synd-1) seems to interact with the viral capsid protein L1 of HPV16 [27,28]. Synd-1 belongs to the heparan sulfate GAGs (HSGAGs) family which regulates cell proliferation and cell-matrix and cell-cell adhesion by modulating the ligand-dependent activation of GAG receptors at the cell surface, interacting with components of the actin-based cytoskeleton with intracellular domain [29]. A direct interaction between the extracellular portion of Synd-1 and HPV-L1 is involved in L1 binding to sperm membrane, as confirmed by heparinasemediated abolition of the co-localization of Synd-1 and HPV-1 [25,26].
In the present study, the L1 binding would act as an anchor complex, preventing further rafts translocation and relocation to the sperm head. This hypothesis is consistent with the results of Chen et al. [30] who demonstrated that Synd-1, clustered upon L1 ligand binding, induced recruitment and binding to cortactin, a protein involved in the membrane-cytoskeleton formation and modulation through actin-cortactin net interaction. From these considerations, we postulate that the probable binding site in human sperm membrane may be included in raft domains, thus being regulated and regulating, in turn, further L1 binding/inclusion and raft relocation. Asta, binding to plasma membranes, probably alters raft composition by sterically filling the loci dedicated to the interaction between external viral L1 and internal Synd.1. The possible mechanism underlying Asta-related induction of capacitation would rely on its ability to insert into the lipid bilayer, disengaging rafts from the blocks that maintain the membrane in non-capacitated form. Consistently, Asta would also release the protein links mediated by L1 and cytoskeleton proteins, both allowing/facilitating the correct engagement of proteins in rafts, such as in the case of actin polymerization [31], and the correct relocation of enzymes, such as Lyn, to the acrosome region. To initiate a successful infection, HPV must bind to heparan sulfate proteoglycan (HSPG) by exposing L1 from its capside structure and through L1 form an endocytic complex [25], that is also responsible for virus internalization [26].
It has been recently confirmed that also in human sperm, GAGs mediate the viral binding to the cell surface, in particular the GAG syndecan-1 (Synd-1) seems to interact with the viral capsid protein L1 of HPV16 [27,28]. Synd-1 belongs to the heparan sulfate GAGs (HSGAGs) family which regulates cell proliferation and cell-matrix and cell-cell adhesion by modulating the ligand-dependent activation of GAG receptors at the cell surface, interacting with components of the actin-based cytoskeleton with intracellular domain [29]. A direct interaction between the extracellular portion of Synd-1 and HPV-L1 is involved in L1 binding to sperm membrane, as confirmed by heparinase-mediated abolition of the co-localization of Synd-1 and HPV-1 [25,26].
In the present study, the L1 binding would act as an anchor complex, preventing further rafts translocation and relocation to the sperm head. This hypothesis is consistent with the results of Chen et al. [30] who demonstrated that Synd-1, clustered upon L1 ligand binding, induced recruitment and binding to cortactin, a protein involved in the membrane-cytoskeleton formation and modulation through actin-cortactin net interaction. From these considerations, we postulate that the probable binding site in human sperm membrane may be included in raft domains, thus being regulated and regulating, in turn, further L1 binding/inclusion and raft relocation. Asta, binding to plasma membranes, probably alters raft composition by sterically filling the loci dedicated to the interaction between external viral L1 and internal Synd.1. The possible mechanism underlying Asta-related induction of capacitation would rely on its ability to insert into the lipid bilayer, disengaging rafts from the blocks that maintain the membrane in non-capacitated form. Consistently, Asta would also release the protein links mediated by L1 and cytoskeleton proteins, both allowing/facilitating the correct engagement of proteins in rafts, such as in the case of actin polymerization [31], and the correct relocation of enzymes, such as Lyn, to the acrosome region.

Semen Collection and Analysis
Forty healthy male donors (age range: 25-43 years, average age: 34.6 years) were enrolled at the Centre of Assisted Reproduction-U.O.C. Obstetrics and Gynecology Clinic-Padua, Italy. After 3 days of abstinence, semen samples were collected by masturbation in a sterile container and then assessed for sperm parameters. All sperm samples used in this study were normal in terms of sperm count, motility, morphology, volume, and pH, according to the World Health Organization criteria [32]. All samples presenting any kind of contamination were discarded. This study was approved by the Ethics Committee for Research and Clinical Trials of our University (Code: 4400-AO18), and all recruited donors gave their informed written consent and provided detailed lifestyle histories.

Sample Preparation
After semen analysis, samples were laid on a discontinuous gradient (Pure Sperm 40%/80%) and centrifuged at 500× g for 30 min at room temperature. The seminal plasma and sperm from the 40% gradient interface were discarded, and the sperm cells from the bottom pellet (80% gradient) were gathered. After gradient separation, sperm samples were washed with PSW and collected three by three in a single pool (stock sample) to obtain a sufficient number of cells to perform all tests. Stock samples (concentration adjusted to 80 × 10 6 sperm cells/mL in PSW) were divided in aliquots, analyzed immediately (T 0 ), or incubated for up to 180 min in capacitating conditions, in the absence (C) or presence of Asta 2 µM (Asta) from stock solutions of 100 mM dissolved in dimethyl sulphoxide (DMSO), L1 1, 10 or 13 µg/mL (L1), or together with 30 min-Asta pre-incubation (Asta+L1 10).

Computer Assisted Sperm Analysis (CASA)
Sperm motility and hyperactivation were analyzed using a computer-assisted sperm analyzer (CASA, Ab scientific, London, UK). For each sample, the following parameters were evaluated: the percentage of motile spermatozoa and VAP (average path velocity), VSL (straight-line velocity), and ALH (amplitude of lateral head displacement) to determine the percentage of hyper-activated (HA) cells [33]. All measurements were performed at 37 • C. A minimum of 100 cells and 5 fields were analyzed for each aliquot.

Anti-L1, Anti-P-Tyr and Anti-Lyn Evaluations with Confocal Microscopy
Aliquots of sperm (15 × 10 6 cells) from each sample were accurately washed with phosphate buffer saline (PBS) containing vanadate 1 mM and protease inhibitor cocktail, fixed with 2% (w/v) paraformaldehyde and incubated overnight at 4 • C on slides pre-coated with poly-L-lysine [20][21][22]. Slides were rinsed twice with PBS and sperm cells were permeabilized with 0.2% (v/v) Triton X-100 for 15 min at 4 • C and then incubated with anti-L1, anti-P-Tyr, or anti-Lyn antibodies for 1 h at 37 • C in a humid chamber. Slides were washed with PBS, stained with anti-mouse or anti-rabbit IgG-FITC conjugate for 1 h at 37 • C in a humid chamber and then rinsed with PBS and mounted. Staining without primary antibody was used as negative control. Fluorescence was detected with the UltraView LCI confocal system (Perkin Elmer, Waltham, MA, USA).

Evaluation of Membrane Rafts
GM1 membrane raft marker was visualized in live human spermatozoa by staining with the cholera toxin subunit B (CTB)-FITC [22,34]. For this purpose, suspensions of cells (15 × 10 6 cells) from each sample were mixed with an equal volume of CTB (50 µg/mL) and incubated for 15 min at 37 • C. The sperm cells were then washed twice in PBS before being fixed in 2% paraformaldehyde for 30 min, mounted on poly-L-lysine coated glass microscope slides, and viewed using the confocal microscope as described above. For each treatment, at least 200 cells were counted.

Evaluation of Acrosome Reaction
Acrosome status was monitored with acrosome-specific FITC-labeled peanut (Arachis hypogaea) agglutinin (FITC-PNA) in conjunction with DNA-specific fluorochrome propidium iodide (PI) as a viability test [23,24]. Briefly, in order to induce AR, aliquots (15 × 10 6 cells) of each sample were incubated for 30 min at 37 • C, in the presence of 10 µM Ca 2+ ionophore A23187. Samples containing DMSO, but not ionophore, were used as control. After incubation sperm cells were centrifuged, resuspended in PBS, and treated for 10 min at room temperature with 12 µM PI. Sperm was washed with PBS, fixed with 2% (w/v) paraformaldehyde and incubated overnight at 4 • C on poly-L-lysine-treated slides. Permeabilized sperm cells, as described above, were stained with 1 mg FITC-PNA/mL for 15 min at 37 • C in the dark, washed and mounted. At least 200 cells were evaluated for each sample, and fluorescence was detected as described above. Only sperm cells showing evenly distributed fluorescence over the acrosomal region were considered acrosome-intact.

Protein L1 Distribution
To determine whether and how much L1 was present in the plasma membrane, head or flagellum, intact spermatozoa (30 × 10 6 cells) from each treatment were accurately washed in PBS-VI (containing vanadate 1 mM and protease inhibitor cocktail), resuspended in 200 µL of the same PBS-VI, and sonicated 3 times (30 s followed by a 10-s rest period each) on ice. Heads and flagellar fragments were then separated by a 15-min centrifugation (700× g) at 4 • C through a 75% Percoll layer in PSW. Flagellar fragments were recovered at the surface of the Percoll layer while the heads were found in the pellet. The purity of each fraction was assessed by microscopy prior to proceeding to analysis. The supernatant was centrifuged for 10 min (10,000× g, 4 • C) and the resulting supernatant was further centrifuged (100,000× g) to separate the membrane from the cytosol [21]. Each resulting fraction (membranes M, cytosol C, head H and flagella F) was diluted with PBS to the initial volume except for membranes, which were resuspended in 30 µL of PBS-VI. The presence of L1 protein was investigated by Western blotting and immuno-revealed with anti-L1 antibody.

Western Blotting
Cell fractions (membrane, heads and tails) or corresponding cytosol ( Figure S1) were solubilized by adding SDS and β-mercaptoethanol (2% final concentration), boiled at 100 • C for 5 min, and subjected to SDS/PAGE (10% polyacrylamide gels). Proteins were electrotransferred to a nitrocellulose membrane and immunorevealed with anti-L1 antibodies. Loading control was performed with anti-actin antibodies [35].
Densitometric analysis. Bands corresponding to L1 protein bands were counted by ImageJ software 1.48v.

ROS Enhanced Chemiluminescence (ECL)
Production of ROS was measured by the chemiluminescence assay method with luminol (5-amino-2,3-dihydro-1,4-phthalazinedione) as probe [23,24]. Briefly, 2 µL of 25 mM luminol and 4 µL of 10 mg/mL horseradish peroxidase, both prepared in DMSO, were added to 200 µL of a sperm suspension at a concentration of 10 × 10 6 cells/mL. ROS levels were determined by a luminometer (Fluoroskan Ascent FL, Labsystems, Helsinki, Finland) in the integrated mode for 180 min at 37 • C. Results are expressed as Relative Luminescence Units (RLU) per 2 × 10 6 sperm cells. Lastly, 2 µL of a 10 mM N-formylmethionyl-leucyl-phenylalanine (FMLP) stock was added and, after a further 10 min of incubation, 4 µL of a 1 nM stock solution of PMA was added, to exclude leukocyte contamination. Only samples with negative response to FLMP and PMA were processed.

Statistical Analysis
Results are expressed as means ± SD. Comparisons among multiple groups were obtained with ANOVA followed by Dunnett's test. Statistical significance was set at p < 0.05 (two-tailed). Statistical analysis were performed with JMP ® 13 software (SAS Institute, Cary, NC, USA).

Conclusions
Recent evidence substantially strengthens the possibility that HPV is responsible not only for HPV-associated cancers, but also for idiopathic infertility due to its detrimental effect on sperm parameters [36][37][38] and the potential stage arrest of embryo development [39]. In this study we showed that L1, at 10 µg/mL, binds to quit 100% of sperm membrane, locating at the sides of mid-piece and head. This protein, blocks rafts translocation, Lyn displacement, and Tyr-P of the head, thus resulting in a dramatic decrease of the percentage of ARC. Asta treatment prevented L1 binding to the membranes reducing by more than 50% the presence of L1, thus protecting cells from HPV L1 binding. Further studies are needed to better clarify the mechanism of HPV L1 entry, and Asta could be considered for its potential antiviral effect.