Combined proteomic and miRNome endocrine …...contributing equally with 50μg of total protein....

Molecular Human Reproduction, pp. 1–14, 2019

doi:10.1093/molehr/gaz003

ORIGINAL ARTICLE

Combined proteomic and miRNomeanalyses of mouse testis exposed to anendocrine disruptors chemicalsmixture reveals altered toxicologicalpathways involved in male infertilityJulio Buñay 1, Eduardo Larriba3, Daniel Patiño-Garcia1,Paulina Urriola-Muñoz1,2, Ricardo D. Moreno1,*,†, and Jesús delMazo3,*,†1Department of Physiology, Faculty of Biological Sciences, Pontificia Universidad Catolica de Chile (PUC), Santiago 8331150, Chile2Chemistry Institute, Pontificia Universidad Católica de Valparaíso, Valparaíso 2340000, Chile 3Department of Cellular and MolecularBiology, Centro de Investigaciones Biologicas (CIB-CSIC), 28040 Madrid, Spain

*Correspondence address. Department of Physiology, Faculty of Biological Science, Pontificia Universidad Católica de Chile, Alameda 340,8331150 Santiago, Chile. Tel: +562-2354-2885; E-mail: [email protected] (R.D.M.)/Department of Cellular and Molecular Biology, CIB(CSIC), Ramiro de Maeztu, 9, 28040 Madrid, Spain. Tel: +34-918373112x4324; E-mail: [email protected] (J.d.M.)

Submitted on May 6, 2018; resubmitted on December 23, 2018; editorial decision on January 23, 2019; accepted on January 24, 2019

The increase in male idiopathic infertility has been associated with daily exposure to endocrine disruptors chemicals (EDCs). Nevertheless,the mechanisms of action in relation to dysregulating proteins and regulatory microRNAs are unknown. We combined proteomic andmiRNome analyses of mouse testis chronically exposed to low doses of a define mixture of EDCs [phthalates: bis(2-ethylhexyl), dibutyl andbenzyl-butyl; 4-nonylphenol and 4-tert-octylphenol], administered in the drinking water from conception until adulthood (post-natal Day 60/75) and compared them with no-exposed control mice. We analysed fertility parameters and global changes in the patterns of mice testisproteome by 2D electrophoresis/mass spectrometry, along with bioinformatic analyses of dysregulated microRNAs, and their associationwith published data in human infertile patients. We detected a decrease in the potential fertility of exposed mice associated with changes inthe expression of 18 proteins (10 up-regulated, 8 down-regulated). Functional analysis showed that 89% were involved in cell death.Furthermore, we found a group of 23 microRNAs/isomiRs (down-regulated) correlated with six of the up-regulated target proteins(DIABLO, PGAM1, RTRAF, EIF4E, IVD and CNDP2). Regarding this, PGAM1 up-regulation was validated by Western blot and mainlydetected in Sertoli cells. Some of these microRNA/protein dysregulations were reported in human testis with spermatogenic failure. Overall,a chronic exposure to EDCs mixture in human males could potentially lead to spermatogenic failure through changes in microRNA expres-sion, which could post-transcriptionally dysregulate mRNA targets that encode proteins participating in cell death in testicular cells. Finally,these microRNA/protein dysregulations need to be validated with other EDCs mixtures and concentrations.

Key words: endocrine disruptors / PGAM1 / miRNAs / isomiRs / proteome / spermatogenic failure / infertility

IntroductionEndocrine disruptors chemicals (EDCs) such as phthalates and alkyl-phenols are environmentally widespread man-made chemicals towhich humans and wildlife are chronically exposed from foetal life toadulthood (Bergman et al., 2012). At a toxicological level, the effectson male fertility are a hallmark of the exposure to EDCs. This has

been documented by epidemiological studies, which report correla-tions between idiopathic male infertility (specially deterioration insperm quality and quantity) and high levels of phthalates such as di-(2-ethylhexyl)phthalate (DEHP), dibutyl phthalate (DBP) and benzyl butylphthalate (BBP or BzBP); and alkylphenols such as nonylphenol (NP)and octylphenol (OP) detected in both urine and semen (Pant et al.,2008; Chen et al., 2013). Furthermore, some reports from human

†Equal contribution authorship.

© The Author(s) 2019. Published by Oxford University Press on behalf of the European Society of Human Reproduction and Embryology. All rights reserved.For Permissions, please email: [email protected]

Dow

nloaded from https://academ

ic.oup.com/m

olehr/advance-article-abstract/doi/10.1093/molehr/gaz003/5304489 by guest on 26 February 2019

http://orcid.org/0000-0002-1953-952X

foetal testis and experimental murine models of single exposure tothese compounds demonstrate that they can act as trigger factors ofdeath in testicular cells (de Jager et al., 1999; Yoshida et al., 2001;Boekelheide et al., 2009; Lambrot et al., 2009). However, the molecu-lar mechanisms involved have not been fully disclosed.

Changes in the global proteomic profile have been detected in eja-culated spermatozoa of idiopathic infertile patients, and are mainlyassociated with the folding and degradation of proteins, cytoskeletonand energy metabolism (Jodar et al., 2017; Bracke et al., 2018; Caoet al., 2018). In relation to the changes in spermatogenesis, recently aquantitative proteomic analyses in testis of infertile patients revealedthat over 520 proteins are dysregulated (Alikhani et al., 2017). Overallthis data lead to the conclusion that a complex dysregulation of proteinexpression occurs in testis of patients with idiopathic infertility.However, the causes that originate it remain unsolved.

Inside the regulatory proteins expression mechanisms, the regula-tion of mRNA expression mediated by microRNAs (miRNAs) andtheirs sequence variants or isomiRs have taken relevance. ThemiRNAs/isomiRs are small non-coding endogenous RNA, which areevolutionarily well-conserved between species and implicated in thenegative post-transcriptional regulatory systems in a sequence-specificmanner (Guo et al., 2010; Neilsen et al., 2012). In fact, several dysre-gulated miRNAs in testes of infertile patients have been reported bymultiple studies (Abu-Halima et al., 2014; Munoz et al., 2015; Noveskiet al., 2016), allowing the hypothesis that ‘an aberrant expression ofmiRNAs could promote certain alterations in spermatogenesis and…be a cause of infertility in males’ (Smorag et al., 2012). Additionally,integrative reports of differentially expressed genes and miRNAs intestis of infertile patients have also emerged, which help in exploringmiRNA–mRNA interactions and uncovering the molecular regulatorynetwork and therapeutic targets in male infertility (Zhuang et al., 2015;Li et al., 2016).

Our previous reports showed for first time that exposure to EDCs(phthalates and alkylphenols) mixture alters specific miRNAs/isomiRsinvolved in the control of hormonal status in mice testis (Buñay et al.,2017). However, the consequent effects on protein networks havenot been fully addressed. Therefore, the aim of this work was to evalu-ate a functional association between dysregulated proteins andmiRNAs in mouse testis with impaired fertility due to the chronicexposure to low-doses of mixtures of phthalates and alkylphenols.The results showed that some toxicological proteins/miRNAs path-ways involved in cell death were altered in mouse testis exposed to anEDCs mixture and have potential relationship with idiopathic humanmale infertility.

Materials andMethods

Animals and ethical statementAll procedures relating to the care and handling of animals were carriedout in accordance with the regulations of the Consejo Superior deInvestigaciones Cientificas (CSIC) and the Pontifical Catholic University ofChile (PUC), following the European Commission (EC) guidelines (direct-ive 86/609/EEC), and the guides for the Care and Use of AgriculturalAnimals in Agricultural Research and Teaching by the National ResearchCouncil of Chile, respectively. The General Direction of Environment ofCAM in Spain (Ref. PROEX 054/15) and the National Fund of Science andTechnology (FONDECYT) (No. 1150 532) in Chile reviewed and

approved all the experimental protocols in this work. C57BL/6J micewere bred at the CSIC or PUC animal facilities under specific, pathogen-free (SPF), temperature- and humidity-controlled conditions in 12-h light/dark cycles with ad libitum access to food and water.

EDCs mixture exposureTo emulate chronic human exposure to an environmental EDCs mixture,we designed a murine model of chronic exposure to a defined mixture of0.3 mg/kg-bw/day of each of the following phthalates: bis(2-ethylhexyl)phthalate (DEHP), dibutyl phthalate (DBP), benzyl butyl phthalate (BBP)(Sigma-Aldrich, USA) and 0.05 mg/kg-bw/day of each of the followingalkylphenols: 4-nonylphenol (NP), 4-tert-octylphenol (OP) (Sigma-Aldrich,USA). The total concentration of the EDCs mixture was 1 mg/kg-bw/day.Phthalates were diluted in DMSO (dimethyl sulfoxide) (Sigma-Aldrich,USA) and alkylphenols were diluted in ethanol. For the control group, weused a mixture of DMSO and ethanol (vehicles) with equivalent intake of0.25 and 0.06 g/kg-bw/day, respectively. In this toxicological approach(Buñay et al., 2017, 2018; Patiño-García et al., 2018), the dose of eachEDCs was chosen based on the non-occupational exposure in human,which is estimated to range between 0.3 and 143 mg/kg-bw/day (Jonsson,2006; National Toxicology Program, 2003a, 2003b, 2006; Ademollo et al.,2008; Hines et al., 2011). We ensured that the chosen doses were at least~1000-fold lower than the LOAEL values for reproductive traits effects inexperimental male animals while still being environmentally relevant doses(Chapin et al., 1999; Nagao et al., 2001; Rider et al., 2010).

The EDCs mixture or control (vehicles) were dissolved in the drinkingwater of C57BL/6 J mice in independent bottles covered with foil. Thefinal dose was calculated according to the volume of water ingested by themice and the (bw) recorded in a pilot study and in agreement with data inthe literature referring to these parameters. Water intake was controlledeach day. The global water intake was not affected by the exposure toEDCs or control. Then, the EDCs mixture or control was administeredwith ad libitum access, to pregnant mice randomly selected at post-coitalDay 0.5 (conception) and throughout gestation and lactation. At weaning,only male offspring were selected and maintained at a maximum of fourindividuals per cage. The administration of the EDCs mixture or controlwas continued in these mice until adulthood (endpoint: post-natal Day 60and post-natal Day 75 for fertility assays) (Buñay et al., 2017, 2018). Wechoose this model in order to mimic the human where the problems ofidiopathic fertility appear from the beginning of their reproductive life.Thus, taking account that the fecundity of male mice declines with age, weconsider that the choice of Day 75 of age for male mice allows us to makesure that the male mice have good reproductive performance.

All male litter of one pregnant mouse exposed was considered as a stat-istical unit (n) and we used a minimum of three distinct male litters pergroup. Within a single litter, the male offspring exposed to EDCs mixtureand/or control were considered biological replicates and used in the differ-ent test.

Protein extractionAt the specified endpoints, animals were sacrificed, and testes wereremoved, decapsulated and mechanically homogenized in radioimmuno-precipitation assay buffer (RIPA), along with a protease inhibitor cocktailwith 2 mM AEBSF (4-(2-Aminoethyl) benzenesulfonyl fluoride hydrochlor-ide), 0.3 μM aprotinin, 130 μM bestatin hydrochloride, 14 μM E-64, 1 mMEDTA and 1 μM leupeptin hemisulfate (Sigma-Aldrich, USA). Proteinswere purified by centrifugation at 12 000 g at 4°C for 10 min and subse-quently quantified by the method described by Bradford (1976).

For 2D protein electrophoresis, 150 μg of total protein was isolatedfrom testis of adult mice. Sample pooling of three biological replicates fromthe control groups and the exposed groups were used, each one

2 Buñay et al.

Dow


ic.oup.com/m


contributing equally with 50 μg of total protein. Protein purification isdescribed above. These protein extracts were precipitated using a metha-nol/chloroform protocol (Wessel and Flügge, 1984). Briefly, cold metha-nol and chloroform were added to the sample tubes previously incubatedat 4°C and then centrifuged at 13 000 g at 4°C for 15 min. Protein-interfaces were washed and centrifuged twice with cold methanol whereasprotein pellets were dried and resuspended in 2× buffer (7 M urea, 2 Mthiourea, 4% [w/v] CHAPS and 0.0003% [w/v] bromophenol blue).Protein concentration quantification was made with the RC-DC ProteinAssay (Bio-Rad, USA).

2D electrophoresis, image acquisition andidentification of the peptide footprintFor 2D electrophoresis, aliquots of 150 μg of protein from both controland exposed groups were diluted in a total volume of 350 μl with 2× buffercontaining 18.2 M DTT and 0.5% of IPG buffer solution of ampholytes (pH3–10) (Bio-Rad, USA) as final concentrations. The first dimension was runon IPG strips (pH 3–10 NL, 17 cm) (Bio-Rad, USA) in a Protean IEF Cellsystem (Bio-Rad, USA) at 20°C. Active rehydration was performed at 50 Vfor at least 16 h. The IEF steps program was: 300 V for 45 min, linear ramp;3500 V for 22 h 45 min, rapid ramp; 5000 V for 30 min, rapid ramp, and100 V, rapid ramp, until total voltage time reached 80 000 V h; maximumcurrent limits: 99 μA/strip. Each gel strip was equilibrated in 5 ml of equili-bration buffer (50 mM Tris:HCl pH 8.8, 2% [w/v] SDS; 6 M urea; 30% [v/v] glycerol) containing 52 mM DTT for 15 min and then in 5 ml of equilibra-tion buffer containing 130 mM iodoacetamide for 15 min.

The second dimension was run in 1.5 mm-thick, 18 × 20 cm2 SDS-PAGE gels (12% acrylamide). Gels were cast according to the manufac-tures protocol (Bio-Rad, USA), prepared the day before, and kept at 4°Cbefore use. Second dimension gels were run at 5 watts/gel for 30 min andthen at 17 watts/gel until the BPB front reached the bottom edge, using acooled Protean II xi Cell (Bio-Rad, USA) and 10 μl of Precision Plus ProteinUnstained Standards solution (Bio-Rad, USA) were used as protein mar-kers. Finally, gels were stained with SYPRO Ruby protein gel stain (Bio-Rad, USA) according to the manufacturer instructions. EXQuest SpotCutter (Bio-Rad, USA) was used to image the gels at different excitation/emission times and to pick the selected spots, which were then subjectedto manual tryptic digestion. For digestion, gel pieces were washed firstwith 50 mM ammonium bicarbonate (Sigma-Aldrich, USA) and then withacetonitrile (ACN) (Scharlau, Spain). Trypsin (Promega, USA) at final con-centration of 12.5 ng/μl in 50 mM ammonium bicarbonate solution wasadded to the gel pieces for 8 h at 37°C. Finally, 100% ACN containing0.5% trifluoroacetic acid (TFA) (Sigma-Aldrich, USA) was added for pep-tide extraction. Tryptic eluted peptides were dried by speed-vacuum cen-trifugation and resuspended in 6 μl of 30% ACN- 0.1% TFA. Then 1 μl ofeach peptide mixture was deposited onto an 800 μm AnchorChip(Bruker-Daltonics, USA) and dried at room temperature, and 1 μl ofmatrix solution (3 mg/ml α-cyano-4-hydroxycinnamic acid) in 33% ACN0.1% TFA was then deposited onto the digest and allowed to dry at roomtemperature.

Mass spectrometry analysisSamples were analysed with an Autoflex III TOF/TOF mass spectrometer(Bruker-Daltonics, USA). Typically, 1000 scans for peptide mass finger-printing (PMF) and 2000 scans for MS/MS were collected. Automated ana-lysis of mass data was performed using FlexAnalysis software (Bruker-Daltonics, USA). Internal calibration of MALDI-TOF mass spectra was per-formed using two trypsin autolysis ions with m/z 842.510 and m/z2211.105; as for MALDI–MS/MS, calibrations were performed with a frag-ment ion spectrum obtained from the proton adducts of a peptide mixture

covering the m/z 700–4000 region. The typical error observed in massaccuracy for calibration was usually below 50 ppm. MALDI–MS and MS/MS data were combined through the BioTools 3.0 program (Bruker-Daltonics, USA) to interrogate the NCBI non-redundant protein databaseSwissProt 2014_03 (542 782 sequences; 193 019 802 residues) usingMASCOT software 2.3 (Matrix Science, UK). Relevant search parameterswere set as follows: enzyme, trypsin; fixed modifications, carbamidomethyl(C); oxidation (M); one missed cleavage allowed; peptide tolerance,50 ppm; MS/MS tolerance, 0.5 Da.

Peptide mass fingerprinting and fragmentation by MS–MALDI-TOF wascarried out in the Proteomics and Genomics Facility (CIB-CSIC), a mem-ber of ProteoRed-ISCIII network.

Western blottingSamples of 20 μg of protein homogenizing were retrieved from the testesof the control groups and the exposed to EDCs groups (n = 3) and sepa-rated by electrophoresis on a 12% polyacrylamide gel (sodium dodecyl sul-phate–polyacrylamide gel electrophoresis) under denaturing and reducingconditions and then transferred to a nitrocellulose membrane (ThermoScientific, USA) at 350 mA for 2 h. Next, membranes were blocked with asolution of 3% (w/v) bovine serum albumin 0.1% (v/v) Tween in Tris-buffered saline, pH 7.4 and incubated overnight with the respective pri-mary antibodies for PGAM1 (0.4 ng/μl) (Abbexa, UK) and ß-ACTIN(0.3 μg/μl) (Sigma-Aldrich, USA) as a loading control. Finally, the secondincubation took place with their respective secondary antibodies conju-gated with horseradish peroxidase (KPL, Gaithersburg, UK) diluted 1:5000in blocking solution for 1 h at room temperature. Peroxidase activity wasdetected by enhanced chemiluminescence (Pierce Biotechnology,Rockford, IL).

ImmunohistochemistryPGAM1 was detected in paraffin-embedded cross-sections of testes (n =3) fixed in Bouin’s solution for each group and treated with sodium citrate0.01 M, pH 6, in heat until it boiled and then kept for 10 min to expose theantigens. After, the samples were first treated with 3% H2O2 for 10 min.Then, to prevent unspecific binding, a standard protein block system (UltraV block, Thermo Scientific, USA) was applied for 10 min. Primary anti-bodies were applied at a concentration of (4 ng/μl); samples were incu-bated overnight at 4°C in a humidified chamber after being washed threetimes for 5 min in a Tris–HCl buffer, pH 7.6, with 0.3 M NaCl and 0.1%Tween-20. Biotinylated secondary antibody, streptavidin–biotinylated–peroxidase complex, amplification reagent (biotinyl tyramide) andperoxidase-conjugated streptavidin were applied step-by-step for 10 mineach (Thermo Scientific, USA). Afterwards, slides were washed threetimes for 5 min in a Tris–HCl buffer, pH 7.6, with 0.3 M NaCl and 0.1%Tween-20. Finally, DAB (3,3-diaminobenzidine tetrahydrochloride) PlusSubstrate and DAB Plus Chromogen (Thermo Scientific, USA) wereapplied for 1 min and washed in distilled water. Samples were stained withhematoxylin and observed under a phase contrast microscope (Optiphot-2, Nikon, Tokyo, Japan) and photographed with a digital camera (CoolPix4500, Nikon, Tokyo, Japan).

Protein functional analyses, expression ofmiRNA/isomiRs and identification ofmiRNA–mRNA–protein targetsProtein ontology analyses were performed firstly at the Mouse GenomeInformatics (MGI) and Uniprot databases. Protein–protein interactionswere analysed by STRING (https://string-db.org/), then functional ana-lysis of GO domains was performed using a Cytoscape plugging ClueGO

3Proteomics/miRNAs in testis exposed to EDCs

Dow


ic.oup.com/m


https://string-db.org/

(Bindea et al., 2009) by a hypergeometric test with a P-value threshold≤0.05.

Small-RNAseq data of testis exposed to EDCs mixture or control wereobtained from NCBI Gene Expression Omnibus (GEO) from our previouswork: GSE8469 (Buñay et al., 2017). Small-RNAseq data were processed asdescribed. Briefly, MicroRNA identification and quantification were performedby aligned trimmed reads to the mouse genome (mm10) using Bowtie aligner(Langmead et al., 2009) and then using HTSeq script (Anders et al., 2015) withGFF file from miRBase v21. Further, isomiRage software were used for isomiRsdetection and quantification (Muller et al., 2014). We carried out data normal-ization and differential expression analysis using the DeSeq tool of the R/Bioconductor software package (Anders et al., 2013).

Correlations between miRNAs and proteins were found using themiRWalk database, combining the searches of three or more databases,along with the validated mRNA targets reporter in miRWalk and a valid-ated miRNA:mRNA database, DIANA-TARbase v7.0 (Vlachos et al.,2015). The interaction map was performed using a Cytoscape.

Fertility assayGestational rate, fertility index and potential fertility were calculatedaccording to Garcia et al. (2012). Representative animals of each ‘n’ pergroup were selected randomly and treated (control and EDCs mixture)until post-natal Day 75 to ensure their potential fertility. After this period,males were placed in individual cages and cohabited with two unexposedadult females for 8 days (to ensure at least two complete female estrouscycles). Vaginal plugs were checked every day early in the morning (08:-00–09:00 h). All females with vaginal plug were set apart and placed in indi-vidual cages. An effective pregnancy was confirmed by the increase of thebody weight.

Between the Days 18 and 20 post-detection of the vaginal plug, a groupof females were sacrificed to quantify the number of implantations andreabsorptions in each uterine horn and the foetal mortality rate (numberof reabsorptions/number of implantations × 100). Then, ovaries wereextracted, fixed in Bouin solution and processed for histological analysis.Serial sections of each ovary were stained with hematoxylin–eosin, and thenumber of corpora lutea per ovary was quantified. Another group of preg-nant females was maintained until the delivery. Between the post-natalDay 1 or 3, the number of newborns was quantified.

Statistical analysisFor the selection and quantification of spots of interest in 2D gels,PDQuest™ V8.0 (Bio-Rad) was used. First, a reference gel was createdfrom images of the analysed gels. Then, each one of these gels were alignedwith the reference gel. Spot quantification in relation to the protein levelswas calculated considering the intensities of all pixels within the definedboundary previously normalized. Finally, the quantitative data of each spotwere normalized using a non-parametric linear regression analysis (LOESS)to compare each expression level. This approach combined with MS,enabled us to identify the PPIA protein levels with no significant changesbetween the two groups (control and exposed to EDCs). As others stud-ies suggest, PPIA is suitable to be used as an internal control (Kim et al.,2014). Consequently, PPIA was used as an endogenous reference protein(housekeeping) in order to perform a second normalization/comparisonof the proteins identified by MS.

In mass spectrometry (MS) analysis, mascot total scores >75 were con-sidered significant (P < 0.05).

Reproductive outcome in male mice exposed to EDCs mixture was ana-lysed by Chi-square test, umpaired t-test and one-way ANOVA along withDunnett’s post hoc test, using GraphPad Prism version 5.0. (GraphPadSoftware, USA).

Results

Chronic exposure to a EDCs mixture andfertility in male miceIn previous works, we showed that exposure to a single EDC or mix-ture of EDCs induces an increase in the body weight, a decrease in tes-tis relative weight, testis lesions and hormonal status changes (Buñayet al., 2017, 2018). In the present work, while assessing the fertility ofmales exposed to a specific EDCs mixture, we found that there wereno significant differences in several reproductive parameters such asgestational rate, fertility index and potential fertility, when comparedto control groups (Supplementary Table SI). Interestingly, in theexposed groups, not all males seem to be equally affected by the treat-ment. We identified four males from different litters which, once ana-lysed by ANOVA, presented a decrease in potential fertility.Furthermore, when mating resulted in pregnancy, the number of reab-sorptions and the rates of pre-implantation loss and foetal mortalitywere increased (Fig. 1 and Supplementary Table SI).

Changes in the proteome profile of mousetestis exposed to EDCs mixtureFirst, to detect changes of testicular proteins levels in mice exposed toEDCs, we performed a global approach and evaluated the mouse testisproteome using 2D electrophoresis comparing exposed mice with con-trols. We analysed a total of 246 spots/proteins matched by PDQuestsoftware and normalized by LOESS. Based on the differential spot inten-sities observed, only those proteins that change more than 1.2-fold inexposed versus control groups were taken into account. This approachallowed us to select 21 spots of unequal signals, plus one spot (8102) ran-domly selected among those of equal relative abundance in all gels(Supplementary Fig. S1). These 22 spots/proteins were excised from thegels for further identification of peptide mass fingerprint by MS–MALDI-TOF. Three spots (904, 1302 and 9104) could not be identified mainly dueto low signals. But this experimental approach allowed us to correctly iden-tify 19 spots/proteins. Of these identified proteins, 10 were classified asup-regulated and 8 as down-regulated in testis exposed to EDCs mixture(Fig. 2A and B, Table I); the spot/protein named as 8102 was identified aspeptidyl-prolyl cis–trans isomerase A (PPIA) and no significant changeswere detected (Fig. 2C, Table I). It is worthy to note that in our previouswork, with the same exposure model, we did not detect changes in thelevels of PpiamRNA (Buñay et al., 2017), and therefore used this transcriptas an endogenous reference gene (Radonić et al., 2004). A second normal-ization using PPIA found that the up-regulated/down-regulated proteinspreviously identified in the EDCs group had a minimum of 1.2-fold change(Supplementary Figure 2). Then we decided to validate our 2D electro-phoresis/MS results by using western blot and immunohistochemistry. Tothis end we chose PGAM1 as it was identified among those proteins up-regulated by 2D electrophoresis/MS. Through Western blot, we found asignificant increase in the protein levels of PGAM1 in testes of miceexposed to EDCs mixture. This result was similar to the one observed with2D electrophoresis/MS. In addition, by immunohistochemistry, PGAM1localized preferably in Sertoli and interstitial cells both in control and EDCstreated mouse testes. No obvious change in localization was detectedbetween these populations (Fig. 3B).

4 Buñay et al.

Dow


ic.oup.com/m


Figure 1 Populations of mice exposed to EDCs mixture have compromised fertility. White diamonds and gray circles indicate malepopulation of control mice (n = 9) and mice exposed to EDCs mixture, (n = 12), respectively. Gray circles inside white background indicate popula-tions of mice with a normal potential fertility and without induced pre-implantation losses. Green and red backgrounds indicate the populations of ani-mals in which the parameters were detected as decreased or increased, respectively.

Figure 2 Proteins that changes its expression in mouse testis exposed to EDCs. (A) Up-regulated proteins, (B) down-regulated proteinsand (C) protein showing uniform expression in mouse testis exposed to EDCs compared with control. Each chart represents the relative levels byregression lineal test (LOESS) for PDQuest, significant changes by 1.2-fold. Images are representative picture of 2D electrophoresis gels in eachcondition.


Dow


ic.oup.com/m


.............................................................................................................................................................................................

Table I Changes in the proteome of mice testis exposed to EDCs mixture.

SpotNo.a

Ac.codeb

Symbol Protein name MM (Da)/pItheorc,d

Mat’dbpepse

Totalscoref

Sequencecoverage(%)g

Molecularfunctionh

Biological processh

A. Proteins up-regulated by the exposure to EDCs mixture and involved in germ cell death

203Q9JIQ3 DIABLO Diablo homolog,

mitochondrial26 975/6.3 7 305 32 Activating

caspases.Inhibitory activity(IAP)

Intrinsic apoptotic signallingpathway

5101P70349 HINT1 Histidine triad

nucleotide-binding protein 1

13 882/6.36 5 157 58 Hydrolase activity Intrinsic apoptotic signallingpathway by p53 class mediator,positive regulation of calcium-mediated signalling

4301P63073 EIF4E Eukaryotic

translationinitiation factor 4E

25 266/5.79 4 236 18 Eukaryoticinitiation factor4 G binding,

Regulation of translation, positiveregulation of mitotic cell cycle,behavioural fear response

3202P99026 PSMB4 Proteasome

subunit beta type-4

29 211/5.47 7 162 37 Threonine-typeendopeptidaseactivity

Proteasome-mediated ubiquitin-dependent protein catabolicprocess

3702Q9D1A2 CNDP2 Cytosolic non-

specificdipeptidase

53 190/5.43 6 131 12 Dipeptidaseactivity

Proteolysis

7101Q01768 NME2 Nucleoside

diphosphatekinase B

17 466/6.97 9 505 51 Nucleosidediphosphatekinase activity,fatty acid binding

Mitophagy in response tomitochondrial depolarization,cellular response to fatty acid,glucose stimulus and oxidativestress

6401Q9DBJ1 PGAM1 Phosphoglycerate

mutase 128 928/6.67 7 365 40 Phosphoglycerate

mutase activity,protein kinasebinding

Glycolytic process

5605Q9JHI5 IVD Isovaleryl-CoA

dehydrogenase,mitochondrial

46 695/8.53 5 200 12 Isovaleryl-CoAdehydrogenaseactivity, fatty-acyl-CoA binding

Lipid homoeostasis

B. Proteins down-regulated by the exposure to EDCs mixture and involved in germ cell death

0 Q8CHP8 PGP phosphoglycolatephosphatase

34 975/5.21 6 170 42 Magnesium ionbinding,phosphoglycolatephosphataseactivity,

Carbohydrate metabolic process,dephosphorylation, peptidyl-tyrosine dephosphorylation

2901P63017 HSPA8 Heat shock

cognate 71 kDaprotein

71 055/5.37 8 362 18 ATPase activity,poly(A) RNAbinding, unfoldedprotein binding

ATP metabolic process, chaperone-mediated protein folding requiringcofactor

3804P30416 FKBP4 Peptidyl-prolyl

cis–transisomerase FKBP4

51 939/5.54 14 367 38 ATP binding,glucocorticoidreceptor binding,poly(A) RNAbinding

Androgen receptor signallingpathway, embryo implantation,male sex differentiation,reproductive structuredevelopment

3807P27773 PDIA3 Protein

disulphide-isomerase A3

57 099/5.88 14 400 30 Poly(A) RNAbinding, proteindisulphideisomerase activity

Cell redox homoeostasis, positiveregulation of extrinsic apoptoticsignalling pathway, protein folding

6302P15626 GSTM2 Glutathione S-

transferase Mu 225 871/6.9 12 455 65 Glutathione

transferaseactivity

Glutathione metabolic process,xenobiotic catabolic process

Q80W21 GSTM7 25 864/6.34 4 71 20

Continued

6 Buñay et al.

Dow


ic.oup.com/m


A bioinformatic analysis of dysregulated proteins detected by 2Delectrophoresis/MS was performed, disclosing some protein–proteininteraction networks (Supplementary Fig. S3). Gene ontology annota-tion of differentially expressed proteins showed an enrichment in theprocess of cellular response to stress, in metabolic processes and inprotein folding pathways (Fig. 4A). Furthermore, we found that eightup-regulated proteins (DIABLO, HINT1, EIF4E, PSMB4, CNDP2,NME2, PGAM1 and IVD) and all down-regulated proteins (VDAC2,PGP, HSPA8, FKBP4, PDIA3, GSTM2, GSTM7 and CCT2), whichrepresent 89% of all dysregulated proteins (16 out of 18) wereinvolved in different steps of cell death (Fig. 4B). As for the other twoup-regulated proteins, MP1 is involved in metabolic processes andRTRAF in the regulation of transcription as an RNA binding protein(Table I). Therefore, cell death seems to be the major dysregulatedpathway at the protein level.

Association between proteins and miRNAsdifferentially expressed in mouse testisexposed to EDCs mixtureIntegrative in-silico analyses revealed that 23 miRNAs/isomiRs (differen-tially down-regulated in our previous studies; Buñay et al., 2017) may beassociated with six up-regulated proteins (33.3% of total dysregulatedproteins), considering the post-transcriptional regulatory roles of identi-fied miRNAs/isomiRs over the corresponding target transcripts encod-ing identified proteins (Fig. 5A). The analyses did not show correlationbetween down-regulated proteins and up-regulated miRNAs. An inte-grative map showed that transcripts for DIABLO, PGAM1, RTRAF,EIF4E, IVD and CNDP2 were targets of more than one down-regulatedmiRNA/isomiRs, which suggested higher potential efficiency in theirexpected post-transcriptional regulation (Fig. 5B and Supplementary

.............................................................................................................................................................................................

Table I Continued

SpotNo.a

Ac.codeb

Symbol Protein name MM (Da)/pItheorc,d

Mat’dbpepse

Totalscoref

Sequencecoverage(%)g

Molecularfunctionh

Biological processh

5308Glutathione S-transferase Mu 7

Glutathionetransferaseactivity, receptorbinding

Glutathione metabolic process,regulation of release of sequesteredcalcium ion into cytosol bysarcoplasmic reticulum, xenobioticcatabolic process

5705P80314 CCT2 T-complex

protein 1 subunitbeta

57 783/5.97 12 541 28 ATP binding,ubiquitin proteinligase binding

Binding of sperm to zona pellucida,chaperone-mediated proteincomplex assembly, toxin transport

7401Q60930 VDAC2 Voltage-

dependent anion-selective channelprotein 2

32 340/7.44 6 180 31 Voltage-gatedanion channelactivity,nucleotide binding

Negative regulation of intrinsicapoptotic signalling pathway,negative regulation of proteinpolymerization

C. Other up-regulated proteins identified

3601Q924M7 MPI Mannose-6-

phosphateisomerase

47 229/5.62 7 211 23 Mannose-6-phosphateisomerase activity,

Mannose to fructose-6-phosphatemetabolic process

6303Q9CQE8 RTRAF RNA

transcription,translation andtransport factorprotein

28 249/6.4 5 215 26 Poly(A) RNAbinding, identicalprotein binding

Positive regulation of transcriptionfrom RNA polymerase II promoter,tRNA splicing, negative regulationof protein kinase activity

D. Other proteins identified

8102P17742 Ppia Peptidyl-prolyl

cis–transisomerase A

18 131/7.74 6 320 53 Poly(A) RNAbinding, peptidebinding

Positive regulation of proteinsecretion, lipid particle organization,protein folding

Data depicted correspond to the most statistically significant candidates encoded inMus musculus.aSpot numbering as shown in the 2D gels depicted in Supplementary Fig. S1.bProtein accession code fromMus musculus database.cTheoretical molecular weight (Da).dTheoretical isoelectric point (pI).eNumber of matched peptides.fMascot Total score is −10*Log(P), where, P is the probability that the observed match is a random event.gSequence coverage is the amino acids ratio (no. identified in peptides/no. in theoretical peptides from sequence data) expressed as a percentage.hOntology classification according Mouse Genome Informatics (M.G.I.) and UniprotKB.


Dow


ic.oup.com/m


Table SII). For example, the overexpression of DIABLO and EIF4E couldbe induced as consequence of down-regulation of five and eightmiRNAs/isomiRs, respectively (Fig. 5B). In addition, we have identifiedtwo isomiRs (miR-30c-5p_trim2 and miR-497a-5p_UA_3prime) whosecanonical miRNAs are reported in the DIANA-TARbase v7.0 as valid-ated for specific transcripts: miR-30c-5p: Diablo and miR-497a-5p: IvdmRNAs. As the detected isomiRs have only different 3′ ends (the corre-sponding 5′ seed regions sequence of these isomiRs are the same thatthe canonical miRNAs). Consequently, it can be expected that the tar-get should be the same for each dysregulated miRNA/isomiR.Furthermore, we showed that the 3′ UTR of Eif4e mRNA can be tar-geted by both the miR-15b-5p (canonical miRNA) and its isomiR miR-15b-5p_AA_3prime, and this interaction is conserved in humans(Fig. 5C). This finding might explain the correlation between six identi-fied proteins and the loss of miRNAs/isomiRs with targets in the corre-sponding mRNAs encoding such proteins, as was detected in testisfrom mice exposed to the EDCs mixture.

Changes in the expression of proteins andmiRNAs inmouse testis exposed to EDCsmixture are linkedwith idiopathicmale infertilitySince some animals exposed to the EDCs mixture showed a decreasefertility, we hypothesized that some of the deregulated protein and

miRNA/isomiRs could be also dysregulated in human testis of infertileman.

To assess this comparative analysis, we selected a recent study oflabel-free quantitative shotgun proteomics on testicular tissue frompatients with non-obstructive azoospermia, including maturationarrest (MA) and Sertoli cell only syndrome (SCOS) (Alikhani et al.,2017). The results revealed eight common dysregulated proteins(Fig. 6A) in testes from azoospermic men and those of mice exposedto the EDCs mixture. These eight common proteins represented 44%of proteins dysregulated by EDCs in mice, all of them involved in celldeath (EIF4E, PGAM1, DIABLO, PSMB4, PDIA3, CCT, HSPA andGST).

Then, we selected from the literature three similar reports of signifi-cant changes in the expression of miRNAs (at least 2-fold changes) intesticular biopsies of patients with MA and SCOS syndromes (Abu-Halima et al., 2014; Munoz et al., 2015; Noveski et al., 2016). Thesedata were contrasted with the list of miRNAs that were differentiallyexpressed by the exposure to EDCs mixture, as previously detected(Buñay et al., 2017). We found two miRNAs (miRNA-15b and miRNA-34b) that were reported dysregulated in every work on human testiswith spermatogenic failure and in our previous work in mice exposedto EDCs. Moreover, three miRNAs were coincidently found only inpatients with MA and SCOS reported by Abu-Halima et al., 2014

Figure 3 Immunolocalization and up-regulation of PGAM1 in mice testis by the exposure to EDCs mixture. (A) PGAM1 proteinlevels in mice testis exposed to EDCs mixture and/or control were determined by Western blot and normalized with β-ACTIN. All graphics representthe mean ± SEM, n = 4. Unpaired t test, *P < 0.05. Abbreviation: AU, arbitrary units. (B) Representative pictures of PGAM1 in mouse testes exposedto EDCs mixture and controls, bar = 50 μm.

8 Buñay et al.

Dow


ic.oup.com/m


(Abu-Halima et al., 2014) and mice exposed to EDCs mixture (miR-382, miR-18a and miR-378) (Fig. 6B). These observations suggestedthat proteins and miRNAs differently expressed by the exposure toEDCs mixture might be related to spermatogenic failure and asso-ciated with the risk of exposure to EDCs of human populations.

DiscussionThis work shows a global integrative profile of dysregulated proteins/miRNAs in testes of mice exposed to a mixture of phthalates and alkyl-phenols, suggesting that chronic exposure to an EDCs mixture duringtheir development might be a plausible cause for idiopathic maleinfertility.

Male mice are sexually mature when they are around 42–49 daysold. At this age, each spermatogenesis cycle takes ~35 days to becompleted (Griswold, 2016). Although protein/miRNAs expressionanalyses and fertility tests were performed in 60 and 75 days old ani-mals, respectively, we believe that this 15-day difference should notaffect the correlation of results since both are in the adult fertileperiod.

As a whole, mice exposed to EDC mixtures did not present signifi-cant changes in global fertility. According to this preliminary study, wecan not be certain that subfertility is more likely present in the exposedgroup than in the control group. However, the potential fertility basedon the increase of the number of pre-implantation losses and reab-sorptions was significantly compromised within a sub-group of

exposed animals which could be due to: (i) differences in the relativeindividual susceptibility at the exposure level, from early periods ofgonadal development, (ii) maternal physiology, (iii) differences in theamount of milk intake during lactation and (iv) differences in metabolicrate, accumulation, elimination and/or global detoxification rates ofEDCs between animals. These observations might mirror the discrep-ancy in epidemiological data reports in human adult males, pregnantwomen and children and the association of adverse effects due to thesingle and mixed exposure to EDCs (Hou et al., 2015; Birks et al.,2016; Sifakis et al., 2017). Therefore, future fertility trials should beperformed in greater number of mice covering longer exposure timesand different aging to validate the tendency shown in this work.Furthermore, we indicate that new specific epidemiological studies arenecessary in the human population with different grades of exposureto EDCs in order to evaluate the impact on male reproductive health.

By 2D/MS proteomic analysis, we detected deregulation of 18 pro-teins in the testes of mice exposed to the EDCs mixture. In this prote-omic approach, we used a sample pooling, which is a validated andcommonly used option (Weinkauf et al., 2006; Diz et al., 2009; Karpand Lilley, 2009). However, it is important to state that this proteomicstudy should be seen as a biological averaging (Karp and Lilley, 2009)and that the expression of several proteins observed as dysregulatedmust be validated through different approaches in future studies.

Our previous reports and other studies have indicated that expo-sures to EDCs mixtures induce apoptosis of germ cells which is alsowell correlated with the decrease in fertility reported in this work

Figure 4 Cell death is the main molecular function related with dysregulated proteins in testis of mice exposed to EDCsmixture.(A) GO enrichment analysis of dysregulated proteins was performed using GlueGO. Enrichment was obtained using hypergeometric distribution test,with a P-value threshold >0.05. Each colour represents a different metabolic pathway. (B) Rate of identified proteins that are involved in cell death.Molecular function was obtained according to the information available at the MGI and UniProt database.


Dow


ic.oup.com/m


Figure 5 Integrative analyses of dysregulated proteins andmiRNAs in mouse testis exposed to EDCsmixture. (A) Hierarchical clus-ter analysis of differentially expressed miRNAs/isomiRs involved in the post-transcriptional control of dysregulated proteins. Expression levels corres-pond to log 2 normalized read counts using DeSeq tool of the R/Bioconductor software package, in testes of mice exposed to EDCs mixturecomparing with control mouse testes. (B) Integrative map of interactions between down-regulated miRNAs/isomiRs and overexpressed proteins inmouse testis exposed to EDCs mixture; red colour indicates overexpression and gray to green scale under-expression. (C) Prediction of Eif4E as targetof miR-15b-5p/miR-15b-5p_AA_3 prime in human and mouse. IsomiRs nomenclature according to Muller et al. (2014).

Figure 6 Common proteins and miRNAs differentially expressed by the exposure of EDCs on mouse testis and in idiopathicinfertile humanmales. (A) Comparative analyses of dysregulated proteins in humans with spermatogenic failure (Alikhani et al., 2017) and in mousetestis exposed to EDCs. (B) Comparative analyses of dysregulated microRNAs in humans with spermatogenic failure (Abu-Halima et al., 2014; Munozet al., 2015; Noveski et al., 2016) and in mouse testis exposed to EDCs (Buñay et al., 2017). The name in parenthesis refers to the reference of thestudy.

10 Buñay et al.

Dow


ic.oup.com/m


(Manikkam et al., 2013; Buñay et al., 2017, 2018). We found, throughdevelopmental analysis, that some forms of GST, e.g. GSTM2, aredown-regulated in normal adults compared to prepuberal mice (Pazet al., 2006), which was in agreement with the higher expression ofthis GST isoform in Sertoli and spermatogonia cells than in spermato-cytes and spermatids (Yu et al., 2003). The decrease in the expressionof two glutathione S-transferases (GSTM2 and GSTM7) detected inour 2D proteomic analysis, suggests that the EDCs mixture is affectingthe testis during development since the early differentiation stages. Infact, the depletion of antioxidant enzymes is a known mechanism ofoxidative stress and cell death. Thus, the decrease in these GSTs levelsin testis after exposure to the EDCs mixture would be in opposition toits cytoprotective role and directly involved in mechanism of oxidativestress to induce sperm DNA damage and male infertility.

New in-vitro evidence show that endoplasmic reticulum (ER) stressis an early driver of the EDC-mediated perturbations (Rajamani et al.,2017). In this way, our in-vivo approach in mouse testis exposed tothe EDCs mixture identified a group of dysregulated proteinsinvolved in ER stress. Firstly, the overexpression of the proteasomesubunit beta type-4 (PSMB4) responsible of proteasome assemblyand protein degradation (Hirano et al., 2008) was correlated with thedecrease in at least three proteins expression. One of them, PDIA3is a chaperone located at the ER that modulates the thiol-disulphidestatus of proteins and is considered a survival factor. In-vitro assaysfound that some EDCs affect the PDIA activity (Klett et al., 2010),and further evidence shows that PDIA3 has the capacity of bindingwith 17β-estradiol (Primm and Gilbert, 2001; Wong et al., 2017), ahormone that was previously reported as decreased in mouse testisexposed to EDCs mixture (Buñay et al., 2017). Another down-regulated protein was the cytosolic chaperone CCT2, a protein par-ticularly involved in the folding of cytoskeletal and cell cycle proteins(actin, tubulin and cyclin-E), which is related with PSMB4 (up-regu-lated). Consequently, altered folding of these proteins may stop thecell cycle, cause an ER stress response, disrupt the mitochondria andinduce apoptosis in somatic or germ cells. In addition, CCT2increases its expression from prepuberal to adult mice, suggesting anassociation with late spermatogenesis (Paz et al., 2006). Finally, thedown-regulation of heat-shock cognate protein HSPA8 was also pre-sent. HSPA8 participates in the response to ER stress with the pivotalrole of correcting the folding of nascent polypeptides and then coop-erates in the re-folding of misfolded proteins. At the same timePSMB4, PDIA3, CCT and HSP are found dysregulated in testicularbiopsies from patients with SCOS and MA (Alikhani et al., 2017).Therefore, the interplay of these proteins of ER stress could be impli-cated in germ-cell death induced by EDCs.

Many efforts have been made to identify miRNAs involved in sper-matogenic failure (Abu-Halima et al., 2014; Munoz et al., 2015;Noveski et al., 2016). Currently, new approaches with small non-coding RNA sequencing (sncRNA-Seq) suggest that changes ofmiRNA variants or isomiRs (more than canonical miRNAs) could bet-ter describe a pathological state (Guo et al., 2011; Kozlowska et al.,2013; Telonis et al., 2015). In this sense, at the reproductive level, thechanges in isomiRs expression due to the EDCs exposure could have amore relevant role than canonical miRNAs. Thus, we suggest thatfuture studies focused on changes in the miRNome of infertile subjects(male and female) should include analyses of isomiRs.

DIABLO (SMAC) is a relevant pro-apoptotic protein with ubiqui-tous expression in testicular cells that showed, by 2D electrophoresis/MS, an increased expression in the testes of mice exposed to EDCsmixture. Interestingly, preliminary reports by immunohistochemistryhave not showed differences in the expression of DIABLO in azoo-spermic patients compared to subjects with normal testicular histology(Bozec et al., 2008). However, Jaiswal et al. (2015) demonstrated thatprotein expression of DIABLO increased in infertile human testes withhypospermatogenesis, MA and SCOS. And recently, the study of glo-bal proteome profile in human testis affected by MA or SCOS havecorroborated that DIABLO expression is dysregulated (Alikhani et al.,2017). Thus, the increase in DIABLO is directly related with the dis-order of regulation of apoptosis in its pathogenesis (Takagi et al.,2001). In addition, it has been detected that a lack of gonadotropinsand androgens causes the increase of DIABLO expression and itstranslocation to the cytoplasm in spermatocytes (Vera et al., 2006).Moreover, exposure to phthalates metabolites with antiandrogeniceffect such as MEHP induces (via decrease of PI3K/AKT and increaseof NF-Kβ) germ cell apoptosis by increasing the levels of DIABLO(Rogers et al., 2008). Here, we suggest that the increase of DIABLO islinked to the exposure to EDCs mixture in a new mechanism thatinvolves the down-regulation of miRNA/isomiRs.

Another up-regulated protein was EIF4E, whose overexpression isassociated with oncogenesis and metastatic progression in humans(Hsieh and Ruggero, 2010; Siddiqui and Sonenberg, 2015).Furthermore, some studies have reported that somatic cell viability,chromosome condensation, cytokinesis during the meiotic divisionand normal production of functional sperm in spermatogenesisrequires EIF4E activity (Amiri et al., 2001; Ghosh and Lasko, 2015). Inaddition, in a previous study, we found that changes in the expressionof EIF4F complex including Eif4E depend on the type of EDCs and thetime of exposure (López-Casas et al., 2012). The increase of the levelof EIF4E could be a consequence of the loss of post-transcriptionalcontrol due to a decrease in the level of one canonical miRNA (miR-15b-5p) and seven doubly adenylated 3′ end isomiRs, targeting its 3′-UTR. This miRNA regulation is also preserved in the human EIF4E 3′UTR. At the same time, both EIF4E and miR-15b are recurrently dysre-gulated in testicular biopsies of patients with spermatogenic failure,which suggests a novel mechanism whereby the exposure to EDCsmight decreases fertility and should be studied in depth in infertilemale patients.

Interestingly, the mitochondrial protein Isovaleryl-CoA dehydrogen-ase (IVD) was found overexpressed and it increases in response tolipid overload, which is correlated with the obesogenic effect and cho-lesterol increase reported by the exposure to EDCs (Heindel et al.,2015). Our in-silico data show that the Ivd transcript is targeted by onecanonical miRNA (miR-15b-5p) and the isomiRs miR-15b-5p_AA_3prime along to three more down-regulated isomiRs.Therefore, an overexpression of IVD due to the loss of the post-transcriptional control by miRNAs/isomiRs might promote high β-oxi-dation of free fatty acids and catabolism of proteins that are related tostress in the ER and increased mitochondrial ROS (Tumova et al.,2016), thus generating mitochondrial dysfunctions commonly detectedin spermatozoa of infertile patients.

Similarly, an important enzyme in the glycolytic pathway, PGAM1,was found to be up-regulated in mouse testis exposed to EDCs. In


Dow


ic.oup.com/m


addition, by immunohistochemistry we found that PGAM1 was prefer-entially expressed in Sertoli and interstitial cells. However, in relationto the protein levels, IHQ analysis could not validate the data obtainedby WB and 2D/MS, possibly due to differences in the treatments ofthe samples related to the fixation method or the epitope retrieval(Uhlen et al., 2016). The increase in PGAM1 suggests that the expos-ure to EDCs modified the glycolytic profile in somatic cells, particularlyin Sertoli cells to subsequently alter the energetic metabolic role sup-plying germ cells and induce germ cell death. In testicular biopsies frominfertile men dysregulated levels of PGAM1 were found (Alikhani et al.,2017), specifically down-regulated in severe hypospermatogenesis andoverexpressed in SCOS, which is associated with an inhibition of pro-liferation and apoptosis (Zhang et al., 2015). Here, we presented thefirst evidence of post-transcriptional control of PGAM1 mediated byisomiRs, suggesting that these interactions could also be used as bio-markers in male infertility.

Another up-regulated protein was the cytosolic nonspecific dipepti-dase 2 (CNDP2). However, its function and implication in human fer-tility is not clear. Apparently, CNDP2 has more than one enzymaticactivity and when increased, it activates p38 and JNK/MAPK pathwaysto induce cell apoptosis (Zhang et al., 2014). Moreover, recentresearch on the enzymatic activity of CNDP2 in the cellular metabo-lome has indicated that this protein can degrade lactate to its metabo-lites N-lactoil-amino acids (N-lac-Phe) (Jansen et al., 2015). In previouswork, we indicated that lactate is a survival factor of spermatocytes(Bustamante-Marín et al., 2012). Although the function of lactate meta-bolites is unknown, it suggests that the degradation of lactate via anincrease in the expression and enzymatic activity of CNDP2 could be anew mechanism that modulates germ cell apoptosis. Our analyses indi-cated that the increase in CNDP2 expression would be associatedwith the loss of the post-transcriptional control mediated by miR-23a-3p_AA_3prime, miR-320-3p_AAA_3prime, and the joint action of miR-3085-3p/miR-3085-3p_UA_3prime. This supports our hypothesis thatisomiRs could participate along with the corresponding canonicalmiRNAs in the regulation of mRNA targets during spermatogenesis,which can be altered by the exposure to EDCs, triggering changes inmetabolic and cell stress pathways.

Finally, it is important to remark the possible central role of thedown-regulation of miRNA-15b-5p and their isomiR (miR-15b-5p_AA_3prime) in male infertility, which suggests that a decrease ofmiRNA-15b-5p and/or their isomiR could be a novel biomarker to beused to diagnose patients with subfertility potentially induced by theexposure to EDCs. However, this remains to be epidemiologically andexperimentally tested.

A group of dysregulated proteins that correlate with somatic and/or germ cell death includes the down-regulated VDAC2, PGP,GSTM2, GSTM7 and the up-regulation of HINT1. We did not find anyassociated miRNAs that could explain the dysregulation of these pro-teins, except HINT1 that is a bone fide target of hsa-miR-15b-5p(DIANA-TarBase v7.0) in human cells lines. Therefore, there must beadditional mechanisms and regulatory molecules that work coopera-tively and are mediating the adverse effects of EDCs on male fertility.This matter should be addressed in future studies.

In conclusion, this work is the first report of integrative data fromproteomics and miRNAs/isomiRs, which aims to clarify the molecularmechanisms that are affecting male fertility due to the exposure tomixture of phthalates and alkylphenols. Also, it is tempting to envision

the possibility of establishing miRNAs pathways as a specific and effect-ive pharmacological therapy for certain types of primary testicularfailure.

Supplementary dataSupplementary data are available at Molecular Human Reproductiononline.

AcknowledgementsWe would like to thank MSc. Leonor Cruz Fernandes for the Englishediting.

Author’s rolesJ.B., R.D.M. and J.d.M. designed the research; J.B., D.P-G. and P.U-M.performed the experiments; J.B. and E.L. contributed new analytictools; J.B., E.L., R.D.M. and J.d.M. analysed data; and J.B., E.L., R.D.M.and J.d.M. wrote the article.

FundingGrants from FONDECYT 1150352 to R.D.M., FONDECYT 3160273to P.U.-M. and CONICYT 21120505 to J.B., Chile, and MINECOBFU2013-42164-R and BFU2017-87095-R to J.d.M., Spain.

Conflicts of interestNone declared.

ReferencesAbu-Halima M, Backes C, Leidinger P, Keller A, Lubbad AM, HammadehM, Meese E. MicroRNA expression profiles in human testicular tissuesof infertile men with different histopathologic patterns. Fertil Steril 2014;101:78–86.

Ademollo N, Ferrara F, Delise M, Fabietti F, Funari E. Nonylphenol andoctylphenol in human breast milk. Environ Int 2008;34:984–987.

Alikhani M, Mirzaei M, Sabbaghian M, Parsamatin P, Karamzadeh R, Adib S,Sodeifi N, Gilani MAS, Zabet-Moghaddam M, Parker L et al.Quantitative proteomic analysis of human testis reveals system-widemolecular and cellular pathways associated with non-obstructive azoo-spermia. J Proteomics 2017;162:141–154.

Amiri A, Keiper BD, Kawasaki I, Fan Y, Kohara Y, Rhoads RE, Strome S.An isoform of eIF4E is a component of germ granules and is required forspermatogenesis in C. elegans. Development 2001;128:3899–3912.

Anders S, McCarthy D, Chen Y, Okoniewski M, Smyth G, Huber W,Robinson M. Count-based differential expression analysis of RNAsequencing data using R and bioconductor. Nat Protoc 2013;8:1765–1786.

Anders S, Pyl PT, Huber W. HTSeq-A Python framework to work withhigh-throughput sequencing data. Bioinformatics 2015;31:166–169.

Bergman Å, Heindel J, Jobling S, Kidd K, Zoeller RT. State of the science ofendocrine disrupting chemicals. Toxicol Lett 2012;211:S3.

Bindea G, Mlecnik B, Hackl H, Charoentong P, Tosolini M, Kirilovsky A,Fridman WH, Pagès F, Trajanoski Z, Galon J. ClueGO: a Cytoscapeplug-in to decipher functionally grouped gene ontology and pathwayannotation networks. Bioinformatics 2009;25:1091–1093.

Birks L, Casas M, Garcia AM, Alexander J, Barros H, Bergström A, BondeJP, Burdorf A, Costet N, Danileviciute A et al. Occupational exposure to

12 Buñay et al.

Dow


ic.oup.com/m


endocrine-disrupting chemicals and birth weight and length of gestation:a European meta-analysis. Environ Health Perspect 2016;124:1785–1793.

Boekelheide K, Kleymenova E, Liu K, Swanson C, Gaido KW. Dose-dependent effects on cell proliferation, seminiferous tubules, and malegerm cells in the fetal rat testis following exposure to di(n-butyl) phthal-ate.Microsc Res Tech 2009;72:629–638.

Bozec A, Amara S, Guarmit B, Selva J, Albert M, Rollet J, El Sirkasi M,Vialard F, Bailly M, Benahmed M et al. Status of the executioner step ofapoptosis in human with normal spermatogenesis and azoospermia.Fertil Steril 2008;90:1723–1731.

Bracke A, Peeters K, Punjabi U, Hoogewijs D, Dewilde S. A search formolecular mechanisms underlying male idiopathic infertility. ReprodBiomed Online 2018;36:327–339.

Bradford M. A rapid and sensitive method for the quantitcation of micro-gram quantities of protein utilizing the principle of protein–dye binding.Anal Biochem 1976;72:248–254.

Bustamante-Marín X, Quiroga C, Lavandero S, Reyes JG, Moreno RD.Apoptosis, necrosis and autophagy are influenced by metabolic energysources in cultured rat spermatocytes. Apoptosis 2012;17:539–550.

Buñay J, Larriba E, Moreno RD, del Mazo J. Chronic low-dose exposure toa mixture of environmental endocrine disruptors induces microRNAs/isomiRs deregulation in mouse concomitant with intratesticular estradiolreduction. Sci Rep 2017;7:3373.

Buñay J, Larriba E, Patiño-Garcia D, Cruz-Fernandes L, Castañeda-Zegarra S,Rodriguez-Fernandez M, del Mazo J, Moreno RD. Differential effects ofexposure to single versus a mixture of endocrine-disrupting chemicals onsteroidogenesis pathway in mouse testes. Toxicol Sci 2018;161:76–86.

Cao X, Cui Y, Zhang X, Lou J, Zhou J, Bei H, Wei R. Proteomic profile ofhuman spermatozoa in healthy and asthenozoospermic individuals.Reprod Biol Endocrinol 2018;16(1):16.

Chapin RE, Delaney J, Wang Y, Lanning L, Davis B, Collins B, Mintz N,Wolfe G. The effects of 4-nonylphenol in rats: a multigeneration repro-duction study. Toxicol Sci 1999;52:80–91.

Chen M, Tang R, Fu G, Xu B, Zhu P, Qiao S, Chen X, Xu B, Qin Y, Lu Cet al. Association of exposure to phenols and idiopathic male infertility.J Hazard Mater 2013;250–251:115–121.

de Jager C, Bornman MS, van der Horst G. The effect of p-nonylphenol, anenvironmental toxic ant with oestrogenic properties, on fertility poten-tial in adult male rats. Andrologia 1999;31:99–106.

Diz AP, Truebano M, Skibinski DO. The consequences of sample poolingin proteomics: an empirical study. Electrophoresis 2009;30:2967–2975.

Garcia PV, Barbieri MF, Perobelli JE, Consonni SR, Mesquita SdeF,Kempinas WdeG, Pereira LA. Morphometric-stereological and func-tional epididymal alterations and a decrease in fertility in rats treatedwith finasteride and after a 30-day post-treatment recovery periodFertilSteril 2012;971444–1451.

Ghosh S, Lasko P. Loss-of-function analysis reveals distinct requirementsof the translation initiation factors eIF4E, eIF4E-3, eIF4G and eIF4G2 inDrosophila spermatogenesis. PLoS One 2015;10:e0122519.

Griswold MD. Spermatogenesis: the commitment to meiosis. Physiol Rev2016;96:1–17.

Guo H, Ingolia NT,Weissman JS, Bartel DP. Mammalian microRNAs predom-inantly act to decrease target mRNA levels. Nature 2010;466:835–840.

Guo L, Yang Q, Lu J, Li H, Ge Q, GuW, Bai Y, Lu Z. A comprehensive sur-vey of miRNA repertoire and 3′ addition events in the placentas ofpatients with pre-eclampsia from high-throughput sequencing. PLoS One2011;6:e21072.

Heindel JJ, Newbold R, Schug TT. Endocrine disruptors and obesity. NatRev Endocrinol 2015;11:653–661.

Hines CJ, Hopf NB, Deddens JA, Silva MJ, Calafat AM. Estimated dailyintake of phthalates in occupationally exposed groups. J Expo Sci EnvironEpidemiol 2011;21:133–141.

Hirano Y, Kaneko T, Okamoto K, Bai M, Yashiroda H, Furuyama K, KatoK, Tanaka K, Murata S. Dissecting beta-ring assembly pathway of themammalian 20S proteasome. EMBO J 2008;27:2204–2213.

Hou JW, Lin CL, Tsai YA, Chang CH, Liao KW, Yu CJ, Yang W, Lee MJ,Huang PC, Sun CW et al. The effects of phthalate and nonylphenolexposure on body size and secondary sexual characteristics duringpuberty. Int J Hyg Environ Health 2015;218:603–615.

Hsieh AC, Ruggero D. Targeting eukaryotic translation initiation factor 4E(eIF4E) in cancer. Clin Cancer Res 2010;16:4914–4920.

Jaiswal D, Trivedi S, Agrawal NK, Singh K. Dysregulation of apoptotic path-way candidate genes and proteins in infertile azoospermia patients. FertilSteril 2015;104:736–743.e6.

Jansen RS, Addie R, Merkx R, Fish A, Mahakena S, Bleijerveld OB, Altelaar M,IJlst L, Wanders RJ, Borst P et al. N-lactoyl-amino acids are ubiquitousmetabolites that originate from CNDP2-mediated reverse proteolysis oflactate and amino acids. Proc Natl Acad Sci USA 2015;112:6601–6606.

Jodar M, Soler-Ventura A, Oliva R. Semen proteomics and male infertility.J Proteomics 2017;162:125–134.

Jonsson B Risk Assessment on Butylphenol, Octylphenol and Nonylphenol, andEstimated Human Exposure of Alkylphenols From Swedish Fish. Uppsala,Sweden: Uppsala University Press, 2006. Available from: http://www.uu.se/digitalAssets/177/c_177024-l_3-k_jonsson-beatrice-report.pdf.

Karp NA, Lilley KS. Investigating sample pooling strategies for DIGE experi-ments to address biological variability. Proteomics 2009;9:388–397.

Kim HJ, Na JI, Min BW, Na JY, Lee KH, Lee JH, Lee YJ, Kim HS, Park JT.Evaluation of protein expression in housekeeping genes across multipletissues in rats. Korean J Pathol 2014;48:193–200.

Klett D, Cahoreau C, Villeret M, Combarnous Y. Effect of pharmaceuticalpotential endocrine disruptor compounds on protein disulfide isomerasereductase activity using di-eosin-oxidized-glutathion. PLoS One 2010;5:e9507.

Kozlowska E, Krzyzosiak WJ, Koscianska E. Regulation of Huntingtin geneexpression by miRNA-137, -214, -148a, and their respective isomiRs.Int J Mol Sci 2013;14:16999–17016.

Lambrot R, Muczynski V, Lécureuil C, Angenard G, Coffigny H, Pairault C,Moison D, Frydman R, Habert R, Rouiller-Fabre V. Phthalates impairgerm cell development in the human Fetal testis in vitro without changein testosterone production. Environ Health Perspect 2009;117:32–37.

Langmead B, Trapnell C, Pop M, Salzberg SL. Ultrafast and memory-efficient alignment of short DNA sequences to the human genome.Genome Biol 2009;10:R25.

Li Z, Zheng Z, Ruan J, Li Z, Zhuang X, Tzeng CM. Integrated analysis miRNAand mRNA profiling in patients with severe oligozoospermia reveals miR-34c-3p downregulates PLCXD3 expression.Oncotarget 2016;7:52781–52796.

López-Casas PP, Mizrak SC, López-Fernández LA, Paz M, de Rooij DG, delMazo J. The effects of different endocrine disruptors definingcompound-specific alterations of gene expression profiles in the devel-oping testis. Reprod Toxicol 2012;33:106–115.

Manikkam M, Tracey R, Guerrero-Bosagna C, Skinner MK. Plastics derivedendocrine disruptors (BPA, DEHP and DBP) induce epigenetic transge-nerational inheritance of obesity, reproductive disease and sperm epi-mutations. PLoS One 2013;8:e55387.

Muller H, Marzi MJ, Nicassio F. IsomiRage: from functional classification to dif-ferential expression of miRNA isoforms. Front Bioeng Biotechnol 2014;2:38.

Munoz X, Mata A, Bassas LL, Larriba S. Altered miRNA signature of devel-oping germ-cells in infertile patients relates to the severity of spermato-genic failure and persists in spermatozoa. Sci Rep 2015;5:17991.

Nagao T, Yoshimura S, Saito Y, Nakagomi M, Usumi K, Ono H.Reproductive effects in male and female rats from neonatal exposure top-octylphenol. Reprod Toxicol 2001;15:683–692.

National Toxicology Program. NTP-CERHRmonograph on the potential humanreproductive and developmental effects of di-n-butyl phthalate (DBP). NTPCerhr Mon 2003a;i-III90. http://www.ncbi.nlm.nih.gov/pubmed/15995736.


Dow


ic.oup.com/m


http://www.uu.se/digitalAssets/177/c_177024-l_3-k_jonsson-beatrice-report.pdf

http://www.uu.se/digitalAssets/177/c_177024-l_3-k_jonsson-beatrice-report.pdf

http://www.ncbi.nlm.nih.gov/pubmed/15995736

National Toxicology Program. NTP-CERHR monograph on the potentialhuman reproductive and developmental effects of butyl benzyl phthalate(BBP). NTP Cerhr Mon 2003b;i-III90. http://www.ncbi.nlm.nih.gov/pubmed/15995737.

National Toxicology Program. NTP-CERHR monograph on the potentialhuman reproductive and developmental effects of di (2-ethylhexyl)phthalate (DEHP). National Institutes of Health 2006; NIH PublicationNo. 06–4476; v, vii-7, II–iii-xiii passim. http://www.ncbi.nlm.nih.gov/pubmed/19407857.

Neilsen CT, Goodall GJ, Bracken CP. IsomiRs—the overlooked repertoirein the dynamic microRNAome. Trends Genet 2012;28:544–549.

Noveski P, Popovska-Jankovic K, Kubelka-Sabit K, Filipovski V, LazarevskiS, Plaseski T, Plaseska-Karanfilska D. MicroRNA expression profiles intesticular biopsies of patients with impaired spermatogenesis. Andrology2016;4:1020–1027.

Pant N, Shukla M, Kumar Patel D, Shukla Y, Mathur N, Kumar Gupta Y,Saxena DK. Correlation of phthalate exposures with semen quality.Toxicol Appl Pharmacol 2008;231:112–116.

Patiño-García D, Cruz-Fernandes L, Buñay J, Palomino J, Moreno RD.Reproductive alterations in chronically exposed female mice to environ-mentally relevant doses of a mixture of phthalates and alkylphenols.Endocrinology 2018;159:1050–1061.

Paz M, Morín M, del Mazo J. Proteome profile changes during mouse testis devel-opment. Comp Biochem Physiol Part D Genomics Proteomics 2006;1:404–415.

Primm TP, Gilbert HF. Hormone binding by protein disulfide isomerase, ahigh capacity hormone reservoir of the endoplasmic reticulum. J BiolChem 2001;276:281–286.

Radonić A, Thulke S, Mackay IM, Landt O, Siegert W, Nitsche A.Guideline to reference gene selection for quantitative real-time PCR.Biochem Biophys Res Commun 2004;313:856–862.

Rajamani U, Gross AR, Ocampo C, Andres AM, Gottlieb RA, Sareen D.Endocrine disruptors induce perturbations in endoplasmic reticulum andmitochondria of human pluripotent stem cell derivatives. Nat Commun2017;8:219.

Rider CV, Furr JR, Wilson VS, Gray LE. Cumulative effects of in utero admin-istration of mixtures of reproductive toxicants that disrupt common targettissues via diverse mechanisms of toxicity. Int J Androl 2010;33:443–462.

Rogers R, Ouellet G, Brown C, Moyer B, Rasoulpour T, Hixon M. Cross-talk between the Akt and NF-kappaB signaling pathways inhibits MEHP-induced germ cell apoptosis. Toxicol Sci 2008;106:497–508.

Siddiqui N, Sonenberg N. Signalling to eIF4E in cancer. Biochem Soc Trans2015;43:763–772.

Sifakis S, Androutsopoulos VP, Tsatsakis AM, Spandidos DA. Humanexposure to endocrine disrupting chemicals: effects on the male andfemale reproductive systems. Environ Toxicol Pharmacol 2017;51:56–70.

Smorag L, Zheng Y, Nolte J, Zechner U, Engel W, Pantakani DV.MicroRNA signature in various cell types of mouse spermatogenesis:

evidence for stage-specifically expressed miRNA-221, -203 and -34b-5pmediated spermatogenesis regulation. Biol Cell 2012;104:677–692.

Takagi S, Itoh N, Kimura M, Sasao T, Tsukamoto T. Spermatogonial prolif-eration and apoptosis in hypospermatogenesis associated with nonob-structive azoospermia. Fertil Steril 2001;76:901–907.

Telonis AG, Loher P, Jing Y, Londin E, Rigoutsos I. Beyond the one-locus-one-miRNA paradigm: microRNA isoforms enable deeper insights intobreast cancer heterogeneity. Nucleic Acids Res 2015;43:9158–9175.

Tumova J, Andel M, Trnka J. Excess of free fatty acids as a cause of meta-bolic dysfunction in skeletal muscle. Physiol Res 2016;65:193–207.

Uhlen M, Bandrowski A, Carr S, Edwards A, Ellenberg J, Lundberg E, RimmDL, Rodriguez H, Hiltke T, Snyder M et al. A proposal for validation ofantibodies. Nat Methods 2016;10:823–827.

Vera Y, Erkkila K, Wang C, Nunez C, Kyttanen S, Lue Y, Dunkel L,Swerdloff RS, Sinha Hikim AP. Involvement of p38 mitogen-activatedprotein kinase and inducible nitric oxide synthase in apoptotic signalingof murine and human male germ cells after hormone deprivation. MolEndocrinol 2006;20:1597–1609.

Vlachos IS, Paraskevopoulou MD, Karagkouni D, Georgakilas G, Vergoulis T,Kanellos I, Anastasopoulos IL, Maniou S, Karathanou K, Kalfakakou D et al.DIANA-TarBase v7.0: indexing more than half a million experimentally sup-ported miRNA:mRNA interactions. Nucleic Acids Res 2015;43:D153–D159.

Weinkauf M, Hiddemann W, Dreyling M. Sample pooling in 2-D gel elec-trophoresis: a new approach to reduce nonspecific expression back-ground. Electrophoresis 2006;27:4555–4558.

Wessel D, Flügge UI. A method for the quantitative recovery of protein indilute solution in the presence of detergents and lipids. Anal Biochem1984;138:141–143.

Wong CW, Lam KKW, Lee CL, Yeung WSB, Zhao WE, Ho PC et al. Theroles of protein disulphide isomerase family A, member 3 (ERp57) andsurface thiol/disulphide exchange in human spermatozoa-zona pellucidabinding. Hum Reprod 2017;32(4):733–742.

Yoshida M, Katsuda SI, Takenaka A, Watanabe G, Taya K, Maekawa A.Effects of neonatal exposure to a high-dose p-tert-octylphenol on themale reproductive tract in rats. Toxicol Lett 2001;121:21–33.

Yu Z, Guo R, Ge Y, Ma J, Guan J, Li S, Sun X, Xue S, Han D. Gene expres-sion profiles in different stages of mouse spermatogenic cells duringspermatogenesis. Biol Reprod 2003;69:37–47.

Zhang Z, Miao L, Xin X, Zhang J, Yang S, Miao M, Kong X, Jiao B.Underexpressed CNDP2 participates in gastric cancer growth inhibitionthrough activating the MAPK signaling pathway.Mol Med 2014;20:17–28.

Zhang S, Zhao Y, Lei B, Li C, Mao X. PGAM1 is involved in spermatogenicdysfunction and affects cell proliferation, apoptosis, and migration.Reprod Sci 2015;22:1236–1242.

Zhuang X, Li Z, Lin H, Gu L, Lin Q, Lu Z, Tzeng CM. Integrated miRNAand mRNA expression profiling to identify mRNA targets of dysregu-lated miRNAs in non-obstructive azoospermia. Sci Rep 2015;5:7922.

14 Buñay et al.

Dow


ic.oup.com/m






Combined proteomic and miRNome endocrine …...contributing equally with 50μg of total protein....

Documents

Transcript of Combined proteomic and miRNome endocrine …...contributing equally with 50μg of total protein....