Separation of somatic and germ cells is requir

ORIGINAL ARTICLE Reproductive biologySeparation of somatic and germ cellsis required to establish primatespermatogonial culturesDaniel Langenstroth 1,Nina Kossack 1,Birgit Westernstro ¨er 1,Joachim Wistuba 1,Ru ¨diger Behr 2,Jo ¨rg Gromoll 1,and Stefan Schlatt 1,*1Institute of Reproduction and Regenerative Biology,Centre of Reproductive Medicine and Andrology,Albert-Schweitzer-Campus 1,Building D11,48149Mu ¨nster,Germany 2Stem Cell Biology Unit,German Primate Center,Kellnerweg 4,37077Go ¨ttingen,Germany*Corresponding Author:E-mail:stefan.schlatt@ukmuenster.deSubmitted on January 23,2014;resubmitted on April 17,2014;accepted on May 27,2014study question:Can primate spermatogonial cultures be optimized by application of separation steps and well defined cultureconditions?summary answer:We identified the cell fraction which provides the best source for primate spermatogonia when prolonged culture is desired.what is known already:Man and marmoset show similar characteristics in regard to germ cell development and function.Several protocols for isolation and culture of human testis-derived germline stem cells have been described.Subsequent analysis revealed doubts on the germline origin of these cells and characterized them as mesenchymal stem cells or fibroblasts.Studies using marmosets as preclinical model con-firmed that the published isolation protocols did not lead to propagation of germline cells.study design,size,duration:Testicular cells derived from nine adult marmoset monkeys (Callithrix jacchus )were cultured for 1,3,6and 11days and consecutively analyzed for the presence of spermatogonia,differentiating germ cells and testicular somatic cells.participants/materials,setting,methods:Testicular tissue of nine adult marmoset monkeys was enzymatically disso-ciated and subjected to two different cell culture approaches.In the first approach all cells were kept in the same dish (non-separate culture,n ¼5).In the second approach the supernatant cells were transferred into a new dish 24h after seeding and subsequently supernatant and attached cells were cultured separately (separate culture,n ¼4).Real-time quantitative PCR and immunofluorescence were used to analyze the expression of reliable germ cell and somatic markers throughout the culture period.Germ cell transplantation assays and subsequent whole-mount analyses were performed to functionally evaluate the colonization of spermatogonial cells.main results and the role of chance:This is the first report revealing an efficient isolation and culture of putative marmoset spermatogonial stem cells with colonization ability.Our results indicate that a separation of spermatogonia from testicular somatic cells is a crucial step during cell preparation.We identified the overgrowth of more rapidly expanding somatic cells to be a major problem when establishing spermatogonial cultures.Initiating germ cell cultures from the supernatant and maintaining germ cells in suspension cultures minimized the somatic cell contamination and provided enriched germ cell fractions which displayed after 11days of culture a significantly higher expression of germ cell markers genes (DDX-4,MAGE A-4;P ,0.05)compared with separately cultured attached cells.Additionally,germ cell transplantation experiments demonstrated a significantly higher absolute number of cells with colonization ability (P ,0.001)in supernatant cells after 11days of separate culture.limitations,reasons for caution:This study presents a relevant aspect for the successful setup of spermatogonial cultures but provides limited data regarding the question of whether the long-term maintenance of spermatogonia can be achieved.Transfer of these preclin-ical data to man may require modifications of the protocol.wider implications of the findings:Spermatogonial cultures from rodents have become important and innovative tools for basic and applied research in reproductive biology and veterinary medicine.It is expected that spermatogonia-based strategies will be transformed into clinical applications for the treatment of male infertility.Our data in the marmoset monkey may be highly relevant to establish spermatogonial cultures of human testes.study funding/competing interest(s):Funding was provided by the DFG-Research Unit FOR 1041Germ Cell Potential &The Author 2014.Published by Oxford University Press on behalf of the European Society of Human Reproduction and Embryology.All rights reserved.For Permissions,please email:journals.permissions@ Human Reproduction,Vol.29,No.9pp.2018–2031,2014Advanced Access publication on June 24,2014doi:10.1093/humrep/deu157at National Taiwan Univ. Hospital on November 14, 2014/Downloaded from(SCHL394/11-2)and by the Graduate Program Cell Dynamics and Disease (CEDAD)together with the International Max Planck Research School –Molecular Biomedicine (IMPRS-MBM).The authors declare that there is no conflict of interest.trial registration number:Not applicable.Key words:spermatogonial stem cell /marmoset monkey /xenotransplantation /testicular cell culture /germ cellIntroduction Spermatogonial stem cells (SSCs)have the lowest degree of differenti-ation among germ cells in the adult testis.They are located within stem cell niches at the basal lamina of the seminiferous tubules.Niche factors provided by Sertoli cells,peritubular cells and the interstitial com-partment determine the fate decision to either differentiate into sperm-atozoa or undergo SSC self-renewal (Meng et al.,2000;Yoshida et al.,2007;Spinnler et al.,2010;Oatley et al.,2011).Thus,the pool of uni-potent SSCs represents the germ cell reserve to sustain fertility through-out adulthood.Unraveling the properties and the regulation of SSCs may reveal reasons for impaired fertility in men.SSCs and their progenitors may also be considered as targets for fertility preservation in men exposed to gonadotoxic chemo-or radiotherapies.Prior to the clinical use of SSCs,safety and feasibility of in vitro procedures need to be assessed in preclinical studies on adequate human-like model organisms (Ehmcke et al.,2006).Marmoset monkeys (Callithrix jacchus )are non-human primates which are used as model organisms in reproductive research.In respect to germ cells and the organization of the seminiferous epithelium,rodents exhibit remarkable differences while marmosets share many relevant aspects with humans.Among these are the embryonic and fetal development of male germ cells,the microenvironment and the organization of the seminiferous epithelium,the spermatogonial subtypes and the expres-sion of spermatogonial markers (Millar et al.,2000;Weinbauer et al.,2001;Wistuba et al.,2003;Luetjens et al.,2005;Ehmcke and Schlatt 2006;Mitchell et al.,2008;Albert et al.,2010;Aeckerle et al.,2012;Eildermann et al.,2012a ;McKinnell et al.,2013).Additionally,transgenic marmosets were generated providing innovative options for future studies (Sasaki et al.,2009).Since the availability of healthy human testicu-lar tissue for research is restricted,marmosets present valid,but thus far rarely used models for preclinical studies on male fertility and SSC research.During the last decades numerous studies on in vitro cultures of male germ cells were performed (for review see Reuter et al.,2012).Sperm-atogonial cultures in mouse models were established using magnetic acti-vated cell sorting and pre-plating selection strategies and have helped to elucidate many aspects of spermatogonial physiology (Kanatsu-Shinohara et al.,2003,2004,2005;Kubota et al.,2004;Ogawa et al.,2004).However,several studies which used similar strategies for the iso-lation and culture of human spermatogonia and which reported an in vitro propagation of human SSCs (Conrad et al.,2008;Golestaneh et al.,2009;Kossack et al.,2009;Sadri-Ardekani et al.,2009,2011;He et al.,2010;Mizrak et al.,2010)were called into question.Subsequent analyses revealed a somatic expression profile for some of these cultures (Koet al.,2010;Chikhovskaya et al.,2012)which illustrated that improved culture systems are needed for human spermatogonia.Similar to studies using human testicular material,studies using mar-moset testicular tissue showed that primary testicular cell cultures are dominated by rapidly expanding somatic cells (Albert et al.,2012;Eilder-mann et al.,2012b ).These findings prompted us to define appropriate invitro conditions to enrich primate spermatogonia during isolation and to promote the selective expansion of germ cells from adult marmosets.Successful establishment of spermatogonial cultures in marmosets could be the basis for efficient in vitro expansion of human spermatogonia.To validate the true nature of primate germ cells in the culture dish Kossack et al.(2013)suggested monitoring both the expression of vali-dated germ cell markers and the expression of somatic markers.For mar-moset monkeys,germ cell markers which are specific for different subpopulations of germ cells have been described.DDX-4was charac-terized as a rather general germ cell marker including differentiating germ cells up to round spermatids (Albert et al.,2010),MAGE A-4was shown to be expressed by spermatogonia and early spermatocytes (Eildermann et al.,2012b )and SALL4was detected in a subpopulation of spermato-gonia in adult marmoset testes (Eildermann et al.,2012a ).In addition markers for a rare spermatogonial subpopulation (LIN28,Aeckerle et al.,2012)and for spermatogonia and gonocytes,which are the devel-opmental progenitors of spermatogonia (LIN28,Aeckerle et al.,2012,POU5F1,TFAP2C,Albert et al.,2010),have been described in adult and newborn marmosets,respectively.To date,no validated SSC markers have been identified,therefore the only proof for the presence of SSCs in cell preparations remains the germ cell transplantation assay which was developed by Brinster et al.(1994).This assay provides an estimate of SSCs in the cell suspension used for transplantation by counting the number of colonies re-established several weeks after transfer (Kanatsu-Shinohara et al.,2003).In mouse-to-mouse and mouse-to-rat transfers recolonizing SSCs generate areas of complete spermatogenesis.In contrast,the transfer of primate SSCs into germ cell depleted mouse testes does not lead to a re-establishment of complete spermatogenesis due to a blockage of the germ cell differentiation program and therefore,the limitedly proliferating SSCs appear as small clones and groups in the recipient testes (Reis et al.,2000;Nagano et al.,2001,2002).Consequently,the analysis of SSCs in cell preparations is relying on the detection and counting of small colonies of primate spermatogonia after xenotransplantation into mouse testes.In this study,we used the marmoset as a model organism for primate germ cell culture experiments.We aimed to establish improved culture conditions for primate spermatogonia by systematically analyzing primary testicular cell cultures using reliable molecular markers for germ cells and somatic cells and using germ cell transplantations.Material and MethodsAnimals,organ retrieval and processing of testicular tissueIn total nine adult marmosets (C.jacchus )from the institutional breeding facility (Supplementary data,Table SI )were used for this study.The Establishment of marmoset spermatogonial cultures 2019 at National Taiwan Univ. Hospital on November 14, 2014/Downloaded frommarmosets were kept in pairs/families under a 12h light/12h darkness regimen and fed food pellets from Altromin (Lage,Germany)developed for marmosets,together with daily supplement of fresh fruits and vegetables.They had unlimited access to tap water.Housing and exercise conditionswere identical for all animals.Prior to organ retrieval the monkeys were anesthetized with Ketaject (equivalent to 100mg/ml ketamine per animal)and killed by decapitation.Testes were dissected and transferred to chilled Minimal Essential Medium (MEM)a (22561-021,Life Technologies GmbH,Gibco,Darmstadt,Germany)and immediately transported to the cell culture laboratory where the enzymatic dissociation was performed as re-cently described for human testicular tissue (Kossack et al.,2013).For germ cell transplantation experiments in total 48male NRMI-Foxn nu /Foxn nu mice were obtained from Janvier Labs (Saint Berthevin,France)and kept in the institutional breeding facility with pellet food and water ad libitum .At the age of 5–8weeks,nude mice were injected i.p.with 38mg/kg busulfan.Four weeks later mice were anesthetized with ketamin/xylazine (K-113,Sigma Aldrich,Steinheim,Germany)and cultured marmoset testicular cells were xenotransplanted into their testes (see below).After an additional 15–18weeks,mice were killed and testes were dissected for subsequent analyses.To control for the success of the transplantation procedure,testes of four 9-day-old EGFP-transgenic mice from the institutional colony were enzymati-cally digested,cryopreserved (freezing medium:Dulbecco’s modified Eagle’s medium +70%fetal bovine serum (FBS)+10%dimethylsulphoxide)and thawed according to established protocols (Alipoor et al.2009),before they were transplanted into depleted nude mouse testes.All procedures were conducted according to the German Federal Law on the Care and Use ofLaboratory Animals (LANUV NRW license numbers:8.87–50.10.46.09.016,87–51.04.2010.A243,8.87–50.10.46.09.018,84–02.05.20.12.0.018).Selection and validation of testicular cell type markers using immunohistochemistry Based on immunohistochemical (IHC)stainings the validity of testicular cell type-specific markers was confirmed for marmoset monkeys.IHC stainings were performed in a routine procedure as described before (Albert et al.,2012).Briefly,primary antibodies which were directed against the DEAD (Asp-Glu-Ala-Asp)box polypeptide 4(DDX-4,also VASA),melanoma antigen family A4(MAGE A-4),boule-like (BOLL),sal-like protein 4(SALL4),LIN28,octamer-binding transcription factor 4(POU5F1),transcrip-tion factor AP-2g (TFAP2C),vimentin (VIM),smooth muscle actin (SMA)and CD44epitopes (listed in Supplementary data,Table SII )were used and incu-bated overnight at 48C.Secondaryhorse radish peroxidase coupled antibodies were incubated for 1h at room temperature before diaminobenzidine was added to visualize the immunostaining.For the morphological identification of the different cell types a counterstain with hematoxylin was performed.Analysis of two testicular cell culture approaches For cell cultures 5million cells were seeded per culture dish (10cm,Sarstedt 83.1802)and cultured in MEM a medium containing 10%FBS and 1%penicil-lin/streptomycin at 358C and 5%CO 2.For one approach all cells were kept inthe same dish throughout the culture period of 3,6and 11days.Thus,we called this approach the non-separate approach.In the separate approach,the supernatant (SN)cells were transferred into a new dish after 24h of culture.Subsequently,SN and attached (AT)cells were cultured in separate dishes (see Supplementary data,Fig.S1).For both approaches one third ofthe medium was replaced every third day.All SN cells which had been removed from the dish with the old medium were resuspended in fresh medium and retransferred to the culture dish.For subsequent analyses SN and AT cells (detached with 0.25%trypsin,25200-56,Gibco)were collected separately,counted and snap frozen in liquid nitrogen.Additionally,some cells were used for germ cell transplant-ation experiments,as outlined below.Immunofluorescence analysis of testicularcell culturesFor immunofluorescence staining,testicular cells were cultured on 8-well culture slides (50000cells/well,354108,Becton Dickinson GmbH,Germany).On Day 6and 11,cells were washed with phosphate-buffered saline (PBS)before they were fixed with 4%paraformaldehyde for 15min at room temperature.Afterwards the cells were washed twice with PBS and stored at 48C until stainings were performed.For blocking,the cells were incubated for 30min in PBS containing 25%goat serum and 5%bovine serum albumin.Subsequently,primary antibodies were diluted in blocking medium (Supplementary data,Table SII )and incubated with the cells for 1h at room temperature.Incubation with non-specific IgG served as a negative control.Afterwards the cells were washed 3times with PBS and incubated for 45min at room temperature with secondary,fluorescence-labeled antibodies (Goat anti-mouse IgG Alexa488,A11001;Goat anti rabbit IgG Alexa546,A11035,Life Technologies GmbH,Invitro-gen)which were diluted 1:100in blocking medium.To remove residual anti-bodies the cells were washed twice with PBS before they were incubated with Hoechst 33258(0.01mg/ml in PBS;861405,Sigma Aldrich)for 30min at room temperature.After a final washing step the cells were mounted with Faramount Aqueous Mounting Medium (S3025,Dako,Denmark).Real-time quantitative PCR analysis of testicular cell culture fractionsRNA purification was carriedout from frozencellswith the RNeasy w PlusMicro Kit (74004,Qiagen,Germany).Subsequent cDNA synthesis was performedwith 200ng RNA using the iScript TM cDNA Synthesis Kit (170-8890,Bio-Rad Laboratories GmbH,Germany)according to the manufacturer’s instructions.Afterwards cDNA was diluted 1:2with Nuclease-Free Water (129115,Qiagen,Germany).The SYBR green-based real-time quantitative PCR (qPCR)analyses were performed in reaction volumes of 15m l using 1.5m l cDNA,1.5–4.5pmol of the respective primers and 7.5m l of Power SYBR Green Master Mix (4367659,Life Technologies GmbH,Applied Biosys-tems).The primers are listed in Supplementary data,Table SIII .Primers were designed specific to marmoset genome sequences being available online [/gene/?term=Callithrix (11June 2014,date last accessed)].When genes were not annotated a trace archive search [/Blast.cgi?PROGRAM ¼blastn&BLAST_SPEC ¼Trace Archive&PAGE_TYPE ¼BlastSearch&PROG_DEFAULTS ¼on (11June 2014,date last accessed)]and a subsequent alignment with the respective homolo-gous human and mouse sequences was performed to identify the mostsimilar marmoset genome sequence.Primer design and optimization of theqPCR assays were performed according to the instructions of the SYBR Select Master Mix User Guide (Life Technologies GmbH,Applied Biosys-tems).qPCRs were run on the StepOnePlus TM with the following program:One cycle of 958C for 10min,40cycles of 958C for 15s and 608C for 1min.Analyses of qPCR data were performed with the StepOne TM software2.2(Life Technologies GmbH,Applied Biosystems).For normalizing the results the constitutively expressed TOP1(DNA topoisomerase I)gene was used as reference.Results are shown as 22D Ct values according to Schmittgen and Livak (Livak and Schmittgen,2001;Schmittgen and Livak,2008).Xenotransplantation of cultured cells into nude miceIt has been shown that post xenotransplantation primate spermatogonia are able to colonize seminiferous tubules of mice and that this approach can be used as a tool to determine numbers of primate spermatogonia 2020Langenstroth et al. at National Taiwan Univ. Hospital on November 14, 2014/Downloaded fromwith colonization ability(Reis et al.,2000;Nagano et al.,2002;Hermann et al.,2007).In this study germ cell transplantation experiments were per-formed to determine the number of spermatogonia with colonization po-tential in the different culture fractions.For this,single-cell suspensions from the enzymatic dissociation of full adult marmoset testes(0days of culture)and cells from the separately cultured SN fraction(SN|SN)as well as from the trypsinized AT cells which had remained in the original dish after transfer of the SN cells on Day1(AT|AT)were used after11 days of culture.From each fraction10m l with100000marmoset cells and10%Trypan blue were injected into the seminiferous tubules via the efferent duct under pressure(100–400hPa,FemtoJet,Eppendorf, Hamburg,Germany)as previously described(Nettersheim et al.,2012). This procedure was repeated2–5times per fraction of each culture experiment.All in all seminiferous tubules of35nude mice were infused with marmoset cells.Testicular cells from9days old EGFP-mice were injected in the same way as technical positive controls(Kanatsu-Shinohara et al.,2003).At15–18weeks post transplantation,nude mouse testes were retrieved.One part of the testis was snap frozen in liquid nitrogen for marmoset-specific qualitative PCR analyses and the rest wasfixed for wholemount analyses.Wholemount analysis of nude mouse testes transplanted with cultured marmoset cells Testes were collected in ice-cold PBS,the tunica albuginea was removed and seminiferous tubules were gently separated withfine tweezers beforetheFigure1Morphology and dynamics of cell growth differ in the supernatant and the attached fractions of adult marmoset testicular cell cultures. Representative images of adult marmoset testicular cells after1,3,6and11days of culture using the non-separate(A–D,n¼3)and the separateculture approach(E–K,n¼4).Cell numbers in the respective cell fractions of non-separate(L)and separate cultures(M).SN:supernatant fraction;AT:attached fraction of non-separate cultures;SN|SN:SN fraction in the dish with SN cells after transfer;AT|AT:AT fraction in the dish with AT cellsafter transfer;h:hours.Values are means+SEM;scale bars represent50m m.Establishment of marmoset spermatogonial cultures2021at National Taiwan Univ. Hospital on November 14, 2014/Downloaded fromtubules werefixed as wholemounts in4%PFA for2h at room temperature. Afterwards the tubules were washed3times with PBS before they were dehydrated in ethanol(25%,50%,75%,95%,2×100%for12min each) and stored in70%ethanol until the stainings were performed.Prior to the staining wholemounts were rehydrated(50%,25%ethanol,2×PBS; 12min each).For permeabilization tubules were treated for10–20min at 378C with1mg/ml collagenase,before they were washed with PBS and further permeabilized and blocked in PBS+10%Triton X-100+5%goat serum for1h at room temperature.The monoclonal MAGE A-4antibody (Supplementary data,Table SII)which specifically binds to primate spermato-gonia and spermatocytes,but not to murine germ cells(Van Saen et al.,2013) was diluted1:20in PBS+1%Triton X-100+0.5%goat serum and incu-bated with the tubules over night at48C.After four15min washes in PBS, the tubules were incubated at room temperature for2h with the secondary antibody(Goat anti-mouse Alexa488).Finally the tubules were washed in PBS(3×PBS+Hoechst,1×PBS only),before they were mounted on slides using Vectashield(H1000,Vector Laboratories Inc.,USA).The number of MAGE A-4positive colonies could be evaluated in33whole-mounts which contained donor cells from the different culture fractions. The number of colonies was determined blinded using afluorescence micro-scope.Since a small part of each testis was used for PCR analyses the number of MAGE A-4positive colonies for each testis was calculated as follows:number of detected MAGE A-4positive colonies x weight of the full testis/weight of testicular tissue which was used for wholemount analysis. The number of total SSCs in the culture fractions was calculated in the follow-ing way:MAGE A-4colonies per testis x total cell number in the respective culture fraction/number of transplanted cells(105cells). Qualitative PCR analysis of nude mouse testes transplanted with culturedmarmoset cellsRNA isolation was performed as described above.For MAGE A-4PCRs cDNA synthesis was performed according to the manufacturer’s instructions for the gene-specific primer protocol with SuperScript II Reverse Transcript-ase(18064-022,Invitrogen Life Technologies,Germany)using1000ng RNA and marmoset-specific MAGE A-4reverse primers.Subsequently two steps of PCR amplification were performed.Thefirst PCR was carried out with 5m l cDNA and marmoset-specific MAGE A-4forward and reverse primers.The second PCR was conducted using1.5m l of product from the first PCR,a marmoset-specific MAGE A-4nested forward primer which binds within thefirst PCR product and the marmoset-specific MAGE A-4 reverse primer.For glyceraldehyde-3-phosphate dehydrogenase(GAPDH)Figure2Somatic marker genes are expressed at high levels in attached and at low levels in supernatant cells of marmoset testicular cell cultures.(A,Dand G)Immunohistochemical detection of vimentin(VIM),a-smooth muscle actin(SMA)and CD44in adult marmoset testis sections.In corresponding immunoglobulin G controls no staining was detected(not shown).Scale bars represent100m m.(B,C,E,F,H and I)Consecutive quantification of mRNA expression for VIM,ACTA2and CD44in non-separate(n¼3)and separate(n¼4)cultures relative to the reference gene TOP1.Note that in the SN fractionof the non-separate culture approach there is somatic marker gene expression after11days of culture.Values are means+SEM;asterisks indicate signifi-cant differences on the same day(P,0.05;Mann–Whitney test).2022Langenstroth et al.at National Taiwan Univ. Hospital on November 14, 2014/Downloaded fromPCRs cDNA synthesis was performed with250ng RNA using the iScript TM cDNA Synthesis Kit according to the manufacturer’s instructions.Subse-quently one step of PCR amplification was performed using mouse GAPDH primers.Primers are listed in Supplementary data,Tables SIII and SIV.PCR products were separated in2%agarose gels,visualized by UV-light and documented photographically.Statistical analysesTo evaluate statistical differences between different groups the non-parametric unpaired Mann-Whitney test was performed.Significant differ-ences are marked with*P,0.05,***P,0.001.GraphPad Prism w Software (Version5.0,GraphPad Software,Inc.,San Diego,USA)was used for statis-tical analyses.ResultsTwo approaches for the enrichment and culture of adult marmoset germ cells Following enzymatic digestion of nine adult decapsulated marmoset testes the cells were subjected to primary cell culture for1,3,6and11days ap-plying two different isolation and culture protocols.In thefirst approach the entire cell preparation was cultured in the same dish(non-separate ap-proach).In the second approach supernatant(SN)and attached(AT)cells were separated after24h and cultured separately(separate approach) (Supplementary data,Fig.S1).We collected cells from bothapproachesFigure3Attached cells mostly express somatic markers after6and11days in vitro.(A and B)Micrographs show the combined detection of the somaticmarker VIM and the germ cell marker MAGE A-4and(C–F)of somatic markers(VIM,a SMA).Arrows indicate nuclei of adherent cells positive for somaticmarkers and negative for germ cell markers(A–F).Filled arrowheads indicate cells that are negative for somatic markers and positive for germ cell markers(A,B and G).Empty arrowheads indicate cells being negative for somatic markers(C,D).Corresponding immunoglobulin G control stainings did not showany signals(Supplementary data,Fig.S2).Scale bars represent10m m.Establishment of marmoset spermatogonial cultures2023at National Taiwan Univ. Hospital on November 14, 2014/Downloaded fromand analyzed four fractions(SN;AT;SN|SN;AT|AT)as indicated in Fig.1 and Supplementary data,Fig.S1.Following enzymatic digestion20.8(+4.1)and23.6(+1.0)million single cells were obtained per100mg testicular tissue(Supplementary data,Table SI).In both protocols only a few cells attached to the culture dish after24h(Fig.1A and E).Cells in the SN and the SN|SN frac-tions were small and round,while the AT cells were larger(≥50m m)and revealed an irregular appearance(Fig.1A–K).Throughout the culture period of11days,cells in the SN fractions of both approaches formed clusters.In both approaches the number of viable cells in the SN and the SN|SN fractions decreased while the number of AT and AT|AT cells increased during11days of culture(Fig.1L and M). Detection of somatic markers in testicular tissue sections and cell culturesA detailed consecutive analysis at Days3,6and11of somatic and germ cell marker gene expression was used to identify different testicular cell types in the different culture fractions.For IHC,markers specific for Sertoli cells(VIM),peritubular cells(VIM,Fig.2A and a SMA,Fig.2D) and mesenchymal interstitial cells(CD44,Fig.2G)were established. These markers were also analyzed at the mRNA level(VIM,ACTA2 (gene for a SMA),CD44)in all four fractions at all three time points (Fig.2).In general adherent cell fractions showed intense expression of somatic marker genes.Floating cell fractions revealed low expression of somatic marker genes.Both approaches showed quite similar dynamics except at11days of culture when all three somatic cell marker genes became clearly detectable in the SN fraction of the non-separate approach. Identification of somatic cells byimmunofluorescence analysesImmunofluorescence stainings of non-separate cultures showed that AT cells with nuclei of.10m m diameter either expressed VIM or a SMA and were negative for germ cell markers(Fig.3A–F).In contrast,AT cells with smaller nuclei were immunopositive for MAGE A-4and were adjacent to cells expressing somatic markers(Fig.3A).Additionally, isolated cells with small nuclei and small clusters of cells,which were located on top of the somatic cells,were negative for somatic markers (Fig.3C and D),but positive for germ cell markers(Fig.3B). Detection of germ cell markers in testicular tissue sections and cell culturesMarkers specific for spermatogonia as well as differentiating germ cells (DDX-4,Fig4A),spermatogonia and spermatocytes(MAGE A-4 Fig.4D)and a subpopulation of spermatogonia(SALL4,Fig.4G)were used for IHC and were specific for the respective cellstypes.Figure4Germ cell marker gene expression dynamics differ in supernatant and attached cells.(A,D and G)Immunohistochemical detection of DDX-4,MAGE A-4and SALL4in adult marmoset testicular tissue sections.In corresponding immunoglobulin G controls no staining was detected(not shown).Scale bars represent100m m.(B,C,E,F,H,I)Consecutive quantification of DDX-4,MAGE A-4and SALL4expression in non-separate(n¼3)and separate(n¼4)cultures relative to the reference gene TOP1.Note that after11days of culture the expression of germ cell marker genes(DDX-4,MAGE A-4)is higherin the supernatant than in the attached fractions of the separate culture approach and thatthe expression dynamics are similar in both approaches.Values aremeans+SEM;asterisks indicate significant differences on the same day(P,0.05;Mann–Whitney test).2024Langenstroth et al.at National Taiwan Univ. Hospital on November 14, 2014/Downloaded from。