dear author, please note that changes made in the …...patients with sertoli cell only (sco)...

Dear author,

Please note that changes made in the online proofing system will be added to the article before publication but are not reflected in this PDF.

We also ask that this file not be used for submitting corrections.

Characterization of the stem cellniche components within theseminiferous tubules in testicularbiopsies of Klinefelter patients

Q4 Dorien Van Saen, Ph.D.,a Veerle Vloeberghs, M.D.,b Inge Gies, Ph.D.,c Jean De Schepper, Ph.D.,a,c,d

Herman Tournaye, Ph.D.,a,b,e and Ellen Goossens, Ph.D.a

a Biology of the Testis, Laboratory for Reproduction, Genetics and Regenerative Medicine, Vrije Universiteit Brussel (VUB),Brussels; b Centre for Reproductive Medicine, Universitair Ziekenhuis Brussel (UZ Brussel), Brussels; c Department ofPediatrics, Division of Pediatric Endocrinology, Universitair Ziekenhuis Brussel, Brussels; d Pediatric Endocrinology,Universitair Ziekenhuis Gent, Gent; e Department of Surgical and Clinical Sciences, Vrije Universiteit Brussel (VUB),Brussels, Belgium

Objective: To characterize the tubular environment in testicular biopsy tissues from patients with Klinefelter syndrome (KS).Design: Observational immunohistochemical study.Setting: Academic research unit.Patient(s): Males with KS and controls at different developmental time points: fetal, prepubertal, peripubertal, and adult.Intervention(s): Immunohistochemical analysis of testicular biopsies samples to characterize maturation of Sertoli cells and tubularwall components—peritubular myoid cells (PTMC) and extracellular matrix (ECM) proteins.Main Outcome Measure(s): Intensity of antim€ullerian hormone staining; proportion of Sertoli cells expressing androgen receptor(AR); and expression of tubular wall markers as characterized by identifying abnormal staining patterns.Result(s): Decreased expression for alpha smooth muscle actin 2 (ACTA2) was observed in peripubertal and adult KS as well as in Ser-toli cell only (SCO) patients. Altered expression patterns for all ECM proteins were observed in SCO and KS biopsy tissues compared withcontrols. Only for collagen I and IV were altered expression patterns observed between KS and SCO patients. In peripubertal samples, nostatistically significant differences were observed in the maturation markers, but altered ECM patterns were already present in somesamples.Conclusion(s): The role of loss of ACTA2 expression in PTMC in the disintegration of tubules in KS patients should be further inves-tigated. Future research is necessary to identify the causes of testicular fibrosis in KS patients. If the mechanism behind this fibroticprocess could be identified, this process might be altered toward increasing the chances of fertility in KS patients. (Fertil Steril�2020;-:-–-. �2020 by American Society for Reproductive Medicine.)Key Words: Extracellular matrix, infertility, Klinefelter syndrome, peritubular myoid cells, tubular integrity

Discuss: You can discuss this article with its authors and other readers at https://www.fertstertdialog.com/users/16110-fertility-and-sterility/posts/58399-29126

K linefelter syndrome (KS), themost frequent cause ofnonobstructive azoospermia, is

characterized by early germ cell loss.The condition is already initiated dur-ing early childhood (1), and may begin

as early as fetal life as suggested by arecent study (2). However, the generalassumption that these men arecompletely sterile is outdated; testicularsperm extraction (TESE) in combina-tion with intracytoplasmic sperm injec-tion (ICSI) in KS patients has been usedfor treatment since the late 1990s (3–8).

Focal spermatogenesis can befound in testicular biopsy samples,and the sperm retrieval rate has beenabout 40% in KS patients undergoingTESE (9). The origin of these foci ofspermatogenesis is still a matter ofdebate. Two contradictory hypotheses

Received October 9, 2019; revised January 9, 2020; accepted January 13, 2020.D.V. received grants from Research Fund of Flanders (FWO) during the conduct of the study. V.V. has

nothing to disclose. I.G. has nothing to disclose. J.D. has nothing to disclose. H.T. received grantsfrom Research Fund of Flanders (FWO) and Ferring during the conduct of the study; grants fromRoche, MSD, Cook, Gedeon Richter, Abbott, Obseva, and Theramex; and consultant fees fromAbbott, Obseva, and Gedeon Richter outside the submitted work. E.G. has nothing to disclose.

Supported by the scientific FundWilly Gepts from the UZ Brussel (D.V., J.D.) and grants from the VrijeUniversiteit Brussel (E.G.). D.V. is a postdoctoral fellow of theQ3 Fonds Wetenschappelijk Onder-zoek (12M2819N).

Reprint requests: Dorien Van Saen, Ph.D., Biology of the Testis, Laboratory for Reproduction, Geneticsand Regenerative Medicine, Vrije Universiteit Brussel (VUB), Brussels, Belgium (E-mail: [email protected]).

Fertility and Sterility® Vol. -, No. -, - 2020 0015-0282/$36.00Copyright ©2020 American Society for Reproductive Medicine, Published by Elsevier Inc.https://doi.org/10.1016/j.fertnstert.2020.01.018

VOL. - NO. - / - 2020 1

ORIGINAL ARTICLE: ANDROLOGY

FLA 5.6.0 DTD � FNS32324_proof � 4 March 2020 � 3:29 am � ce E

123456789

1011121314151617181920212223242526272829303132333435363738394041424344454647484950515253545556575859

60616263646566676869707172737475767778798081828384858687888990919293949596979899

100101102103104105106107108109110111112113114115116117118

https://www.fertstertdialog.com/users/16110-fertility-and-sterility/posts/58399-29126

https://www.fertstertdialog.com/users/16110-fertility-and-sterility/posts/58399-29126

mailto:[email protected]

mailto:[email protected]

https://doi.org/10.1016/j.fertnstert.2020.01.018

are related to the fact that 47,XXY spermatogonia can com-plete meiosis (10–12) or only arises from rarespermatogonia having or acquiring a normal karyotype,thus suggesting testicular mosaicism (13–15). Because ICSIwith testicular sperm from KS patients has resulted in thebirth of children with a normal karyotype (16, 17), moreand more support has accumulated for the secondhypothesis, but it cannot explain the increased rate ofaneuploidy in KS patients compared with controls (18). Thehigher rate of aneuploidy in KS sperm could result from theoccurrence of meiotic errors due to a deficient testicularenvironment.

Spermatogenesis is tightly regulated by factors producedby the stem cell niche that define both spermatogonialrenewal and differentiation. The germ cells are anchored be-tween Sertoli cells and are surrounded by the lamina propria(LP). The LP consists of a basement membrane (BM)composed of laminin and collagen I and IV; external tothis are three to five layers of peritubular myoid cells(PTMC) surrounded by two layers of fibroblasts. The PTMCsproduce several extracellular matrix (ECM) proteins such asfibronectin, collagen I, collagen IV, and laminin, which aredeposited between the different layers of PTMCs and fibro-blasts (19, 20) The PTMCs have a smooth muscle phenotypeand thus intensively express alpha smooth muscle actin 2(ACTA2), whereas fibroblasts do not express smooth musclecharacteristics (21).

Germ cell loss in KS patients is followed by hyalinizationof the seminiferous tubules, already starting in adolescence.The mechanism of this hyalinization process in KS is notknown. A variable degree of hyalinization of the tubuleshas been observed in adolescent and adult KS testes, fromthe presence of normal to degenerating tubules between areaswith hyalinized tubules, up to a total loss of tubular structures(1). It is however not clear whether initiation of this destruc-tive process is initiated from inside the seminiferous tubulesor caused by factors in the interstitial area.

It is common knowledge that the tubular wall isfrequently altered in tubules with impaired spermatogenesis.The tubular wall in azoospermic patients is often thickened,characterized by accumulated deposits of ECM (20). Despitethe thickened lamina propria, the tubular structure is mostlystill recognizable. However, in hyalinized tubules, Sertoli cellsand germ cells are completely lost and replaced by fibrocytesand collagen fibers. This is also associated with a disorganizedlamina propria, for which the thickness is increased by depo-sition of membrane-like material (22). Testicular fibrosis isnot common for all azoospermic patients, in addition to KSpatients, patients with undescended testes and severe oligo-zoospermia or azoospermia may also show associated tubularhyalinization (22).

To gain more insight in the integrity of the remainingseminiferous tubules in KS biopsy samples we characterizedthe different cellular structures within seminiferous tubulesin KS biopsied tissues at different time points of developmentand compared them with other pathologic conditions associ-ated to primary testicular dysfunction as well as with normalcontrols. Because PTMC act in cooperation with Sertoli cells,the maturation status of Sertoli cells was studied in parallel.

MATERIALS AND METHODSOur study was approved by the internal review board of theUZ Brussel (B.U.N. 143201524312).

Tissue Sampling

Adult testicular tissue from non-mosaic KS men (n ¼ 27) andpatients with Sertoli cell only (SCO) syndrome (n ¼ 21) wascollected during diagnostic TESE procedures as part of theirfertility treatment. The TESE was performed under generalanesthesia by an open-excisional technique (23). Duringthis TESE procedure, a small biopsy sample was sent to thepathology department for histologic analysis, and theparaffin-embedded tissue was obtained from the pathologydepartment. After written informed consent, testicular tissuefrom the controls (C) was collected from patients who wereundergoing vasectomy reversal (n ¼ 7). During surgery, asmall biopsy sample was collected and immediately broughtto our laboratory. The testicular tissue was fixed in acidifiedalcoholic formalin (AFA; 81009.362; VWR) for maximum 2hours and then transported to the pathology department forfurther dehydration and fixation, using the TISSUE TEK-VIP (Sakura), and paraffin embedding.

Peripubertal (n ¼ 16, aged 12–18 years) KS testicular tis-sue was obtained from patients participating in our testicularcryopreservation program, which was started in 2009 on aresearch basis (24, 25). All patients included had a non-mosaic 47,XXY karyotype except for one who had a48,XXYY karyotype. In these patients, written informed con-sent was obtained from both the parents and teenager to use10% of the taken biopsy for research purposes. This small bi-opsy sample was fixed as described before for adult controltissue.

In addition, cryopreserved testicular tissue from fournon-mosaic 47,XXY prepubertal boys (aged 4–7 years) wasstudied. Two boys had a prenatal diagnosis (advancedmother’s age) while in two others, karyotyping was performedin the neonatal period because of ICSI follow-up or neonatalhypotonia. Control testicular samples from boys undergoingtesticular tissue preservation before the start of a gonadotoxiccancer therapy were available for comparison. Eight prepu-bertal (aged 8 months to 10 years) and five peripubertal(aged 11–15 years) controls were included. All materialswere included in our study after informed consent from theparents to use part of the testicular tissue for research.

Paraffin embedded testicular tissue from five fetuses witha 47,XXY karyotype (gestational age of 19–21 weeks) was ob-tained from the pathology department of the UZ Brussel. Forall but one fetus the pregnancy had been terminated after pre-natal diagnosis. For the remaining fetus, a cytomegalovirusinfection together with prenatal diagnosis of 47,XXY wasthe reason for abortion. Testicular tissues from five abortedcontrol fetuses with normal macroscopic testes at autopsywere also obtained from the same department.

Immunohistochemistry

Formol-fixed (fetus, adult KS, and SCO) and AFA-fixed(pre- and peripubertal KS and vasectomy-reversal)

2 VOL. - NO. - / - 2020



119120121122123124125126127128129130131132133134135136137138139140141142143144145146147148149150151152153154155156157158159160161162163164165166167168169170171172173174175176177

178179180181182183184185186187188189190191192193194195196197198199200201202203204205206207208209210211212213214215216217218219220221222223224225226227228229230231232233234235236

paraffin-embedded tissue were cut in 5-mm-thick sections atthree different depths with an interval of 100 mm and stainedwith different antibodies to evaluate the niche cells and testic-ular ECM deposition.

Generally, sections were deparaffinized by rinsing inxylene (3 � 10 minutes) and subsequently rehydrated in adecreasing series of ethanol (2 � 100%, 90%, 70%). Afterwashing with phosphate-buffered saline (PBS) (70011-036;Life Technologies) for 5 minutes, tissue sections were incu-bated with a hydrogen peroxide solution (H202 diluted inmethanol) for 30minutes to eliminate endogenous peroxidaseactivity.

After washing, antigen retrieval was performed, and theslides were incubated with normal goat serum (NGS; B304;Tebu-bio). After the blocking steps, primary antibodies wereapplied to the sections except the negative control in whichprimary antibody was omitted. Antibodies for antim€ullerianhormone (AMH) (MCA2246; AbD Serotec) and androgen re-ceptor (AR) (sc-816; BioConnect) were used to evaluate thematuration state of Sertoli cells. The PTMCs were stainedwith alpha smooth muscle actin 2 (ACTA2) (A2547; Sigma-Aldrich) and ECM proteins laminin (Ab76164; Abcam),fibronectin (A0245; Dako, Heverlee, Belgium), collagen I(Ab34710; Abcam), and collagen IV (Ab6586; Abcam).

The slides were incubated overnight at 4�C in a humidi-fied chamber. After three wash steps, tissue sections wereincubated with peroxidase-labeled secondary antibody(K5007; Dako Real Envision Detection System) for 1 hour atroom temperature. The sections were washed, visualizedwith 3,3ʹ-diaminobenzidine (DAB; Dako Real Envision Detec-tion System) and counterstained with hematoxylin. For moredetailed information about the protocol, see SupplementalTable 1 (available online).

Histologic Analysis

Histologic examination was performed on an inverted lightand fluorescence microscope (Olympus IX81). Digital imageswere made using a digital camera (CC12 Soft Imaging System)and evaluated with the image processing program Fiji (26).Two or three sections at different depths (50 mm betweendifferent depths) were evaluated per patient. For analysiswith Fiji, five image fields at �10 magnification were takenfrom each section. All tubules within these five image fieldswere evaluated and scored.

For the evaluation of the Sertoli cell marker AMH, a scorefrom 0 to 3 was used to interpret the intensity of the expres-sion profiles in testicular biopsies. Tubules were scored as 0 ifstaining was absent, 1 for weak expression, 2 for moderatestaining, and 3 for strong staining. For the adult biopsies,seminiferous tubules were divided into tubules with presenceof germ cells (GC tubules) or only Sertoli cells (SCO tubules).

For evaluation of the other Sertoli cell marker AR, thepresence or absence of AR expression was evaluated. Forthe tubules in which ARwas expressed, the percentage of pos-itive Sertoli cells was determined. In the adult biopsies, sem-iniferous tubules were again divided into GC and SCO tubules.Supplemental Figure 1 (available online) gives an overview ofthe different staining patterns for ECM proteins.

For fibronectin staining, tubules were scored as F-0 if theBM was stained and the LP contained several stained layers;as F-1 if the BM was stained and the LP was thickened andintensely stained; as F-2 if the BM was stained and the LPconsisted of less intensely stained ECM and an obvious outerlayer; and as F-3 if the BM was stained and the LP was notstained and the presence of an outer layer was not obvious.For laminin staining, tubules were scored as L-0 if the BMwas stained and the LP was composed of several stainedlayers; as L-1 if the BM was stained and the LP was intenselystained; as L-2 if the BM was stained and the LP was notstained; and as L-3 if both the BM and the LPwere not stained.

For collagen IV staining, tubules were scored as CIV-0 ifthe BM was stained and the LP contained several stainedlayers; as CIV-1 if the BM was stained and the LP wasintensely stained; as CIV-2 if the BM was stained and the LPconsisted of an obvious inner and outer layer with a thickenedlayer of ECM; and as CIV-3 if the BM was stained and the LPconsisted of an obvious inner layer and a thickened ECM butthe presence of the outer layer is not obvious. For collagen Istaining, tubules were scored as CI-0 if the BM was stainedand the LP was composed of several stained layers; as CI-1if the BM was stained and the LP was less intensely stainedand had an obvious outer layer; and as CI-2 if the BM wasstained and the LP was not stained and the presence of anouter layer was not obvious.

Statistical Analysis

Data are expressed as mean � standard deviation. Statisticalanalysis was performed using GraphPad Prism 6.0 (GraphPadSoftware). Statistical differences between prepubertal andperipubertal KS and control samples were determined usingthe Mann-Whitney U test. Kruskal-Wallis test with Dunn’smultiple comparisons test as a post hoc test was used tocompare adult KS, controls and SCO. P< .05 were consideredstatistically significant.

RESULTSThe expression of the different markers was analyzed intesticular biopsies at different time points of development.The fetal age ranged from 19 to 23 weeks with a mean ageof 22.4 � 1.3 weeks for the control group, and the fetal ageof the KS group was 20.6 � 1.1 weeks. For the prepubertalgroup, the mean age was 5.1� 3.6 years for the control groupand 5.5� 1.3 years for the KS group. The peripubertal controlgroup had a mean age of 12.8 � 1.3 years, and the KS group14.7 � 1.7 years. The adult biopsy samples were retrievedfrom patients older than 18 years.

Characterization of Sertoli Cell Maturation

Normally AMH is expressed in the Sertoli cells from the fetaluntil peripuberty age. During puberty, the intensity of AMHexpression by Sertoli cells decreases, as visualized in a testic-ular biopsy of a 15-year-old control boy (Fig. 1A–D). In KSpatients, intense AMH expression is observed in fetal and pre-pubertal testicular biopsy tissues.

VOL. - NO. - / - 2020 3


Fertility and Sterility®

237238239240241242243244245246247248249250251252253254255256257258259260261262263264265266267268269270271272273274275276277278279280281282283284285286287288289290291292293294295

296297298299300301302303304305306307308309310311312313314315316317318319320321322323324325326327328329330331332333334335336337338339340341342343344345346347348349350351352353354

In biopsy tissues from peripubertal KS patients, tubuleswith different AMH intensities are present within one sample,with intensity 1 (34% of tubules) being the most observed. Incontrol peripubertal samples, tubules with different AMH in-tensities are also observed; intensity 0 (37% of the tubules)was the most observed (see Fig. 1G–J; Supplemental Fig. 2,available online). Although there is a difference in meanage between the control (13 years) and KS group (16 years),no statistically significant difference in AMH staining wasobserved between both groups (see Fig. 1G–J). The 15-year-old control boy showed absence of AMH staining in all tu-bules, but absence of AMH staining in the KS samples wasonly observed in a limited number of tubules although theKS boys are older.

Ideally, the control group should be aged-matched butunfortunately, but there were no more biopsy samples avail-able from control patients older than 14 years because thesepatients can perform sperm freezing and hence do not enrolin the prepubertal fertility preservation program. No statisti-cally significant differences were obtained between the con-trol and KS groups in the fetal, prepubertal, andperipubertal testicular biopsies (see Supplemental Fig. 2).

In the adult controls, AMH expression was absent inmost of the tubules (see Fig. 1E). In KS samples, there wasa discrepancy between tubules having germ cell differentia-tion and tubules without germ cell differentiation. In the tu-bules with germ cell differentiation, AMH expression waslacking in most of them (see Fig. 1L), whereas different

FIGURE 1

(A–L) Antim€ullerian hormone (AMH) and (M–X) androgen receptor (AR) expression during testicular development in biopsies from controls andKlinefelter syndrome (KS) patients. AMH and AR staining are shown from the same patient sample. AMH and AR staining in a 23-week-oldcontrol fetus (A, M) and a 21-week-old KS fetus (G, S). All tubules express AMH in the same intensity as well in control as in KS samples. ARexpression in Sertoli cells is lacking in fetal samples. AMH and AR staining in a 4-year-old boy (B, N) and a 6-year-old KS patient (H, T). IntenseAMH expression is observed in prepubertal biopsies, while AR is still absent in Sertoli cells. Interstitial staining of AR is observed. AMH and ARstaining in a 12-year-old (C, O, P) and a 15-year old control boy (D) and a 13-year-old (I, U) and 15-year-old KS boy (J,V). AMH staining is faintor lacking in the 15-year-old control boy while AMH is still expressed in the KS boys with the same age. AR staining is present in peripubertalSertoli cells, but not all Sertoli cells express AR. AMH and AR staining in adult control (E,Q) and KS (K, L, W, X) biopsy. In the control biopsy,AMH expression is absent, while Sertoli cells express AR. In adult KS samples, AMH expression is still present in some tubules (K). However,AMH expression is absent in tubules containing differentiation (L). AR is expressed in most tubules, but not in all Sertoli cells within one tubule.In SCO samples, AMH and AR expression is observed (F,R). An overview of AMH and AR expression in adult samples is visualized in the graphs.Prolonged AMH expression was observed in KS and SCO samples. *P<.05. **P<.01, different from control.Van Saen. Tubular niche in Klinefelter testes. Fertil Steril 2020.

4 VOL. - NO. - / - 2020



355356357358359360361362363364365366367368369370371372373374375376377378379380381382383384385386387388389390391392393394395396397398399400401402403404405406407408409410411412413

414415416417418419420421422423424425426427428429430431432433434435436437438439440441442443444445446447448449450451452453454455456457458459460461462463464465466467468469470471472

AMH expression intensities were present in tubules withoutgerm cell differentiation (see Fig. 1K). Of course, only a fewtubules with germ cell differentiation were present in the KSbiopsies.

In SCO patients, the mixture of tubules with different in-tensities of AMH staining were similar to tubules withoutgerm cell differentiation in KS patients (see Fig. 1F). Whenthe three groups were compared, a statistically significant dif-ference was obtained (P¼ .0026). The number of tubuleswithout differentiation and scored with intensity 0 was statis-tically significantly lower in KS samples compared with con-trols (P¼ .0047). In the SCO group, statistically significantlyfewer tubules showed AMH intensity 0 (P¼ .0024), whilemore tubules with intensity 1 (P¼ .0187) and intensity 2(P¼ .0350) were present compared with controls, but notcompared with KS (see Fig. 1).

The start of AR expression around puberty was confirmedin our control samples (see Fig. 1M–P). In peripubertal controlsamples, few tubules showed expression of AR in all Sertolicells (see Fig. 1O), while most Sertoli cells showed AR expres-sion in the oldest peripubertal sample and adult samples (seeFig. 1P–Q). In KS patients, AR expression was also observedfrom the peripubertal stage onward (see Fig. 1S–X). Althoughtubules without AR expression or tubules containing

AR-negative Sertoli cells were observed in KS and SCO sam-ples, no statistically significant difference was reached withcontrol samples (see Fig. 1; Supplemental Fig. 2). In tubuleswith ongoing spermatogenesis in the KS samples, AR expres-sion was present in all Sertoli cells.

Characterization of the Tubular Wall Changes

Peritubular myoid cells. The expression of ACTA2 by PTMCswas evaluated in prepubertal up to adult testicular biopsysamples. Absent expression of ACTA2 in PTMCs formingthe tubular wall of the seminiferous tubule was characterizedas interrupted LP. In prepubertal controls, ACTA2 expressionwas not observed in the PTMCs, except in two patients inwhom interrupted LP was observed. Expression of ACTA2was not observed in PTMCs of KS prepubertal testicular bi-opsies. In both control and KS biopsy samples, ACTA2 waspresent in the smooth muscle cells surrounding blood vessels(Fig. 2A–C).

In peripubertal biopsy samples, ACTA2 expression wasdeveloping in the tubular wall but was often incomplete (in-terrupted membrane). In the 15-year-old control boy, expres-sion was complete, showing ACTA2 expression lining thewhole tubular wall. No statistically significant difference

FIGURE 1 Continued

print&web4C=FPO

Van Saen. Tubular niche in Klinefelter testes. Fertil Steril 2020.

VOL. - NO. - / - 2020 5



473474475476477478479480481482483484485486487488489490491492493494495496497498499500501502503504505506507508509510511512513514515516517518519520521522523524525526527528529530531

532533534535536537538539540541542543544545546547548549550551552553554555556557558559560561562563564565566567568569570571572573574575576577578579580581582583584585586587588589590

FIGURE 2

print&web4C=FPO

Alpha smooth muscle actin 2 (ACTA2) expression during testicular development in biopsies from controls and Klinefelter syndrome (KS) patients.(A–C) ACTA2 Q5expression absent in the peritubular myoid cells before puberty in both controls (A) and KS patients (B). Expression is seen in theendothelial cells from blood vessels. (D–H) ACTA2 expression in a 12-year-old (D) and a 15-year-old (F) control boy and a 12-year-old (E) and a16-year-old (G) KS boy. During puberty, expression of ACTA2 becomes obvious in the peritubular myoid cells. However, not all tubules showACTA2 expression (asterisk), and ACTA2 expression is also not completely surrounding the seminiferous tubules (arrows). Expression iscomplete in the 15-year-old control, but in KS boys there are tubules with interrupted membrane staining present. (I–M) ACTA2 expression inan adult control testicular biopsy (I), a testicular biopsy from an Sertoli cell only (SCO) patient (K) and KS patients (J, L). In adult control biopsies,ACTA2 expression is present in the peritubular cells. In Klinefelter biopsies, expression is interrupted in some tubules or even completelylacking. In hyalinized tubules, some remaining expression can be observed, while others completely lack ACTA2 expression. In SCO biopsies,ACTA2 expression is observed without interruption in most tubules. However, in some biopsies, interrupted membrane staining was observed.HT ¼ hyalinized tubules; LP ¼ lamina propria; NT ¼ normal tubules. Black asterisks denote different from control; white asterisks denotedifferent from SCO. *P<.05. **P<.01. ***P<.001.Van Saen. Tubular niche in Klinefelter testes. Fertil Steril 2020.

6 VOL. - NO. - / - 2020



591592593594595596597598599600601602603604605606607608609610611612613614615616617618619620621622623624625626627628629630631632633634635636637638639640641642643644645646647648649

650651652653654655656657658659660661662663664665666667668669670671672673674675676677678679680681682683684685686687688689690691692693694695696697698699700701702703704705706707708

was observed in ACTA2 expression in the tubular wall be-tween control and KS biopsies (see Fig. 2D–H).

In the adult control testicular biopsy samples, ACTA2expression lined the tubular wall in all seminiferous tubules(see Fig. 2I). However, in KS biopsies interrupted membraneswere often visible. If the tubular wall was thickened, two ringsof ACTA2 expression were observed, and both rings showedinterruptions (see Fig. 2J). In SCO biopsies, ACTA2 expressionlining the tubular wall was mostly similar to the controls, butinterrupted membranes were sometimes present (see Fig. 2K).A statistically significant difference in the ACTA2 expressionin normal tubules was found (P< .001) between the threegroups. Fewer tubules with a normal staining pattern werepresent in KS biopsies compared with control (P¼ .0001)and SCO (P¼ .0136).

Hyalinized tubules in KS biopsies showed remnants ofACTA2 staining or no staining at all (see Fig. 2L). Hyalinizedtubules in KS biopsies showed statistically significantly lessACTA2 staining compared with hyalinized tubules in SCOsamples (P¼ .0097) (see Fig. 2M).

Extracellular matrix. We investigated the distribution of theECM proteins (fibronectin, laminin, collagen IV and I) withinthe tubular wall. Expression of fibronectin and collagen I andIVwas present in the interstitium and the peritubular area, butlaminin expression was only visible in the BM. In fetal sam-ples ECM proteins were not expressed, but expression wasalready observed before puberty (Supplemental Fig. 3A–H,available online). Tubules in the prepubertal control sampleswere clearly aligned with layers of peritubular cells express-ing the ECM proteins (see Supplemental Fig. 3Aʹ–Dʹ). Expres-sion in the prepubertal KS samples was similar to that of thecontrols (see Supplemental Fig. 3Eʹ–Gʹ), except for collagen Istaining. In three out of the four KS prepubertal samples,collagen I staining did not show a clear alignment of the

tubules (see Supplemental Fig. 3Hʹ). In the peripubertal sam-ples, tubules with normal expression patterns for ECM pro-teins were present next to tubules with increased thicknessof the LP. Hyalinized tubules were not yet present in all sam-ples, but their presence seems to correlate with the age of thepatients and the presence of tubules with normal expressionpatterns (Table 1; see Supplemental Fig. 3M–X). Normalstaining patterns were present in the control samples.Table 1 gives an overview of the different staining patternspresent in normal tubules, tubules with thickened membrane,and hyalinized tubules in the peripubertal KS samples.

In general, the samples from the younger patientsmostly had normal tubules present (see SupplementalFig. 3M–P). In other samples, tubules with thickened LPand hyalinized tubules are represented (see SupplementalFig. 3Q–T). In the samples from the oldest peripubertal pa-tients, mainly hyalinized tubules were present. The thick-ened LP was positive for fibronectin and collagen I.Laminin expression was present in the BM. However,some tubules only showed faint expression (seeSupplemental Fig. 3N). Laminin was also expressed bythe Sertoli cells in the remaining tubules, which made itdifficult to evaluate expression in the BM. Collagen IVstaining was also observed in the BM, while the thickenedLP showed less intense staining. The hyalinized tubules arepositive for fibronectin and collagen I, while faint or absentexpression of collagen IV and laminin was observed.

In the adult groups, statistically significant differentexpression of all ECM proteins was observed in KS and SCObiopsies compared with control biopsies. Although in controlsamples (see Supplemental Fig. 1, pattern 0) all four examinedECM proteins lined the BM of the seminiferous tubules,different expression patterns were observed in KS and SCO bi-opsies (see Supplemental Fig. 1). The distribution of these

TABLE 1

Staining patterns for extracellular matrix proteins in peripubertal Klinefelter syndrome samples.

ID Age

Staining patterns

Normal tubules Tubules with thickened membrane Hyalinized tubules

Fibronectin LamininCollagen

IVCollagen

I Fibronectin LamininCollagen

IVCollagen

I Fibronectin LamininCollagen

IVCollagen

I

T47 12 NA 0 0 0 NA 2 1/2/3 1 NA 2/3 2 1T53 12 0 0 0 0 — — — — — — —

T52 13 0 0 0 0 1 2 1 1T75 14 — — — — 1 2/3 3 1T50 14 0 0 0 0 1 2 1 1T60 14 — — — — 1 2 3 2T78 14 — — — — — — — — 1 3 3 1T40 14 — — — — 1 2 2/3 NA — — — —

T87 14 0 OS 0 0 1 OS 3 1 — — — —

T92 14 0 OS 0 0 1 OS 3 1 1 2 3 1T77 15 0 OS 0 0 1 OS 3 1 — — — —

T64 15 — — — — 1 2 1 1 — — — —

T61 16 — — — — 1 OS 3 NA 1 3 3/NEG 1/NEGT76 16 — — — — 1 OS 1 1 1 2/3 3/NEG 1T82 16 — — — — 1 OS 3 1 1 2/3 3 1T42 16 1 2 1 1 1 2/3 1 1Note: — ¼ tubules not present; NA ¼ not available; NEG ¼ no expression; OS ¼ overexpression in Sertoli cells.


VOL. - NO. - / - 2020 7



709710711712713714715716717718719720721722723724725726727728729730731732733734735736737738739740741742743744745746747748749750751752753754755756757758759760761762763764765766767

768769770771772773774775776777778779780781782783784785786787788789790791792793794795796797798799800801802803804805806807808809810811812813814815816817818819820821822823824825826

different expression patterns for the different ECM proteins isshown in Figure 3A.

For fibronectin staining, statistically significantly fewernormal tubules were observed in KS (P¼ .0003) and SCO bi-opsies (P¼0016).Q1 The most observed staining pattern in KSand SCO biopsies was intense staining in the thickened wall(pattern F-1; see Fig. 3Aa,b), and this was statistically signif-icantly different compared with controls (KS: P¼ .0013; SCO:P¼ .0048). Patterns F-2 and F-3 were not observed in controlsamples and thus were only represented in SCO and KS bi-opsies. No statistically significant difference was observed be-tween SCO and KS.

For laminin staining, statistically significantly fewernormal tubules were observed in KS (P¼ .0053) and SCO(P¼ .0003) compared with controls. The most observed stain-ing pattern in KS (P¼ .0109) and SCO (P¼ .0252) biopsies waspattern L-3, actually referring to the absence of laminin in the

tubular wall (see Fig. 3A,c,d). The other staining patterns werealso statistically significantly different between SCO and con-trols (P¼ .0007 for pattern L-1 and P¼ .0480 for pattern L-2),while in KS samples only pattern L-1 was different from con-trols (P¼ .0017). No statistically significant difference wasobserved between KS and SCO.

For collagen IV staining, statistically significantly fewernormal tubules were present in KS (P¼ .0022) and SCO(P¼ .0006) biopsies. The most represented staining patternin KS was pattern CIV-3 (see Fig. 3Ae; P¼0041), Q2while thedifferent staining patterns were almost equally distributedin SCO biopsies (see Fig. 3Af). Only staining pattern CIV-2was statistically significantly different between SCO and KS(P¼ .0464).

For collagen I, statistically significantly fewer normal tu-bules were present in KS (P¼ .0100). No differences wereobserved between control and SCO biopsies. The most

FIGURE 3

(A) Distribution of extracellular matrix (ECM) staining patterns for fibronectin, laminin, collagen IV, and collagen I in adult testicular biopsies. Picturesunderneath the graphs show the most representative staining pattern for each ECM protein within the Klinefelter syndrome (KS) and Sertoli cellonly (SCO) groups. (B) Schematic overview of the expression of the proteins evaluated in this study in the seminiferous tubules from control (C) andKS testicular biopsies. Antim€ullerian hormone (AMH) expression is still observed in adult KS samples, while a decrease in alpha smoothmuscle actin2 (ACTA2) expression is seen in KS biopsies. The distribution of the ECM proteins in the seminiferous wall is clearly affected in seminiferous tubuleswith a thickened wall. In prepubertal samples, the ECM proteins align the seminiferous tubules, except for collagen I staining in which alignment ofthe tubules was not clear. Black asterisks denote different from control; white asterisks denote different from SCO. *P<.05. **P<.01. ***P<.001.Van Saen. Tubular niche in Klinefelter testes. Fertil Steril 2020.

8 VOL. - NO. - / - 2020



827828829830831832833834835836837838839840841842843844845846847848849850851852853854855856857858859860861862863864865866867868869870871872873874875876877878879880881882883884885

886887888889890891892893894895896897898899900901902903904905906907908909910911912913914915916917918919920921922923924925926927928929930931932933934935936937938939940941942943944

represented pattern was staining pattern CI-1, but there wasno difference with control and SCO (see Fig. 3Ag,h). PatternCI-2 was statistically significantly increased in KS comparedwith control (P¼ .0014) and SCO (P¼ .0009).

In general, KS and SCO samples showed a thickenedtubular wall that was composed of the BM aligning theepithelial cells and concentric layers of PTMCs. In betweenthese layers was a homogenous thickened layer that con-tained fibronectin. Laminin was present in the BM but notin the outer layers, and in many tubules staining was notobserved. Collagen I and IV patterns were different for SCOand KS. In SCO biopsies, collagen I and IV expression wasfound in the inner and outer circle, but in KS biopsies theouter circle was often absent.

DISCUSSIONKlinefelter syndrome is mainly diagnosed in adults whenazoospermia is found to be present during an infertilityworkup. Testicular degeneration is a hallmark in KS patientsand hampers the efficiency of retrieving spermatozoa fromthe rare remaining foci with active spermatogenesis in thetestis. We have shown previously that fibrosis is already initi-ated in peripubertal but not in prepubertal patients (1). Thecause of this fibrotic initiation is not yet clear. Although KStestes show highly fibrotic areas, in some KS patients

seminiferous tubules are still present with active spermato-genesis. It is not clear why degeneration of the tubular struc-ture occurs in KS patients while the tubular structure remainsintact in SCO patients. Therefore, in this study we examinedthe expression of Sertoli-cell maturationmarkers andmarkersfor functional PTMCs and ECM proteins in seminiferous tu-bules of KS patients at different stages of development.

The Sertoli cell maturation status was evaluated by AMHand AR expression. In prepubertal and peripubertal controlbiopsies, a decrease in AMH intensity was observed withadvancing age. The expression of AMH in this limited numberof control patients was consistent with the AMH expressionreported in a large cohort of prepubertal and peripubertal pa-tients (71 patients aged 0–16 years) (27). The investigators re-ported strong AMH expression in more than 90% of thetubules in patients until the age of 6. In biopsies from patientsfrom 6 years until 16 years old, a gradual decrease in tubuleswith strong AMH intensity was observed with the majority oftubules lacking AMH expression in the age category of 12–16years.

In our prepubertal control group, more than 60% of tu-bules showed high AMH intensity, and in the peripubertalgroup more that 60% showed absent or low intensity stainingof AMH. In prepubertal and peripubertal KS samples, normalexpression of these Sertoli cell maturation markers wasobserved. This is consistent with the normal secretion level

FIGURE 3 Continued

print&web4C=FPO


VOL. - NO. - / - 2020 9



945946947948949950951952953954955956957958959960961962963964965966967968969970971972973974975976977978979980981982983984985986987988989990991992993994995996997998999

1000100110021003

10041005100610071008100910101011101210131014101510161017101810191020102110221023102410251026102710281029103010311032103310341035103610371038103910401041104210431044104510461047104810491050105110521053105410551056105710581059106010611062

of two secretion products of Sertoli cells, AMH and inhibin B,in KS infants (28). Although no statistically significant differ-ence was present in the peripubertal group in this study, a de-layed disappearance in expression of AMH and a decreasedAR expression was already reported in KS peripubertal bi-opsies (29).

In the current study, no statistically significant differ-ences were observed in AR expression in adult KS testis, butprolonged AMH expression was observed, indicating thepresence of immature Sertoli cells in the adult KS testis.Similar to the KS group, delayed expression of AMH wasalso observed in the SCO group. Expression of other markersfor immature Sertoli cells (DACH2) and Leydig cells (FAM9A)was reported to be up-regulated in KS testis (30). As alsoobserved in the present study, immature Sertoli cells couldstill be detected in biopsies from SCO patients and patientswith maturation arrest. By contrast, SCO biopsies from menwith deletion in the azoospermia factor or AZF region didnot show expression of the immature marker cytokertin-18.This indicates that the absence of spermatogenesis in thesepatients is germ cell-related (31) whereas in biopsies fromKS and idiopathic SCO patients a deficient somatic environ-ment might be the cause of germ cell loss. However, SCO sam-ples show prolonged AMH expression with intact tubularstructure, so delayed maturation on its own cannot explainthe degeneration of seminiferous tubules observed in KSpatients.

Expression of ACTA2 starts around puberty and isinduced by androgens, whose production is enhanced byfollicle-stimulating hormone (32). In this study, we confirmedthe absence of ACTA2 staining in PTMCs in the prepubertalhuman testis. In peripubertal biopsies, expression of ACTA2starts and increases with age in our control group, obtainingthe highest level at adulthood. In the adult KS testis, however,we observed a loss of expression of the contractile markerACTA2 in PTMCs. By contrast, no statistically significantloss was found in SCO biopsies compared with controls exceptwhen only hyalinized tubules, which are not present in allSCO biopsies, were considered. The expression pattern in pre-pubertal and peripubertal KS patients did not statisticallysignificantly differ from controls.

Expression of ACTA2 in the tubular wall was interruptedin the tissue samples from the youngest patients in the peripu-bertal control group, and expression was found around theentire tubules in the 15-year-old control patient. From thesedata, we can hypothesize that ACTA2 expression increaseswith age and testicular development, although only one olderparticipant (15 years) could be included in this study. In theperipubertal KS group, none of the samples showed completeexpression in all tubules.

Although the mean age of the KS group is higher, it isdifficult to say whether expression is still not complete or isalready starting to get lost. Because ACTA2 expression isstimulated by androgens, it might be possible that the acqui-sition of the contractile phenotype of peritubular cells ishampered in KS patients when having low testosterone levels.However, testosterone levels in adolescent KS patients aremostly in the normal range. Also, tubules with properACTA2 staining in the peritubular cells are present in adult

biopsies, indicating that differentiation in contractile myoidcells is feasible. Rather than a hampered differentiation,PTMCs might change their phenotype. Change of phenotypeswitches in response to changes in the local environment iscommon for smooth muscle cells (33).

In the tubular wall, PTMCs can evolve to a more secretorycell type resembling fibroblasts, and this is commonly associ-ated with a loss of contractile function (34). We observed theloss of the contractile marker ACTA2 in KS adult tissue sam-ples, while the loss of other contractile markers—myosinheavy chain (MYH11) and calponin—was reported in patientswith mixed atrophy. As in our study, normal as well as partialand complete loss of peritubular staining was observed withinone biopsy, excluding methodological factors (34). When thededifferentiation potential of peritubular cells was investi-gated in infertile men, no evidence for loss of contractilemarkers was observed. Loss of MYH11 was observed in partsof the biopsies but was attributed to methodological issuesand did not correlate with pathologies. They included patientswith cryptorchidism, testicular feminization, prostate cancer,and vasectomy (21). In the KS group from the current study,ACTA2 expression was different from control and SCO. Lossof ACTA2 was also observed in patients with mixed atrophy.

The transformation of PTMCs to fibroblasts which areresponsible for the production of collagen fibers and ECM(22) could explain the fibrotic appearance of the KS testiculartestis. Because loss of ACTA2 is less frequently observed in theSCO testis, local factors produced in the KS testis could beresponsible for the phenotype switch and initiation of thefibrotic process. It has been proven that the phenotype of my-ofibroblasts is highly dependent on the environmental fac-tors; the difference in expression of contractile markersin vivo was not detected if peritubular cells (isolated from bi-opsies with normal spermatogenesis and mixed atrophy) wereisolated and cultured in vitro. The investigators hypothesizedthat myofibroblasts adapt to a common phenotype because ofthe exposure to identical culture conditions (34). However,which factors are responsible for the possible dedifferentia-tion of PTMCs in KS patients have yet to be identified. Therole of immune cells in this phenotype switch can be consid-ered. Increased numbers of mast cells and macrophages in thetestis of infertile patients have been described (35, 36). Theprostaglandin (PG) D2 metabolite 15-deoxy-delta-12,14-PGJ2 (15dPGJ2), which is expressed in infertile patients and isstimulated by mast cells, had a strong influence on culturedPTMCs. Stimulation with 15dPGJ2 induced loss of contractilemarkers in the cultured PTMCs (34).

Next to their contractile role, PTMCs have a secretoryrole. They secrete glial cell line–derived neurotrophic factor(GDNF), which plays an important role in spermatogonialstem cell (SSC) self-renewal and has long been thought tobe a Sertoli cell-specific product (37). The production ofGDNF strengthens the assumption that PTMCs are part ofthe SSC niche. Next to GDNF, PTMCs are responsible forECM protein production. All four ECM products in our studyare produced by PTMCs, while collagen IV and laminin arealso secreted by Sertoli cells (38). The distribution patternfor laminin and fibronectin was similar between KS andSCO biopsies in our study. Increased expression of laminin

10 VOL. - NO. - / - 2020



10631064106510661067106810691070107110721073107410751076107710781079108010811082108310841085108610871088108910901091109210931094109510961097109810991100110111021103110411051106110711081109111011111112111311141115111611171118111911201121

11221123112411251126112711281129113011311132113311341135113611371138113911401141114211431144114511461147114811491150115111521153115411551156115711581159116011611162116311641165116611671168116911701171117211731174117511761177117811791180

b2 chain has been reported in patients with SCO syndrome, incontrast to the increased absence of laminin staining in SCOand KS biopsies in this study (39).

Laminin contributes to the attachment of Sertoli cells andspermatogonia to the BM by binding to the integrin receptor,which is expressed on SSCs and Sertoli cells (40, 41). Migra-tion of gonocytes to the BM occurs during the fetal andneonatal period in humans (42). Deficient laminin expressioncould result in hampered migration of the gonocytes to theBM. Recently, Winge et al. (2) suggested that the Klinefelterphenotype is already initiated at fetal age. Differentiation ofgonocytes in prespermatogonia is hampered in the KS fetusdue to aberrant expression of long noncoding RNAs. Whethermigration of the surviving prespermatogonia occurs normallyshould be investigated.

The most common staining pattern in KS biopsies in tu-bules with a thickened lamina propria was the presence offibronectin in the acellular zone of the thickened membrane,while collagen I and IV form two circular rings separated bythe acellular zone. The presence of different staining patternsin azoospermic patients was already described by Volkmannet al. (21). However, in KS biopsies we saw that the outer layerwas often absent, as it was in peripubertal and adult testicularbiopsies. This could indicate that the lamina propria starts todisintegrate in KS samples. This could be the initial phase ofhyalinization of the tubules.

In KS patients, most tubular structures completely disin-tegrate, or cellular content is replaced by BM-like materialreferred to as hyalinized tubules (43). The absence of (differ-entiating) germ cells cannot explain this tubular degenerationbecause patients having no germ cells at all (SCO patients) orpatients with maturation arrest do not show the testiculardegeneration seen in KS patients. Furthermore, targeted abla-tion of the germ cells by Diphtheria toxin injections in post-natal mice resulted in a normal testicular architecture atadult age, although without germ cells (44).

Whether tubular integrity is lost due to loss of Sertoli cellsor disintegration of the tubular wall is difficult to concludefrom our observations. Studies inmice showed that PTMC dif-ferentiation requires the presence of prepubertal Sertoli cells.Ablation of Sertoli cells at fetal age or early postnatal age re-sulted in the absence of recognizable seminiferous tubules orexpression of PTMC markers at an adult age (44). However,when Sertoli cells were ablated at adult age, no impact onthe tubular wall was observed (45). Because Sertoli cells arepresent in fetal and prepubertal KS biopsies, it could be hy-pothesized that the disintegration of the tubular wall is notcaused by loss of Sertoli cells. Laminin expression was oftenlost in seminiferous tubules and could thus result in detach-ment of Sertoli cells from the BM. Although absent or faintlaminin expression was observed in most KS biopsies, tubulessometimes showed high expression of laminin in Sertoli cells.

Laminins are heterotrimers of a-, b-, and g-chains (46).The laminin antibody used in this study was directed to thefull-length protein. However, we cannot rule out methodo-logical problems with the antibody because the same incuba-tion time to produce the color reaction with DAB sometimesresulted in understained or overstained samples. However,the presence of tubules with and without laminin staining

in the same cross-sections suggests that laminin expressionis truly hampered in KS and SCO samples.

The limitations of this study included the descriptive dataobtained by immunostaining. Variability between differentstainings, even within the same sample, represents a largelimitation in the evaluation of immunostainings. The fixativebetween samples coming from the pathology department andsamples that were fixed in the research laboratory alsodiffered. However, both were formol-based and gave similarstaining patterns in the immunostainings. Confirmationwith more quantifiable techniques as Western blot orreverse-transcription polymerase chain reaction could addmore value to the obtained data, but these were not performeddue to the scarcity of testicular biopsy samples from KS pa-tients. We cannot draw conclusions about the causal factorsrelated to the degeneration of the seminiferous tubules inKS patients. More research is necessary to reveal the causeof tubular degeneration in KS patients.

In KS samples, AMH expression is prolonged and ACTA2expression is decreased. The distribution of the ECM proteinsin the seminiferous wall is clearly affected in seminiferous tu-bules with a thickened wall. These tubules appear at puberty,while in prepubertal samples, the alignment of the tubules isnot as clear as in control samples. These conclusions are visu-ally presented in Figure 3B.

We observed an altered expression pattern between SCOand KS samples for collagen I, collagen IV, and ACTA2.More research is necessary to identify causes of these alter-ations and the associated testicular fibrosis in KS patients. Ifthe mechanism behind this fibrotic process could be identi-fied, this process could be altered to increase the fertilitychances of KS patients.

Acknowledgments: The authors thank Fabian Van Haelstand Pierre Hilven for their technical assistance with theimmunohistochemistry procedure. The authors also thank totwo master students, Laura Landwehr and Irena Gripshi,who performed their master thesis in the laboratory andhelped with the experiments and analysis of the results, andthe pathology department for providing testicular tissuesamples.

REFERENCES1. Van Saen D, Vloeberghs V, Gies I, Mateizel I, Sermon K, De Schepper J, et al.

When does germ cell loss and fibrosis occur in patients with Klinefelter syn-drome? Hum Reprod 2018;33:1009–22.

2. Winge SB, Dalgaard MD, Jensen JM, Graem N, Schierup MH, Juul A, et al.Transcriptome profiling of fetal Klinefelter testis tissue reveals a possibleinvolvement of long non-coding RNAs in gonocyte maturation. Hum MolGenet 2018;27:430–9.

3. Tournaye H, Staessen C, Liebaers I, Van Assche E, Devroey P, Bonduelle M,et al. Testicular sperm recovery in nine 47,XXY Klinefelter patients. Hum Re-prod 1996;11:1644–9.

4. Tournaye H, Camus M, Vandervorst M, Nagy Z, Joris H, Van Steirteghem A,et al. Surgical sperm retrieval for intracytoplasmic sperm injection. Int J An-drol 1997;20:69–73.

5. Bourne H, Stern K, Clarke G, Pertile M, Speirs A, Baker HW. Delivery ofnormal twins following the intracytoplasmic injection of spermatozoafrom a patient with 47,XXY Klinefelter’s syndrome. Hum Reprod 1997;12:2447–50.

VOL. - NO. - / - 2020 11



11811182118311841185118611871188118911901191119211931194119511961197119811991200120112021203120412051206120712081209121012111212121312141215121612171218121912201221122212231224122512261227122812291230123112321233123412351236123712381239

12401241124212431244124512461247124812491250125112521253125412551256125712581259126012611262126312641265126612671268126912701271127212731274127512761277127812791280128112821283128412851286128712881289129012911292129312941295129612971298

6. Reubinoff BE, Abeliovich D, Werner M, Schenker JG, Safran A, Lewin A. Abirth in non-mosaic Klinefelter’s syndrome after testicular fine needle aspi-ration, intracytoplasmic sperm injection and preimplantation genetic diag-nosis. Hum Reprod 1998;13:1887–92.

7. Nodar F, De Vincentiis S, Olmedo SB, Papier S, Urrutia F, Acosta AA. Birth oftwin males with normal karyotype after intracytoplasmic sperm injectionwith use of testicular spermatozoa from a nonmosaic patient with Klinefel-ter’s syndrome. Fertil Steril 1999;71:1149–52.

8. Palermo GD, Schlegel PN, Hariprashad JJ, Ergun B, Mielnik A, Zaninovic N,et al. Fertilization and pregnancy outcome with intracytoplasmic sperm in-jection for azoospermic men. Hum Reprod 1999;14:741–8.

9. Corona G, Pizzocaro A, Lanfranco F, Garolla A, Pelliccione F, Vignozzi L, et al.Sperm recovery and ICSI outcomes in Klinefelter syndrome: a systematic re-view and meta-analysis. Hum Reprod Update 2017;23:265–75.

10. Foresta C, Galeazzi C, Bettella A, Marin P, Rossato M, Garolla A, et al. Anal-ysis of meiosis in intratesticular germ cells from subjects affected by classicKlinefelter’s syndrome. J Clin Endocrinol Metab 1999;84:3807–10.

11. Yamamoto Y, Sofikitis N, Mio Y, Loutradis D, Kaponis A, Miyagawa I.Morphometric and cytogenetic characteristics of testicular germ cells andSertoli cell secretory function in men with non-mosaic Klinefelter’s syn-drome. Hum Reprod 2002;17:886–96.

12. Gonsalves J, Turek PJ, Schlegel PN, Hopps CV, Weier JF, Pera RA. Recombi-nation in men with Klinefelter syndrome. Reproduction 2005;130:223–9.

13. Blanco J, Egozcue J, Vidal F. Meiotic behaviour of the sex chromosomes inthree patients with sex chromosome anomalies (47,XXY, mosaic 46,XY/47,XXY and 47,XYY) assessed by fluorescence in-situ hybridization. HumReprod 2001;16:887–92.

14. Bergere M, Wainer R, Nataf V, Bailly M, Gombault M, Ville Y, et al. Biopsiedtestis cells of four 47,XXY patients: fluorescence in-situ hybridization andICSI results. Hum Reprod 2002;17:32–7.

15. Sciurano RB, Luna Hisano CV, Rahn MI, Brugo Olmedo S, Rey Valzacchi G,Coco R, et al. Focal spermatogenesis originates in euploid germ cells in clas-sical Klinefelter patients. Hum Reprod 2009;24:2353–60.

16. Staessen C, Coonen E, Van Assche E, Tournaye H, Joris H, Devroey P, et al.Preimplantation diagnosis for X and Y normality in embryos from three Kli-nefelter patients. Hum Reprod 1996;11:1650–3.

17. Greco E, Scarselli F, Minasi MG, Casciani V, Zavaglia D, Dente D, et al. Birthof 16 healthy children after ICSI in cases of nonmosaic Klinefelter syndrome.Hum Reprod 2013;28:1155–60.

18. Maiburg M, Repping S, Giltay J. The genetic origin of Klinefelter syndromeand its effect on spermatogenesis. Fertil Steril 2012;98:253–60.

19. Pop OT, Cotoi CG, Plesea IE, Gherghiceanu M, Enache SD, Mandache E,et al. Histological and ultrastructural analysis of the seminiferous tubulewall in ageing testis. Rom J Morphol Embryol 2011;52:241–8.

20. Mayerhofer A. Human testicular peritubular cells: more than meets the eye.Reproduction 2013;145:R107–16.

21. Volkmann J, Muller D, Feuerstacke C, Kliesch S, Bergmann M, Muhlfeld C,et al. Disturbed spermatogenesis associated with thickened lamina propriaof seminiferous tubules is not caused by dedifferentiation of myofibroblasts.Hum Reprod 2011;26:1450–61.

22. Haider SG, Talati J, Servos G. Ultrastructure of peritubular tissue in associa-tion with tubular hyalinization in human testis. Tissue Cell 1999;31:90–8.

23. Tournaye H, Donoso P, Osorio F, Donoso P, Osorio F. Sperm recovery tech-niques: clinical aspects. London: Taylor & Francis; 2004.

24. Gies I, De Schepper J, Van Saen D, Anckaert E, Goossens E, Tournaye H. Fail-ure of a combined clinical- and hormonal-based strategy to detect earlyspermatogenesis and retrieve spermatogonial stem cells in 47,XXY boysby single testicular biopsy. Hum Reprod 2012;27:998–1004.

25. Van Saen D, Gies I, De Schepper J, Tournaye H, Goossens E. Can pubertalboys with Klinefelter syndrome benefit from spermatogonial stem cellbanking? Hum Reprod 2012;27:323–30.

26. Schindelin J, Arganda-Carreras I, Frise E, Kaynig V, Longair M, Pietzsch T,et al. Fiji: an open-source platform for biological-image analysis. NatMethods 2012;9:676–82.

27. De Michele F, Poels J, Giudice MG, De Smedt F, Ambroise J, Vermeulen M,et al. In vitro formation of the blood-testis barrier during long-term

organotypic culture of human prepubertal tissue: comparison with a largecohort of pre/peripubertal boys. Mol Hum Reprod 2018;24:271–82.

28. Lahlou N, Fennoy I, Carel JC, Roger M. Inhibin B and anti-Mullerian hor-mone, but not testosterone levels, are normal in infants with nonmosaic Kli-nefelter syndrome. J Clin Endocrinol Metab 2004;89:1864–8.

29. Wikstrom AM, Hoei-Hansen CE, Dunkel L, Rajpert-De Meyts E. Immunoex-pression of androgen receptor and nine markers of maturation in the testesof adolescent boys with Klinefelter syndrome: evidence for degeneration ofgerm cells at the onset of meiosis. J Clin Endocrinol Metab 2007;92:714–9.

30. Winge SB, Dalgaard MD, Belling KG, Jensen JM, Nielsen JE, Aksglaede L,et al. Transcriptome analysis of the adult human Klinefelter testis andcellularity-matched controls reveals disturbed differentiation of Sertoli-and Leydig cells. Cell Death Dis 2018;9:586.

31. Bar-Shira Maymon B, Paz G, Elliott DJ, Hammel I, Kleiman SE, Yogev L, et al.Maturation phenotype of Sertoli cells in testicular biopsies of azoospermicmen. Hum Reprod 2000;15:1537–42.

32. Schlatt S, Weinbauer GF, Arslan M, Nieschlag E. Appearance of alpha-smooth muscle actin in peritubular cells of monkey testes is induced by an-drogens, modulated by follicle-stimulating hormone, and maintained afterhormonal withdrawal. J Androl 1993;14:340–50.

33. Owens GK, Kumar MS, Wamhoff BR. Molecular regulation of vascularsmooth muscle cell differentiation in development and disease. PhysiolRev 2004;84:767–801.

34. Welter H, Kampfer C, Lauf S, Feil R, Schwarzer JU, Kohn FM, et al. Partial lossof contractile marker proteins in human testicular peritubular cells in infer-tility patients. Andrology 2013;1:318–24.

35. Meineke V, Frungieri MB, Jessberger B, Vogt H, Mayerhofer A. Humantesticular mast cells contain tryptase: increasedmast cell number and altereddistribution in the testes of infertile men. Fertil Steril 2000;74:239–44.

36. Frungieri MB, Calandra RS, Lustig L, Meineke V, Kohn FM, Vogt HJ, et al.Number, distribution pattern, and identification of macrophages in thetestes of infertile men. Fertil Steril 2002;78:298–306.

37. Spinnler K, Kohn FM, Schwarzer U, Mayerhofer A. Glial cell line-derived neu-rotrophic factor is constitutively produced by human testicular peritubularcells and may contribute to the spermatogonial stem cell niche in man.Hum Reprod 2010;25:2181–7.

38. Skinner MK, Tung PS, Fritz IB. Cooperativity between Sertoli cells and testic-ular peritubular cells in the production and deposition of extracellular matrixcomponents. J Cell Biol 1985;100:1941–7.

39. Ooba T, Ishikawa T, Yamaguchi K, Kondo Y, Sakamoto Y, Fujisawa M.Expression and distribution of laminin chains in the testis for patients withazoospermia. J Androl 2008;29:147–52.

40. Colognato H, Yurchenco PD. Form and function: the laminin family of het-erotrimers. Dev Dyn 2000;218:213–34.

41. He Z, Kokkinaki M, Jiang J, Dobrinski I, Dym M. Isolation, characterization,and culture of human spermatogonia. Biol Reprod 2010;82:363–72.

42. Mitchell RT, Cowan G, Morris KD, Anderson RA, Fraser HM, McKenzie KJ,et al. Germ cell differentiation in the marmoset (Callithrix jacchus) duringfetal and neonatal life closely parallels that in the human. Hum Reprod2008;23:2755–65.

43. McLachlan RI, Rajpert-De Meyts E, Hoei-Hansen CE, de Kretser DM,Skakkebaek NE. Histological evaluation of the human testis–approachesto optimizing the clinical value of the assessment: mini review. Hum Reprod2007;22:2–16.

44. Rebourcet D, O’Shaughnessy PJ, Pitetti JL, Monteiro A, O’Hara L, Milne L,et al. Sertoli cells control peritubular myoid cell fate and support adult Leydigcell development in the prepubertal testis. Development 2014;141:2139–49.

45. Rebourcet D, O’Shaughnessy PJ, Monteiro A, Milne L, Cruickshanks L,Jeffrey N, et al. Sertoli cells maintain Leydig cell number and peritubu-lar myoid cell activity in the adult mouse testis. PLoS One 2014;9:e105687.

46. Hager M, Gawlik K, Nystrom A, Sasaki T, Durbeej M. Laminin {alpha}1 chaincorrects male infertility caused by absence of laminin {alpha}2 chain. Am JPathol 2005;167:823–33.

12 VOL. - NO. - / - 2020



12991300130113021303130413051306130713081309131013111312131313141315131613171318131913201321132213231324132513261327132813291330133113321333133413351336133713381339134013411342134313441345134613471348134913501351135213531354135513561357

13581359136013611362136313641365136613671368136913701371137213731374137513761377137813791380138113821383138413851386138713881389139013911392139313941395139613971398139914001401140214031404140514061407140814091410141114121413141414151416

SUPPLEMENTAL FIGURE 1

web4C=FPO

Different staining patterns defined for extracellular matrix (ECM) proteins.Van Saen. Tubular niche in Klinefelter testes. Fertil Steril 2020.

VOL. - NO. - / - 2020 12.e1



14171418141914201421142214231424142514261427142814291430143114321433143414351436143714381439144014411442144314441445144614471448144914501451145214531454145514561457145814591460146114621463146414651466146714681469147014711472147314741475

14761477147814791480148114821483148414851486148714881489149014911492149314941495149614971498149915001501150215031504150515061507150815091510151115121513151415151516151715181519152015211522152315241525152615271528152915301531153215331534


web4C=FPO

Overview of different profiles of antim€ullerian hormone (AMH) and androgen receptor (AR) expression during testicular development from fetaluntil peripubertal age in testicular biopsies from control (C) and Klinefelter syndrome (KS) biopsies.Van Saen. Tubular niche in Klinefelter testes. Fertil Steril 2020.

12.e2 VOL. - NO. - / - 2020



15351536153715381539154015411542154315441545154615471548154915501551155215531554155515561557155815591560156115621563156415651566156715681569157015711572157315741575157615771578157915801581158215831584158515861587158815891590159115921593

15941595159615971598159916001601160216031604160516061607160816091610161116121613161416151616161716181619162016211622162316241625162616271628162916301631163216331634163516361637163816391640164116421643164416451646164716481649165016511652


web4C=FPO

Extracellular matrix (ECM) protein expression in (A–D) prepubertal control and (F–H) Klinefelter syndrome (KS) biopsies. ECM protein expression in(I–L) peripubertal control and (M–X) KS biopsies. (M–P) Testicular biopsy from a 14-year-old boy with KS shows mostly normal tubules Q6with normalECM expression patterns. (Q–T) Testicular biopsy from another 14-year-old boy with KS shows tubules with thickened membrane and hyalinizedtubules. (U–X) Testicular biopsy from a 16-year-old boywith KSwithmainly hyalinized tubules. Inserts markedwith represent amagnification of therespective picture. All ECM proteins were already expressed before puberty. C ¼ control; NT ¼ normal tubule; TM ¼ tubule with thickenedmembrane. *Hyalinized tubule.Van Saen. Tubular niche in Klinefelter testes. Fertil Steril 2020.

VOL. - NO. - / - 2020 12.e3



16531654165516561657165816591660166116621663166416651666166716681669167016711672167316741675167616771678167916801681168216831684168516861687168816891690169116921693169416951696169716981699170017011702170317041705170617071708170917101711

17121713171417151716171717181719172017211722172317241725172617271728172917301731173217331734173517361737173817391740174117421743174417451746174717481749175017511752175317541755175617571758175917601761176217631764176517661767176817691770

dear author, please note that changes made in the …...patients with sertoli cell only (sco)...

Documents