Differences in the seminal plasma proteome are associated with oxidative stress levels in men with normal semen parameters

Differences in the seminal plasma proteome are associated with oxidative stress levels in men with normal semen parameters

ORIGINAL ARTICLE: ANDROLOGY Differences in the seminal plasma proteome are associated with oxidative stress levels in men with normal semen parameter...

653KB Sizes 5 Downloads 29 Views


Differences in the seminal plasma proteome are associated with oxidative stress levels in men with normal semen parameters Paula Intasqui, M.Sc.,a Mariana Pereira Antoniassi, M.Sc.,a Mariana Camargo, M.Sc.,a Marcílio Nichi, Ph.D., D.V.M.,b Valdemir Melechco Carvalho, Ph.D.,c Karina Helena Morais Cardozo, Ph.D.,c Daniel Suslik Zylbersztejn, M.D., Ph.D.,a and Ricardo Pimenta Bertolla, Ph.D.a ~ o Paulo Federal University–Sa ~o Paulo Human Reproduction Section, Division of Urology, Department of Surgery, Sa ~o Paulo; and c Fleury Hospital; b Department of Animal Reproduction, School of Veterinary Medicine, University of Sa ~o Paulo, Brazil Group, Sa a

Objective: To study the seminal plasma proteome in association with semen lipid peroxidation levels in men with normal semen parameters. Design: Cross-sectional study. Setting: University andrology and research laboratories. Patient(s): A total of 156 normozoospermic men. Intervention(s): Seminal lipid peroxidation levels were assessed in individual samples through thiobarbituric acid reactive substances quantification. Subsequently, lipid peroxidation data were used to divide the samples into the experimental groups: low lipid peroxidation levels (control group, bottom 15%, n ¼ 23) and high lipid peroxidation levels (study group, top 15%, n ¼ 23). Seminal plasma proteins from these groups were pooled (four pools per group, with biological variation between the pools) and used for a shotgun proteomic analysis using a liquid chromatography–tandem mass spectrometry approach. Quantitative data were used for univariate (unpaired Student's t test) and multivariate (partial least-squares discriminant analysis, logistic regression, and discriminant analyses) statistical analyses. Significant proteins were also used for functional enrichment analysis. Main Outcome Measure(s): Seminal plasma protein profile and postgenomic pathways of seminal plasma are associated with seminal lipid peroxidation levels. Result(s): In total, 629 proteins were quantified in seminal plasma. Of these, 23 proteins were absent or underexpressed and 71 were exclusive or overexpressed in the study group. The main enriched functions in association with seminal lipid peroxidation were unsaturated fatty acids biosynthesis, oxidants and antioxidants activity, cellular response to heat stress, and immune response. Moreover, we suggested mucin-5B as a potential biomarker of semen oxidative stress. Conclusion(s): The seminal plasma proteome does reflect semen lipid peroxidation status and, Use your smartphone thus, oxidative stress. (Fertil SterilÒ 2015;-:-–-. Ó2015 by American Society for Reproducto scan this QR code tive Medicine.) and connect to the Key Words: Lipid peroxidation, oxidative stress, proteomics, seminal plasma, sperm Discuss: You can discuss this article with its authors and with other ASRM members at http:// fertstertforum.com/intasquip-proteomics-seminal-oxidative-stress/

Received January 20, 2015; revised and accepted April 29, 2015. P.I. has nothing to disclose. M.P.A. has nothing to disclose. M.C. has nothing to disclose. M.N. has nothing to disclose. V.M.C. has nothing to disclose. K.H.M.C. has nothing to disclose. D.S.Z. has nothing to disclose. R.P.B. has nothing to disclose. This work was supported by Fleury projects funding, the National Council for Scientific and Techno~o Paulo Research Foundation logical Development (CNPq, process 472941/2012-7), and the Sa (scholarship for P.I., process 2012/14631-7). ~o Paulo Federal University, R. Embau, 231, Reprint requests: Daniel Suslik Zylbersztejn, M.D., Ph.D., Sa ~o Paulo, Sa ~o Paulo, Brazil (E-mail: [email protected]). 04039-060 Sa Fertility and Sterility® Vol. -, No. -, - 2015 0015-0282/$36.00 Copyright ©2015 American Society for Reproductive Medicine, Published by Elsevier Inc. http://dx.doi.org/10.1016/j.fertnstert.2015.04.037 VOL. - NO. - / - 2015

discussion forum for this article now.*

* Download a free QR code scanner by searching for “QR scanner” in your smartphone’s app store or app marketplace.


xidative stress is detected in semen of up to 40% of infertile men (1), and thus it is widely considered one of the main cellular mechanisms of male infertility (2–6). It has also been associated with idiopathic infertility, even in men with normal semen quality (7). Therefore, it is suggested that seminal oxidative stress may affect sperm physiology, especially through impairment of sperm functions (6, 8, 9). Human sperm 1

ORIGINAL ARTICLE: ANDROLOGY are notably sensitive to oxidative stress, owing to the high membrane concentration of unsaturated fatty acids (8) and to its limited antioxidant and membrane repair capacity (10, 11). Oxidative stress promotes its negative effects on sperm functions particularly by oxidation of seminal plasma and sperm lipids and proteins (9, 12–19). Oxidation of cellular and mitochondrial membrane lipids, known as lipid peroxidation, is a self-propagating process that ultimately leads to the loss of membrane integrity, permeability, and stability and to enzymes inactivation (12). Consequently, lipid peroxidation can lead to reduced sperm mitochondrial activity, acrosome damage, and DNA fragmentation (9, 12, 13). Additionally, oxidation of seminal plasma and sperm proteins produces carbonyl groups, rendering them highly susceptible to proteolysis (16). Furthermore, sulfhydryl groups of cysteine residues are highly affected by oxidation by reactive oxygen species (ROS), leading to accumulation of oxidized proteins and decreased protein function (14, 15). Thus, loss of free sulfhydryl groups is considered a marker of oxidative stress (20). Several studies have demonstrated a significant reduction in free sulfhydryl levels in seminal plasma of infertile men (18, 19), which was also associated with decreased sperm quality and function (17). It is suggested that the effects of oxidative stress are more pronounced in extracellular proteins, owing to their longer half-life and lower damage repair capacity (15). Therefore, we hypothesized that seminal oxidative stress may alter the seminal plasma proteome, especially because of [1] oxidation of seminal plasma proteins, [2] oxidation of sperm proteins, which are released to seminal plasma, [3] peptides derived from proteolysis of seminal plasma and sperm proteins, and [4] released proteins from dead or altered sperm associated with oxidative stress in the male reproductive tract. To test this hypothesis, a proteomic analysis of seminal plasma in association with seminal lipid peroxidation levels was performed. Our results demonstrated several altered proteins, mostly related to ROS production and metabolism.

MATERIALS AND METHODS Study Design This study was approved by the S~ao Paulo Federal University (UNIFESP, Brazil) institutional review board (Research Ethics Committee approval no. 1184/11). Informed written consent was provided by all included patients. We performed a cross-sectional study with men undergoing semen analysis at the Andrology Laboratory of the Human Reproduction Section, UNIFESP. Semen was retrieved by masturbation after a 2- to 5-day period of ejaculatory abstinence. Only men aged 20–50 years and with sperm concentration >15  106 sperm/mL were included. Subjects presenting ejaculate volume <1.5 mL, progressive motility <32%, morphology <4%, or leukocyte concentration R1  106 leukocytes/mL were excluded. Only men with normal semen parameters were included, to avoid bias in our proteomics findings, because some studies have already shown a difference in the seminal plasma proteome in asthenozoospermic and teratozoospermic men (21, 22). Additionally, men with leukocytospermia were also excluded, to avoid 2

the contamination of seminal plasma with leukocytes proteins. Therefore, we tried to isolate our study factor (oxidative stress), to detect the proteome changes indeed associated with it. Between July 2012 and November 2013, 233 samples complied with the inclusion criteria and were prospectively collected. Of these, 77 samples presented alterations in semen analysis and were, thus, excluded. Therefore, this study was carried out including 156 subjects (n ¼ 156). For all semen samples, an aliquot was used for semen analysis, according to the World Health Organization 2010 manual (23), and the remaining volume was centrifuged at 800  g for 30 minutes. The obtained seminal plasma was directly frozen and kept at 20 C to be further used for lipid peroxidation evaluation and proteomic analysis. Before these analyses, seminal plasma was thawed and centrifuged at 16,100  g for 1 hour at 4 C to remove cellular debris. All reagents used in this study were obtained from Sigma-Aldrich, unless otherwise described.

Seminal Lipid Peroxidation Evaluation Seminal peroxidation levels were assessed in individual samples by measuring the thiobarbituric acid reactive substances (TBARS) levels in seminal plasma, as previously described by Oborna et al. (24), with a few modifications. Thiobarbituric acid reactive substances is a red complex with absorbance peak at 540 nm, formed by the reaction of malondialdehyde (MDA), a by-product of lipid peroxidation, with two molecules of thiobarbituric acid (24). Semen lipid peroxidation levels were evaluated instead of sperm ROS levels for the following reasons. [1] ROS in sperm demonstrate only ROS levels generated by sperm, not providing any information regarding the sperm damage caused by ROS (that might be or not generated in sperm), whereas lipid peroxidation levels directly reflect this damage. [2] Reactive oxygen species are capable of oxidizing sperm and seminal plasma lipids (measured by semen lipid peroxidation levels), and proteins, which might lead to alterations on the seminal plasma proteome. Therefore, semen lipid peroxidation levels measured directly in the seminal plasma might be better correlated to seminal plasma proteomics data than ROS levels measured in sperm. [3] Some by-products of lipid peroxidation, such as MDA (measured in our study), are also capable of oxidizing sperm and seminal plasma lipids and proteins. Thus, this measure is more informative about the real effects of oxidative stress on the seminal plasma proteomics than measuring ROS levels. Briefly, 100 mL of a solution containing 8.1% (wt/vol) sodium dodecyl sulfate (GE Healthcare), 0.8% (wt/vol) thiobarbituric acid, and 20% (vol/vol) acetic acid (Carlo Erba Reagents) were added to 100 mL of seminal plasma. Samples were then incubated in a water bath at 100 C for 1 hour and cooled in ice for 5 minutes to stop the reaction. To separate the TBARS molecules, 250 mL of N-buthanol (LabSynth) were added to each sample, which were then homogenized in vortex for 1 minute and centrifuged at 16,100  g for 15 minutes at 15 C. Finally, 100 mL of the aqueous phase were transferred to a microplate, in duplicate. A standard curve VOL. - NO. - / - 2015

Fertility and Sterility® with known MDA concentrations (0–50 ng/mL; Acros Organics) and submitted to the same protocol was also added to the microplate, in duplicate. Quantification of TBARS was performed by spectrophotometry, in a microplate reader (ELx800 Absorbance Microplate Reader, Biotek), at 540-nm wavelength. Lipid peroxidation data were described as ng TBARS/mL semen. Grouping. Statistical analyses were performed using PASW (SPSS) 18.0 software for Windows. For experimental groups' separation, initially the distribution normality of the lipid peroxidation variable from the 156 included samples was evaluated, using histograms and the Kolmogorov-Smirnov test, and outlier samples were excluded. Then samples were divided into percentiles of 15%, because this cutoff enhanced the differences between the groups and minimized the differences within each group. Thus, samples from the bottom and top percentiles were selected as the experimental groups: low (control group, bottom 15%, n ¼ 23) and high (study group, top 15%, n ¼ 23) seminal lipid peroxidation levels. For comparison of semen variables between the experimental groups, the Kolmogorov-Smirnov test was used, and variables that did not obey the normality premises were transformed to their logarithmic values. Groups were then compared using an unpaired Student's t test, and the statistical significance was set to 5%.

Seminal Plasma Proteomic Analysis Sample preparation. Total protein concentration in the seminal plasma from patients included in the groups was evaluated using the bicinchoninic acid protein assay (25). Quantification was performed by measuring absorbance at 540 nm, using a microplate reader. Each sample was quantified in triplicate, and a quantification curve of different concentrations of bovine serum albumin (0–1,000 mg/mL) was generated in duplicate. Samples quantified with coefficients of variation over 5% were quantified again in another run. After protein quantification, four pools of each group were prepared, assuring biological variation between the pools (different samples in each pool). Samples were randomized into the pools, taking care that each sample contributed the same amount of total protein to the final pools (200 mg). Because each group was composed of 23 samples, we prepared three pools containing six samples each and one pool with five samples for each group. These pools were then quantified in triplicate using the assay mentioned above. Thereafter, the volume corresponding to 50 mg of proteins of each pool was separately diluted in Milli-Q water to a final volume of 50 mL, to which 10 mL of a 50-mM NH4HCO3 solution were added. The protein samples were then denatured with 25 mL of a 2-mg/mL RapiGest SF surfactant (Waters) solution in water for 15 minutes at 80 C. Then, disulfide bonds of the sulfur residues from proteins were reduced in the presence of 2.5 mL of 100 mM dithiothreitol (GE Healthcare) at 60 C for 30 minutes and alkylated with 2.5 mL of 300 mM iodoacetamide (GE Healthcare) at room temperature. Enzymatic digestion occurred at 37 C overnight with trypsin (Sequencing Grade Modified Trypsin, Promega) at 1:100 VOL. - NO. - / - 2015

(wt/wt) enzyme/protein ratio. Next, 10 mL of 5% (vol/vol) trifluoroacetic acid (Merck Schuchardt OHG) were added to the digestion mixture to hydrolyze the RapiGest, and the samples were incubated at 37 C for 90 minutes. The tryptic peptide solution was then centrifuged at 16,000  g for 30 minutes at 4 C, and the supernatant was transferred to Waters Total Recovery Vials. Each pool was analyzed in triplicate by Liquid chromatography–tandem mass spectrometry (LC-MS/MS). Liquid chromatography–tandem mass spectrometry. The LC-MS/MS experiments were conducted using a nanoACQUITY UPLC (Waters) system coupled with a hybrid QuadrupoleOrbitrap mass spectrometer (Thermo Fisher Scientific). For chromatographic runs, 2 mL (equivalent to 0.5 mg of tryptic digest) were loaded onto a PST C18 nanoACQUITY Trap column (180 mm  20 mm) with a flow rate set to 15 mL/min of 0.1% (vol/vol) TFA during 5 minutes. Analytic separation of the peptides was performed using a nanoACQUITY UPLC HSS C18 Column (1.8 mm, 75 mm  150 mm) with a flow rate of 0.4 mL/min and a temperature of 60 C. Reversed-phase chromatography was performed using a binary system consisting of 2% (vol/vol) dimethylsulfoxide and 0.1% (vol/vol) formic acid in water (solution A) and 5% (vol/vol) dimethylsulfoxide and 0.1% formic acid in acetonitrile (solution B). Chromatographic separation was performed as follows: from 0 to 5 minutes: 2% solution B; from 5 to 60 minutes: linear gradient from 2% to 20% solution B; from 60 to 100 minutes: linear gradient from 20% to 40% solution B; from 100 to 110 minutes: linear gradient from 40% to 90% solution B. The chromatography was coupled to a Q-Exactive mass spectrometer via the nanoelectrospray source (Proxeon Nanospray Flex, Thermo Fisher Scientific) operating in positive mode. The Q-Exactive instrument was operated using a data-dependent top-12 experiment. For each cycle, one full MS scan (m/z 390–1,650) was acquired in the Orbitrap at a resolution of 70,000 at m/z 200 with an automatic gain control target of 3e6. Each full scan was followed by the quadrupole selection and isolation of the 12 most intense ions at a window m/z 4, which were dissociated through higher-energy collisional dissociation, using normalized collision energy of 26 (MS/MS). The product ions were analyzed by the Orbitrap at a resolution of 17,500 at m/z 200 with an automatic gain control target of 5e4. Ions with an unassigned, þ1 and >þ8 charges were rejected, and dynamic exclusion was set to 30 seconds. Mass spectrometry raw files were analyzed by MaxQuant software (version and the Andromeda search engine for protein identification and quantification. A maximum of two missed trypsin cleavages were allowed in the database search. Cysteine carbamidomethylation was included as a fixed modification and protein N-terminal acetylation and methionine oxidation as variable modifications. The chromatograms were aligned by the software, allowing up to 2 minutes of distortion (in each direction), to maximize the MS/ MS spectra identification. Peak lists were searched against the revised human SwissProt Uniprot database (version 2014_02_19, downloaded February 19, 2014 with 20,263 entries and 13,081 kB). The false discovery rate was set to 1% 3

ORIGINAL ARTICLE: ANDROLOGY and was determined by searching in a reverse database. Quantification was performed through label-free quantification, using intensity-based absolute quantification (iBAQ), in which protein content is normalized to the total number of potential peptides (26). Data analysis. For each group, four biological replicates were formed, which were acquired in triplicate (technical replicates) during LC-MS/MS experiments. To maximize the number of observations in each study and because MS data are highly variable, each technical replicate of the pools was considered as a different observation. Therefore, each experimental group presented a total of 12 observations (n ¼ 12 per group). Contaminant proteins and those identified only by site and/or reverse sequence database were excluded. Statistical analyses were performed separately for each study. The iBAQ data, generated by MaxQuant software for each protein, were normalized by total iBAQ value of each observation (%iBAQ). Proteins were only considered present in the group when quantified in at least three observations. Therefore, proteins quantified in one or two observations in the group were excluded. For univariate statistical analyses, Microsoft Excel (Office 365) was used. Initially, a descriptive analysis was performed to calculate the fold-change of %iBAQ (ratio between the study group and the control group means). For groups' comparison, an unpaired Student's t test was carried out, and an a of 5% was used. Therefore, differentially expressed proteins were identified, as well as the conserved proteins (those equally expressed in both groups). Protein data were presented according to international standards (UniProt Accession Number–UniProt AC, gene symbol, and protein name). For statistical validation of our findings, multivariate statistical analyses were performed, using the online platform Metaboanalyst (27) and the PASW (SPSS) 18.0 software for Windows. Data were transformed to their logarithmic values, and a partial least-squares discriminant analysis (PLS-DA) was performed. Using this method the components (groups of correlated variables) were extracted, and in each component the complete data matrix received a Variable Importance in the Projection score. For PLS-DA results, the tridimensional graph demonstrating the separation of the groups was reported. Thereafter, proteins with Variable Importance in the Projection score >2 were used for logistic regression and discriminant analysis, to identify the most important proteins to differentiate the groups. For these analyses, nontransformed %iBAQ data of each protein were used as independent variable, and group variable was used as dependent variable. For logistic regression analysis, independent variables were inserted in the model using the enter method. The results were presented as the model's predictive values and the receiver operating characteristics (ROC) curve. For discriminant analysis, a linear model was constructed. Therefore, suggested biomarkers from univariate analysis were confirmed through the use of multivariate analyses. Functional enrichment analysis. The quantified proteins were used, initially, in Cytoscape 3.1.0 software (28) for Venn diagrams construction, using the PINA4MS plugin 4

(29). Then, differentially expressed proteins were also used for functional enrichment analysis of Gene Ontology categories, Kyoto Encyclopedia of Genes and Genomes and Reactome, using the Cytoscape software and the ClueGO 2.1.1 plugin (30). Only enriched functions with P value < .05 were considered.

RESULTS Mean (SD) lipid peroxidation levels for the control and study groups were 14.9  1.60 ng/mL and 32.0  2.68 ng/ mL, respectively. Both groups differed significantly (P< .0001) for lipid peroxidation. Threshold values were <17.39 ng/mL (bottom 15%) for the control group and >29.34 ng/mL (top 15%) for the study group. Clinical data of each group are presented in Table 1. Progressive motility and total motility were statistically lower in the study group, although those means were still within normal values, because only normozoospermic men were included in this study. In total, 629 proteins were quantified in the seminal plasma. In the univariate analysis, 4 proteins were absent, 19 were underexpressed, 8 were exclusive, and 63 were overexpressed in the study group (Fig. 1, Table 2). A total of 535 proteins were conserved between the groups (Supplemental Table 1). Main enriched functions associated with lipid peroxidation were cellular response to heat stress, to superoxide anion, and to transition metal ions, chemokine production, immune response, ROS detoxification, glutathione and hydrogen peroxide metabolism, loss of proteins required for centrosome microtubule organization, homeostasis, and unsaturated fatty acids biosynthesis (Fig. 1). In the multivariate analysis, a complete separation between the groups was obtained in PLS-DA analysis using 37 proteins (Fig. 2). These proteins were used in logistic regression and discriminant analyses. In the logistic regression analysis, 1 protein (mucin-5B) was capable of separating the control and study groups. The final model presented positive, negative, and total predictive values of 83.3%. The obtained ROC curve presented an area under the curve of 94.4%, P¼ .0002, sensitivity of 100%, and specificity of 83.3% (Fig. 2). The cutoff value for the %iBAQ of mucin-5B was 0.001242; values lower than the cutoff predict the control group, whereas values greater than that predict the study group. In the discriminant analysis, 20 ,30 -cyclic-nucleotide 30 -phosphodiesterase was able to predict low seminal lipid peroxidation levels, whereas five proteins were able to predict high seminal lipid peroxidation levels (arginine/serine rich coiled-coil protein 2; mucin-5B; multifunctional protein ADE2, procollagen-lysine; 2-oxoglutarate 5-dioxygenase 2; and SH3 domain-binding glutamic acid-rich-like protein). Of these, mucin-5B was identified in univariate analysis and further validated by multivariate analysis. Therefore, it was selected as a biomarker of seminal lipid peroxidation.

DISCUSSION Oxidative stress is one of the major causes of sperm functional abnormalities and, consequently, may lead to male infertility. The mechanisms of oxidative stress–derived VOL. - NO. - / - 2015

Fertility and Sterility®

TABLE 1 Clinical data from control (low lipid peroxidation levels) and study groups (high lipid peroxidation levels). Variable Lipid peroxidation levels (ng/mL) Mean  SD 95% CI Age (y) Mean  SD 95% CI Semen volume (mL) Mean  SD 95% CI Concentration (106/mL) Mean  SD 95% CI Progressive motility (%) Mean  SD 95% CI Total motility (%) Mean  SD 95% CI Morphology (% normal) Mean  SD 95% CI Round cells (106/mL) Mean  SD 95% CI Neutrophils (106/mL) Mean  SD 95% CI

Control (n [ 23)

Study (n [ 23)

14.9  1.60 14.2–15.6

32.0  2.68 30.8–33.2

35.1  5.41 32.8–37.5

36.3  6.55 33.5–39.1

3.8  1.51 3.1–4.5

3.4  1.61 2.7–4.1

70.9  54.19 47.5–94.4

100.0  58.30 74.8–125.2

54.5  7.82 51.5–57.9

48.0  8.26 44.4–51.5

59.7  7.09 56.6–62.7

53.1  8.98 49.2–57.0

7.5  2.97 6.2–8.8

7.3  2.46 6.3–8.4

1.2  0.97 0.8–1.7

1.8  1.41 1.2–2.4

0.1  0.21 0.04–0.2

0.1  0.13 0.03–0.1

P value < .0001a .511 .398 .058 .008a .008a .988 .213 .505

Note: Groups were compared using an unpaired Student's t test (P< .05). Nonnormally distributed variables were transformed to their logarithmic values before statistical analysis. CI ¼ confidence interval. a Statistically significant. Intasqui. Proteomics and semen oxidative stress. Fertil Steril 2015.

sperm alterations involve lipid peroxidation and proteins oxidation (3, 31). Therefore, we hypothesized that the seminal protein profile may reflect the seminal lipid peroxidation status, an indirect measure of oxidative stress. To test this hypothesis, we performed a shotgun

proteomic approach to compare the seminal plasma proteome between pooled semen samples with low and high lipid peroxidation levels, evaluated by TBARS quantification. Our major aims were to identify [1] the altered proteins and postgenomic pathways, to unravel the


Proteomic data and functional enrichment analysis. Left: Venn diagram of quantified proteins; control group (green), low seminal lipid peroxidation levels; study group (red), high seminal lipid peroxidation levels. Right: Main enriched functions in association with semen oxidative stress and the number of proteins identified in this study related to each of these functions. Intasqui. Proteomics and semen oxidative stress. Fertil Steril 2015.

VOL. - NO. - / - 2015



TABLE 2 Differentially expressed proteins in the study group (high seminal lipid peroxidation levels). UniProt AC

Gene symbol

P31943 Q8N335 Q9N2K0 Q9UJA9 P51888 P26572 Q8N104 Q9UIK5 P06865 P54289 Q6UXI9 P24593 P21246 P06858 Q9NZ08 Q9NQ79 O60635 Q66K79 Q08629 P09668 P28907 P12273 Q9H173 P01596 P01743 P09960 P30153


P54578 Q03013 Q96TA1 Q9UBC9 Q12841 P62328 P54107 P18827 Q06830 P13639 Q99519 P55072 P07858 O60888 Q8NBJ4 P00918 P13489 P02647 P62937 P01857 P01765 P07998 P07900 Q8IZP9 Q9UHI8 P22314 P00441 P09211 P60660 P50991 P15311 P01859 P21217 Q07654 P0CG06 Q9H1M3 P01833 O15296 Q9NZH0


Protein name

P value


Heterogeneous nuclear ribonucleoprotein H Glycerol-3-phosphate dehydrogenase 1-like protein HERV-H_2q24.3 provirus ancestral Env polyprotein Ectonucleotide pyrophosphatase/phosphodiesterase family member 5 Prolargin Alpha-1,3-mannosyl-glycoprotein 2-b-N-acetylglucosaminyltransferase Beta-defensin 106 Tomoregulin-2 Beta-hexosaminidase subunit a Voltage-dependent calcium channel subunit a-2/delta-1 Nephronectin Insulin-like growth factor-binding protein 5 Pleiotrophin Lipoprotein lipase Endoplasmic reticulum aminopeptidase 1 Cartilage acidic protein 1 Tetraspanin-1 Carboxypeptidase Z Testican-1 Pro-cathepsin H ADP-ribosyl cyclase 1 Prolactin-inducible protein Nucleotide exchange factor SIL1 Ig k chain V-I region CAR Ig heavy chain V-I region HG3 Leukotriene A-4 hydrolase Serine/threonine-protein phosphatase 2A 65 kD regulatory subunit A a isoform Ubiquitin carboxyl-terminal hydrolase 14 Glutathione S-transferase Mu 4 Niban-like protein 1 Small proline-rich protein 3 Follistatin-related protein 1 Thymosin b-4 Cysteine-rich secretory protein 1 Syndecan-1 Peroxiredoxin-1 Elongation factor 2 Sialidase-1 Transitional endoplasmic reticulum ATPase Cathepsin B Protein CutA Golgi membrane protein 1 Carbonic anhydrase 2 Ribonuclease inhibitor Apolipoprotein A-I Peptidyl-prolyl cis-trans isomerase A Ig g-1 chain C region Ig heavy chain V-III region TIL Ribonuclease pancreatic Heat shock protein HSP 90-a G-protein coupled receptor 64 A disintegrin and metalloproteinase with thrombospondin motifs 1 Ubiquitin-like modifier-activating enzyme 1 Superoxide dismutase [Cu-Zn] Glutathione S-transferase P Myosin light polypeptide 6 T-complex protein 1 subunit delta Ezrin Ig g-2 chain C region Galactoside 3(4)-L-fucosyltransferase Trefoil factor 3 Ig l-3 chain C regions Beta-defensin 129 Polymeric immunoglobulin receptor Arachidonate 15-lipoxygenase B G-protein coupled receptor family C group 5 member B

– – – – .001 .005 .011 .029 < .001 .038 .030 .049 .020 .020 < .001 .030 .026 .003 .033 .041 .013 .024 .010 – – – –

Absent Absent Absent Absent 0.165 0.549 0.592 0.596 0.643 0.690 0.718 0.724 0.733 0.742 0.743 0.753 0.768 0.770 0.773 0.796 0.820 0.829 0.831 Exclusive Exclusive Exclusive Exclusive

– – – – .019 .025 .048 .029 .014 .048 .041 .011 < .001 .041 .036 .038 .028 .022 .025 .014 .034 .020 .001 .019 .037 .018 .031 .023 .005 .048 .006 .001 .022 .035 .003 .005 .037 .018 .029

Exclusive Exclusive Exclusive Exclusive 1.210 1.253 1.259 1.260 1.263 1.266 1.266 1.271 1.290 1.293 1.311 1.312 1.314 1.316 1.321 1.324 1.352 1.359 1.372 1.377 1.387 1.393 1.395 1.396 1.405 1.406 1.413 1.419 1.428 1.462 1.464 1.479 1.494 1.494 1.502

Intasqui. Proteomics and semen oxidative stress. Fertil Steril 2015.


VOL. - NO. - / - 2015

Fertility and Sterility®

TABLE 2 Continued. UniProt AC

Gene symbol

P06727 P15121 P62258 Q9HC38 Q96BQ1 P01834 P0CG04 P01876 P02787 P01623 P06702 P01860 Q96BH3 P01764 P01766 P61626 P55259 O43692 Q13228 P14550 P84077 O95994 P01037 P01617 P01009 Q8N4F0 P31025 Q9HC84


Protein name Apolipoprotein A-IV Aldose reductase 14-3-3 protein epsilon Glyoxalase domain-containing protein 4 Protein FAM3D Ig k chain C region Ig l-1 chain C regions Ig a-1 chain C region Serotransferrin Ig k chain V-III region WOL Protein S100-A9 Ig g-3 chain C region Epididymal sperm-binding protein 1 Ig heavy chain V-III region VH26 Ig heavy chain V-III region BRO Lysozyme C Pancreatic secretory granule membrane major glycoprotein GP2 Peptidase inhibitor 15 Selenium-binding protein 1 Alcohol dehydrogenase [NADP(þ)] ADP-ribosylation factor 1 Anterior gradient protein 2 homolog Cystatin-SN Ig k chain V-II region TEW Alpha-1-antitrypsin BPI fold-containing family B member 2 Lipocalin-1 Mucin-5B

P value


.005 .001 .001 .008 .014 < .001 .037 .007 .010 .047 .007 .031 .024 .019 .024 .041 .017 .009 .020 .001 .007 .004 .042 .016 .008 .001 .003 .001

1.527 1.531 1.532 1.552 1.568 1.571 1.578 1.580 1.589 1.592 1.596 1.642 1.682 1.715 1.731 1.735 1.736 1.760 1.784 1.839 1.878 2.020 2.052 2.060 2.160 3.079 3.144 3.595

Note: Control (low seminal lipid peroxidation levels) and study groups were compared using an unpaired Student's t test (P< .05). Fold-changes lower than 1 represent underexpressed proteins, whereas fold-changes greater than 1 represent overexpressed proteins in the study group. Intasqui. Proteomics and semen oxidative stress. Fertil Steril 2015.

mechanisms associated with semen oxidative stress, and [2] potential protein biomarkers for semen oxidative stress. Recent studies have already evaluated the effects of oxidative stress on seminal plasma proteome. Sharma et al. (32) used semen samples from 32 infertile men and 20 healthy controls, separating them into ROS-positive (ROSþ) or negative (ROS) groups and comparing the protein profiles of these groups using LC-MS/MS. They identified a total of 14 proteins, of which proteins related to the protection against oxidative stress damage were exclusively expressed in the ROS group. In another study (22), LC-MS/MS was used to compare the seminal plasma proteomes between 11 fertile men and 11 infertile men presenting idiopathic oligoasthenoteratozoospermia. Moreover, the seminal plasma protein carbonyl levels, a marker of oxidative stress, were also assessed and compared between the groups. In total, 2,489 proteins were identified, of which 46 proteins were overexpressed in the infertile group. These proteins are involved in cellular organization and modification, metabolism, inflammation, and immune and stress responses. Carbonyl proteins were overexpressed in the infertile groups, demonstrating the association between oligoasthenoteratozoospermia, seminal plasma proteomic alterations, and semen oxidative stress. In our study the main enriched functions associated with lipid peroxidation in normozoospermic men were cellular response to superoxide anion, glutathione and hydrogen peroxide metabolism, response to transition metals, ROS VOL. - NO. - / - 2015

detoxification, homeostasis, unsaturated fatty acid biosynthesis, cellular response to heat stress, and immune response. Interestingly, the above functions are closely related to oxidative stress. Indeed, the semen oxidative stress pathway is well known and starts with superoxide anion production during oxidative metabolism (8). It can be converted to hydrogen peroxide by superoxide dismutase (SOD), both in mitochondria (MnSOD) and cytoplasm (Cu-Zn-SOD) (33). Furthermore, Cu-Zn-SOD can be secreted by the epididymal epithelium, and its level and activity increase from the caput to the cauda epididymis (34, 35). The formed hydrogen peroxide may be converted into water and oxygen by catalase and glutathione peroxidase. Alternatively, it may easily cross the sperm mitochondrial and cellular membranes and receive a donated electron from iron or copper ions, originating hydroxyl radicals, the most harmful ROS (33). The major targets of ROS are unsaturated fatty acids, mostly those polyunsaturated, owing to double bounds between their carbon atoms (6, 36). Human sperm membrane is particularly rich in unsaturated fatty acids, which are responsible for its fluidity and flexibility, which are essential features for fertilization (8). Thus, enzymatic antioxidants, such as superoxide dismutase, catalase, and glutathione peroxidase, and nonenzymatic antioxidants, such as albumin, glutathione, and taurine, are important for ROS detoxification and, consequently, for homeostasis maintenance (37). Therefore, the enrichment of cellular 7



Multivariate statistical analyses. Top: Tridimensional graphic obtained in the PLS-DA analysis, demonstrating the groups' separation. Bottom: The ROC curve obtained in the logistic regression analysis, with an area under the curve of 94.4%, P¼.0002, sensitivity of 100%, and specificity of 83.3%. Intasqui. Proteomics and semen oxidative stress. Fertil Steril 2015.

response to superoxide anion, glutathione and hydrogen peroxide metabolism, response to transition metals, ROS detoxification, and homeostasis indicates the ROS overproduction in semen from normozoospermic men and a consequent production of antioxidants, as an attempt to restore the redox balance in seminal plasma. Furthermore, the enrichment of unsaturated fatty acids biosynthesis might demonstrate [1] an attempt of sperm membrane repair after ROS-derived damage and lipid peroxides formation, or [2] lipid peroxidation self-propagation, with the formation of novel unsaturated fatty acids and their incorporation into the sperm membrane, which becomes even more vulnerable to the negative effects of oxidative stress. These functions are thus closely associated with the higher lipid peroxidation levels observed in the seminal plasma from the study group. 8

Studies have demonstrated that the main causes of semen oxidative stress are heat stress and immune response (9, 38–44). Heat stress is caused by an increase in scrotal temperature, which may be due to varicocele (38), obesity (45), or lifestyle factors, such as work seated for long periods (46). Heat stress leads to ROS production, which is suggested to act as a cytotoxic agent to germ cells and sperm (47). Additionally, a growing body of evidence has suggested that male reproductive tract infections and inflammation are directly associated with semen oxidative stress (9, 40–44). More than 15% of infertile men have genitourinary infections or inflammation (48). A recent study has also shown that hepatitis C virus infection leads to an increase in ROS production, which is associated with mitochondrial dysfunction (43). Indirectly, the infiltration of activated macrophages and neutrophils in the injured tissue also increases ROS levels, inducing sperm damage (40–42). Moreover, inflammation in the male reproductive tract may also cause oxidative stress due to the production of ROS and proinflammatory cytokines, leading to decreased semen quality and altered sperm function (9, 44, 49–51). In addition, studies have also demonstrated a correlation between hydrogen peroxide levels and male genitourinary tract inflammation in infertile men (52). Thus, the enrichment of cellular response to heat stress and immune response in the seminal plasma from normozoospermic men suggests that the higher seminal lipid peroxidation levels observed in the study group may be caused by an increased testicular temperature and/or by immune response activation. Finally, we performed a multivariate statistical analysis to confirm our results and identified mucin-5B as a biomarker of semen oxidative stress. Mucins are highly glycosylated proteins involved with mucosal innate immune response (53, 54). Specifically, mucin-5B is a high-molecular-weight, gel-forming protein (55). Russo et al. (56) were the first to demonstrate the presence of mucin-5B in the seminal plasma. Thereafter, Piludu et al. (55) demonstrated that this protein is secreted by the bulbourethral glands, where it seems to protect the epithelium against infection. The function of mucin-5B in semen remains unknown, but it is suggested that it might modulate sperm transport in the male and female reproductive tracts (55, 57). In lungs, oxidative stress promotes an up-regulation of mucin-5B in the respiratory epithelium, both in vivo and in vitro (58). Our results strongly suggest that the same might occur in semen. A limitation of our study should be addressed referred to the samples used in this work. Because we included only men with normal semen parameters, the higher lipid peroxidation levels observed in the study group might still reflect the normal physiologic ranges of ROS in the seminal plasma. Consequently, the proteomic alterations and the biomarker identified here might be applicable only to normozoospermic men, and therefore a direct conclusion cannot be made for men presenting significant semen alterations. Accordingly, the statistically significant differences in sperm motility between control and study groups might not have any clinical significance, because both sperm motility means are still within normal limits. VOL. - NO. - / - 2015

Fertility and Sterility® 21.

In conclusion, high seminal lipid peroxidation levels are associated with alterations in the seminal plasma proteomic profile and postgenomic pathways in normozoospermic men. Main altered biological mechanisms associated with oxidative stress are ROS detoxification and metabolism, heat stress, and an exacerbated immune response. Moreover, we suggest the mucin-5B protein as a potential biomarker of semen oxidative stress in these men.




1. 2.

3. 4. 5.

6. 7.



10. 11. 12.



15. 16.





Iwasaki A, Gagnon C. Formation of reactive oxygen species in spermatozoa of infertile patients. Fertil Steril 1992;57:409–16. Ferramosca A, Pinto Provenzano S, Montagna DD, Coppola L, Zara V. Oxidative stress negatively affects human sperm mitochondrial respiration. Urology 2013;82:78–83. Aitken RJ, Smith TB, Jobling MS, Baker MA, De Iuliis GN. Oxidative stress and male reproductive health. Asian J Androl 2014;16:31–8. Agarwal A, Saleh RA. Role of oxidants in male infertility: rationale, significance, and treatment. Urol Clin North Am 2002;29:817–27. Hsu PC, Hsu CC, Guo YL. Hydrogen peroxide induces premature acrosome reaction in rat sperm and reduces their penetration of the zona pellucida. Toxicology 1999;139:93–101. Agarwal A, Makker K, Sharma R. Clinical relevance of oxidative stress in male factor infertility: an update. Am J Reprod Immunol 2008;59:2–11. ru-Abbasog lu S, Ku lu A, Ozdemirler€c¸u €kgergin C, Kadıog Aktan G, Dog Erata G, Koc¸ak-Toker N. Mystery of idiopathic male infertility: is oxidative stress an actual risk? Fertil Steril 2013;99:1211–5. Aitken RJ, Clarkson JS. Cellular basis of defective sperm function and its association with the genesis of reactive oxygen species by human spermatozoa. J Reprod Fertil 1987;81:459–69. Henkel R, Kierspel E, Stalf T, Mehnert C, Menkveld R, Tinneberg HR, et al. Effect of reactive oxygen species produced by spermatozoa and leukocytes on sperm functions in non-leukocytospermic patients. Fertil Steril 2005;83:635–42. Rajender S, Rahul P, Mahdi AA. Mitochondria, spermatogenesis and male infertility. Mitochondrion 2010;10:419–28. Aitken RJ. The Amoroso Lecture. The human spermatozoon–a cell in crisis? J Reprod Fertil 1999;115:1–7. Wang X, Sharma RK, Gupta A, George V, Thomas AJ, Falcone T, et al. Alterations in mitochondria membrane potential and oxidative stress in infertile men: a prospective observational study. Fertil Steril 2003;80(Suppl 2):844–50. Fariello RM, Del Giudice PT, Spaine DM, Fraietta R, Bertolla RP, Cedenho AP. Effect of leukocytospermia and processing by discontinuous density gradient on sperm nuclear DNA fragmentation and mitochondrial activity. J Assist Reprod Genet 2009;26:151–7. Janero DR, Hreniuk D, Sharif HM. Hydroperoxide-induced oxidative stress impairs heart muscle cell carbohydrate metabolism. Am J Physiol 1994; 266(1 Pt 1):C179–88. Linton S, Davies MJ, Dean RT. Protein oxidation and ageing. Exp Gerontol 2001;36:1503–18. Vittorini S, Paradiso C, Donati A, Cavallini G, Masini M, Gori Z, et al. The agerelated accumulation of protein carbonyl in rat liver correlates with the agerelated decline in liver proteolytic activities. J Gerontol A Biol Sci Med Sci 1999;54:B318–23. Eskiocak S, Gozen AS, Yapar SB, Tavas F, Kilic AS, Eskiocak M. Glutathione and free sulphydryl content of seminal plasma in healthy medical students during and after exam stress. Hum Reprod 2005;20:2595–600. Chen SS, Chang LS, Wei YH. Oxidative damage to proteins and decrease of antioxidant capacity in patients with varicocele. Free Radic Biol Med 2001; 30:1328–34. Piomboni P, Stendardi A, Gambera L, Tatone C, Coppola L, De Leo V, et al. Protein modification as oxidative stress marker in normal and pathological human seminal plasma. Redox Rep 2012;17:227–32. Tatone C, Amicarelli F, Carbone MC, Monteleone P, Caserta D, Marci R, et al. Cellular and molecular aspects of ovarian follicle ageing. Hum Reprod Update 2008;14:131–42.

VOL. - NO. - / - 2015



26. 27.

28. 29.




33. 34.





39. 40. 41.



Wang J, Wang J, Zhang HR, Shi HJ, Ma D, Zhao HX, et al. Proteomic analysis of seminal plasma from asthenozoospermia patients reveals proteins that affect oxidative stress responses and semen quality. Asian J Androl 2009; 11:484–91. Herwig R, Knoll C, Planyavsky M, Pourbiabany A, Greilberger J, Bennett KL. Proteomic analysis of seminal plasma from infertile patients with oligoasthenoteratozoospermia due to oxidative stress and comparison with fertile volunteers. Fertil Steril 2013;100:355–66.e2. World Health Organization. Department of Reproductive Health and Research. WHO laboratory manual for the examination and processing of human semen. Geneva: WHO; 2010. Oborna I, Wojewodka G, De Sanctis JB, Fingerova H, Svobodova M, Brezinova J, et al. Increased lipid peroxidation and abnormal fatty acid profiles in seminal and blood plasma of normozoospermic males from infertile couples. Hum Reprod 2010;25:308–16. Smith PK, Krohn RI, Hermanson GT, Mallia AK, Gartner FH, Provenzano MD, et al. Measurement of protein using bicinchoninic acid. Anal Biochem 1985; 150:76–85. Mann K, Mann M. The proteome of the calcified layer organic matrix of turkey (Meleagris gallopavo) eggshell. Proteome Sci 2013;11:40. Xia J, Mandal R, Sinelnikov IV, Broadhurst D, Wishart DS. MetaboAnalyst 2.0–a comprehensive server for metabolomic data analysis. Nucleic Acids Res 2012;40(Web Server issue):W127–33. Saito R, Smoot ME, Ono K, Ruscheinski J, Wang PL, Lotia S, et al. A travel guide to Cytoscape plugins. Nat Methods 2012;9:1069–76. Wu J, Vallenius T, Ovaska K, Westermarck J, M€akel€a TP, Hautaniemi S. Integrated network analysis platform for protein-protein interactions. Nat Methods 2009;6:75–7. Bindea G, Mlecnik B, Hackl H, Charoentong P, Tosolini M, Kirilovsky A, et al. ClueGO: a Cytoscape plug-in to decipher functionally grouped gene ontology and pathway annotation networks. Bioinformatics 2009;25:1091–3. Shiva M, Gautam AK, Verma Y, Shivgotra V, Doshi H, Kumar S. Association between sperm quality, oxidative stress, and seminal antioxidant activity. Clin Biochem 2011;44:319–24. Sharma R, Agarwal A, Mohanty G, Du Plessis SS, Gopalan B, Willard B, et al. Proteomic analysis of seminal fluid from men exhibiting oxidative stress. Reprod Biol Endocrinol 2013;11:85. Halliwell B, Gutteridge JMC. Free radicals in biology and medicine. Oxford: Clarendon; 1989. Koziorowska-Gilun M, Koziorowski M, Fraser L, Strzez_ ek J. Antioxidant defence system of boar cauda epididymidal spermatozoa and reproductive tract fluids. Reprod Domest Anim 2011;46:527–33. Park K, Jeon S, Song YJ, Yi LSH. Proteomic analysis of boar spermatozoa and quantity changes of superoxide dismutase 1, glutathione peroxidase, and peroxiredoxin 5 during epididymal maturation. Anim Reprod Sci 2012;135:53–61.  P, Dondero F, et al. Fatty acid Lenzi A, Gandini L, Maresca V, Rago R, Sgro composition of spermatozoa and immature germ cells. Mol Hum Reprod 2000;6:226–31. Lewis SE, Boyle PM, McKinney KA, Young IS, Thompson W. Total antioxidant capacity of seminal plasma is different in fertile and infertile men. Fertil Steril 1995;64:868–70. Shiraishi K, Takihara H, Matsuyama H. Elevated scrotal temperature, but not varicocele grade, reflects testicular oxidative stress-mediated apoptosis. World J Urol 2010;28:359–64. Peltola V, Huhtaniemi I, Ahotupa M. Abdominal position of the rat testis is associated with high level of lipid peroxidation. Biol Reprod 1995;53:1146–50. Tremellen K, Tunc O. Macrophage activity in semen is significantly correlated with sperm quality in infertile men. Int J Androl 2010;33:823–31. Eley A, Hosseinzadeh S, Hakimi H, Geary I, Pacey AA. Apoptosis of ejaculated human sperm is induced by co-incubation with Chlamydia trachomatis lipopolysaccharide. Hum Reprod 2005;20:2601–7. Henkel R, Maass G, Hajimohammad M, Menkveld R, Stalf T, Villegas J, et al. Urogenital inflammation: changes of leucocytes and ROS. Andrologia 2003; 35:309–13. Quarato G, Scrima R, Agriesti F, Moradpour D, Capitanio N, Piccoli C. Targeting mitochondria in the infection strategy of the hepatitis C virus. Int J Biochem Cell Biol 2013;45:156–66.



45. 46. 47.

48. 49.




Moretti E, Cosci I, Spreafico A, Serchi T, Cuppone AM, Collodel G. Semen characteristics and inflammatory mediators in infertile men with different clinical diagnoses. Int J Androl 2009;32:637–46. Pasquali R, Patton L, Gambineri A. Obesity and infertility. Curr Opin Endocrinol Diabetes Obes 2007;14:482–7. Bujan L, Daudin M, Charlet JP, Thonneau P, Mieusset R. Increase in scrotal temperature in car drivers. Hum Reprod 2000;15:1355–7. Ishii T, Matsuki S, Iuchi Y, Okada F, Toyosaki S, Tomita Y, et al. Accelerated impairment of spermatogenic cells in SOD1-knockout mice under heat stress. Free Radic Res 2005;39:697–705. Haidl G, Allam JP, Schuppe HC. Chronic epididymitis: impact on semen parameters and therapeutic options. Andrologia 2008;40:92–6. Perdichizzi A, Nicoletti F, Vignera SL, Barone N, D'Agata R, Vicari E, et al. Effects of tumour necrosis factor-a on human sperm motility and apoptosis. J Clin Immunol 2007;27:152–62. Kopa Z, Wenzel J, Papp GK, Haidl G. Role of granulocyte elastase and interleukin-6 in the diagnosis of male genital tract inflammation. Andrologia 2005;37:188–94. Gruschwitz MS, Brezinschek R, Brezinschek HP. Cytokine levels in the seminal plasma of infertile males. J Androl 1996;17:158–63.


53. 54. 55.




Kullisaar T, T€ urk S, Kilk K, Ausmees K, Punab M, M€andar R. Increased levels of hydrogen peroxide and nitric oxide in male partners of infertile couples. Andrology 2013;1:850–8. Parker D, Prince A. Innate immunity in the respiratory epithelium. Am J Respir Cell Mol Biol 2011;45:189–201. Rose MC, Voynow JA. Respiratory tract mucin genes and mucin glycoproteins in health and disease. Physiol Rev 2006;86:245–78. Piludu M, Hand AR, Cossu M, Piras M. Immunocytochemical localization of MG1 mucin in human bulbourethral glands. J Anat 2009;214: 179–82. Russo CL, Spurr-Michaud S, Tisdale A, Pudney J, Anderson D, Gipson IK. Mucin gene expression in human male urogenital tract epithelia. Hum Reprod 2006;21:2783–93. €eso P, Gargiulo AR, Hill JA, et al. Gipson IK, Moccia R, Spurr-Michaud S, Argu The amount of MUC5B mucin in cervical mucus peaks at midcycle. J Clin Endocrinol Metab 2001;86:594–600. Casalino-Matsuda SM, Monzon ME, Day AJ, Forteza RM. Hyaluronan fragments/CD44 mediate oxidative stress-induced MUC5B upregulation in airway epithelium. Am J Respir Cell Mol Biol 2009;40: 277–85.

VOL. - NO. - / - 2015