perturbation of epigenetic processes by doxorubicin in the ... uni - akinjo et al... · shown to...

Toxicology Research

PAPER

Cite this: Toxicol. Res., 2016, 5, 1229

Received 29th February 2016,Accepted 30th May 2016

DOI: 10.1039/c6tx00078a

www.rsc.org/toxicology

Perturbation of epigenetic processes bydoxorubicin in the mouse testis†

Oluwajoba O. Akinjo,‡ Timothy W. Gant and Emma L. Marczylo*

Epigenetic processes play a major role in normal mammalian development, particularly during gameto-

genesis and early embryogenesis. Thus, perturbation of epigenetic processes in the testis by xenobiotics

could have a major impact on testicular function and fertility, and potentially affect the development and

health of subsequent generations. There has been substantial research into the epigenetic toxicity of

environmental exposures over the last decade. However, few studies have focussed on pharmaceutical

drugs, which due to the nature of their use are typically found at much higher concentrations within

exposed individuals than environmental chemicals. Here, we investigated genome-wide changes in testi-

cular mRNA transcription, microRNA expression and DNA methylation to assess the contribution of epi-

genetic mechanisms to the testicular toxicity induced by doxorubicin (DOX) as a representative, widely

used and well-characterised anti-cancer drug. We demonstrated that DOX is able to induce transcrip-

tional, microRNA and DNA methylation changes, which perturb pathways involved in stress/cell death and

survival and testicular function and lead to germ cell loss and reproductive organ damage. This identified

potential novel mechanisms of DOX-induced testicular toxicity for further focussed investigations. Such

work is required to fully assess the role of epigenetics in toxicity, determine whether single and/or multi-

generational epigenetic toxicity is a real public health concern, and begin to develop and incorporate

relevant epigenetic endpoints into regulatory toxicology.

Introduction

Cancer is an important Public Health issue, not only in termsof diagnosis and treatment but also the continued medicalsupport of survivors. It is now estimated that over 2 millionpeople in the UK are living with and beyond a diagnosis ofcancer,1 and that there are 14.5 million cancer survivors in theUS.2 Survivors of childhood cancers require careful follow-upas they have a greater risk of developing secondary cancers andchronic conditions in later life as a result of their previoustreatments, and are more likely to want to go on to have theirown children. The most common types of cancers in the UKand US in both childhood (0–15 years) and teenagers/youngadults (15–24 years) are leukaemias and lymphomas,1,2 thetreatment of which normally involves DOX or another anthra-cycline.3 With the overall 5 year survival rate of cancers diag-nosed before the age of 25 now greater than 80%, there arecurrently 35 000 cancer survivors in the UK that were diag-nosed before the age of 15,1 and 388 501 cancer survivors in

the US that were diagnosed before the age of 21.2 It is esti-mated that around half of these will have been treated with ananthracycline.3

The long term cardiotoxicity associated with anthracyclineuse is widely recognised in cancer survivors and has beenextensively investigated mechanistically in various modelsystems (both by ourselves4,5 and others6). However, much lessis known about the long term effects of DOX, or other anthra-cyclines, on subsequent reproduction and development. Thereis some epidemiological evidence that chemotherapeutic treat-ment can be detrimental to later life reproductive function.7–9

It is therefore important to investigate the testicular toxicityinduced by chemotherapeutics like DOX. Such pharmaceuti-cals are widely prescribed to younger male cancer patients thatmay go on to father subsequent generations. To-date DOX hasbeen associated with reduced testicular blood flow andvolume, testicular atrophy, increased cellular apoptosis,altered spermatogenesis, increased sperm abnormalities andreduced sperm counts and fertility.10–17 Nevertheless, themolecular mechanisms underlying DOX-induced testicular toxi-city, particularly epigenetic processes, remain underexplored.

Epigenetics encompasses the mechanisms that regulategene expression without altering the underlying DNAsequence, including DNA methylation, histone modificationand non-coding RNAs (ncRNAs) (such as microRNAs

†Electronic supplementary information (ESI) available. See DOI: 10.1039/c6tx00078a‡Now at Reckitt Benckiser (Brands) Ltd, Dansom Lane, Hull, HU8 7DS.

Toxicology Department, CRCE, PHE, Chilton, Oxfordshire OX11 0RQ, UK.

E-mail: [email protected]

This journal is © The Royal Society of Chemistry 2016 Toxicol. Res., 2016, 5, 1229–1243 | 1229

Publ

ishe

d on

01

June

201

6. D

ownl

oade

d by

RSC

Int

erna

l on

14/0

6/20

18 0

8:47

:15.

View Article OnlineView Journal | View Issue

www.rsc.org/toxicology

http://crossmark.crossref.org/dialog/?doi=10.1039/c6tx00078a&domain=pdf&date_stamp=2016-06-20

http://dx.doi.org/10.1039/c6tx00078a

http://pubs.rsc.org/en/journals/journal/TX

http://pubs.rsc.org/en/journals/journal/TX?issueid=TX005004

(miRNAs), piwi-interacting RNAs (piRNAs) and endogenoussiRNAs (esiRNAs)). These epigenetic processes play vital rolesduring mammalian development.18 Briefly, the mammalianlife cycle involves two genome-wide epigenetic re-programmingand programming events. The first event occurs immediatelyafter fertilisation to reset the embryonic genome so that it iscapable of producing all of the different cell types necessary toform a new individual, and subsequently drive the cellular pro-liferation and differentiation processes involved in develop-ment of the offspring through embryogenesis and into amature adult. The second event involves the primordial germcells (PGCs) only. It functions to reset all of the future germcells according to the gender of the embryo, (parentallyimprinted genes must be correctly reset to enable the pro-duction of future generations), and subsequently drive gameto-genesis. Both of these events involve genome-wide DNAdemethylation and changes in histone modifications and non-coding RNA expression to effectively ‘wipe clean’ the embryo-nic and germ cell epigenomes, followed by genome-wide DNAremethylation, and further changes in histone modificationsand non-coding RNA expression to redirect and regulate theproceeding development.

These epigenetic mechanisms do not function indepen-dently; they interact to regulate gene expression at multiplelevels. This helps to confer efficiency, redundancy, robustnessand fidelity onto epigenetic re-programming/programmingduring development, and thus the successful and continuedproduction of multiple generations.18 While such complex net-works exist to maintain cellular homeostasis, they are open toperturbation. A wide range of environmental factors have beenshown to induce epigenetically-mediated adverse phenotypesin rodent models and can be linked with epigenetic changesand adverse outcomes in human epidemiological cohorts,leading to the “Developmental Origins of Health and Disease”(DOHaD) hypothesis.18 Indeed, a number of recent indepen-dent studies have demonstrated that early life trauma or par-ental stress can induce altered miRNA expression in sperm,which in turn can transmit behavioural and metabolicabnormalities through to the next generation.19–21 There isalso evidence that DNA methylation changes and histonemodifications in the male germline could transmit unique epi-genetic information to activate and regulate early embryodevelopment in the next generation.22–25 Any change thatbecomes permanently established in the germline could con-tinue to adversely affect subsequent generations, even in theabsence of the original exposure, leading to so-called trans-generational toxicity. There is concern that pharmaceuticalscould also induce similar epigenetic toxicity.26 Furthermore, ithas been demonstrated in rodent models that paternal anti-cancer drug exposure can increase genomic instability in thetreated animals and their resulting offspring,27 and paternalDOX exposures can induce adverse developmental phenotypesin subsequent zygotes.13,28–30

In order to improve our understanding of the epigeneticmechanism(s) involved in pharmaceutical-induced testiculartoxicity, we selected DOX as a representative, widely-used and

well-characterised anti-cancer drug. We used a dosing method(3 doses of 3 mg−1 kg wk−1) previously shown to induce testi-cular toxicity in rodents.11,13,31 A total dose of 9 mg kg−1

equates to 27 mg m−2 in a typical 60 kg human (using the rec-ommended body surface area normalisation method for con-version).32 This is 17–20× lower than the maximum cumulativedose of 450–550 mg m−2 used in cancer treatment protocols.Since epigenetic processes interact to regulate gene expressionat multiple system levels, we analysed genome-wide changes inmRNA and miRNA expression, and DNA methylation, along-side toxicological endpoints in mice treated with DOX. Thesedata provided a greater understanding of the epigenetic andtranscriptional processes that contribute to DOX-induced testi-cular toxicity, and related specific epigenetic and transcrip-tional changes to actual phenotypic endpoints. Such changesare useful for informing further focused studies on DOX-induced epigenetic testicular toxicity.

Materials & methodsMaterials

All materials, including DOX (as doxorubicin hydrochloride)were purchased from Sigma Aldrich, UK unless otherwisestated.

In vivo model

Animal experimentation, transportation and care were per-formed in compliance with the Animal Scientific ProceduresAct (1986) under project licenses 80/2126 and 30/3021, andapproved by local ethical review committees at the Universityof Leicester or the Centre for Radiation, Chemical andEnvironmental Hazards, Public Health England. C57BL/6Jmale mice (8 week of age and weighing 18–25 g) were housedand fed ad libitum food and water with a 12 h light/dark cycle,and acclimatised for 1 week prior to dosing. DOX was dis-solved in 0.85% (v/v) saline to a working dilution of 0.3 mgml−1 and administered via intraperitoneal injection (ip).Animals received either 3 mg kg−1 DOX (n = 6) or vehicle-only(0.85% (v/v) saline) (n = 6) once per week for 3 week, resultingin a total dose of 9 mg kg−1. Animals were then sacrificed at 1,4 or 7 week following the final drug administration asdescribed below. Twelve animals were used for each of thetime points, 6 control and 6 treated in each group.

Animals were anaesthetised with isofluorane, blood was col-lected by cardiac puncture and organs were perfused with10 ml of PBS (Life Technlogies, UK). Animals were then sacri-ficed by cervical dislocation and organs (testes, caudal epididy-mis and seminal vesicles) were removed, weighed,immediately snap frozen in liquid nitrogen and stored at−80 °C until further analysis. Separate animal experimentswere conducted to generate testis samples for histopathologi-cal analysis. Paired testes were pierced at each end with aneedle and one testis was placed in Bouin’s fixative, whilst thesecond testis (from the same animal) was placed in Zenker’sfixative for 24 h prior to histological processing.

Paper Toxicology Research

1230 | Toxicol. Res., 2016, 5, 1229–1243 This journal is © The Royal Society of Chemistry 2016

Publ

ishe

d on

01

June

201

6. D

ownl

oade

d by

RSC

Int

erna

l on

14/0

6/20

18 0

8:47

:15.

View Article Online


Histopathological analysis

Paired Bouin’s and Zenker’s fixed testes were washed directlyunder running water for 24 h, followed by dehydration in 70%alcohol for a further 24 h. Dehydrated testes were thenembedded in paraffin wax and stored at 4 °C until sectioning.Paraffin-embedded testes were cut into 5 μm sections on aRM2155 microtome (Leica Microsystems, Switzerland),mounted onto glass slides, stained with H&E or PAS and viewedand imaged using Leica Application Suite Version 3.8 softwareon a DM2000 microscope connected to a DFC450C camera(Leica Microsystems, Switzerland). Following identification andconfirmation of each type of germ cell population in the semini-ferous tubule (spermatogonia, spermatocytes, round spermatidsand elongated spermatids), the presence of each germ cellpopulation per seminiferous tubule within the field of view(40× magnification) was calculated as a percentage.

Sperm count

Caudal epididymides from individual animals were finelychopped in a petri dish (Sarstedt, Germany), collected in 1 mlPBS and passed through a CD-1TM 190 μm mesh screen. AC-Chip disposable haemocytometer (PAA, UK) was used tocount 10 μl of the resulting sperm suspension. The totalnumber of sperm per caudal epididymides was calculated asnumber of sperm heads per ml.

Nucleic acid profiling

RNA or DNA were extracted from frozen testis using either Trireagent or a DNeasy Blood and Tissue kit (Qiagen, Germany),respectively, according to the manufacturer’s instructions.RNA was stored at −80 °C and DNA at −20 °C until microarrayanalysis. RNA integrity was determined using a 2100 Bioanaly-ser and RNA 6000 LabChip® kit (Agilent Technologies, USA)according to the manufacturer’s instructions. Only RNAsamples with an RNA integrity number (RIN) ≥ 8.0 were usedfor downstream analysis.

mRNA (transcriptional) microarrays

Genome-wide changes in testicular mRNA expression were pro-filed using microarrays obtained from the Medical ResearchCouncil Toxicology Unit (Leicester, UK). These microarrayswere printed in-house using an inkjet style ArrayJet Ultra-Mara-thon II microarray printer (ArrayJet, Roslin, Scotland) and theMouse Exonic Evidence-Based Oligonucleotide (MEEBO) probeset (Life Technologies, UK). The MEEBO microarrays included atotal of 39 000 probes that provided complete coverage of themouse genome. RNA (200 ng) was amplified and labelled with aTwo-Colour Low Input Quick Amp Labelling Kit (Agilent Techno-logies, USA) according to the manufacturer’s instructions andincluded a dye-swap to avoid dye bias (n ≥ 3 for both forwardand reverse labelling). cDNA concentration and efficiency ofcDNA labelling was measured with a NanoDrop ND-1000spectrophotometer. Only cDNA samples with a concentration≥200 ng μl−1 and Cy3/Cy5 labelling ≥5 pmol μl−1 were used forhybridisation. Control and treated samples were randomly

paired and pooled so that the concentrations of Cy3 and Cy5cDNA were equal and dried down to a final volume of 9 μlprior to the addition of 1 μl fragmentation reagent (Ambion,UK). Samples were incubated for 15 min at 70 °C. The frag-mentation reaction was terminated by the addition of 1 μl stopsolution (Ambion, UK) and incubation for 5 min on ice, fol-lowed by addition of 28 μl ultra-pure water (Thermofisher, UK),1 μl yeast tRNA (4 mg ml−1) and 40 μl 2× enhanced hybridis-ation buffer (Genisphere, USA). Samples were then denaturedfor 5 min at 100 °C and incubated for 30 min at 42 °C prior tohybridisation on the MEEBO microarrays overnight at 42 °C.Microarray slides were washed on a rocking platform (Scienti-fic Laboratory Supplies, UK) set at 50 oscillations per min for5 min each at RT, firstly in 1× SSC containing 0.03% (w/v) SDS,secondly in 0.2× SSC and finally in 0.05× SSC. Slides weredried by centrifugation at 1200g for 4 min at RT and scannedon a 4200A Axon scanner (Molecular Devices, CA) at wave-lengths of 532 and 635 nm and a resolution of 5 μm usingGenePix 6.0 software (Molecular Devices, CA).

miRNA microarrays

Genome-wide changes in testicular miRNA expression wereprofiled using microarrays also obtained from the MedicalResearch Council Toxicology Unit (Leicester, UK), printed in-house using the printer described above and the miRCURYLNA™ probe set (Exiqon, Denmark). The miRCURY probe setincluded a total of 1500 probes that provided complete cover-age of version 8.0 of miRBase. RNA (500 ng) was labelled witheither the Flashtag™ RNA Labelling Kit (Genisphere, USA) orthe Label IT® miRNA labelling kit (Mirus, USA) according tothe manufacturer’s instructions (except that all reactions werehalved) and included a dye-swap to avoid dye bias (n ≥ 6 forboth forward and reverse labelling). Control and treatedsamples were randomly paired and pooled so that the concen-trations of Cy3 and Cy5 cDNA were equal and treated accordingto the labelling kit’s manufacturer’s instructions. Sampleswere then denatured for 3–10 min at 65 °C and incubated for30 min at 37 °C prior to hybridisation on the miRCURY micro-arrays overnight at 37 °C. Microarray slides were washed firstlyin 2× SCC containing 0.05% (w/v) SDS, secondly in 1× SSC andfinally in 0.5× SSC, dried by centrifugation, and scanned asdescribed above.

DNA methylation arrays

Genome-wide changes in testicular DNA methylation were pro-filed using MM9 CpG Refseq Prom MeDIP microarrays pur-chased from Roche NimbleGen (Roche, USA). These microarraysincluded a total of 720 000 probes designed to cover all CpGislands within gene bodies and the known promoter regions ofgenes in the mouse genome version mm9. Immunoprecipita-tion of methylated DNA was carried out according to Nimble-Gen DNA Methylation Services: sample preparationinstructions v4.1 (Roche, USA). Labelling and hybridisation ofinput and immunoprecipitated samples, followed by washingand scanning of microarrays was carried out according toNimbleGen Arrays User’s Guide: Epigenetics Arrays v1.0 using a

Toxicology Research Paper


Publ

ishe

d on

01

June

201

6. D

ownl

oade

d by

RSC

Int

erna

l on

14/0

6/20

18 0

8:47

:15.

View Article Online


NimbleGen Hybridisation System 4, a MS 200 Microarrayscanner and the MS Data Collection Software (Roche, USA).

Bioinformatic analysis

mRNA and miRNA expression. Pixel intensities were anno-tated to mouse reference files (mRNAs and miRNAs) andextracted using GenePix pro 6.0 (Molecular Devices, CA). Theseraw data files were normalised using the LOESS within arrayfollowed by the quantile between array commands in the Bio-conductor limma library package of R version 2.15.2 (http://www.r-project.org/). Normalised log2[Cy5/Cy3] ratios were com-piled into a spreadsheet and imported into Qlucore OmicsExplorer (QOE) version 2.3 (Qlucore, Sweden) to identify stat-istically significant differentially expressed transcripts ormiRNAs (p < 0.001, q < 0.01 significance level, where q rep-resents a multiple comparison correction with a false discoveryrate of 1%) at each time point following DOX treatment. Listscontaining the top 100 most significantly differentiallyexpressed transcripts or miRNAs over time were imported intoIngenuity® Pathway Analysis Software ((IPA®), Ingenuity®Systems http://www.ingenuity.com/products/ipa) to explore themRNA and miRNA networks and pathways involved in DOX-induced testicular toxicity.

DNA methylation changes. Data was analysed using DEVASoftware User Guide version 1.2 (RocheNimbleGen, USA)according to the manufacturer’s instructions. Briefly, log2-[Immunoprecipitated/Input] ratios were calculated and assigneda probability score (p value) by performing a Kolmogorov–Smirnov test on several adjacent probes using a sliding windowset at 750 bp with overlapping 500 bp spacing. From thisp value, the peak score was calculated by combining four con-secutive probes per peak and values ≥2 were identified asmethylated peaks. These peaks were mapped to annotationfiles provided by Roche NimbleGen to generate a spreadsheetof peak values mapped to their associated gene. To obtain avalue for the difference between control and treated samples,the ratio of the peak score (peak score DOX–peak scorecontrol) was calculated. These ratios were compiled into aspreadsheet and imported into QOE to identify statistically sig-nificant differentially methylated genes (p < 0.001, q < 0.01 sig-nificance level, where q represents a multiple comparisoncorrection with a false discovery rate of 1%) at each time pointfollowing DOX treatment. A list containing the top 100 mostsignificantly differentially methylated regions (DMRs) over timewere imported into IPA® to explore the differentially methy-lated gene networks and pathways involved in DOX-induced tes-ticular toxicity, and correlated with the list of the top 100 mostsignificantly differentially expressed mRNA transcripts.

Raw data files for all three datasets have been submitted toGEO: mRNA (GEO***), miRNA (GEO***) and DNA methylation(GEO***).

Real-time PCR validation. Microarray results were validatedby qRT-PCR using either a 7500 Fast (Life Technologies, UK)or a Rotor-Gene Q (Qiagen, Germany) qRT-PCR System.

mRNA qRT-PCR. Synthesis of cDNA was performed usingeither a High Capacity RNA to cDNA Kit (Life Technologies,

UK) or a QuantiTecT Reverse Transcription Kit (Qiagen,Germany) according to the manufacturer’s instructions. Theresulting cDNA was diluted with ultrapure water/1× TE (10 mMTris-HCl (pH 8.0), 0.1 mM EDTA) to a final concentration equi-valent to 10–100 ng RNA input. 1 µl diluted cDNA was analysedby qPCR in a 20–25 µl reaction containing 200 nM primersand 10–12.5 µl of either 2× Power Sybr Green Master Mix (LifeTechnologies, UK) or 2× Perfecta Sybr Green Supermix (QuantaBiosciences, USA) according to the manufacturer’s instruc-tions. Relative mRNA expression was calculated using the deltaCt method and Gapdh as an endogenous control. All primersequences are shown in ESI Table 1.†

miRNA qRT-PCR. Synthesis of cDNA was performed using aqScript microRNA cDNA Synthesis Kit (Quanta BioSciences,USA) according to the manufacturer’s instructions. The result-ing cDNA was diluted with 1× TE to a final concentration equi-valent to 10 ng RNA input. 4 μl of diluted cDNA was analysedby qPCR in a 20 µl reaction containing 250 nM microRNAprimer (Integrated DNA Technologies, UK), 250 nM perfectauniversal pcr primer (Integrated DNA Technologies, UK) and10 μl 2× perfecta SYBR green supermix (Quanta BioSciences,USA) according to the manufacturer’s instructions. RelativemiRNA expression was calculated using the delta Ct methodand RNU6 as an endogenous control.

ResultsDOX induced testicular toxicity

Reductions in the weight of testes, caudal epididymides andseminal vesicles, and the concentration of sperm within thecaudal epididymides were observed following DOX treatment(Fig. 1). A number of these decreases were time-dependent and

Fig. 1 Weight of reproductive organs and sperm count over time fol-lowing treatment with DOX. Values represent mean ± SD (n = 3). *p <0.05, **p < 0.01 and ***p < 0.001 (T-test).



Publ

ishe

d on

01

June

201

6. D

ownl

oade

d by

RSC

Int

erna

l on

14/0

6/20

18 0

8:47

:15.

View Article Online


were reflected in the histopathological analysis of the testes.Testicular sections were examined using two different fixatives(Bouin’s and Zenker’s), each with two different staining tech-niques (H&E and PAS) to ensure that all relevant cell typescould be clearly identified and counted (for further infor-mation see ESI Fig. 1†). A representative image showing theidentification of different testicular cell populations is pres-ented in Fig. 2A. Control testis exhibited normal morphologyof the seminiferous and interstitial tubules at all time points(Fig. 2C, E & G). Seminiferous tubules consisted of Sertoli cellsand well organised germ cells, with progressive differentiationfrom spermatogonia in the basal compartment movinginwards through spermatocytes and round spermatids toelongated spermatids close to release into the lumen.Elongated spermatids demonstrated intact maturation withformation of the acrosomal cap and Leydig cells were presentwithin the interstitial tubules. In contrast, testis treated withDOX displayed disorganised seminiferous tubules with shrink-age, vacuolisation and some level of germ cell depletion at alltime points (Fig. 2D, F & H). At 1 week post-treatment germcells at earlier stages of spermatogenesis (spermatogonia, sper-matocytes and even round spermatids) were absent; whilstmore mature elongated spermatids close to an abnormal, butdiscernible, lumen were present (Fig. 2D). As in controlanimals, these elongated spermatids showed intact maturationwith formation of the acrosomal cap. However, unlike those inthe control animals, the DOX exposed elongated spermatidsappeared to be arrested in the germinal epithelium, with noevidence of release by any of the seminiferous tubules. Sper-matogenesis is a continual process. Thus, at any point in time,different seminiferous tubules will be at different stages ofsperm production. This is reflected in the control animalswith the release, and therefore absence, of elongated sperma-tids from a proportion of the seminiferous tubules (Fig. 2C). Incontrast, all of the seminiferous tubules in the treated animalscontain elongated spermatids, which, together with the lack ofspermatogonia, spermatocytes and round spermatids, suggestsarrest of spermatogenesis at 1 week post-treatment (Fig. 2D).By 4 week post-treatment seminiferous tubules were distinctlyatrophied and showed a Sertoli-only syndrome with a lack oflumen and all germ cells (Fig. 2F). However, at 7 week post-treatment, although the majority of seminiferous tubulesremained atrophied and showed a Sertoli-only syndrome,some had begun to show signs of recovery with the presence ofspermatogonia and spermatocytes (Fig. 2H). This increasedsensitivity of germ cells at earlier stages of spermatogenesis toDOX-induced toxicity compared to the more mature elongatedspermatids with an intact acrosomal cap, followed by somesubsequent recovery, is particularly well highlighted by pres-enting the number of spermatogonia, spermatocytes, roundspermatids or elongated spermatids per seminiferous tubuleover time (Fig. 2B). The apparent increased number ofelongated spermatids in the seminiferous tubules of theanimals at 1 week post-treatment reflects the arrest of sperma-togenesis. It is also worth noting that while Leydig cells werepresent in both control and DOX treated testes, there appeared

to be signs of Leydig cell hyperplasia following DOX exposure,particularly at 4 and 7 week post-treatment.

In summary, treatment with DOX induced significant testi-cular atrophy, (involving progressive loss of germ cells leadingto reduced sperm counts and Leydig cell hyperplasia), with evi-dence of some recovery at 7 week.

DOX perturbed both transcriptomic and epigenetic testicularprofiles

DOX induced the differential expression of approximately 700transcripts and 200 miRNAs at the p < 0.001, q < 0.01 signifi-cance level (q represents a multiple comparison correctionwith a false discovery rate of 1%), with the 4 week post-treat-ment group demonstrating the greatest level of change. Thetop 100 most statistically significant changes over time, corres-ponding to 96 mRNAs and 75 miRNAs, are shown in Fig. 3 & 4.

Those mRNAs and miRNAs of particular interest, due toeither differential expression at more than one time point andfrom literature searches identifying their role in male infertilityand tissue atrophy, were validated by qRT-PCR (Table 1). Vali-dated mRNA transcripts were classified into groups based ontheir biological and physiological roles; testicular function,stress/cell death and survival, or DNA methylation. Of the 9transcripts with roles in testicular functions, 7 (Catsper1 & 3,Crem, Hspa2, Nme5, Stra8 and Tnp2) were downregulated and 2(Cyp17a1 and Insl3) were upregulated following treatment withDOX. In contrast, of the 10 stress/cell death and survival tran-scripts, 9 (Bcl2, Casp3, Casp6, Ctsb, Ctsd, Ctss, Ctsw, Gstm3 andTcf21) were upregulated and 1 (Hspa1a) was downregulatedpost DOX exposure. DOX also induced downregulation of theDNA methyltransferase (Dnmt ) family transcripts, with Dnmt3ldemonstrating the greatest level of change. A total of5 miRNAs (miRs-26a, 29a–c and 145) were subsequently vali-dated due to their common targeting of the DNA methyl-transferase family (mRNA targets of the 5 miRNAs wereidentified using miRwalk33) and their pro-apoptoticfunctions.34–38 All 5 of these miRNAs were upregulated at twoor more time points following DOX exposure, with the greatestlevel of change at 4 week post-treatment. All of the qRT-PCRresults were in agreement with the microarray data (ESITable 2†).

The inverse correlation in expression of the Dnmt familytranscripts with that of miRNAs known to target themsuggested that DOX may disrupt methylation patterns withinthe testis. Therefore, genome-wide changes in testicular DNAmethylation were also analysed. DOX induced the differentialmethylation of over 450 regions at the p < 0.001, q < 0.01 sig-nificance level (q represents a multiple comparison correctionwith a false discovery rate of 1%). The top 126 most signifi-cantly DMRs over time are shown in Fig. 5A. Of these 126DMRs, 99 were associated with known genes, mapping toregions within the promoter or gene body (Fig. 5B). Distri-bution of the top 126 DMRs appeared to be largely random,being located on the majority of chromosomes with thehighest proportion on chromosome 2 and the X chromosome(Fig. 5C).



Publ

ishe

d on

01

June

201

6. D

ownl

oade

d by

RSC

Int

erna

l on

14/0

6/20

18 0

8:47

:15.

View Article Online


Fig. 2 Histological analysis of testis over time following treatment with DOX. (A) Representative image of control animals demonstrating identifi-cation of different cell populations using H&E staining following fixation in Bouin’s solution. SG = spermatogonia, SC = spermatocyte, rSP = roundspermatid and eSP = elongated spermatid. Scale bar = 100 µm. (B) Counts of different germ cell types per seminiferous tubule presented as a per-centage of control. Values represent mean ± SD (n = 3). *p < 0.001 (T-test). (C), (E) & (G) Representative images of control animals at 1, 4 and 7 weekpost-treatment. ● = lumen. Scale bar = 50 µm. (D), (F) & (H) Representative images of treated animals at 1, 4 and 7 week post-treatment. Disorga-nised seminiferous tubule with spermatogonia and spermatocytes ↑, elongated spermatids », lumen ●, and seminiferous tubule vacuolisation →.Scale bars = 50 µm.



Publ

ishe

d on

01

June

201

6. D

ownl

oade

d by

RSC

Int

erna

l on

14/0

6/20

18 0

8:47

:15.

View Article Online


Finally, significantly DMRs associated with known genesand significantly differentially expressed mRNA transcripts(p ≤ 0.001) were compared. This identified 6 inverse corre-lations, 3 of which (Hook1, Spag16 and Stam) were identified inthe literature as being involved in testicular and reproductivefunction. All 3 displayed a general increase in DNA methylationand decrease in mRNA expression over time (ESI Table 3†).

In summary, treatment with DOX induced significantchanges in mRNA and miRNA expression and DNA methyl-ation that were associated with biological pathways relevant tothe observed testicular toxicity.

DOX induced transcriptomic and epigenetic changes inpathways associated with developmental and reproductivesystem disease

The most statistically significant transcriptional and epigeneticchanges, corresponding to 96 mRNAs, 75 miRNAs and99 gene-associated DMRs (p ≤ 0.001) were selected for morefocussed functional analyses using IPA software. This revealedroles for altered transcripts and miRNAs in various molecularand cellular functions including three major networks con-cerning cancer and cellular, tissue, organ and embryonicdevelopment, developmental disorders and reproductivesystem disease. Of the 36 canonical pathways identified, thetop two were involved in oxidative stress related pathways suchas glutathione-mediated detoxification and NRF-2 mediatedoxidative stress response. Furthermore, 15 of the altered tran-scripts were also particularly relevant for the proper develop-ment and function of the reproductive system and thesuccessful progression of spermatogenesis (Table 2); 9 of the12 altered miRNAs important in reproductive system abnorm-alities (let-7 family, and miRs-141-5p, 145, 190b, 767, 539-5p,339-5p, 29b and 9-3p) have also been shown to be altered inmen with non-obstructive azoospermia; and 4 of the DMRs(Cdkn2a, Foxm1, Vezt and Zfx) were associated with genesinvolved in reproductive system development and disease.

In summary, treatment with DOX induced transcriptomicand epigenetic changes in pathways associated with develop-mental disorders and reproductive system disease.

Discussion

Spermatogenesis is a complex developmental process thatinvolves proliferation and differentiation of diploid spermato-gonia to haploid spermatozoa, with the support of endocrinesecretion and other cellular signals from Leydig and Sertolicells (reviewed in ref. 39–42). Compound-induced damage toany of the testicular cell types could affect the production ofhealthy spermatozoa and therefore impede fertility and poten-tially adversely affect the resulting embryo. Previous studies inrodents have demonstrated that treatment with DOX causesdamage to the male reproductive system with further impli-cations in fertility and developmental processes. For example,DOX has been shown to induce decreased sperm counts withaltered sperm morphology, germ cell death, vacuolated semini-

Fig. 3 Heatmap of the top 100 most significantly altered mRNA tran-scripts (p ≤ 0.001) over time following treatment with DOX. Data rep-resent forward and reverse labelling (n = 3). Expression levels representlog2[DOX/Control] (Log2 R), with downregulations shown in green andupregulations in red.



Publ

ishe

d on

01

June

201

6. D

ownl

oade

d by

RSC

Int

erna

l on

14/0

6/20

18 0

8:47

:15.

View Article Online


Fig. 4 Heatmap of the top 100 most significantly altered miRNAs (p ≤ 0.001) over time following treatment with DOX. Data represent forward andreverse labelling (n = 6). Expression levels represent log2[DOX/Control] (Log2 R), with downregulations shown in green and upregulations in red.



Publ

ishe

d on

01

June

201

6. D

ownl

oade

d by

RSC

Int

erna

l on

14/0

6/20

18 0

8:47

:15.

View Article Online


ferous tubules, infertility and damage to the developingembryo.28,29,43–45 To date, limited data is available on studiescomparing genome-wide toxicogenomic data (transcriptional,miRNA or DNA methylation changes) with pathological altera-tions in the testis following DOX exposure. The present studywas therefore conducted to link molecular transcriptomic andepigenetic changes in the testes following DOX treatment withphenotypic outcome, thereby exploring novel mechanisms ofDOX-induced testicular toxicity and highlighting those forfurther focussed study.

In keeping with the current literature, treatment with DOXinduced organ atrophy, reduced sperm counts and long termdamage to the seminiferous tubules within the testes(Fig. 1–2). Even 7 week post-treatment, when enough time hadelapsed since DOX exposure for a complete cycle of spermato-genesis (35 days in the mouse), only 3–8% of seminiferoustubules showed repopulation with differentiating spermatogo-nia and spermatocytes. This does suggest, however, that somespermatogonial stem cells survive and eventually regain theirdifferentiating capacity. Similar long term murine DOX-induced testicular toxicity has previously been reported, where

the rate of fertility was decreased up to 32 weeks post-treat-ment, and sperm counts or morphology did not return tonormal until 1 year post DOX-treatment.28

Atrophy of the testes and caudal epididymides correlatedwith, and thus appeared to be largely due to, testicular germcell loss and decreased sperm production. Furthermore, germcells at the earlier stages of spermatogenesis (spermatogonia,spermatocytes and round spermatids) appeared to be moresensitive to DOX exposure than those in the latter stages(elongated spermatids). In contrast, Leydig and Sertoli cellsshowed greater resistance to DOX exposure. While there wasno evidence of Sertoli cell toxicity, there did appear to be someLeydig cell hyperplasia, particularly at 4 and 7 week post-treat-ment. Such hyperplasia has also been reported in rats follow-ing treatment with other testicular toxicants such as di-(2-ethylhexyl) phthalate,46 and may represent a compensatoryresponse to the loss of germ cells.

Differing sensitivities to DOX, and thus the ultimate toxici-ties observed following DOX treatment, are likely to be thecombined result of three main factors; (1) the presence ofphysiological barriers such as the blood-testes barrier (BTB),the blood-epididymis barrier (BEB) and the acrosomal cap, (2)the specific expression of efflux pump transporters on thesurface of different germ/testicular cell types, and (3) the pro-liferative/differentiation status of different germ/testicular celltypes. The BTB separates the seminiferous tubule into basaland adluminal compartments and prevents the accumulationof xenobiotics in the latter. Spermatogenesis begins with divi-sion of the testicular stem cells (spermatogonia) within thebasal layer. As the spermatogonia divide and differentiate intospermatocytes and spermatids they move through the BTB intothe protected adluminal compartment. Thus, those spermato-cytes and spermatids within the adluminal compartment ofthe testes, and the spermatozoa within the BEB of the caudalepididymides, are provided greater protection from circulatingDOX. Formation of the acrosomal cap during spermatid differ-entiation provides another physiological barrier that mayprotect the elongated spermatids and spermatozoa from DOX-induced toxicity. In addition, different germ/testicular cellshave subtle variations in the type and level of expression oftransporters such as efflux pumps.47–49 The type and numberof such pumps within each of the different cell types will there-fore also contribute to the final intracellular levels of DOX.Finally, the proliferative/differentiation status of each germ/tes-ticular cell type varies. Unlike spermatids, spermatogonia andspermatocytes undergo active proliferation and DNA synthesisin addition to differentiation. A predominant mechanism ofaction/toxicity of DOX is DNA intercalation and inhibition ofDNA synthesis (hence its use as an anti-cancer agent). DOXcan therefore impair the mitotic activity of, and DNA synthesiswithin, spermatogonia and spermatocytes, thus preventing thedownstream progression of spermatogenesis and inducing celldeath. In contrast, Leydig and Sertoli cells are terminallydifferentiated and do not normally undergo mitotic division inthe sexually mature testis, thereby limiting DOX intercalationwith DNA and reducing any cytotoxic effects.

Table 1 qRT-PCR validation of mRNAs and miRNAs of interest

mRNAs/miRNAs were selected due to either significant differentialexpression (p ≤ 0.001) at more than one time point and from literaturesearches identifying their role in male infertility and tissue atrophy,and classified into groups based on their biological and physiologicalroles. Values represent mean log2[DOX/Control] (n ≥ 3), with down-regulations shown in green and upregulations in red. *p < 0.05,**p < 0.01 and *** p < 0.001.



Publ

ishe

d on

01

June

201

6. D

ownl

oade

d by

RSC

Int

erna

l on

14/0

6/20

18 0

8:47

:15.

View Article Online


The subsequent genome-wide analysis of transcriptionaland epigenetic changes induced by treatment with DOX com-bined with integrative bioinformatic analyses proved effective

for providing further insights into molecular mechanisms ofthe DOX-induced testicular toxicity described above. In par-ticular, DOX exposure altered the expression of mRNAs and

Fig. 5 (A) Heatmap of the top 126 most significantly DMRs (p ≤ 0.001) over time following treatment with DOX. Data represent forward and reverselabelling (n = 3). Expression levels represent peak score ratio (peak score DOX–peak score control) (Peak score R), with decreased DNA methylationshown in green and increased DNA methylation in red. (B) The percentage of the top 126 DMRs associated with known genes (either within the pro-moter region or gene body) or other CpGs not associated with a known gene. (C) The chromosomal distribution of the top 126 DMRs.



Publ

ishe

d on

01

June

201

6. D

ownl

oade

d by

RSC

Int

erna

l on

14/0

6/20

18 0

8:47

:15.

View Article Online


miRNAs involved in testicular function, stress/cell death andsurvival, and DNA methylation (Table 1).

Transcripts expressed in germ cells during different stagesof spermatogenesis (Catsper1, Catsper3, Crem, Hspa2, Nme5,Stra8 and Tnp2) were all downregulated following DOXexposure. Such changes, particularly those that occur withinthe earlier stages of spermatogenesis (in spermatogonia, sper-matocytes and round spermatids), are likely to be associatedwith the loss of germ cells and long term infertility resultingfrom DOX-induced testicular toxicity. In contrast, the 2 tran-scripts associated with hormonal functions within the testis(Cyp17a1 and Insl3) were upregulated. Likewise, 2 additionaltranscripts with roles in steroid biosynthesis (Fads1 andFads2), also identified in the top 100 differentially expressedtranscripts post DOX exposure but not further validated, wereupregulated. Protocols using higher doses of DOX to inducetesticular toxicity, where measurements were taken within afew hours or days following treatment, have been shown toreduce serum testosterone levels.50–53 However, other studiesemploying similar treatment procedures to the one used here,with lower doses of DOX and a delay between DOX exposureand effect measurement, have reported no change in serumtestosterone levels during DOX-induced testicular toxicity(hence no serum measurements were performed here).10,54,55

The increased expression of transcripts involved in steroid andandrogen biosynthesis could thus be a compensatory mechan-ism to maintain testicular hormone homeostasis and/or stimu-late production of more germ cells to replace those that havebeen lost or damaged as a result of DOX-induced toxicity, and/or a simple reflection of an increase in the number of Leydigcells due to the observed Leydig cell hyperplasia. Changes inthe expression of Cyp17a1 and/or Insl3 have been reported innumerous studies investigating Leydig cell toxicity and theeffects of chemicals on testicular function.46,56–60

DOX also induced upregulation of mRNAs and miRNAsinvolved in apoptotic cell death pathways, including caspases

(Casp3 and Casp6), cathespins (Ctsb, Ctsd, Ctss and Ctsw), andmiRs 145, 26a and 29a–b. The greater upregulation of cathe-spins compared to caspases provides evidence for a redoxstress related mechanism of DOX-induced testicular cyto-toxicity. This is in keeping with the altered expression ofmRNAs involved in cellular responses to oxidative stressreported here (Gstm3 and Nme5) and previous studies on themechanisms of DOX-induced toxicity in both the testis andother organs.61–65 In addition, miRs-145, 26a and 29a–c havealso been shown to act as pro-apoptotic regulators.34–38 It istherefore possible that induction of miR-145, miR-26a and themiR-29 family regulate DOX-induced germ cell death. This isunlikely to involve Bcl-2 repression as Bcl-2 mRNA was shownto be upregulated in the DOX samples. However, other pro-apoptotic mechanisms may be involved.

The same miRNAs (miRs-145, 26a, and 29a–c) are alsoknown to target the Dnmt family (miRwalk33). Upregulation ofthese miRNAs inversely correlated with the downregulation ofthe Dnmt family transcripts following DOX exposure. This is inagreement with a similar study in rats where members of themiR-29 family were upregulated and Dnmt transcripts weredownregulated following exposure to xenoestrogens.38 SuchmiRNA-mediated reduction of Dnmt transcripts could poten-tially lead to abnormal DNA methylation within the testis, thusDOX-induced genome-wide changes in testicular DNA methyl-ation were also analysed. A significant change in the methylomeof the testis was observed over time following treatment withDOX (Fig. 5). The highest proportions of these DOX-inducedDMRs were located on chromosome 2 and the X chromosome,whilst no DMRs were observed on the X chromosome in controlanimals over time (ESI Fig. 2†). This may indicate a particularsusceptibility of the X chromosome to aberrant methylationoccurring from DOX-induced toxicity. A large number of genesexpressed in germ/testicular cells are located on the X chromo-some and the DMRs associated with these genes tend to behypomethylated.66 In general, DOX induced increased methyl-

Table 2 Reproductive system functions/diseases associated with the top 96 most significantly differentially expressed mRNAs

Function/disease p Value Transcript(s)

Acrosome reaction 1.94 × 10−4 Atp8b3, Oaz3, Tnp2Bilateral cryptorchidism 5.63 × 10−3 Insl3Binding of Sertoli cells 8.44 × 10−3 Vcam1Degeneration of testis 2.51 × 10−2 Insl3Descent of testis 5.63 × 10−3 Insl3Development of reproductive system 1.48 × 10−4 Insl3, Meig1, Nme5, Oaz3, Prm2, Spa17, Tnp1, Tnp2, Vcam1Fertilisation 9.89 × 10−4 Atp8b3, Cyp17a1, Insl3, Lcn2, Prm2, Spa17, Tnp1, Tnp2Infertility 2.01 × 10−4 Cyp17a1, Insl3, Meig1, Oaz3, Prm2, Tnp1, Tnp2Metastatic castration-resistant prostate cancer 5.63 × 10−3 Cyp17a1Morphology of germ cells 1.40 × 10−2 Nme5, Tnp1, Tnp2Morphology of gonad 1.32 × 10−3 Akt2, Fabp9, Meig1, Nme5, Tnp1, Tnp2Penetration of zona pellucida 1.96 × 10−2 Tnp2Premature senescence of reproductive system 3.33 × 10−2 Cyp17a1Quantity of spermatocytes and spermatids 2.51 × 10−2 Insl3Sperm morphology and function 5.88 × 10−5 Atp8b3, Cyp17a1, Fabp9, Meig1, Nme5, Prm2, Spa17, Tnp1, Tnp2Spermatogenesis 1.06 × 10−5 Insl3, Meig1, Nme5, Oaz3, Prm2, Spa17, Tnp1, Tnp2

The top 96 most significantly differentially expressed mRNAs (p ≤ 0.001) were subjected to pathway analysis using IPA software, identifying 15transcripts involved in the proper development and function of the reproductive system and the successful progression of spermatogenesis.



Publ

ishe

d on

01

June

201

6. D

ownl

oade

d by

RSC

Int

erna

l on

14/0

6/20

18 0

8:47

:15.

View Article Online


ation of X-linked DMRs over time, which could lead to therepression of genes involved in testicular function and ulti-mately contribute to the observed germ cell loss.

One of the main mechanisms by which DOX has beenshown to exert cytotoxicity is via p53-mediated apoptosis,67,68

and a number of apoptosis related transcripts and miRNAshave been identified here. In keeping with this, p53 signallingwas the most significantly differentially methylated canonicalpathway following DOX treatment. Furthermore, two geneswithin this pathway that also play significant roles in sperma-togenesis (Cdkn2a and Pcna) demonstrated increased methyl-ation post DOX exposure. Increased methylation and thusrepression of genes involved in early spermatogenesis could beone of the key factors driving the germ cell loss following DOXtreatment.

Comparison of the DNA methylation and transcriptionalprofiles revealed some correlation between methylation statusand transcript expression following treatment with DOX. Allshowed increased methylation with a corresponding decreasedlevel of expression over time post DOX exposure (ESI Table 3†).Amongst those identified, Hook1, Spag16 and Stam are knownto play significant roles in testicular and reproductive function.Such changes, particularly Hook1 and Spag16, may also con-tribute to DOX-induced germ cell loss. It is worth noting,however, that less than 2% of the gene-associated DMRs corre-lated with the differentially altered transcripts. This couldreflect the complexities of correlating large genome-wide datasets performed on different platforms, or suggest that DNAmethylation is not a predominant regulator of gene expressionduring DOX-induced testicular toxicity. Indeed, a recent studyin vitro has shown that genome-wide changes in DNA methyl-ation can have limited transcriptional consequences.69 Thus,DNA methylation (or miRNA expression) changes may notalways be mechanistically involved in toxicity, instead they maysimply act as markers of exposure/toxicity. Both causal andconsequential epigenetic markers could be of potential benefitin the clinic. For example, the 9 miRNAs altered by DOX thathave also been shown to be differentially expressed in menwith non-obstructive azoospermia may represent useful novelclinical markers of testicular/reproductive toxicity.

A number of the testicular transcriptional and epigeneticchanges following DOX treatment were also associated withdevelopmental and reproductive system disease. This is inkeeping with the observed adverse developmental phenotypesin the offspring of male mice treated with DOX.28–30 and thewider literature demonstrating that epigenetic changesinduced by environmental factors and pharmaceutical drugscan cause harmful outcomes in future progeny via the malegermline.18–20,27 This could be of particular concern for malecancer survivors that want to go on to have children. Treatmentmay not only affect their own reproductive function, but couldpotentially also have detrimental effects on the development oftheir future offspring.

The systems biology approach employed in this study hasenabled the successful correlation of multiple datasets toexplore the associations between epigenetic processes and

toxic end points (Fig. 6). Such genome-wide analysis is challen-ging and highlights the complexity involved in the regulationof gene expression. Nevertheless, potential epigenetic mecha-nisms and/or markers of DOX-induced testicular toxicity wereidentified from the correlation of large datasets. Furtherinsight into which of these are the key players in DOX-inducedtesticular toxicity could begin with Western blot validation ofmRNA changes of interest and/or validation of the DNAmethylation status of genes demonstrating a negative corre-lation between methylation status and mRNA expression,before moving onto focussed mechanistic studies in appropri-ate in vitro models. Such applied work is required to determinerelevance to humans and fully assess whether single and/ormultigenerational epigenetic toxicity is a real public healthconcern.

Conclusions

Perturbation of epigenetic processes is a concern, particularlyin relation to reproductive toxicity and early development,

Fig. 6 Correlation of multiple genome-wide datasets. Gene expressionis regulated by multiple mechanisms at multiple levels, including DNAmethylation (transcriptional level) and miRNAs (post-transcriptionallevel). DNA methylation generally represses the transcription of genesand miRNAs; whilst miRNAs induce mRNA degradation or inhibit mRNAtranslation. Analysis of genome-wide changes in miRNA expression (themiRome), DNA methylation (the methylome) and mRNA levels (the tran-scriptome) identified potential epigenetic mechanisms and/or markersof DOX-induced testicular toxicity. DOX induced changes in themiRome, methylome and transcriptome that appeared to be related tothe observed testicular toxicity (germ cell loss) and warrant furtherinvestigation.



Publ

ishe

d on

01

June

201

6. D

ownl

oade

d by

RSC

Int

erna

l on

14/0

6/20

18 0

8:47

:15.

View Article Online


since these are particularly vulnerable epigenetic windows andchanges could affect future generations. Here we have shownthat DOX is able to induce transcriptional, miRNA and DNAmethylation changes, which perturb pathways involved instress/cell death and survival and testicular function and leadto germ cell loss and reproductive organ damage. There aremany levels of control involved in gene expression. Consider-ing them together in a more systems based approach is criticalfor understanding the complex mechanisms involved in com-pound-induced toxicity, and for informing further focusedstudies. Like many others we therefore propose that toxicitytesting should include an epigenetic element. Further elucida-tion of relevant epigenetic mechanisms are required in orderfor such tests to be developed and incorporated into regulatorytoxicology.

Conflicts of interest

The authors have no conflicts of interest to declare.

Acknowledgements

This works was supported by Public Health England (PHE) andthe Medical Research Council (MRC) Toxicology Unit. Theauthors would also like to thank Kate Dudek for printing themicroarrays and Kelvin Lau for help using R. The viewsexpressed are those of the authors and do not necessarily reflectthose of PHE, the NHS, the Department of Health or the MRC.

References

1 NCIN, National Registry of Childhood Tumours ProgressReport, http://www.ncin.org.uk/publications/).

2 NCI, Childhood Cancer Survivor Study: An Overview, http://www.cancer.gov/types/childhood-cancers/ccss).

3 J. Bryant, J. Picot, L. Baxter, G. Levitt, I. Sullivan andA. Clegg, Clinical and cost-effectiveness of cardioprotectionagainst the toxic effects of anthracyclines given to childrenwith cancer: a systematic review, Br. J. Cancer, 2007, 96,226–230.

4 J. D. Parry, A. V. Pointon, U. Lutz, F. Teichert,J. K. Charlwood, P. H. Chan, T. J. Athersuch, E. L. Taylor,R. Singh, J. Luo, K. M. Phillips, A. Vetillard, J. J. Lyon,H. C. Keun, W. K. Lutz and T. W. Gant, Pivotal role for twoelectron reduction in 2,3-dimethoxy-1,4-naphthoquinoneand 2-methyl-1,4-naphthoquinone metabolism and kine-tics in vivo that prevents liver redox stress, Chem. Res.Toxicol., 2009, 22, 717–725.

5 A. V. Pointon, T. M. Walker, K. M. Phillips, J. Luo, J. Riley,S. D. Zhang, J. D. Parry, J. J. Lyon, E. L. Marczylo andT. W. Gant, Doxorubicin in vivo rapidly alters expressionand translation of myocardial electron transport chaingenes, leads to ATP loss and caspase 3 activation, PLoS One,2010, 5, e12733.

6 F. S. Carvalho, A. Burgeiro, R. Garcia, A. J. Moreno,R. A. Carvalho and P. J. Oliveira, Doxorubicin-induced car-diotoxicity: from bioenergetic failure and cell death to car-diomyopathy, Med. Res. Rev., 2014, 34, 106–135.

7 E. J. Chow, K. L. Stratton, W. M. Leisenring, K. C. Oeffinger,C. A. Sklar, S. S. Donaldson, J. P. Ginsberg, L. B. Kenney,J. M. Levine, L. L. Robison, M. Shnorhavorian, M. Stovall,G. T. Armstrong and D. M. Green, Pregnancy after chemo-therapy in male and female survivors of childhood cancertreated between 1970 and 1999: a report from the Child-hood Cancer Survivor Study cohort, Lancet Oncol., 2016,DOI: 10.1016/s1470-2045(16)00086-3.

8 D. M. Green, W. Liu, W. H. Kutteh, R. W. Ke, K. C. Shelton,C. A. Sklar, W. Chemaitilly, C. H. Pui, J. L. Klosky,S. L. Spunt, M. L. Metzger, D. Srivastava, K. K. Ness,L. L. Robison and M. M. Hudson, Cumulative alkylatingagent exposure and semen parameters in adult survivors ofchildhood cancer: a report from the St Jude LifetimeCohort Study, Lancet Oncol., 2014, 15, 1215–1223.

9 O. Stahl, H. A. Boyd, A. Giwercman, M. Lindholm,A. Jensen, S. K. Kjaer, H. Anderson, E. Cavallin-Stahl andL. Rylander, Risk of birth abnormalities in the offspring ofmen with a history of cancer: a cohort study using Danishand Swedish national registries, J. Natl. Cancer Inst., 2011,103, 398–406.

10 J. A. Ward, C. W. Bardin, M. Knight, J. Robinson,G. Gunsalus and I. D. Morris, Delayed effects of doxo-rubicin on spermatogenesis and endocrine function inrats, Reprod. Toxicol., 1988, 2, 117–126.

11 H. Imahie, T. Adachi, Y. Nakagawa, T. Nagasaki,T. Yamamura and M. Hori, Effects of adriamycin, an anti-cancer drug showing testicular toxicity, on fertility in malerats, J. Toxicol. Sci., 1995, 20, 183–193.

12 K. Shinoda, K. Mitsumori, K. Yasuhara, C. Uneyama,H. Onodera, M. Hirose and M. Uehara, Doxorubicininduces male germ cell apoptosis in rats, Arch. Toxicol.,1999, 73, 274–281.

13 M. Kato, S. Makino, H. Kimura, T. Ota, T. Furuhashi andY. Nagamura, Sperm motion analysis in rats treated withadriamycin and its applicability to male reproductive toxi-city studies, J. Toxicol. Sci., 2001, 26, 51–59.

14 T. Adachi, T. Nishimura, H. Imahie and T. Yamamura, Col-laborative work to evaluate toxicity on male reproductiveorgans by repeated dose studies in rats 9). Testicular toxi-city in male rats given adriamycin for two or four weeks,J. Toxicol. Sci., 2000, 25(Spec No), 95–101.

15 I. Tsunenari, M. Kawachi, T. Matsumaru and S. Katsuki, Colla-borative work to evaluate toxicity on male reproductive organsby repeated dose studies in rats 10). Testicular toxicity ofadriamycin observed 2 and 4 weeks after a single intravenousadministration, J. Toxicol. Sci., 2000, 25(Spec No), 103–115.

16 R. C. Lui, M. C. Laregina, D. R. Herbold and F. E. Johnson,Testicular cytotoxicity of intravenous doxorubicin in rats,J. Urol., 1986, 136, 940–943.

17 H. Bar-Joseph, I. Ben-Aharon, M. Tzabari, G. Tsarfaty,S. M. Stemmer and R. Shalgi, In vivo bioimaging as a novel



Publ

ishe

d on

01

June

201

6. D

ownl

oade

d by

RSC

Int

erna

l on

14/0

6/20

18 0

8:47

:15.

View Article Online


strategy to detect doxorubicin-induced damage to gonadalblood vessels, PLoS One, 2011, 6, e23492.

18 E. L. Marczylo, M. N. Jacobs and T. W. Gant, Environmen-tally-Induced Epigenetic Toxicity: Potential Public HealthConcerns, Crit. Rev. Toxicol., 2016, DOI: 10.1080/10408444.2016.1175417.

19 K. Gapp, A. Jawaid, P. Sarkies, J. Bohacek, P. Pelczar,J. Prados, L. Farinelli, E. Miska and I. M. Mansuy, Impli-cation of sperm RNAs in transgenerational inheritance ofthe effects of early trauma in mice, Nat. Neurosci., 2014, 17,667–669.

20 A. B. Rodgers, C. P. Morgan, S. L. Bronson, S. Revello andT. L. Bale, Paternal stress exposure alters sperm microRNAcontent and reprograms offspring HPA stress axis regu-lation, J. Neurosci., 2013, 33, 9003–9012.

21 A. B. Rodgers, C. P. Morgan, N. A. Leu and T. L. Bale, Trans-generational epigenetic programming via sperm microRNArecapitulates effects of paternal stress, Proc. Natl. Acad.Sci. U. S. A., 2015, 112, 13699–13704.

22 D. Miller, M. Brinkworth and D. Iles, Paternal DNA packa-ging in spermatozoa: more than the sum of its parts? DNA,histones, protamines and epigenetics, Reproduction, 2010,139, 287–301.

23 D. T. Carrell, Epigenetics of the male gamete, Fertil. Steril.,2012, 97, 267–274.

24 M. A. Surani, Human Germline: A New Research Frontier,Stem Cell Rep., 2015, 4, 955–960.

25 A. M. O’Doherty and P. A. McGettigan, Epigenetic processesin the male germline, Reprod., Fertil. Dev., 2015, 27(5), 725–738.

26 A. B. Csoka and M. Szyf, Epigenetic side-effects of commonpharmaceuticals: a potential new field in medicine andpharmacology, Med. hypotheses, 2009, 73, 770–780.

27 C. D. Glen and Y. E. Dubrova, Exposure to anticancer drugscan result in transgenerational genomic instability in mice,Proc. Natl. Acad. Sci. U. S. A., 2012, 109, 2984–2988.

28 M. L. Meistrich, L. S. Goldstein and A. J. Wyrobek, Long-term infertility and dominant lethal mutations in male micetreated with adriamycin,Mutat. Res., 1985, 152, 53–65.

29 V. Vendramini, B. Robaire and S. M. Miraglia, Amifostine-doxorubicin association causes long-term prepubertal sper-matogonia DNA damage and early developmental arrest,Hum. Reprod., 2012, 27, 2457–2466.

30 E. D. Gol’dberg, T. G. Borovskaya and M. E. Poluektova,Effects of antitumor drugs on offspring, Bull. Exp. Biol.Med., 2000, 129, 367–369.

31 J. Kang, Y. Lee, K. No, E. Jung, J. Sung, Y. Kim and S. Nam,Ginseng intestinal metabolite-I (GIM-I) reduces doxo-rubicin toxicity in the mouse testis, Reprod. Toxicol., 2002,16, 291–298.

32 S. Reagan-Shaw, M. Nihal and N. Ahmad, Dose translationfrom animal to human studies revisited, FASEB J., 2008, 22,659–661.

33 H. Dweep and N. Gretz, miRWalk2.0: a comprehensiveatlas of microRNA-target interactions, Nat. Methods, 2015,12, 697.

34 N. Kliese, P. Gobrecht, D. Pachow, N. Andrae, A. Wilisch-Neumann, E. Kirches, M. Riek-Burchardt, F. Angenstein,G. Reifenberger, M. J. Riemenschneider, E. Meese,D. Panayotova-Dimitrova, D. H. Gutmann and C. Mawrin,miRNA-145 is downregulated in atypical and anaplasticmeningiomas and negatively regulates motility and pro-liferation of meningioma cells, Oncogene, 2013, 32, 4712–4720.

35 B. Zhang, X. X. Liu, J. R. He, C. X. Zhou, M. Guo, M. He,M. F. Li, G. Q. Chen and Q. Zhao, Pathologically decreasedmiR-26a antagonizes apoptosis and facilitates carcinogen-esis by targeting MTDH and EZH2 in breast cancer, Car-cinogenesis, 2011, 32, 2–9.

36 Y. Xiong, J. H. Fang, J. P. Yun, J. Yang, Y. Zhang, W. H. Jiaand S. M. Zhuang, Effects of microRNA-29 on apoptosis,tumorigenicity, and prognosis of hepatocellular carcinoma,Hepatology, 2010, 51, 836–845.

37 J. L. Mott, S. Kobayashi, S. F. Bronk and G. J. Gores, mir-29regulates Mcl-1 protein expression and apoptosis, Onco-gene, 2007, 26, 6133–6140.

38 L. Meunier, B. Siddeek, A. Vega, N. Lakhdari, L. Inoubli,R. P. Bellon, G. Lemaire, C. Mauduit and M. Benahmed,Perinatal programming of adult rat germ cell death afterexposure to xenoestrogens: role of microRNA miR-29 familyin the down-regulation of DNA methyltransferases andMcl-1, Endocrinology, 2012, 153, 1936–1947.

39 R. A. Hess and L. Renato de Franca, Spermatogenesis andcycle of the seminiferous epithelium, Adv. Exp. Med. Biol.,2008, 636, 1–15.

40 P. J. O’Shaughnessy, Hormonal control of germ cell devel-opment and spermatogenesis, Semin. Cell Dev. Biol., 2014,29, 55–65.

41 S. Ramaswamy and G. F. Weinbauer, Endocrine control ofspermatogenesis: Role of FSH and LH/ testosterone, Sper-matogenesis, 2014, 4, e996025.

42 M. D. Griswold, Spermatogenesis: The Commitment toMeiosis, Physiol. Rev., 2016, 96, 1–17.

43 V. Vendramini, E. Sasso-Cerri and S. M. Miraglia, Amifos-tine reduces the seminiferous epithelium damage in doxo-rubicin-treated prepubertal rats without improving thefertility status, Reprod. Biol. Endocrinol., 2010, 8, 3.

44 M. L. Meistrich, M. E. van Beek, J. C. Liang, S. L. Johnsonand J. Lu, Low levels of chromosomal mutations in germcells derived from doxorubicin-treated stem spermatogoniain the mouse, Cancer Res., 1990, 50, 370–374.

45 O. Brilhante, T. Stumpp and S. M. Miraglia, Long-term tes-ticular toxicity caused by doxorubicin treatment duringpre-pubertal phase, Int. J. Med. Sci., 2011, 3, 52–60.

46 B. T. Akingbemi, R. Ge, G. R. Klinefelter, B. R. Zirkin andM. P. Hardy, Phthalate-induced Leydig cell hyperplasia isassociated with multiple endocrine disturbances, Proc.Natl. Acad. Sci. U. S. A., 2004, 101, 775–780.

47 J. Bart, H. Hollema, H. J. Groen, E. G. de Vries,N. H. Hendrikse, D. T. Sleijfer, T. D. Wegman, W. Vaalburgand W. T. van der Graaf, The distribution of drug-effluxpumps, P-gp, BCRP, MRP1 and MRP2, in the normal



Publ

ishe

d on

01

June

201

6. D

ownl

oade

d by

RSC

Int

erna

l on

14/0

6/20

18 0

8:47

:15.

View Article Online


blood-testis barrier and in primary testicular tumours,Eur. J. Cancer, 2004, 40, 2064–2070.

48 N. Melaine, M. O. Lienard, I. Dorval, C. Le Goascogne,H. Lejeune and B. Jegou, Multidrug resistance genes andp-glycoprotein in the testis of the rat, mouse, Guinea pig,and human, Biol. Reprod., 2002, 67, 1699–1707.

49 N. Melaine, A. P. Satie, J. Lassurguere, S. Desmots, B. Jegouand M. Samson, Molecular cloning of several rat ABC trans-porters including a new ABC transporter, Abcb8, and theirexpression in rat testis, Int. J. Androl., 2006, 29, 392–399.

50 S. M. Rizk, H. F. Zaki and M. A. Mina, Propolis attenuatesdoxorubicin-induced testicular toxicity in rats, Food Chem.Toxicol., 2014, 67, 176–186.

51 A. Shalizar Jalali and S. Hasanzadeh, Crataegus monogynafruit aqueous extract as a protective agent against doxo-rubicin-induced reproductive toxicity in male rats, AvicennaJ. Phytomed., 2013, 3, 159–170.

52 H. Malekinejad, H. Janbaz-Acyabar, M. Razi andS. Varasteh, Preventive and protective effects of silymarinon doxorubicin-induced testicular damages correlate withchanges in c-myc gene expression, Phytomedicine, 2012, 19,1077–1084.

53 A. O. Ceribasi, F. Sakin, G. Turk, M. Sonmez and A. Atessahin,Impact of ellagic acid on adriamycin-induced testicular histo-pathological lesions, apoptosis, lipid peroxidation and spermdamages, Exp. Toxicol. Pathol., 2012, 64, 717–724.

54 H. Takahashi, H. Tainaka, M. Umezawa, K. Takeda,H. Tanaka, Y. Nishimune and S. Oshio, Evaluation of testi-cular toxicology of doxorubicin based on microarray analy-sis of testicular specific gene expression, J. Toxicol. Sci.,2011, 36, 559–567.

55 K. Jahnukainen, T. Jahnukainen, T. T. Salmi,K. Svechnikov, S. Eksborg and O. Soder, Amifostine pro-tects against early but not late toxic effects of doxorubicinin infant rats, Cancer Res., 2001, 61, 6423–6427.

56 R. Ivell, K. Heng, H. Nicholson and R. Anand-Ivell, Briefmaternal exposure of rats to the xenobiotics dibutyl phtha-late or diethylstilbestrol alters adult-type Leydig cell develop-ment in male offspring, Asian J. Androl., 2013, 15, 261–268.

57 B. R. Hannas, C. S. Lambright, J. Furr, N. Evans,P. M. Foster, E. L. Gray and V. S. Wilson, Genomic bio-markers of phthalate-induced male reproductive develop-mental toxicity: a targeted RT-PCR array approach fordefining relative potency, Toxicol. Sci., 2012, 125, 544–557.

58 V. Kumar, A. Chakraborty, M. R. Kural and P. Roy, Altera-tion of testicular steroidogenesis and histopathology ofreproductive system in male rats treated with triclosan,Reprod. Toxicol., 2009, 27, 177–185.

59 K. Pogrmic-Majkic, S. Fa, V. Dakic, S. Kaisarevic andR. Kovacevic, Upregulation of peripubertal rat Leydig cell

steroidogenesis following 24 h in vitro and in vivo exposureto atrazine, Toxicol. Sci., 2010, 118, 52–60.

60 H. Zhang, Y. Lu, B. Luo, S. Yan, X. Guo and J. Dai, Proteo-mic analysis of mouse testis reveals perfluorooctanoic acid-induced reproductive dysfunction via direct disturbance oftesticular steroidogenic machinery, J. Proteome Res., 2014,13, 3370–3385.

61 C. Y. Lien, T. Y. Chuang, C. H. Hsu, C. L. Lin, S. E. Wang,S. J. Sheu, C. T. Chien and C. H. Wu, Oral treatment withthe herbal formula B307 alleviates cardiac toxicity in doxo-rubicin-treated mice via suppressing oxidative stress,inflammation, and apoptosis, OncoTargets Ther., 2015, 8,1193–1210.

62 G. Joshi, R. Sultana, J. Tangpong, M. P. Cole, D. K. St Clair,M. Vore, S. Estus and D. A. Butterfield, Free radicalmediated oxidative stress and toxic side effects in braininduced by the anti cancer drug adriamycin: insight intochemobrain, Free Radical Res., 2005, 39, 1147–1154.

63 S. Singla, N. R. Kumar and J. Kaur, In vivo Studies on theProtective Effect of Propolis on Doxorubicin-Induced Toxi-city in Liver of Male Rats, Toxicol. Int., 2014, 21, 191–195.

64 B. Srdjenovic, V. Milic-Torres, N. Grujic, K. Stankov,A. Djordjevic and V. Vasovic, Antioxidant properties of full-erenol C60(OH)24 in rat kidneys, testes, and lungs treatedwith doxorubicin, Toxicol. Mech. Methods, 2010, 20, 298–305.

65 J. Das, J. Ghosh, P. Manna and P. C. Sil, Taurine protectsrat testes against doxorubicin-induced oxidative stress aswell as p53, Fas and caspase 12-mediated apoptosis, AminoAcids, 2012, 42, 1839–1855.

66 R. Ikeda, H. Shiura, K. Numata, M. Sugimoto, M. Kondo,N. Mise, M. Suzuki, J. M. Greally and K. Abe, Large, malegerm cell-specific hypomethylated DNA domains withunique genomic and epigenomic features on the mouse Xchromosome, DNA Res., 2013, 20, 549–565.

67 M. Yoshida, I. Shiojima, H. Ikeda and I. Komuro, Chronicdoxorubicin cardiotoxicity is mediated by oxidative DNAdamage-ATM-p53-apoptosis pathway and attenuated bypitavastatin through the inhibition of Rac1 activity, J. Mol.Cell. Cardiol., 2009, 47, 698–705.

68 J. Huang, J. Yang, B. Maity, D. Mayuzumi and R. A. Fisher,Regulator of G protein signaling 6 mediates doxorubicin-induced ATM and p53 activation by a reactive oxygenspecies-dependent mechanism, Cancer Res., 2011, 71,6310–6319.

69 M. P. Ramos, N. A. Wijetunga, A. S. McLellan, M. Suzukiand J. M. Greally, DNA demethylation by 5-aza-2′-deoxycyti-dine is imprinted, targeted to euchromatin, and haslimited transcriptional consequences, Epigenet. chromatin,2015, 8, 11.



Publ

ishe

d on

01

June

201

6. D

ownl

oade

d by

RSC

Int

erna

l on

14/0

6/20

18 0

8:47

:15.

View Article Online


perturbation of epigenetic processes by doxorubicin in the ... uni - akinjo et al... · shown to...

Documents