Differential seminal plasma proteome according to semen retrieval in men with spinal cord injury

ORIGINAL ARTICLE: ANDROLOGY

Differential seminal plasma proteomeaccording to semen retrieval in menwith spinal cord injury

Edson Guimar~aes lo Turco, D.V.M., Ph.D.,a Paula Toni Del Giudice, M.Sc.,a

Thiesa Butterby Soler, D.V.M., Ph.D.,a Deborah Montagnini Spaine, Ph.D.,a Milton Borrelli Junior, M.D.,a

F�abio C�esar Gozzo, Ph.D.,d,e Eduardo Jorge Pilau, Ph.D.,d,e,f Jerusa Simone Garcia, Ph.D.,c,g

Christina Ramires Ferreira, D.V.M., Ph.D.,c Marcos Nogueira Eberlin, Ph.D.,c

and Ricardo Pimenta Bertolla, D.V.M., Ph.D.a

a Human Reproduction Section, Division of Urology, Department of Surgery, Sao Paulo Federal University, Sao Paulo; bMassSpectrometry Applications Research and Development Laboratory, Waters Corporation, Sao Paulo; c Thomson MassSpectrometry Laboratory, d Institute of Chemistry, University of Campinas, Campinas; e National Institute ofBioanalytical Science and Technology, Sao Paulo; f Department of Chemistry, Center for Exact Sciences, Maringa StateUniversity, Maring�a; and g Institute of Exact Sciences, Alfenas Federal University, Alfenas, Brazil

Objective: To evaluate protein expression profile and to quantify proteins present in seminal plasma from men with spinal cord injury(SCI) and healthy men without SCI.Design: Experimental study.Setting: University hospital.Patient(s): Twelve SCI patients divided in to two groups, six who underwent electroejaculation (EEJ) and six who underwent penilevibratory stimulation (PVS); and ten control subjects presenting normal sperm motility and concentration.Intervention(s): EEJ and PVS.Main OutcomeMeasure(s): The seminal plasma protein profile was analyzed by two proteomic strategies: data-independent label-freequantitative proteomics (MSE) and two-dimensional sodium dodecyl sulfate–polyacrylamide gel electrophoresis (2D SDS-PAGE).Result(s): A total of 638 different proteins were identified by MSE and 18 by 2D SDS-PAGE followed by tandem mass spectrometry.Interactome analysis showed key reproductive biologic processes—insemination, sperm and oocyte fusion, and acrosomereaction—related to all groups, as were triglyceride stimuli. Processes related to actin and muscle function and to iron oxidation,transportation, and homeostasis were found only in the EEJ and PVS groups; response to hydrogen peroxide and increased immuneresponse was found only in the PVS group.Conclusion(s): This studywas able to demonstrate differential protein expression among control, PVS, andEEJgroups; SCI is responsible

Use your smartphone

for alterations in seminal plasmaprotein profile leading to a deviation fromhomeostasis; proteinsreported in both PVS andEEJ groups correlatewith the pathophysiology of SCI-related infertility.(Fertil Steril� 2013;-:-–-. �2013 by American Society for Reproductive Medicine.)KeyWords: Biomarkers, spinal cord injury, seminal plasma, proteomics, functional enrichment

Discuss:You can discuss this article with its authors and with other ASRM members at http://fertstertforum.com/silvabfd-seminal-plasma-proteomics-spinal-cord-injury/

to scan this QR codeand connect to thediscussion forum forthis article now.*

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

S pinal cord injury (SCI) is a uniquemedical condition, defined as anacute traumatic injury to the

Received April 24, 2012; revised and accepted June 6B.F.d.S. has nothing to disclose. G.H.M.F.S. has noth

P.T.D.G. has nothing to disclose. T.B.S. has notM.B.J. has nothing to disclose. F.C.G. has nothinhas nothing to disclose. C.R.F. has nothing to dnothing to disclose.

Supported by the Conselho Nacional de DesenvolvimReprint requests: Ricardo Pimenta Bertolla, D.V.M., P

04039–060, Sao Paulo, Sao Paulo, Brazil (E-mail

Fertility and Sterility® Vol. -, No. -, - 2013 0015-Copyright ©2013 American Society for Reproductivehttp://dx.doi.org/10.1016/j.fertnstert.2013.06.009

VOL. - NO. - / - 2013

spinal cord that can lead to varyingdegrees of motor and/or sensorydeficits, paralysis, and reduced life

, 2013.ing to disclose. E.G.l.T. has nothing to disclose.hing to disclose. D.M.S. has nothing to disclose.g to disclose. E.J.P. has nothing to disclose. J.S.G.isclose. M.N.E. has nothing to disclose. R.P.B. has

ento Científico e Tecnol�ogico.h.D., Sao Paulo Federal University, R Embau, 231,: [email protected]).

0282/$36.00Medicine, Published by Elsevier Inc.

quality (1). Despite great advances inthe field of rehabilitation medicine,infertility remains a significant compli-cation due to SCI; only 5% of SCIpatients can father a child with-out medical intervention (2). Fertilityproblems in SCI men are mainly dueto: 1) erectile or ejaculatory distur-bances; and 2) poor semen quality (3).

Various techniques have been usedto overcome the erectile and ejacula-tory dysfunction. Using a combinationof two different semen retrieval

1

ORIGINAL ARTICLE: ANDROLOGY

procedures, electroejaculation (EEJ) and penile vibratorystimulation (PVS), semen samples can be obtained in up to97% of SCI patients (4). However, the retrieved semenspecimens often present an unusual profile, characterizedby normal sperm concentration but extremely impaired spermmotility and viability (5). The most important reasons leadingto this condition remain unknown and do not seem to be dueto lack of ejaculation, testicular hyperthermia, genitourinarytract infections, histologic testicular abnormalities, or yearsafter injury, as previously assumed (6).

Therefore, researchers have turned attention directly tothe molecular changes that recognizably occur in the seminalplasma after establishment of the SCI. In an interesting study,Brackett et al. showed that when sperm obtained from controlsubjects were added to seminal plasma obtained from SCIpatients, motility declined, whereas the addition of spermretrieved from SCI men to seminal plasma from uninjuredmen led to an improvement in motility (7). Furthermore,corroborating with these findings, another study carried outby the same group demonstrated that sperm aspirated directlyfrom the vas deferens of SCI patients presented significantlyhigher motility than those obtained in the ejaculate of thesame patients. Even though vas-aspirated sperm from SCImen presented lower motility and viability compared withcontrol men, the study suggested that the great decrease inthe parameters was due to contact with seminal plasma.

These conclusions clearly indicate that SCI men haveseminal plasma abnormalities that are somehow deleteriousto sperm cells (8). It is known that human seminal plasma isdesigned to assure not only sperm survival but also thesuccess of fertilization process, and its complex proteincontent is important in essential steps such as spermcapacitation, immune response inside the uterus, formationof the tubal sperm reservoir (7), sperm–zona pellucidainteraction, and sperm and oocyte fusion (9–11).

Considering the importance of the proteins in seminalplasma, the present study aimed for full coverage of theseminal plasma proteome from SCI patients. By using state-of-the-art mass spectrometry (MS) methods, we comparedthe proteome of SCI patients with that of healthy men withoutSCI in an attempt to elucidate the molecular mechanismsleading to poor semen quality in this particular group ofpatients. We also compared the seminal plasma collectedby PVS to that collected by EEJ to explore their possibledifferential effects on protein profile.

MATERIALS AND METHODSPatients and Sample Collection

Institutional Review Board approval was obtained from SaoPaulo Federal University Research Ethics Committee, andsigned informed written consent was obtained from eachparticipant. We included in this study 12 seminal plasmasamples from SCI patients and 10 seminal plasma samplesfrom healthy donors (presenting normal values of spermmotility and concentration) undergoing the clinical andlaboratory evaluation in our assisted reproduction program.Fertility status and/or proved paternity were not taken intoconsideration for these individuals. Patient samples were

2

obtained from volunteers presenting any level of SCI, andgrouped according to semen retrieval method (six patientscollected semen through PVS and six through EEJ). Theexclusion criterion for both groups was azoospermia, and inthe case of SCI group, only antegrade specimens were used.

Because the majority of SCI men are not able to produceantegrade ejaculation by masturbation, the same urologist,specialized in the treatment of SCI, performed either PVS orEEJ in our patients. Because PVS is a simple and less invasivemethod of semen retrieval, it was our first choice of treatmentand performed with the use of a 100-Hz-frequency and2.5-mm-amplitude vibrator, in 2–3-minute periods until theoccurrence of ejaculation. However, not all patients couldcollect antegrade ejaculation with PVS. In these cases EEJ,which is successful in obtaining ejaculate from men with alllevels of SCI (including those who do not have majorcomponents of the ejaculatory reflex arc), was the methodof choice. For performing EEJ, an electric probe delivering a12–15-V (100–200-mOhm) stimulation range was chosen.Around 2–3 stimulus cycles (1–2 s/cycle), in short timeintervals, were performed until the occurrence of ejaculation.Control subjects collected samples bymasturbation in specificsterile containers after at least 4 days, but no longer than7 days, of sexual abstinence.

All samples were kept warm (22–37�C) and evaluatedwithin 1 hour of collection. After liquefaction of thesemen, seminal parameters (volume, total sperm count,concentration, sperm motility, round cells, neutrophils,vitality, and red blood cells) were evaluated according toWorld Health Organization criteria (12) and morphologyevaluated by the Kruger strict criteria (13). Normal valueswere sperm motility R50% (aþb), sperm concentrationR20 � 106/mL, and normal sperm forms >14%. For seminalparameters, statistical analysis was performed with the use ofSPSS 13.0 for Windows. First, all variables were tested todetermine variance homogeneity. Heteroscedastic variableswere transformed to their square root. The tested variables(age, time after injury, concentration, total sperm count,morphology, round cells, neutrophils, vitality, and red bloodcells) were compared among groups with the use of eitherthe Student t test (when presented only in SCI groups) oranalysis of variance followed by a least significant differencespost hoc test. All variables that did not present variancehomogeneity even after the transformation were comparedwith the use of either the Mann-Whitney test (when presentedonly in SCI groups) or the Kruskal-Wallis followed byMann-Whitney test. Statistical significance was set at P< .05.

Seminal Plasma Preparation

Immediately after semen analysis, all samples werecentrifuged at 3,500g for 10 minutes for complete separationof seminal plasma from sperm cells and other cell lines.Although an optimal centrifugation at 500g for 15 minutesis routinely used for sperm washing (seminal plasmaseparation) when cellular membranes need to be kept free ofdamage (14), we chose a higher centrifugation speed to assurethat most of cellular debris was pelleted and our final seminalplasma samples would be free of membrane and organelle

VOL. - NO. - / - 2013

Fertility and Sterility®

pieces which could interfere in subsequent preparation steps.However, being aware of the fact that such g forces can causebreakage of cells and leakage of cellular constituentsinto seminal plasma, we subsequently characterized allproteins found in three major gene ontology (GO) categories:1) biologic process; 2) molecular function; and 3) cellularcomponent, which guided us to a better interpretation ofresults. Moreover, because seminal plasma and sperm cellscompose an integrative and functional complex, it might bepart of the physiologic motion of this complex that somecells suffer apoptotic breakdown and cellular componentsend up diluted in seminal plasma. Seminal plasmasamples were stored at �20�C until use. Samples werethen thawed at room temperature, when a serine proteaseinhibitor (1 mmol/L phenylmethylsulfonyl in methanol) wasadded. Samples were immediately centrifuged at 16,000gfor 60 minutes at 4�C, followed by protein concentrationevaluation with the use of the modified Lowry method(bicinchinonic acid) assay (15). To obtain sufficient andhomogeneous material for all the experiments, individualsamples were pooled for nano–u naltra-high-performanceliquid chromatography– mass spectrometry (nUPLC-MSE)analysis and subjected to individual analysis bytwo-dimensional sodium dodecyl sulfate–polyacrylamidegel electrophoresis (2D SDS-PAGE). Because this study alsofocused attention on the differences resulting from EEJ andPVS techniques, we chose to pool SCI patients samplesaccording to the semen retrieval method: Three groups wereformed: 1) control; 2) PVS; and 3) EEJ.

Protein Identification and Quantification bynUPLC Tandem Nano–Elestrospray Ionization MSE

The initial pools were submitted to protein quantification withthe use of the Bradford assay, and 50 mg was used for trypticdigestion. Briefly, the protein (fluid extract) samples weredenatured with 0.1% Rapigest SF Protein Digestion surfactant(Waters), reduced (10 mmol/L dithiothreitol), alkylated(10 mmol/L iodoacetamide), and enzymatically digestedwith trypsin at 1:50 (w/w) enzyme:protein ratio. Afterdigestion, an internal standard (alcohol dehydrogenase,spiked to 50 fmol, Swiss-Prot accession number P00330)was added to the pools to allow quantitative analysis. Qualita-tive and quantitative nUPLC tandem nano–elestrospray ioni-zation (nESI) MSe experiments were conducted with the use ofa 1.5-hour reversed phase gradient from 5% to 40% (v/v)acetonitrile (0.1% v/v formic acid) at 600 nL/min on a Nano-acquity UPLC core system. A Nanoacquity UPLC C18 BEH 1.7mm, 100 mm � 10 cm column was used in conjunction withan SCX 5 mm, 180 mm � 23 mm column. Typical on-column sample loads were 250 ng of total protein digests.For all measurements, the time of flight (TOF) analyzer ofthe mass spectrometer was operated in the ‘‘W’’ mode with atypical resolving power of R20,000. All analyses wereperformed with the use of ESI in the positive ion mode (þ)and a Nanolock spray source. The lock mass channel wassampled every 30 seconds, calibrated with a human (Glu1)-fibrinopeptide B (GFP) solution (100 fmol/mL) deliveredthrough the reference sprayer of the Nanolock spray source.

VOL. - NO. - / - 2013

The doubly-charged ion ([Mþ2H]2þ) was used for initialsingle-point calibration, and MS/MS fragment ions of GFPwere used to obtain the final instrument calibration. Data-independent scanning (MSE) experiments were performedwith a Waters Synapt G1 HDMS mass spectrometer (Waters),which was automatically planned to switch between standardMS (3 eV) and elevated collision energies MSE (12–40 eV)applied to the trap ‘‘T-wave’’ collision-induced dissociation(CID) cell with argon gas; the transfer collision cell wasadjusted for 1 eV, using a scan time of 1.0 s, both in low-energy and in high-energy CID orthogonal acceleration (oa)TOF MSE from m/z 50 to 3,000. The radio frequency offset(MS profile) was adjusted such that the liquid chromatography(LC)/MS data were effectively acquired fromm/z 300 to 3,000,which ensured that any masses observed in the LC/MSE dataless than m/z 300 were known to arise from dissociations inthe collision cell. Protein identifications and quantitativedata packagingwere generatedwith the use of dedicated algo-rithms (16) and searching against a species-specific proteindatabase. The databases used were randomized ‘‘on the fly’’during the database queries and appended to the original data-base to access the false-positive rate of identification. Forproper spectra processing and database searching conditions,a Protein Lynx Global Server v.2.4 with an ExpressionE Infor-matics v.2.4 license installed was used. The search was basedon retention time, ion intensities, charge state, and accuratemasses of both precursor and product ions.

2D SDS-PAGE

The 2D SDS-PAGE experiment was carried out according tothe Laemmli system (17). The necessary volume for obtaining400 mg total protein was estimated, and a 2D gel was built foreach patient. Briefly, the volume of each sample was mixedwith sample buffer (7 mol/L urea, 2 mol/L thiourea, 2%CHAPS, 20 mmol/L dithiothreitol (DTT), 1% immobilized pHgradient [I PG] buffer 3–10, and 0.001% blue bromophenol)to a final volume of 340 mL and rehydrated in IPG (3–10 linearpH gradient, 18 cm) strips overnight. Isoelectric focusing wascarried out for 24,000 VHs with the use of an Ettan IPGphor 3Isoeletric Focusing System (GE Healthcare). The IPG stripswere then soaked in equilibration buffer (50 mmol/LTris-HCL, pH 8.8, 6 mol/L urea, 30% glycerol, and 2% SDS)containing 26 mg/mL DTT for 15 minutes. Thereafter, theywere soaked into equilibration buffer containing 54 mg/mLiodoacetamide for 15 minutes. The second dimensionwas carried out in 10%–17.5% gradient acrylamide-bisacrylamide gels at 2.0 W/gel overnight (Ettan Dalt SixLarge Vertical System; GE Healthcare). The gels were fixedand proteins were visualized by Coomassie Brilliant BlueR-350 according to the manufacturer’s instructions. Subse-quently, the gels were imaged with the use of Scanner III(GE Healthcare) and analyzed with Imagemaster 2D Platinum7.0 software (GE Healthcare). This software makes matchesbetween the spots of different gels of similar samples toshow differences between individuals. To accurately comparethe spots between gels, image spot intensity was normalizedby dividing the raw intensity of each spot in a gel by the totalintensity of all the valid spots in that gel. Spots of interest

3

ORIGINAL ARTICLE: ANDROLOGY

(matches) were manually excised from the gels, trypsindigested, and used for MS/MS analysis. The ESI–quadru-pole-time of flight (Q-TOF) experiment was performed in aSynapt G1 HDMS Q-TOF (Waters). First, the peptides were in-jected into a Nanoacquity UPLC (Waters) system equippedwith a BEH130–C18 (Waters) column, where they were allseparated in 60 minutes by a 0.1%/0.1% acetonitrile/watergradient. The Synapt was operated in ESI þ mode. Peptidemasses obtained fromMS analysis were used for protein iden-tification in the Swiss-Prot 57.2 database with the use of thepeptide search routine Mascot (Matrix Science). All peptidemass values were considered to be monoisotopic, and masstolerance was set at 50 ppm. Mascot scores that led to a Pvalue of identification of <5% were considered to be signifi-cant. For MS/MS analysis, peaks were searched against theSwiss-Prot database, using the same setting of MS analysis,with a fragment tolerance of 0.3 Da.

Data Analysis

All protein data obtained from both proteomics strategieswere used for interactome construction. Protein accessionnumbers were converted to Uniprot ID/AC and loaded intoCytoscape 2.7.0 with the use of two plugins, Bisogenet 1.41and Gene Mania. These plugins work as open-source networkvisualization platforms based on protein-protein interactiondatabases, including the Database of Interacting Proteins,the Biomolecular Interaction Network Database, the HumanProtein Reference Database, and the Molecular InteractionDatabase, among others. The networks from both plugins’analyses were merged, resulting in a single representativenetwork for each group. Only the physical interactions wereconsidered. Finally, the networks were analyzed with theuse of the Bingo plugin, which is efficient in determiningwhich GO categories are significantly overrepresented in agiven set of genes or gene products.

RESULTSSeminal Analysis

The study groups included 12 patients with SCI; a total ofthree patients presented cervical, six thoracic, and three lum-bar injuries. Mean (�SD) time after injury was 4.8 � 2.16years among PVS and 11.8 � 6.94 years among EEJ patients.No significant difference was found when years after injurywas evaluated between PVS and EEJ groups (P¼ .092)(Table 1). As control subjects, ten men without SCI, presentingnormal sperm motility and concentration, were included. Re-sults for age, concentration, total sperm count, morphology,round cells, neutrophils, and vitality are presented inTable 1. No statistically significant differences were foundbetween PVS or EEJ and control groups regarding age,concentration, or total sperm count. As expected, PVS andEEJ groups presented significantly decreased morphology(P¼ .008 and P¼ .0003, respectively) and an increase in roundcells (P¼ .007 and P¼ .017, respectively) and neutrophils(P¼ .001 and P¼ .038, respectively) counts compared withthe control group. When these variables were tested betweenSCI groups, no statistically significant differences were found.

4

The same was observed when the variables red blood cells(P¼ .514) and vitality (P¼ .173) were compared between EEJand PVS groups. Motility results are presented in Table 2. Itwas observed that both PVS and EEJ groups presented loweraþb motility (P¼ .0003) and b motility (P¼ .0003) comparedwith the control group. On the other hand, SCI groupspresented higher values for d motility (P¼ .0003) comparedwith control. No statistically significant differences werefound when these variables were compared between EEJand PVS groups.

Seminal Plasma Proteome—MSE

The MSE analysis resulted in the identification of 637proteins; 35 proteins were exclusively identified in the controlgroup, 88 proteins were exclusively identified in the PVSgroup, and 66 proteins were exclusively identified inEEJ group (Supplemental Table 1; Supplemental Tables 1–7and Supplemental Figs. 1–2 are available online atwww.fertstert.org). Moreover, 52 proteins were identified inall groups, but with significantly different expression(Supplemental Table 2). The remaining 396 proteins wereidentified in all groups and were not differentiallyexpressed (Supplemental Table 3).

Individual Analysis—2D SDS-PAGE

All gels resulted from the 2D SDS-PAGE experiments werescanned, totaling nine gels in the control group, six gels in thePVS group, and six gels in the EEJ group. Protein spotintensities were analyzed both qualitatively (presence orabsence) and quantitatively (high, moderate, or lowprotein expression). After software analysis, one gel of eachgroup was chosen to be group representative and allstatistically different spots among groups were marked(Supplemental Fig. 1) for subsequent gel removal. However, dur-ing the spot removal process, those numbered 102, 103, 104, and105 (EEJ exclusive) could not be separated because of their loca-tion, being considered as a single spot. Thus, 39 protein spots ofinterest were excised from the gels and analyzedfirst by tandemMS and then with the use of the Mascot search algorithm. Morethanoneproteinwas identified per spot (Supplemental Tables 4–6); all peptideswitha scoreabove the significance thresholdwereconsidered to be positively identified. A score was calculated foreach peptide based on the number of fragments and length ofthe peptide in the fixed and variable modifications.Proteins expressed exclusively in the control, PVS, and EEJgroups are presented in Supplemental Table 4, differentiallyexpressed proteins are presented in Supplemental Table 5, andthe proteins identified in all groups are presented inSupplemental Table 6.

Data Analysis

Owing to the fact that we used two different proteomicsstrategies, not all proteins initially identified by the MSE

technique could also be identified by the 2D SDS-PAGEapproach, which, in turn, identified some proteins notidentified by MSE. At a methodologic glance, the MSE

technique proved to be an efficient protein discovery system,

VOL. - NO. - / - 2013

TABLE 1

Comparison of years after injury, age, sperm concentration, total sperm count, sperm morphology, round cells, neutrophils, red blood cells, andvitality among control, penile vibratory stimulation (PVS), and electroejaculation (EEJ) groups.

Control EEJ PVS P Value

Years after injury — 11.83 � 6.94[4.55–19.12]

4.83 � 2.16[2.57–7.10]

.092

Age (y) 35.5 � 5.27[31.72–39.27]

33.7 � 7.71[24.57–40.76]

27 � 6.87[19.79–34.21]

.059

Sperm concentration (�106/mL) 126.77 � 48.10[92.36–161.18]

95.45 � 104.26[�13.96–204.86]

106.13 � 231.04[�136.33–348.59]

.099

Total sperm count (�106) 403.13 � 157.35a

[290.57–515.69]58.98 � 57.04b

[�0.87–118.84]259.58 � 580.18b

[�349.28–868.44].004*

Sperm morphology (% normal) 13.7 � 2.67a

[11.79–15.61]5.33 � 3.08b

[2.10–8.56]6.33 � 5.03b

[�6.17–18.84].0002*

Round cells (�106) 2.15 � 1.52a

[1.06–3.24]8.5 � 9.23b

[�1.19–18.19]10.37 � 10.24b

[�0.37–21.11].007*

Neutrophils (�106) 0.5 � 0.43a

[0.19–0.81]3.88 � 4.52b

[�0.86–8.62]5.6 � 7.65b

[�2.43–13.63].004*

Red blood cells (�106) — 14.4 � 12.30[�16.15–44.95]

8.42 � 10.30[�7.96–24.81]

.514

Vitality (%) — 29.5 � 6.22[52.77–28.55]

52 � 8.47[�7.96–24.81]

.173

Note: Values are presented as mean� SD [95% confidence interval]. Analysis of variance (ANOVA) followed by Mann-Whitney (years after injury), ANOVA followed by least significant differences(age and for morphology), Student's t test (for red blood cells and vitality), and Kruskal-Wallis followed byMann-Whitney (sperm concentration, total sperm count, round cells, and neutrophils) testswere used.* Statistically significant difference (P< .05).a,b Different superscript letters in the same row indicate a statistically significant difference in the post hoc least significant differences test (P< .05).

da Silva. Seminal plasma proteome in SCI men. Fertil Steril 2013.

Fertility and Sterility®

allowing the analysis of even low-abundance proteins.Moreover, because the MSE technique does not requireseparation of proteins by 2D electrophoresis, many experi-mental variations are removed. However, the sample poolingaccording to our experimental design brought a limitationfactor for this technique. Pooling samples is a matter ofconcern among researchers, because it represents the averageof single expression but not of all the proteins present in amixture. Therefore, to overcome any possible loss in proteinidentification we also individually analyzed the same patientsthrough traditional 2D SDS-PAGE followed by MS, aiming toidentify differentially expressed proteins, which brought usinsights into the variability within a group. A great datavolume from both strategies was generated, resulting in adata analysis challenge. To overcome this problem, we choseto cross-compare all protein lists already presented, ending up

TABLE 2

Motility values (%) for control, penile vibratory stimulation (PVS), and ele

Control EE

Motility a 4.2 � 3.58[1.64–6.76]

0.33 �[�0.52

Motility b 58.8 � 4.13a

[55.84–61.75]14.83 �[�2.66

Motility c 3.3 � 1.95[1.91–4.70]

2.17 �[�0.26

Motility d 33.8 � 5.71a

[29.71–37.88]82.67 �[62.84–

Motility aþb 63 � 5.54a

[59.04–66.96]15.17 �[�2.82

Note: Values are presented as mean � SD [95% confidence interval]. Kruskal-Wallis followed by M* Statistically significant differences (P< .05).a,b Different superscript letters in the same row indicate a statistically significant difference in the p

da Silva. Seminal plasma proteome in SCI men. Fertil Steril 2013.

VOL. - NO. - / - 2013

with two relevant and simplified lists, one formed by proteinsexclusively found in the control group (Supplemental Table 7)and another by proteins exclusively shown in the SCI groups(PVS þ EEJ) (Supplemental Table 7). This strategy led us to abetter comparison between control and SCI groups regardingprotein-protein interaction networks and GO functionalenrichment. Furthermore, because we also aimed to assess ifparticular differences in semen retrieval techniques inducedchanges in the seminal plasma proteome, we also comparedthe individual PVS and EEJ protein lists.

Protein Interaction and GO Analysis

To obtain systematic insights into the roles that seminalplasma proteins play when in complex solution, we examinedthe protein-protein interactions between our identified

ctroejaculation (EEJ) groups.

J PVS P Value

0.82–1.20]

1.67 � 4.08[�2.62–5.95]

.061

16.67b

–32.32]10.83 � 16.86b

[�6.86–28.52].0003*

2.32–4.60]

1.67 � 2.87[�1.35–4.68]

.192

18.90b

102.50]85.83 � 22.22b

[62.51–109.15].0003*

17.14b

–33.15]12.5 � 19.94b

[�8.42–33.42].0003*

ann-Whitney test was used.

ost hoc least significant differences test (P< .05).

5

ORIGINAL ARTICLE: ANDROLOGY

proteins with the use of the Cytoscape software as ourbioinformatics analysis platform. The visualization formatof our results allowed us to study protein expression inconjunction with its interactions. Protein-protein interactionsare fundamental to all biologic events, and their overalldetermination enabled us to understand seminal plasma asan integrative system. The construction of protein-proteininteraction networks was based mainly on database search.These databases store genetic information previouslyobtained from human genome studies as well as datafrom protein interactions obtained and validated bydifferent experimental methods. Although the databases arefar from being completed, the data available today allowthe creation of complex and representative proteinmaps (18–20). Therefore, we loaded all the proteinsidentified and quantified for each group separately intoCytoscape. Figure 1 shows the main protein clustersdisplayed within an interaction network. The nodes’ colors

FIGURE 1

Illustrative representation of highlighted clusters found in the interactionvibratory stimulation, and (C) electroejaculation groups. Color representsprotein common to two groups (blue), according only to the ‘‘shotgun prsodium dodecyl sulfate–polyacrylamide gel electrophoresis followed by tfinal gene ontology analysis.da Silva. Seminal plasma proteome in SCI men. Fertil Steril 2013.

6

portray their expression: red for exclusivity, blue forpresence in at least two groups, and green for presence inall groups. The proteins exclusively identified in each groupwere also examined for their biologic association to GOcategories (Supplemental Fig. 2). Moreover, takingadvantage of the Bingo plugin and using human genescurrently carrying human GO annotations as query files, itwas possible to verify the distribution of our proteins in theGO category structure. Although not all identified proteinshave been currently annotated by the GO consortium, theyalready covered most of GO ‘‘biologic process,’’ ‘‘cellularcomponent,’’ and ‘‘molecular function’’ categories. Thus,this study could functionally relate the proteins present inthe seminal plasma of control subjects and SCI patients tocentral reproductive processes such as sperm and oocytefusion, insemination, and acrosome reaction. The acutephase of inflammatory response and response to stimulitriggered by triglycerides were also processes related to all

network of differentially secreted proteins from (A) control, (B) penileunique group protein (red), protein common to all groups (green), oroteomics’’ experiment (MS) results. Results from the two-dimensionalandem MS were after uploaded and merged to these networks for

VOL. - NO. - / - 2013

Fertility and Sterility®

groups. The proteins present in the seminal plasma of SCIpatients relate to actin-related processes such as smoothmuscle function, cytoskeletal binding, and response tocalcium, as well as processes related to homeostasis,transport, and oxidation of iron ion. Interestingly, anindividual close look at the PVS group revealed that thiswas the only group related to response to hydrogen peroxide,hypoxia, and severe inflammatory response.

DISCUSSIONIn the past few years the number of individuals with SCI hasincreased significantly. It is estimated that >80% of all SCIpatients are men at the peak of their reproductive age (5, 6).Once the injury is established, both sexual and reproductivelives are compromised. This clinical scenario arises from thefact that either erection or ejaculation can be achieved onlyif the psychologic and physical stimulation processes can bewell coordinated. The ejaculatory reflex, for example, is acomplex neurobiologic and muscular event coordinated inthe spinal cord. Spinal ejaculatory centers receive stimulifrom supraspinal regions as well as from peripheral sensoryafferents from the genital area (21). The first phase ofejaculation—emission—is dependent on thoracolumbarsympathetic fibers from segments T10–L2 and the lastphase, expulsion, on somatic fibers from segments S2–S4(22). Because SCI leads to a disruption of the nerve supplyresponsible for this complex process, only �9% of SCI mencan ejaculate through masturbation or sexual intercourse(4). Therefore, assisted ejaculation in SCI patients isnecessary, and biologic fatherhood becomes possible, inmost of the cases, only when assisted reproductiontechniques (ART) are applied (23).

However, despite the clinical efforts made in thetreatment of SCI-related infertility, even ART can be impairedby the poor semen quality of these patients. Semen analysis ofharvested samples from SCI men generally present normal tohigh sperm concentration, and the number of cells do notsuffer progressive decline during the postinjury years (5).On the other hand, decrease in sperm motility (24), highincidence of amorphous sperm cells (25), and leukocytosper-mia (R1.0 � 106 neutrophils/mL) (26, 27) are commonlyfound. Confirming these findings, the present study foundno significant differences between SCI and control groupsregarding sperm concentration. As expected, SCI groupspresented a significant decrease in sperm morphology andan increase in neutrophils compared with the control group.Sperm motility also was decreased in the SCI groups.

Some evidence suggests that seminal plasma, which iscomposed of secretions originating from the testes, epididy-mides, and accessory glands, is a major contributor to thelow semen quality of these men. In this regard, research hasbeen focusing attention on the molecular changes that SCIbrings to seminal plasma. Interestingly, Basu et al. (28)obtained seminal samples of SCI and nonlesioned menfor a constitutional comparison. Their results remarkablydemonstrated a differential protein expression between thetwo groups, mainly regarding cytokines and proteins oflow molecular mass released by different cell lines of the

VOL. - NO. - / - 2013

immunologic system. Some proteins, such as interleukin (IL)1b, IL-6, IL-12, interferon-gamma, and tumor necrosis factoralpha (TNF-a), were overexpressed in SCI seminal plasmacompared with control.

Focusing on a deeper molecular understanding ofseminal plasma of SCI men, the present study used shotgunproteomics on pooled samples, enabling a direct proteinprofile screening of seminal plasma. This approach wasfacilitated by the use of MSE technology, and as a result,637 unique proteins could be identified in control, PVS, andEEJ groups. Despite the considerable amount of identifiedproteins, our study also benefited from an individualinvestigation of seminal plasma samples. All samples wereindividually separated with the use of 2D SDS-PAGE followedby ESI-MS/MS analysis. At this step, 18 different proteinscould be identified among the groups. Owing to the overallvolume of functional proteins found in this study, theintegration and further data interpretation was greatlyfacilitated by the use of bioinformatics tools.

Control Versus SCI

As expected, GO analysis of our protein networks showed thatdespite the differential protein expression patterns betweengroups, all groups presented proteins related to reproductiveprocesses, such as insemination and oocyte fertilization.However, a deeper look into the ‘‘biologic processes’’ GOcategory revealed some differences between control and SCIgroups (Supplemental Fig. 2). For example, processes suchas cell adhesion, locomotion, and prostate gland growth areapparently related to proteins found only in the control group.Briefly, the biologic process of cell adhesion links up tomolecules responsible for the attachment of a cell to anothercell or to the extracellular matrix, and the locomotion processrelates to proteins involved in cellular movement from onelocation to another. These findings intrinsically connect toour clinical picture and not only enhance our belief thatsome important molecules in sperm-oocyte interaction couldbe missing, but also support the hypothesis that the decreasein motility may be due to the deficiency in some molecules.Controversially, a process assigned as cell motility anddefined as the controlled self-propelled movement of a cellthat resulting in cellular translocation, was annotated onlyto proteins from SCI groups.

Although assigned to only one protein, an interestingfinding was the process named ‘‘prostate gland growth’’only in control samples. This is a process whose specificoutcome is the progression of the prostate gland over time,from its formation to the mature structure. It could be thatproteins related to this particular process when secreted inseminal plasma indicate an ensured and maintained prostaticdevelopment, which may not be occurring in SCI patients. It isknown that prostatic epithelial cells secrete numerousproteins and other substances into the seminal plasma thatare needed either for optimal fertility or for auto/paracrineregulation of epithelia growth, differentiation, and secretoryfunction. Moreover, the glandular epithelial and secretoryfunction are stimulated and maintained by continuingpresence of serum testosterone, which must be converted by

7

ORIGINAL ARTICLE: ANDROLOGY

prostatic 5a-reductase into dihydrotestosterone. Recently, ithas been shown that pituitary-derived prolactin and growthhormone influence prostatic growth and function (29–32).Several animal studies corroborate this fact and havealready demonstrated that the lack of neurologic control ofthe prostate because of SCI generates atrophy and decreasedfunction of the gland. It was observed in rats that surgicaldenervation of the prostate resulted in a decrease in organweight, epithelial height, and secretory activity (33).Moreover, this decrease in prostate weight after prostaticdenervation was not associated with a decrease in DNAcontent per organ, suggesting that a decrease in syntheticand/or secretory activities, rather than a decrease in cellnumber, might be responsible for the smaller prostate in thedenervated animals (32). On the other hand, the majority ofhuman studies regarding prostatic atrophy and secretorydeficits remain inconclusive owing to small sample sizes,different inclusion and exclusion criteria, different durationof SCI, different age range of patients, and different assaysused for measurement, requiring further evaluation in welldesigned population-based studies (34).

Furthermore, the comparison of control and SCI groupsregarding the GO ‘‘cellular component’’ category showedthat in all groups many proteins are annotated as beingcell, membrane, and organelle parts. This comes from thefact that the cellular and noncellular components of seminalplasma form a tight complex of interactions where secretedproteins and cell membrane proteins interact mutually. Inaddition, it is true that cell disruption and leak of cellcontent into seminal plasma might have happened due toour experimental conditions (high centrifugal force). How-ever, physiologically, components of different cell lines(sperm cells, white and red blood cells) are constantlycontaminating seminal plasma owing to degradation withinthe male reproductive tract. Nonetheless, it is notable thatcompared with control, the SCI groups have more proteinsannotated in cell, membrane, and organelle parts(Supplemental Fig. 2B). An explanation for this fact is thatSCI men present more and different cell lines in seminalplasma, which corroborates with the leukocytospermiaobserved in these patients. Leukocytospermia can be detri-mental in seminal plasma mainly because higher abundanceof leukocytes leads to high production of reactive oxygenspecies (ROS). ROS are unstable reactive molecules thatcan cause toxicity or death to different cells, includingsperm cells. ROS include hydrogen peroxide (H2O2), supero-xide anion (O2�), and hydroxyl radicals ($HO) (5). It isknown that normal concentrations of ROS play an importantrole in the sperm physiology. Experimental data alreadydefined processes such as sperm hyperactivation, capa-citation, and acrosome reaction as ‘‘oxidative’’ or ‘‘redoxregulated’’ (35). However, detrimental effects to spermstructure and functionality can also be attributed to ROS(36). Under certain conditions there is an imbalance betweenthe antioxidant ability of seminal plasma and ROSproduction, leading to a process known as oxidative stress.Sperm are particularly vulnerable to oxidative stress owingto their high concentration of unsaturated fatty acids.Consequently, decreased membrane fluidity and increased

8

mitochondrial damage, and thus decreased sperm motility,are the first notable effects (37–39).

Another finding in our protein networks was theformation of a lactotransferrin (TRFL) and ceruloplasmin(CERU) cluster. CERU was expressed only in the SCI groupsand plays a role in iron metabolism and homeostasis,allowing apotransferrin to incorporate Fe3þ (40). TRFL is aglycoprotein involved in the control of iron levels throughits strong but reversible binding to Fe3þ (41). Interestingly,it has been recently reported that CERU and TRFL interactboth in vivo and in vitro to form a stable complex, whichincreases the CERU ferroxidase activity (42). This is particu-larly important because iron in its ionic form participates inthe formation and release of ROS. Moreover, GO analysis ofthis cluster evidenced its involvement in response to bacterialinfection and humoral immune response. It is clear that ourprotein clusters and GO analyses led us to relate the proteinsfound in this study with different ROS formation processes inthe seminal plasma of SCI patients. However, it is notable thatonly the presence of so many different proteins in the seminalplasma mix can potentially be harmful to the sperm–seminalplasma homeostasis. This happens because the presence ofmore proteins than physiologically expected can somehowdisrupt the balance existing in the protein system, whichcould be observed as different clusters expression and GOanalysis.

In this study, a total of 38 proteins seem to be expressedonly in the control group. This differential expression patternrepresents important potential targets for prognosis andfuture treatment strategies for general male infertility.Although not all proteins correlate directly with reproductionor fertilization processes, their importance in seminal plasmashould not be underestimated. Earlier studies have reportedthat proteins showing no apparently important function inseminal plasma itself have already been established as goodbiomarkers of preserved spermatogenesis in other species,such as prostaglandin D synthase in cows (43–45).However, to validate some of the proteins suggested asfertility biomarkers, future studies should consider theindividual evaluation of higher numbers of patients toensure confidence and avoid conflicting results.

PVS Versus EEJ

Currently, the most widely used methods to stimulateejaculation in SCI patients are PVS and EEJ (6). Bothtechniques have basically the same purpose but differ insome aspects. In general, PVS is an effective and welltolerated procedure that involves placing a vibrator in theperifrenular area of the glans penis. The artificial stimulationtakes a few minutes until antegrade ejaculation occurs.Optimal candidates for this technique are those SCI patientswith injury above T10, because to achieve ejaculation throughPVS, patients need an intact ejaculatory arc reflex, whichmeans that the sensory afferents from the penis, the spinalcord at the levels of sensory input (S2–S4), the efferent exit(T10–L2), and sympathetic outflow must all be functioning(21). If these initial requisites are met, PVS is effective in88% of cases. EEJ is an alternative technique to PVS. In

VOL. - NO. - / - 2013

Fertility and Sterility®

EEJ, a specific electric probe is connected to an external powersupply and is introduced into the rectum. The device voltage ismanually increased in time intervals until the occurrence ofejaculation. The success of EEJ has been reported in 97% ofSCI patients, independent from injury level. In the presentstudy, PVS was applied in patients presenting differentlevels of cervical and thoracic injuries and EEJ in patientspresenting different levels of cervical, thoracic, and lumbarinjuries. Regarding adverse effects, PVS is known to causeskin abrasions and, in some cases, autonomic dysreflexia(AD) (46). EEJ can also cause AD, pain (general anesthesia isoften required), and retrograde ejaculation. Earlier studieswith SCI men have demonstrated that electric currentadministered in vivo (via EEJ) and in vitro decreased spermmotility and increased the formation of ROS (47, 48).However, so far no definitive conclusions regardingwhether PVS is better than EEJ regarding semen qualityhave been attained. To bring new information into thisdiscussion, the present study compared PVS and EEJregarding seminal plasma protein content to access possibledifferential effects on protein expression.

To start with, a result found during GO analysis related toa protein cluster identified both in PVS and EEJ networks.This cluster, almost entirely composed of actin-relatedproteins, consisted mainly of six proteins (ACTC, ACTB,ACTG, NRAP, and ACTN3), with nesprin-1 (SYNE1)exclusively expressed in the EEJ group and spectrin alphachain, nonerythrocytic 1 (SPTA2) exclusively expressed inthe PVS group (Fig. 1). The functional enrichment analysisof the referred cluster reported a relation with calciumresponse, movements based on actin filaments, binding tocytoskeletal components, and ATP and nitric oxide syntheses.Both SYNE1 and SPTA2 are expressed in peripheral bloodleukocytes (which could also explain its presence in theseminal plasma) but have a high expression in skeletal andsmooth muscles. The seminal vesicle in male adults is formedby an epithelium surrounded by stromal cells, including aninner layer composed of contractile smooth muscle. Develop-ment and normal function of the seminal vesicles depend onreciprocal interactions between stroma and epithelia. Studieswith recombinant epithelia demonstrated that the stromalcompartment is the main site for androgen action. Thus, thehormonal action is decisive in the fate and functionalmorphology of this epithelium, regulating their proliferationor apoptosis (49). Similarly, the depth of their ownmaintenance and muscle layer appears to depend onandrogenic hormones (50). Therefore, any change in thehypothalamic-pituitary gonadal axis, such as that presentedby many SCI patients (50), could interfere in the circulatinghormone levels, leading to a consequent reduction in thesmooth muscle layer that forms the seminal vesicles. Thepresent study did not aim to measure the serum levels oftestosterone in SCI patients, but earlier data report that thesepatients have their testosterone levels changed as a result ofthe neurologic trauma (51). Thus, the decrease in the smoothmuscles size may release some of its constituent proteins,which end up being secreted in seminal plasma. Furthermore,the description of such processes in both groups, EEJ andPVS, suggests that these mechanisms might be a direct

VOL. - NO. - / - 2013

consequence of the physiologic changes after SCI. The factis that the decreased muscle layer and epithelium of theseminal vesicles are limiting factors for the correctfunctioning of this gland. Therefore, a detailed functionalstudy not only of the seminal vesicles but also of otheraccessory glands (as previously discussed) could be veryuseful to further investigate that hypothesis. However,whether these different types of actin and muscle proteinshave some direct adverse impact on sperm or somehow shiftthe biochemical cascade of important biologic events whenreleased in seminal plasma is also a matter of futureinvestigation.

The PVS protein network caught our attention becausecluster analysis resulted in many exclusive cluster formations(Fig. 1B). For example, we could observe a cluster functionallyrelated to response to H2O2 and hypoxia. That clusterpresented three of the PVS exclusive proteins: 1) CREB-binding protein (CBP); 2) nuclear factor of activated T cells,cytoplasmic 4 (NFAC4); and 3) histone acetyltransferasep300 (EP300). The presence of such a cluster related to H2O2

and hypoxia response may lead us to suggest that ROSformation and oxidative processes are indeed, and moreintensely, occurring in the seminal plasma of SCI patientswho undergo PVS. This finding corroborates some studiesthat have successfully demonstrated that semen from SCImen have higher ranges of ROS compared with semen frommen with other fertility disorders (37). Moreover, becausewe found this process more intensively occurring in thePVS group, either ROS formation or depletion in seminalplasma antioxidant ability could be a process influenced byPVS technique. However, to state that, further investigationis necessary. Still regarding the PVS protein network, twoother clusters were also found only in this group, both relatingto immunologic system defense.

The exclusive presence of complement C3 protein (CO3)in this group shows that infectious processes may beoccurring more intensively. The cleavage of CO3 is respon-sible for initiating the alternate pathway of the complementsystem, which has the ability to mediate acute inflammatoryreactions. Moreover, an immunoglobulin identified as Igkappa chain C region (IGKC) showed expression only in thePVS group and apparently responds for the secondaryimmune response when interacting with other proteins withinthe same cluster. Urinary tract infections (UTIs) can lead to anincrease not only in immunoglobulin rates but also in otherimmune system components in seminal plasma (52). Howev-er, whether or not these components are deleterious to spermremains unclear. It is true that the specificity betweenimmunoglobulin-antigen binding may ensure some safetyto sperm cells, but a recent study reported high concentrationsof cytotoxic cytokines in the seminal plasma of SCI men whopresented asthenozoospermia (53). Another study, usingseminal plasma of 17 SCI patients and monoclonal antibodiesto IL-6, IL-1b, and TNF-a at the median neutralization doseconcentrations for 1.0–1.5 hours showed improvement insperm motility of all patients (54). According to de Lamirandeet al. (55) cytokines increase the production and release ofROS, reducing not only spermmotility but also, e.g., its abilityto properly penetrate the oocyte. The presence of a great

9

ORIGINAL ARTICLE: ANDROLOGY

amount of proteins related to immunologic process in the PVSgroup suggests that SCI patients in this particular group maysuffer more from recurrent infections. It is true that UTIs arecommon findings in almost all SCI patients and develop asa result of neurogenic bladder and the need for catheterization(56). Even though our semen analysis did not demonstrateany statistical significant difference when the variable‘‘neutrophils’’ was tested between PVS and EEJ groups, GOanalysis demonstrated that somehow these immunologicprocesses are more intense in PVS patients. However,although the continued use of PVS can lead to skin abrasionsand some local skin infection, the technique itself does notseem to be involved in the etiology of UTI. In any case, afollow-up study should be performed to confirm and betterexplain these findings.

In conclusion, proteomics results and subsequentnetwork and GO analyses found important differences inthe protein expression when seminal plasma samples of SCIpatients were compared with control samples. Althoughsimilar processes and functions were described for all groups,the differential protein expression in SCI samples enrichedfunctions not observed in control samples. This suggeststhat a deviation from homeostasis is occurring in the seminalplasma of SCI patients. This fact is particularly importantbecause seminal plasma with altered functions hindersfertilization and probably accounts for the poor semen qualityobserved, as has been previously reported (5). Moreover,despite small differences, proteins reported in both PVS andEEJ groups correlate well with the pathophysiology of SCI-related infertility. However, it was not possible to concludeif the differences found were due to technical effects.

REFERENCES1. Kraus JF, Franti CE, Riggins RS, Richards D, Borhani NO. Incidence of

traumatic spinal cord lesions. J Chronic Dis 1975;28:471–92.2. McKinley WO, Jackson AB, Cardenas DD, DeVivo MJ. Long-term medical

complications after traumatic spinal cord injury: a regional model systemsanalysis. Arch Phys Med Rehabil 1999;80:1402–10.

3. Linsenmeyer TA, Perkash I. Infertility in men with spinal cord injury. ArchPhys Med Rehabil 1991;72:747–54.

4. Brackett NL, Ibrahim E, Iremashvili V, Aballa TC, Lynne CM. Treatment forejaculatory dysfunction in men with spinal cord injury: an 18-year singlecenter experience. J Urol 2010;183:2304–8.

5. Brackett NL, Nash MS, Lynne CM. Male infertility following spinal cordinjury: facts and fiction. Phys Ther 1996;76:1121–231.

6. Sonksen J, Ohl DA. Penile vibratory stimulation and eletroejaculation in thetreatment of ejaculatory dysfunction. J Androl 2002;25:324–32.

7. Brackett NL, Davi RC, Padron OF, Lynne CM. Seminal plasma of spinalcord injured men inhibits sperm motility of normal men. J Urol 1996;155:1632–5.

8. Brackett NL, Lynne CM, Aballa TC, Ferrell SM. Sperm motility from the vasdeferens of spinal cord injured men is higher than from the ejaculate.J Urol 2000;164:712–5.

9. Evans JP, Kopf GS. Molecular mechanisms of sperm-egg interactions andegg activation. Andrologia 1998;30:297–307.

10. Primakoff P, Myles DG. Penetration, adhesion, and fusion in mammaliansperm-egg interaction. Science 2002;296:2183–5.

11. Yi YJ, Manandhar G, Oko RJ, Breed WG, Sutovsky P. Mechanism ofsperm–zona pellucida penetration during mammalian fertilization: 26Sproteasome as a candidate egg coat lysin. Soc Reprod Fertil Suppl 2007;63:385–408.

10

12. World Health Organization. Laboratory manual for the examination ofhuman semen and sperm–cervical mucus interaction. 4th ed. Cambridge,U.K.: Cambridge University Press; 1999.

13. Kruger TF, Menkeveld R, Stander FS, Lombard CJ, van der Merwe JP, vanZyl JA, et al. Sperm morphology features as a prognostic factor in vitrofertilization. Fertil Steril 1986;46:1118–23.

14. Katkov II, Mazur P. Influence of centrifugation regimes on motility,yield, and cell associations of mouse spermatozoa. J Androl 1998;19:232–41.

15. Wiechelman K, Braun R, Fitzpatrick J. Investigation of bicinchoninic acidprotein assay: identification of the groups responsible for color formation.Anal Biochem 1988;175:231–7.

16. Silva JC, Denny R, Dorschel CA, Gorenstein M, Kass IJ, Li GZ, et al.Quantitative proteomic analysis by accurate mass retention time pairs.Anal Chem 2005;77:2187–200.

17. Laemmli UK. Cleavage of structural proteins during the assembly of the headof bacteriophage T4. Nature 1970;227:680–5.

18. Shannon P, Markiel A, Ozier O, Baliga NS, Wang JT, Ramage D, et al.Cytoscape: a software environment for integrated models of biomolecularinteraction networks. Genome Res 2003;13:2498–504.

19. Spirin V, Mirny LA. Protein complexes and functional modules in molecularnetworks. Proc Natl Acad Sci U S A 2003;100:12123–8.

20. Ideker T, Ozier O, Schwikowski B, Siegel AF. Discovering regulatory andsignalling circuits in molecular interaction networks. Bioinformatics 2002;(18 Suppl 1):S233–40.

21. Barazani Y, Stahl PJ, Nagler HM, Stember DS. Management of ejaculatorydisorders in infertile men. Asian J Androl 2012;14:525–9.

22. Giuliano F, Clement P. Neuroanatomy and physiology of ejaculation.Annu Rev Sex Res 2005;16:190–216.

23. Sedor JF, Hirsch IH. Evaluation of sperm morphology of eletroejaculates ofspinal cord injured men by strict criteria. Fertil Steril 1995;63:1125–7.

24. Ohl DA, Sonksen J, Mengue AC, Mc Cabe M, Keller LM. Electroejaculationversus vibratory stimulation in spinal cord injured men: sperm quality andpatient preference. J Urol 1997;157:2147–9.

25. Brackett NL, Lynne CM, Weizman MS, Bloch WE, Abae M. Endocrineprofiles and semen quality of spinal cord injured men. J Urol 1994;151:114–9.

26. Trabulsi EJ, Shupp-Byrne D, Sedor J, Hirsch IH. Leukocyte subtypes inelectroejaculates of spinal cord injured men. Arch Phys Med Rehabil 2002;83:31–4.

27. McDonald JW, Sadowsky C. Spinal-cord injury. Lancet 2002;359:417–25.

28. Basu S, Aballa TC, Ferrell SM, Lynne CM, Brackett NL. Inflammatory cytokineconcentrations are elevated in seminal plasma of men with spinal cordinjuries. J Androl 2004;25:250–4.

29. Colao A, Marzullo P, Ferone D, Spiezia S, Cerbone G, Marin�o V, et al.Prostatic hyperplasia: an unknown feature of acromegaly. J Clin EndocrinolMetab 1998;83:775–9.

30. Costello LC, Franklin RB. Effect of prolactin on the prostate. Prostate 1994;24:162–6.

31. Wennbo H, Kindblom J, Isaksson OG, T€ornell J. Transgenic mice overexpress-ing the prolactin gene develop dramatic enlargement of the prostate gland.Endocrinology 1997;138:4410–5.

32. Untergasser G, Rumpold H, Hermann H, Dirnhofer S, Jilg G, Berger P.Proliferative disorders of the aging human prostate: involvement of proteinhormones and their receptors. Exp Gerontol 1999;34:275–87.

33. Wang JM, McKenna KE, McVary KT, Lee C. Requirement of innervationfor maintenance of structural and functional integrity in the rat prostate.Biol Reprod 1991;44:1171–6.

34. Shim HB, Jung TY, Lee JK, Ku JH. Prostate activity and prostate cancer inspinal cord injury. Prostate Cancer Prostatic Dis 2006;9:115–20.

35. de Lamirande E, Jiang H, Zini A, Kodama H, Gagnon C. Reactive oxygenspecies and sperm physiology. Rev Reprod 1997;2:48–54.

36. Sanocka D, Kurpisz M. Reactive oxygen species and sperm cells. Reprod BiolEndocrinol 2004;2:12.

37. Sharma JK. Agarwal. A Role of reactive oxygen species in male infertility.Urology 1996;48:835–50.

VOL. - NO. - / - 2013

Fertility and Sterility®

38. Aitken RJ, Ryan AL, Baker MA, McLaughlin EA. Redox activity associatedwith the maturation and capacitation of mammalian spermatozoa. FreeRadic Biol Med 2004;36:994–1010.

39. Agarwal A, Sharma RK, Nallella KP, Thomas AJ Jr, Alvarez JG, Sikka SC.Reactive oxygen species as an independent marker of male factor infertility.Fertil Steril 2006;86:878–85.

40. Hellman NE, Gitlin JD. Ceruloplasmin metabolism and function. Annu RevNutr 2002;22:439–58.

41. Lizzi AR, Carnicelli V, Clarkson MM, Di Giulio A, Oratore A. Lactoferrinderived peptides: mechanisms of action and their perspectives asantimicrobial and antitumoral agents. Mini Rev Med Chem 2009;9:687–95.

42. Sokolov AV, Pulina MO, Zakharova ET, Susorova AS, Runova OL,Kolodkin NI, et al. Identification and isolation from breast milk ofceruloplasmin-lactoferrin complex. Biochemistry 2006;71:160–6.

43. Gerena RL, Irikura D, Eguchi N, Urade Y, Killian GJ. Immunocytochemicallocalization of lipocalin-type prostaglandin D synthase in the bull testisand epididymis and on ejaculated sperm. Biol Reprod 2000;62:547–56.

44. Leone MG, Haq HA, Saso L. Lipocalin type prostaglandin D-synthase: whichrole in male fertility? Contraception 2002;65:293–5.

45. Moura AA, Chapman DA, Koc H, Killian GJ. Proteins of the cauda epididymalfluid associatedwith fertility ofmature dairy bulls. J Androl 2006;27:534–41.

46. Sheel AW, Krassioukov AV, Inglis JT, Elliott SL. Autonomic dysreflexia duringsperm retrieval in spinal cord injury: influence of lesion level and sildenafilcitrate. J Appl Physiol 2005;99:53–8.

47. RajasekaranM,HellstromWJ, SparksRL, SikkaSC. Spermdamagingeffects ofelectric current: possible role of free radicals: Reprod Toxicol 1994;8:427–32.

VOL. - NO. - / - 2013

48. Sikka SC, Nang R, Kukuy E, Walker CF, Hellstrom WJ. The detrimentaleffects of electric current on normal human sperm. J Androl 1994;15:145–50.

49. Cunha GR, Alarid ET, Turner T, Donjacour AA, Boutin EL, Foster BA. Normaland abnormal development of the male urogenital tract. Role of androgens,mesenchymal-epithelial interactions, and growth factors. J Androl 1992;13:465–75.

50. Welsh M, Moffat L, Jack L, McNeilly A, Brownstein D, Saunders PT, et al.Deletion of androgen receptor in the smooth muscle of the seminal vesiclesimpairs secretory function and alters its responsiveness to exogenoustestosterone and estradiol. Endocrinology 2010;151:3374–85.

51. Bauman WA, Spungen AM. Metabolic changes in persons after spinal cordinjury. Phys Med Rehabil Clin North Am 2000;11:109–40.

52. Strugnell RA, Wijburg OL. The role of secretory antibodies in infectionimmunity. Nat Rev Microbiol 2010;8:656–67.

53. Brackett NL, Cohen DR, Ibrahim E, Aballa TC, Lynne CM. Neutralization ofcytokine activity at the receptor level improves sperm motility in men withspinal cord injuries. J Androl 2007;28:717–21.

54. Cohen DR, Basu S, Randall JM, Aballa TC, Lynne CM, Brackett NL. Spermmotility in menwith spinal cord injuries is enhanced by inactivating cytokinesin the seminal plasma. J Androl 2004;25:922–5.

55. de Lamirande E, Leduc BE, Iwasaki A, Hassouna M, Gagnon C. Incresasedreactive species formation in semen of pacients with spinal cord injury. FertilSteril 1995;63:637–42.

56. Cardenas DD, Hooton TM. Urinary tract infection in persons with spinal cordinjury. Arch Phys Med Rehabil 1995;76:272–80.

11

SUPPLEMENTAL FIGURE 1

Coomassie-stained two-dimensional sodium dodecyl sulfate–polyacrylamide gel electrophoresis maps of seminal plasma for (A) control, (B) penilevibratory stimulation, and (C) electroejaculation groups. Molecular mass separation is 225–12 kDa (top to bottom). Numbered spots indicateproteins that have statistically significant differential expression between groups according to the Imagemaster 2D Platinum 7.0 software(GE Healthcare), which made matches between the spots of different gels to show differences between individuals. One gel was chosen as arepresentative map of its group. The three groups were compared, and 39 spots of interest (matches) were manually excised from the gels,trypsin digested, and used for tandem mass spectrometry analysis. Identification results for these spots are shown in the Supplemental Tables.da Silva. Seminal plasma proteome in SCI men. Fertil Steril 2013.

ORIGINAL ARTICLE: ANDROLOGY

11.e1 VOL. - NO. - / - 2013

SUPPLEMENTAL FIGURE 2

da Silva. Seminal plasma proteome in SCI men. Fertil Steril 2013.

VOL. - NO. - / - 2013 11.e2

Fertility and Sterility®

SUPPLEMENTAL FIGURE 2 Continued

Some functional annotations for the gene ontology categories (A) ‘‘biologic process,’’ (B) ‘‘cellular component,’’ and (C) ‘‘molecular function’’enriched by proteins exclusively identified in the seminal plasma of controls and spinal cord injury (SCI; penile vibratory stimulation andelectroejaculation ) men.da Silva. Seminal plasma proteome in SCI men. Fertil Steril 2013.

ORIGINAL ARTICLE: ANDROLOGY

11.e3 VOL. - NO. - / - 2013