Dynamics, ultrastructure and gene expression of human in vitro organized testis cells from testicular sperm extraction biopsies

Dynamics, ultrastructure and gene expression of human in vitro organized testis cells from... Abstract STUDY QUESTION Is it possible to induce in vitro reorganization of primary human testis cells from testicular sperm extraction (TESE) biopsies, maintain their long-term cultivation in a 2D system and identify cellular compositions? SUMMARY ANSWER In vitro reorganization of primary human testis cells from TESE biopsies and their long-term cultivation on uncoated cell culture dishes is feasible and the cellular compositions can be uncovered through gene expression and microscopic analyses. WHAT IS KNOWN ALREADY It has been shown in the rodent model that mixtures of testicular cell types are able to reassemble into clusters when cultivated on different kinds of surfaces or three-dimensional matrices. Two recent publications demonstrated the ability of primary human testicular cells to assemble into testicular organoids and their cultivation for a period of 3–4 weeks. STUDY DESIGN SIZE, DURATION Primary human testis cells from TESE biopsies from 16 patients were reorganized in vitro and the clusters were cultivated long term on uncoated cell culture dishes, providing a solid ground for in vitro spermatogenesis. Gene expression analysis as well as fluorescence/transmission electron microscopy (TEM) were employed to uncover the cellular composition of the clusters. PARTICIPANTS/MATERIALS, SETTING, METHODS Testis biopsies from adult, normogonadotropic patients displaying full spermatogenesis (n = 11), hypospermatogenesis (n = 2), predominantly full spermatogenesis with some hypospermatogenic tubules (n = 1), meiotic arrest (n = 1) or mixed atrophy (n = 1) were enzymatically digested and dispersed cells were cultivated on 96-well plates or chamber dishes as aggregate-free cell suspensions. Time-lapse imaging of cluster formation was performed over a period of 48 h. For receptor tyrosine kinase inhibition of cluster formation, cells were treated twice with K252a within 2–3 days. Immunofluorescence staining and confocal microscopy was carried out on clusters after 1–3 weeks of cultivation to identify the presence of Sertoli cells (SC) (SOX9), peritubular myoid cells (SMA), Leydig cells (LC) (STAR), undifferentiated spermatogonia (FGFR3), differentiating spermatogonia/spermatocytes (DDX4) and postmeiotic germ cells (PRM1). Single clusters from four patients and a pool of eight larger clusters from another patient were manually picked and subjected to quantitative real-time PCR to evaluate the presence of SC (SOX9, AR), LC (INSL3, STAR, HSD3B1), peritubular myoid cells (ACTA2), fibroblasts (FSP1), endothelial cells (CD34), macrophages (CD68), undifferentiated spermatogonia (FGFR3), differentiating spermatogonia/spermatocytes (DDX4) and postmeiotic germ cells (PRM1). Finally, an ultrastructural investigation was conducted based on TEM of clusters from six different patients, among them 3-month cultivated large clusters from two patients. MAIN RESULTS AND THE ROLE OF CHANCE Quantitative PCR-based analysis of single-picked testicular cell clusters identified SC, peritubular myoid cells, endothelial cells, fibroblasts, macrophages, spermatids and LC after 1, 2 or 3 weeks or 3 months of cultivation. Immunofluorescence positivity for SC and peritubular myoid cells corroborated the presence of these two kinds of testis niche cells. In addition, round as well as elongated spermatids were frequently encountered in 1 and 2 weeks old clusters. Transmission electron microscopical classification confirmed all these cell types together with a few spermatogonia. Macrophages were found to be of the proinflammatory M1 subtype, as revealed by CD68+/CD163−/IL6+ expression. Time-lapse imaging uncovered the specific dynamics of cluster fusion and enlargement, which could be prevented by addition of protein kinase inhibitor K252a. LARGE SCALE DATA N/A. LIMITATIONS REASON FOR CAUTION Cell composition of the clusters varied based on the spermatogenic state of the TESE patient. Although spermatids could be observed with all applied methods, spermatogonia were only detected by TEM in single cases. Hence, a direct maintenance of these germ cell types by our system in its current state cannot be postulated. Moreover, putative dedifferentiation and malignant degeneration of cells in long-term cluster cultivation needs to be investigated in the future. WIDER IMPLICATIONS OF THE FINDINGS This work demonstrates that the reorganization of testicular cells can be achieved with TESE biopsies obtained from men enroled in a standard clinical assisted reproduction program. The formed clusters can be cultivated for at least 3 months and are composed, to a large extent, of the most important somatic cell types that are essential to support spermatogenesis. These findings may provide the cellular basis for advances in human in vitro spermatogenesis and/or the possibility for propagation of spermatogonia within a natural stem cell niche-like environment. STUDY FUNDING AND COMPETING INTERESTS The project was funded by a DFG grant to K.v.K. (KO 4769/2-1). The authors declare they have no conflicts of interest. testis, cell cluster formation, organoid, time-lapse, in vitro spermatogenesis Introduction Testicular cells reorganizing under cell culture conditions may constitute a viable alternative to organ culture and could form the basis of in vitro spermatogenesis. Morphogenetic dynamics of in vitro cultivated testis cells are known from rodent studies (Gassei et al., 2008; Pan et al., 2013; Yokonishi et al., 2013; Mäkelä et al., 2014; Reuter et al., 2014; Zhang et al., 2014; Alves-Lopes et al., 2017), where mixtures of testicular cell types are able to reassemble into clusters when cultivated on different kinds of surfaces or three-dimensional matrices (matrigel, nanogrids, collagen sponges). Within these de-novo formed aggregates, Sertoli cells (SC), peritubular myoid cells, fibroblasts and germ cells exist up to the meiotic stage. Baert et al. (2017) demonstrated the ability of primary human testicular cells, derived from whole testes of bilateral orchidectomized ± 75-year-old men and biopsies of a pubertal boy, to repopulate a human decellularized cadaveric testicular matrix. In a parallel setup, multi-layered spheroid structures were formed in a scaffold-free, agarose-layer containing system. Moreover, Leydig cells (LC) and SC constitutively secreted testosterone and Inhibin B, respectively, which could not be stimulated by hCG or FSH. In their study, testis somatic cells with preserved functionality as well as proliferating spermatogonia (SPG) were still detectable after a cell culture period of 4 weeks. Also recently, Pendergraft et al. (2017) used a hanging drop cell culture system for human testicular organoid formation to test in vitro gonadotoxicity. For this purpose, spermatogonial (spg) stem cells, as well as SC and LC from testes of three 56–61-year-old patients were isolated, and SC and LC were lentivirally immortalized and subsequently cultured under addition of solubilized extracellular matrix components. After initial formation on cell culture Day 2, organoids were maintained in 96-well ultra-low attachment U-bottom plates. By employing qPCR and immunofluorescent staining on organoid culture Day 23, markers for undifferentiated and differentiating SPG and postmeiotic germ cells as well as somatic cells were detected. In this work, we demonstrate that cluster formation of primary human testis cells can be achieved on simple uncoated cell culture dishes using small biopsies (~2 × 30 mg) obtained from adult (28–58 years), normogonadotropic patients presenting with full spermatogenesis (FS), hypospermatogenesis (HS), meiotic arrest (MA) or mixed atrophy (FS plus Sertoli cell-only (SCO) tubules). The clusters behave extremely dynamically by constantly merging to form new and larger clusters, as revealed by time-lapse imaging, and this formation can be prevented by addition of protein kinase inhibitor K252a. To identify the cellular composition of the clusters, we applied semithin-sectioning, transmission electron microscopy (TEM) and three-dimensional-construction of cluster models from immunofluorescence (IF) laser-scanning confocal microscopy. To validate the identified cell types, a panel of somatic- and germ-cell specific transcripts served for qPCR-based quantitation. Materials and Methods Patients, testicular biopsies and ethical approval Testis biopsies were surgically extracted parallel to therapeutic testicular sperm extraction (TESE) and diagnostic purposes (Ježek et al., 1998; Schulze et al. 1999; Feig et al., 2007) from 16 adult (28–58 years) normogonadotropic (FSH: 2.2–5.8 IU/ml; LH: 2.24–6.49 IU/ml; T: 2.43–13.39 ng/ml; volume per testis: 15–28 ml) patients. Specifically these were (i) 11 patients presenting with FS (clinical indication: postvasectomized, obstructive azoospermia or CBAVD), (ii) two patients with HS, (iii) one patient with predominantly FS and some HS tubules, (iv) one patient with MA and (v) one patient with mixed atrophy (FS and SCO tubules). Spermatogenic stages of the patients were classified microscopically during test-TESE as follows: FS = > 10 spermatozoa/field of view, normal thickness of tubular epithelium and open tubulus lumen; HS = complete spermatogenesis, with only few mature spermatids (1–10 spermatozoa/field of view); MA = no spermatids, many spermatocytes; mixed atrophy = only one patient with tubules presenting FS and SCO (only SC, no germ cells). Informed and written consent as well as Ethical Committee Approval (WF-007/11 and WF-005/13) were obtained, and the study was performed accordant to the Declaration of Helsinki ethical principles. Preparation of cell suspensions from testis biopsies Two 30 mg fragments of testis tissue per patient were collected separately in 1 ml DMEM (Life Technologies Gibco, Paisley, UK) and stored at 37°C, before enzymatic digestion with Collagenase D (1 mg/ml; Roche, Mannheim, Germany) was performed at 37°C for 30 min with shaking every 10 min. The reaction was stopped through the removal of collagenase by pipetting, addition of 600 μl 37°C prewarmed HBSS (Life Technologies Gibco, Paisley, UK) and three 6 min centrifugations at 300g and 37°C. The tissue was transferred to a 1 ml 37°C preheated DMEM-containing glass Petri dish and minced by means of a microscopy scissor. Subsequent filtration through Falcon® 70 μm-cell strainers (BD Biosciences, Franklin Lakes, USA) generated aggregate-free cell suspensions. For volume restriction, cells were centrifuged at 300g for 10 min. Cell culture For cell culture experiments, biopsies were processed fresh directly after surgery. For each of the 16 patients, aggregate-free cell suspensions were obtained by pooling the two biopsies (from one testis or both testes of the same patient) and these were subsequently split into 18 aliquots for cultivation in uncoated wells of 96-plastic-well-plates (F-bottom). Alternatively, eight aliquots were employed in case of cultivation in Permanox® plastic chambers. For both, incubation was performed at 35°C under 5% CO2 in a total volume of 200 or 500 μl DMEM, respectively. Cell culture media were supplemented with 15% Knockout serum replacement (KO SR; Gibco Invitrogen, Carlsbad, CA, USA), 0.1 mM 2-Mercaptoethanol, 2 mM l-Glutamin (both from Gibco Life Technologies, Carlsbad, CA, USA), 0.1 mM 100× MEM non-essential amino acids (Gibco Invitrogen, Carlsbad, CA, USA), 1% Penicillin/Streptamycin (PAA, Pasching, Austria), 40 ng/ml EGF (Life Technologies, Carlsbad, CA, USA), 20 ng/ml FGFa (Peprotech, Rocky Hill, NJ, USA), 20 ng/ml FGFb (Gibco, Invitrogen, Carlsbad, CA, USA), 20 ng/ml FGF9, 100 ng/ml GDNF and 10 ng/ml IGF (all from Peprotech, Rocky Hill, NJ, USA). Cell feeding was carried out under complete removal of the cell culture medium including the non-adherent cells (with exception of qPCR patient 3, TEM for the same patient and qPCR patient 5, where the cell sediment was re-transferred at each time-point of feeding) every second or third day. Photo documentation was performed in the same time interval prior to feeding. Time-lapse imaging Cells were cultured in Lab-Tek® eight well Permanox® chambers (# 177 445, Thermo Fisher Scientific Inc., Waltham, MA, USA) for the first experiment or in Cellstar 35 mm-cell culture dishes (# 627 160; Greiner Bio-One GmbH, Frickenhausen, Germany) for the second experiment. Photos were taken every 10 min over a period of 48 h with a Keyence BZ-9000 microscope using ×40 magnification. During the experiment, samples were maintained in an incubator at 35°C and 5% CO2 under water-saturated atmosphere. Kinase inhibition Cells were cultured on 96-well-plates as described above. For receptor tyrosine kinase inhibition, cells were treated twice with 5 nM–5 μM K252a (# K1639, Sigma, St. Louis, MO, USA), initially at time of plating and then at first feeding 2–3 days later. Finally, cells were cultured without feeding for another 4 days. As a solvent control, cell culture medium without kinase inhibitors but DMSO in the same concentration (0.5%) was used. The experiment was carried out for four different patients with two to four technical replicates. IF staining IF staining was carried out on clusters cultivated in Permanox® chambers after 1 or 2 weeks of cell culture and on cytospin preparations of 1, 2 and 3 weeks 96-well plates-cultivated clusters. Specimens were Bouin-fixed (1 h, 4°C) and the first staining steps were performed employing the whole mount staining protocol of Ehmcke and Schlatt (2008) with some modifications. In short, samples were rinsed with 1× TBS buffer, treated with 1 M HCl for HCl retrieval (15 min, room temperature (RT) each) and washed with distilled water (3 min). To permeabilize cellular membranes, complete dehydration (1 h 85%, 1 h 95%, overnight 100% ethanol) was followed by complete rehydration (1 h 95%, 1 h 85% and overnight or 3–6 h 70% ethanol). Permeabilization (20 min, 1% Triton X-100 in 1× TBS, RT) and rinsing (20 min, 1× TBS, RT) were continued by the blocking of unspecific antibody binding (1 h, 1× TBS plus 2% normal goat serum, humid chamber, RT). Primary antibody staining for DDX4, FGFR3, PRM1, SMA, SOX9 and STAR with antibodies as listed in Supplemental Table S1 was conducted for 60–72 h (4°C, humid chamber), and further permeabilization and rinsing (see above) were carried out. Secondary antibody staining (Alexa Fluorochromes 488 and 555, dilution 1:200) was achieved by incubation for 4 h (RT, humid and dark chamber) prior to another permeabilization/rinsing step. Cell nuclei were displayed by using DNA dye DRAQ5 (#4084 S, Cell Signaling Technology, Danvers, MA, USA; dilution 1:200). Clusters incubated under omission of the primary antibodies served as negative controls. Samples were mounted and stored at 4°C until evaluation on a Zeiss Confocor 2 confocal scanning system consisting of an Axiovert 200 M inverted microscope equipped with LSM/FCS software (Carl Zeiss, Jena, Germany) and a Zeiss Axiocam digital camera. Alternatively, a Leica TCS SP5 confocal scanning system consisting of a Leica DMI 6000 inverted microscope equipped with LEICA LAS AF software was employed. Acquired confocal fluorescence signals were confirmed to be bleed-through-free between channels. Quantitative PCR Single, 1 and 2 weeks cultivated clusters (n = 8–10) from each of four patients (patient 1, HS; patient 2, FS; patient 3, FS in left testis and FS mixed with SCO tubules in right testis; patient 5, FS with some HS tubules) were micromanipulator-picked, transferred into a 0.5 ml Eppendorf DNA LoBind Tube® (Eppendorf AG, Hamburg, Germany) and immediately frozen at minus 20°C. A pool of eight larger clusters (3 weeks culture time on 96-well plates) from another patient (patient 4, FS) was manually picked using a 1000 μl-pipettor attached to a heat-extruded Pasteur pipette and frozen as above. After thawing, 10 μl DNA-free water (Qiagen, Venlo, Netherlands) were added before the clusters were heat-lysed 4 min at 90°. Second strand synthesis was performed using SuperScript™ III First-Strand Synthesis SuperMix for quantitative real-time PCR (qRT-PCR) (Invitrogen, Carlsbad, CA, USA) according to the manufacturer’s protocol. Samples of 2 μl undiluted cDNA were used for each qRT-PCR reaction, which was performed on a LightCycler™ 1.0 (Roche, Basel, Switzerland) employing SYBR® Premix DimerEraser™ (Takara Bio Europe, Saint-Germain-en-Laye, France). Working procedures were conducted according to manufacturer’s protocols, but with primer concentrations of 0.5 μM in a total volume of 20 μl. Primers (Eurofins Genomics, Ebersberg, Germany) were designed using Primer3 (http://www.ncbi.nlm.nih.gov/tools/primer-blast/), intron-overlapping, and captured all reference nucleotide sequences listed on NCBI with the exception of insulin like 3 (INSL3; only for amplification of transcript variant 1). Primer sequences and amplicon sizes are summarized in Supplemental Table 2. PCR cycling conditions were: 60 cycles of 10 s at 95°C, 20 s at 60°C and 30 s at 72°C. As a control transcript, ribosomal protein S27 (RPS27) was employed (Thorrez et al., 2008; von Kopylow et al., 2016) to test (i) whether clusters were correctly deposited in the tube, (ii) if the reverse transcription reaction was successful and (iii) for normalization to relative expression values. Equal-volume amounts of cluster-free cell culture medium were also tested for background transcript contamination. Specificity evaluation of amplicons was carried out by melting point determination, sizing by capillary electrophoresis (Bioanalyzer; Agilent Technologies, Santa Clara, CA, USA) and sequencing (Eurofins Genomics, Ebersberg, Germany). Differential expression between 1 and 2 weeks of culture and corresponding P-values was calculated by unequal variance T-tests on the log2-transformed and normal distributed relative expression values. Epon embedding, semithin section technique and TEM Clusters from Permanox chambers after 1 or 2 weeks culture time were obtained by up-and-down pipetting of the cell culture medium and 15 min centrifugation (300g, 37°C) for cluster sedimentation. Two individual 3-month cultivated 600–700 μm large clusters from two different patients were picked with a pair of tweezers after a cell culture period of 3 months in 96-well plates. The sediment or single large clusters were fixed in 1 ml phosphate-buffered 5.5% glutaraldehyde (pH 7.4) for 1 h at 4°C, centrifuged again (15 min, 300g) and further processed as described in Holstein and Wulfhekel (1971). In short, samples were postfixed in 1% osmium tetroxide in 0.1 m PBS, dehydrated, included in a drop of 2% liquid gelatine before embedding in Epon 812, then cut in semithin sections (1 μm), and stained with toluidine blue and pyronine (9 + 1). For TEM, 600–800 Å thin sections were prepared and double-contrasted (1% uranyl acetate, 20 min; 1% Reynolds lead citrate, 5 min). Specimens were photographed using a Philips CM100 transmission electron microscope (Philips, Amsterdam, Netherlands) equipped with a Quemesa Olympus TEM CCD camera (Olympus, Tokyo, Japan). Results Cluster formation of human primary testis cells and cluster development as investigated by light microscopy and time-lapse imaging After the first cell culture days under the described conditions, signs of cluster formation became apparent, with initiation time points depending on the kind of cell culture dish (e.g. 5 days in uncoated plastic wells of 96-well plates; Fig. 1C). From this time point, the formation of distinct round clusters (Fig. 1D) with subsequent cluster enlargement (Fig. 1E) was observed. To uncover the enlargement of clusters in more detail, time-lapse experiments were performed from cell culture Day 11 to cell culture Day 13 or 14 (Fig. 1, F1–F8). Here, rolling motion dynamics of the clusters became evident, in which clusters in near vicinity fused to larger, but again round shaped, clusters. Sometimes, adherent and extending cells captured clusters in a protruding, lasso-like manner, thereby causing cluster fusion with fast, only minute-elapsing constriction movements (Fig. 1, F6–F7; also compare Supplemental Video 1). Figure 1 View large Download slide Cluster formation chronology of cultured human primary testis cells. (A, B) Starting point (Day 0) of cell culture displaying homogeneous dispersed cells. (C) First signs of cluster formation on Day 5. (D) Appearance of distinct round clusters (arrow) on Day 7. (E) Enlarged clusters (arrows) are visible on Day 11. At this time point, fast cluster fusion could be recorded by time-lapse imaging after 0 (F1), 11 (F2), 16 (F3), 18 (F4), 21.5 (F5), 24.5 (F6), 24.7 (F7) and 31.5 (F8) hours (compare Supplemental Video 1). (A–E) Patient with MA (96-well plate). F1–F8: Patient with FS (8-well Permanox chamber). Figure 1 View large Download slide Cluster formation chronology of cultured human primary testis cells. (A, B) Starting point (Day 0) of cell culture displaying homogeneous dispersed cells. (C) First signs of cluster formation on Day 5. (D) Appearance of distinct round clusters (arrow) on Day 7. (E) Enlarged clusters (arrows) are visible on Day 11. At this time point, fast cluster fusion could be recorded by time-lapse imaging after 0 (F1), 11 (F2), 16 (F3), 18 (F4), 21.5 (F5), 24.5 (F6), 24.7 (F7) and 31.5 (F8) hours (compare Supplemental Video 1). (A–E) Patient with MA (96-well plate). F1–F8: Patient with FS (8-well Permanox chamber). Three-dimensional-cellular organization, transcript- and protein-based analysis and ultrastructure of single testicular clusters Epon-embedded semithin sections of clusters were morphologically analysed to disclose the types and arrangement of cluster-composing cells. Light-microscopic evaluation showed elongated peritubular-like cells at the outer cluster borders, whereas cell nuclei in the cluster interior often resembled those of SC (Fig. 2A, B). For an initial protein-based identification of these two cell types, antibodies against the SC marker SOX9 and the peritubular myoid cell marker smooth-muscle actin (SMA) were employed in IF whole cluster staining. Laser-scanning confocal fluorescence microscopy and Z-stack imaging demonstrated both markers in 1 (Fig. 2C–E), 2 (Fig. 2F, G; also compare Supplemental Video 2) and 3 (data not shown) week old clusters. Based on these results, SC as well as peritubular myoid cells are likely to contribute to cluster formation in samples from patients with FS (Fig. 2C–E), as well as HS (Fig. 2F, G). Clusters from another set of two patients (FS and FS with some HS tubules) were employed for further evaluation with the spermatid marker PRM1, undifferentiated SPG marker FGFR3, differentiating SPG and spermatocyte marker DDX4 and LC marker STAR. These were used in conjunction with SOX9 staining to investigate the localization of these cell types relative to SC. Whereas no FGFR3, DDX4 and STAR protein was detected at the protein level in cluster-included cells (Supplemental Fig. S1A–F), round and/or elongated spermatids were identified by means of PRM1 expression in 1 and 2 weeks old clusters of both patients (Fig. 3). Nevertheless, culture dishes after 1 week of cluster culture contained some DDX4-positive 2D but not three-dimensional germ cell aggregates (Supplemental Fig. S1G, H). IF staining of testis sections for the germ cell markers DDX4, FGFR3 and PRM1, as well as for the LC marker STAR and the SC marker SOX9, showed expected and specific signals in the corresponding cell type, i.e. differentiated germ cells, SC, LC, spermatids and undifferentiated SPG (Supplemental Fig. S2A–D). Figure 2 View largeDownload slide Semithin sections and immunofluorescence staining of whole clusters for smooth-muscle actin (SMA) and SOX9. (A, B) Semithin sections of two different glutaraldehyde-fixed, toluidine blue/pyronine-stained clusters from a patient with MA, cultivated for 11 days on 96-well plates. Note the different morphologies of the cluster-composing cells. Arrows denote elongated peritubular-like cells at the outer border of the clusters. (C) One of the upper layers from a 1-week-old cluster, and (D) the bottom layer of adherent cells (optical sectioning) from the same patient presenting FS. (E) Middle layer of a 1-week-old cluster from another patient with FS. (F, G) One of the upper and lower layers of a 2-week-old cluster from a patient with HS. (H) Negative control. (C–H) DNA staining by DRAQ5. Figure 2 View largeDownload slide Semithin sections and immunofluorescence staining of whole clusters for smooth-muscle actin (SMA) and SOX9. (A, B) Semithin sections of two different glutaraldehyde-fixed, toluidine blue/pyronine-stained clusters from a patient with MA, cultivated for 11 days on 96-well plates. Note the different morphologies of the cluster-composing cells. Arrows denote elongated peritubular-like cells at the outer border of the clusters. (C) One of the upper layers from a 1-week-old cluster, and (D) the bottom layer of adherent cells (optical sectioning) from the same patient presenting FS. (E) Middle layer of a 1-week-old cluster from another patient with FS. (F, G) One of the upper and lower layers of a 2-week-old cluster from a patient with HS. (H) Negative control. (C–H) DNA staining by DRAQ5. Figure 3 View largeDownload slide Immunofluorescence double staining of whole clusters for PRM1 and SOX9. (A, B) Two different 1-week-old clusters from a patient with FS. (C, D) Two different 2-week-old clusters from a patient presenting predominantly FS but also some tubules with HS. DNA co-staining by DRAQ5. Arrows point to two round spermatids (A), longitudinal optical sections of testicular spermatozoa (B, C), and an elongated spermatid head (D). Figure 3 View largeDownload slide Immunofluorescence double staining of whole clusters for PRM1 and SOX9. (A, B) Two different 1-week-old clusters from a patient with FS. (C, D) Two different 2-week-old clusters from a patient presenting predominantly FS but also some tubules with HS. DNA co-staining by DRAQ5. Arrows point to two round spermatids (A), longitudinal optical sections of testicular spermatozoa (B, C), and an elongated spermatid head (D). For a further elucidation of other potential cluster composing cell types, a transcript-based analysis of single-picked clusters was performed. In an initial step, qPCR reactions were conducted for the control transcript RPS27 to check for successful reverse transcription and for normalization. A cell type-specific analysis on 8–10 single clusters per patient/time-point then uncovered the presence of SC from positivity of AR (androgen receptor) and SOX9, peritubular myoid cells by ACTA2, fibroblasts by FSP1 (fibroblast-specific protein-1; official gene name: S100 calcium binding protein A4, S100A4; Mäkelä et al., 2014), endothelial cells by CD34 (Sidney et al., 2014), postmeiotic germ cells by protamine 1 (PRM1), and macrophages by CD68/CD163/IL6 (Martinez et al., 2006; Goluža et al., 2014; Chávez-Galán et al., 2015; Lucket-Chastain et al., 2016) (Fig. 4A–D). Contrasting this, no LC-specific INSL3 expression was detected in single clusters after 1 and 2 weeks cell culture (Fig. 4A–C). In the 3-week cultured eight-cluster pool, INSL3 was detectable only in traces (data not shown). The expression of the two other Leydig cell markers steroidogenic acute regulatory protein (STAR) and hydroxy-delta-5-steroid dehydrogenase, 3 beta- and steroid delta-isomerase 1 (HSD3B1; Fig. 4C, E) indicated the presence of dedifferentiated/immature LC in the clusters (Benton et al., 1995; Raucci et al., 2014), however, the STAR results must be viewed with caution because of possible co-expression in SC (Gregory and DePhilip, 1998; Ishikawa et al., 2005). Spermatogonial and spermatocytic marker DDX4 (Fig. 4A–C) and undifferentiated SPG / potential spg stem cell marker FGFR3 (tested on five 1-week-old clusters; data not shown) were not detectable. SOX9 and AR were present in single clusters at 25–70 and 25–50%, respectively, without (Fig. 4A, B) and at 9–25 vs 50–73% with (Fig. 4C) re-centrifugation. The latter high proportion of SC is explicable by cell-type enrichment in cases of many tubules with the SCO phenotype (mixed atrophy) (Cappallo-Obermann et al., 2013). The lower frequency of SOX9-positive qPCR reactions compared to more positive AR qPCR reactions may partially be explained by the high GC-content of SOX9 (locally up to 95%) that renders cDNA synthesis problematic. ACTA2 expression increased from week 1 to week 2, indicating a proliferative effect on myoid cells. As there was no PRM1, ACTA2 and FSP1 detectable in the three weeks cultivated eight-cluster pool (Fig. 4D), we do not rule out a potential PCR-inhibiting effect of larger cell culture medium remnants from the pooling procedure. In this pool, CD34 was detected, indicating the existence of a few endothelial cells whose transcripts are otherwise under the detection limit. This assumption was supported by endothelial cells in TEM images (see below). PRM1 expression in samples of patient 1 and 2 declined after week 1, pointing to the maintenance of postmeiotic germ cells in the cell culture dish. A longer detectable PRM1 expression (2 weeks) in the two re-centrifuged samples (patients 3 and 5; Fig. 4C, E) correlates well with the presence of sperm heads found in the TEM preparations of the same patient (Supplemental Fig. S4) and the IF stainings (Fig. 3). A more detailed quantitative qPCR analysis of the single clusters from the four patients (Fig. 5) revealed unaltered cell proportions between 1 and 2 weeks of culture for SC (SOX9, AR), fibroblasts (FSP1), LC (STAR) and spermatids (PRM1) (under the assumption that the RPS27-normalized expression values correlate to some extent to the cell numbers). In contrast, peritubular myoid cells (ACTA2) displayed a significant (P < 0.05) increase either in cellular proportions or from differential regulation. Hence, for this transcript, the quantitative data confirm the qualitative observations in Fig. 4, although a rise in cellular proportions need not necessarily coincide with increased overall transcript levels, as is the case for FSP1. Figure 4 View largeDownload slide Cellular composition of 1, 2 and 3 weeks old clusters as revealed by cell-specific mRNA expression. (A) Clusters from patient 1 (HS), (B) patient 2 (FS) and (C) patient 3 (FS, mixed atrophy) were cultured for 1 (left bars) and 2 (right bars) weeks in Permanox chambers. (D) Clusters from patient 4 (FS) were cultured for 3 weeks in 96-well plates. (E) Clusters from patient 5 (FS, HS) were cultured for 1 (left bars) and 2 (right bars) weeks in 96-well plates. Single clusters (n = 8–10) from patients 1–3 and 5 as well as a pool of eight clusters from patient 4 were interrogated by qPCR for cell-specific transcripts (Sertoli cells: SOX9, AR; postmeiotic germ cells: PRM1; peritubular cells: ACTA2; fibroblasts: FSP1; macrophages: CD68; endothelial cells: CD34; spermatogonia / spermatocytes: DDX4; Leydig cells: INSL3, STAR, HSD3B1; housekeeping gene: RPS27). Abbreviations: n.d., not determined; negative, no expression detected. Figure 4 View largeDownload slide Cellular composition of 1, 2 and 3 weeks old clusters as revealed by cell-specific mRNA expression. (A) Clusters from patient 1 (HS), (B) patient 2 (FS) and (C) patient 3 (FS, mixed atrophy) were cultured for 1 (left bars) and 2 (right bars) weeks in Permanox chambers. (D) Clusters from patient 4 (FS) were cultured for 3 weeks in 96-well plates. (E) Clusters from patient 5 (FS, HS) were cultured for 1 (left bars) and 2 (right bars) weeks in 96-well plates. Single clusters (n = 8–10) from patients 1–3 and 5 as well as a pool of eight clusters from patient 4 were interrogated by qPCR for cell-specific transcripts (Sertoli cells: SOX9, AR; postmeiotic germ cells: PRM1; peritubular cells: ACTA2; fibroblasts: FSP1; macrophages: CD68; endothelial cells: CD34; spermatogonia / spermatocytes: DDX4; Leydig cells: INSL3, STAR, HSD3B1; housekeeping gene: RPS27). Abbreviations: n.d., not determined; negative, no expression detected. Figure 5 View largeDownload slide qPCR-based relative expression values of SOX9, AR, PRM1, ACTA2, FSP1 and STAR from 1 and 2 weeks cultured testis clusters. Relative expression was calculated by the delta-Ct method using RPS27 as the normalizing gene to correct for cDNA synthesis bias. P-values were calculated by unequal variance T-tests on the log2-transformed data (significant in case of asterisks) between the two time points. Numbers (x/y) reflect the number of positive qPCR reactions (dots) in relation to all qPCR reactions. These figures include all Cq values obtained from patients 1 to 3 and 5 (compare Fig. 4). Figure 5 View largeDownload slide qPCR-based relative expression values of SOX9, AR, PRM1, ACTA2, FSP1 and STAR from 1 and 2 weeks cultured testis clusters. Relative expression was calculated by the delta-Ct method using RPS27 as the normalizing gene to correct for cDNA synthesis bias. P-values were calculated by unequal variance T-tests on the log2-transformed data (significant in case of asterisks) between the two time points. Numbers (x/y) reflect the number of positive qPCR reactions (dots) in relation to all qPCR reactions. These figures include all Cq values obtained from patients 1 to 3 and 5 (compare Fig. 4). The presence of the macrophage marker CD68 in 25–73% of single clusters (Fig. 4A–C) indicated a possible role of this cell type for either positive or negative cluster integrity. To shed light on the involved macrophage subtypes, we tested for the proinflammatory M1 macrophage cytokine IL6 and the anti-inflammatory M2 macrophage marker CD163 on 1 and 2-week-old clusters of two patients. Here, we found IL6 expression in 16/19 (Cq values: 31.23–38.53) and CD163 expression in 1/16 clusters (Cq value: 37.66), pointing to macrophages of a predominantly M1 subtype. Currently we cannot distinguish between native M1 macrophages or those that arise from the M2 to M1 subtype reprogramming (Mosser and Edwards, 2008), where the latter would imply a severe effect of the mechanical disruption and/or cluster formation. Here, the application of a macrophage marker panel may offer a more robust discrimination of involved subtypes, however, the amount of RNA isolated from single picked clusters does not suffice for this approach. Quantitative PCR analysis of cluster-free cell culture medium was always negative, demonstrating the absence of cell-released transcripts in the cell culture dishes. The identification of the clusters’ cell types via TEM based on their ultrastructure in overview (Fig. 6) and in more detail (Fig. 7, Supplemental Figs S3 and S4) is founded on TEM images depicted in mostly older literature. It is somewhat complicated by putative dedifferentiation of cultured cells after removal from their natural context, e.g. LC that barely preserve their typical lipid droplets. Contrasting this, SC (Holstein and Roosen-Runge, 1981; Schulze, 1984; Fig. 7G, Supplemental Fig. S3) were clearly identifiable in almost all cases based on their typical morphology. They were present in clusters of all patients independent of spermatogenic state, cell culture duration and type of cell culture dish, hence, appearing as the most dominating and stable cell type in the clusters. This outcome fits well to the importance of SC for the formation of rat testicular organoids within a three-layer gradient system (Alves-Lopes et al., 2017). Figure 6 View largeDownload slide TEM depiction of cluster cell typology of three different patients. (A, B) Clusters cultivated for 11 days on 96-well plates from a patient with MA. (C, D) Clusters from two different patients with FS cultivated for 7 days in Permanox chambers. Clusters in (D) were re-transfered to the same well at each time-point of feeding. Abbreviations: LC, Leydig cell; EC, endothelial cell; PC, peritubular cell; SC, Sertoli cell; FB, fibroblast; SPG, spermatogonium; in this case Apale, higher magnification in Fig. 7A. Asterisks denote layers of collagen fibres. Figure 6 View largeDownload slide TEM depiction of cluster cell typology of three different patients. (A, B) Clusters cultivated for 11 days on 96-well plates from a patient with MA. (C, D) Clusters from two different patients with FS cultivated for 7 days in Permanox chambers. Clusters in (D) were re-transfered to the same well at each time-point of feeding. Abbreviations: LC, Leydig cell; EC, endothelial cell; PC, peritubular cell; SC, Sertoli cell; FB, fibroblast; SPG, spermatogonium; in this case Apale, higher magnification in Fig. 7A. Asterisks denote layers of collagen fibres. Figure 7 View largeDownload slide Higher magnification TEM micrographs of cluster cell types from three different patients. (A) Apale SPG, and (B–D) spermatids from 1 week cultivated clusters from a patient with FS (96-well plate). (E) Peritubular cell, (F) endothelial cell, (G) Sertoli cell, (H) Leydig cell from 11 days cultivated clusters from a patient with MA (96-well plate). (I) Macrophage from a 11-week cultivated cluster from a patient with FS (Permanox chamber). White arrow in (A) points to a typical Apale nucleolus, peripherally located to the nuclear membrane. The nucleus of the cell appears homogeneous pale and shows no rarefaction-zone in contrast to Adark SPG (Clermont, 1963, 1966, 1972). Black arrow in (I) points to a primary lysosome as a macrophage-typical cellular structure. Figure 7 View largeDownload slide Higher magnification TEM micrographs of cluster cell types from three different patients. (A) Apale SPG, and (B–D) spermatids from 1 week cultivated clusters from a patient with FS (96-well plate). (E) Peritubular cell, (F) endothelial cell, (G) Sertoli cell, (H) Leydig cell from 11 days cultivated clusters from a patient with MA (96-well plate). (I) Macrophage from a 11-week cultivated cluster from a patient with FS (Permanox chamber). White arrow in (A) points to a typical Apale nucleolus, peripherally located to the nuclear membrane. The nucleus of the cell appears homogeneous pale and shows no rarefaction-zone in contrast to Adark SPG (Clermont, 1963, 1966, 1972). Black arrow in (I) points to a primary lysosome as a macrophage-typical cellular structure. In addition, Fig. 7 and Supplemental Fig. S4 exemplify different cluster-composing cell types, i.e. peritubular myoid cells (PC; Ross, 1967; Maekawa et al., 1995, 1996; Johnson et al., 2010), endothelial cells (EC; Maekawa et al., 1996; Lee and Cheng, 2004; ; Johnson et al., 2010), fibroblasts (FB; Ross, 1967), macrophages (MP; Ross, 1967; Johnson et al., 2010; Goluža et al., 2014; Dempster et al., 2011), and spermatozoa (Holstein and Roosen-Runge, 1981). Infrequently, LC (Mori and Christensen, 1980; Schulze, 1984; Ezeasor, 1985; Johnson et al., 2010), and in the case of a re-centrifuged sample, single SPG (Holstein and Roosen-Runge, 1981) were encountered. It cannot be excluded that these SPG, only observed in the periphery of the cluster cells (Fig. 6D), are a consequence of the experimental setup. Extracellular matrix, as presented by collagen fibrils (Clermont, 1958; Ross, 1967; Lee and Cheng, 2004; Fig. 6B, Supplemental Fig. S5A–D) and typical actin filaments in the cytoplasm of a peritubular myoid cell (Supplemental Fig. S5D–F) were also identified. Prevention of cluster formation by kinase inhibition K252a is an alkaloid from the prokaryote Nonomuraea longicatena that is often used as a selective inhibitor of tyrosine protein kinase activity, e.g. of neurotrophin receptors (Tapley et al., 1992). Here, in analogy to the induced disturbance of rat in vitro SC aggregation (Gassei et al., 2008), we were able to prevent cluster formation of primary adult testis cells with K252a concentrations of 500 nM and higher (Fig. 8A–F). At these concentrations, cluster formation and length/number of protrusions are significantly reduced. The corresponding solvent controls (0.5% DMSO; Fig. 8G) and no-solvent controls (Fig. 8H) displayed normal cluster formation. Figure 8 View largeDownload slide Disruption of cluster formation by protein kinase inhibition (6 days after plating). (A–F) Addition of kinase inhibitor K252a in concentrations ranging from 5 nM to 1 μM to the culture medium. (G) Solvent control (0.5% DMSO), (H) no solvent control (0% DMSO). Cells from the testis of a patient displaying FS. Figure 8 View largeDownload slide Disruption of cluster formation by protein kinase inhibition (6 days after plating). (A–F) Addition of kinase inhibitor K252a in concentrations ranging from 5 nM to 1 μM to the culture medium. (G) Solvent control (0.5% DMSO), (H) no solvent control (0% DMSO). Cells from the testis of a patient displaying FS. Discussion In this work, we describe a stable experimental protocol for in vitro reorganization of human testicular cells obtained from rice grain-sized TESE biopsies. The aggregate-free cell suspensions were cultured in simple cell culture dishes without matrix support, similar to the mouse study of Mäkelä et al. (2014), but in contrast to the agarose-containing cultivation system of a recent human study (Baert et al., 2017). The observed structures display a distinct and intrinsic pattern of cluster formation and fusion that possibly recapitulates the cellular morphogenetic recognition program, a well-known process when organ models consist of multiple cell types (Yin et al., 2016), and which occurs in embryonic testicular cord development (Gassei et al., 2008). Our time-lapse-based investigation, in extension to previous findings in the rodents (Pan et al., 2013; Mäkelä et al., 2014), reveals a transition from an aggregating phase to a fusing phase when clusters acquire a size of ~60–80 μm. This is accompanied by a highly dynamic process where large cellular protrusions capture clusters in near vicinity, suggesting the involvement of chemotactic compounds as observed in embryoid body formation (Jiang et al., 2007). The impairment of cluster formation and absence of protrusions after protein kinase inhibition may therefore (among other potential effects) be attributed to the modulation of the chemotactic machinery. To date, we cannot pinpoint the affected cell type, but the observed protrusions are to some extent analogous to the NGF-induced, K252a-inhibitable neurite-like processes of the PC12 cell line (Tomokiyo et al., 2012). Our results indicate putative dedifferentiation of SC and LC entrapped in the clusters. Along these lines, while in adult SC SOX9-protein localization pertains mostly to nuclei (Morais da Silva et al., 1996), in the investigated clusters, it always appeared cytoplasmic. Expression of SOX9 during cell proliferation (Chakravarty et al., 2011) does not normally occur in the male adult, but can be encountered in the reversion to an immature embryonal-like state of male gonadal somatic cells (De Santa Barbara et al., 2000) or in a disrupted microtubular network (Malki et al., 2005). The latter may be caused from experimental procedures such as the in vitro culture conditions. A recent study, in which mouse spg stem cell culture conditions were tested on human testicular cells, resulted partly in testicular clusters and a similar cytoplasmatic SOX9 localization in SC (Medrano et al., 2016) as described in this work. Dedifferentiation of supporting LC is indicated by the absence of classical lipid droplets, absence of INSL3 mRNA expression, downregulation of HSD3B1 mRNA and the presence of STAR (Medrano et al., 2016), where in our case the latter was only significantly expressed at the mRNA level. A possible cause for the dedifferentiation of these two somatic cell types might be the absence of vital stimulatory factors such as hCG and FSH, and indeed, preliminary cluster stimulations with 50 ng/ml FSH (Gassei et al., 2008) resulted in up to 50% increase in the cluster diameter (data not shown). However, other works revealed an overall hCG stimulatability only until cell culture Day 2 (Pendergraft et al., 2017), or not at all, even under optimized three-dimensional culture conditions (Baert et al., 2017). Here, the addition of spg survival factors may provide more favourable conditions for in vitro spermatogenesis (Sato et al., 2011) or the proliferation of spg stem cells for male infertiliy treatment (Mulder et al., 2016). Equally important as the lack of essential stimulatory factors is an impairment of cluster or organoid function by the presence of proinflammatory molecules. The large proportion of M1 macrophages and corresponding CD68/IL6 expression in our clusters, with the absence of CD163, indicated a proinflammatory situation in the culture system. Similar elevated IL6 secretion was also encountered in one and four week old human testicular organoids (Baert et al., 2017). Moreover, testicular organoid formation in a three-dimensional model in rat was disrupted by addition of proinflammatory interleukin 1 alpha and tumour necrosis factor alpha (Alves-Lopes et al., 2017). These observations suggest that future research on testicular clusters or organoids should focus at least partly on the suppression of proinflammatory signalling (e.g. by non-steroidal anti-inflammatory drugs), as this may promote normal cell function (Sawazaki et al., 2014). Our employed methodology on single picked clusters enables a highly sensitive detection and classification of the clusters’ cellular composition, in which qPCR-based classification with cell-specific transcripts is supplemented by IF stainings and confirmed by the ultrastructure-based classification of single cells by TEM. With respect to germ cells, a few SPG were encountered in border regions of re-centrifuged clusters earlier than 2 weeks, as opposed to a considerably higher quantity found elsewhere (Baert et al., 2017; Pendergraft et al., 2017). Also, some remnant spermatozoa entrapped in the clusters could be identified by PRM1-positive qPCR, IF and TEM, demonstrating, to some extent, the potential of human testis-derived clusters to propagate postmeiotic germ cells. It is worth mentioning that the ability of cluster to form per se was independent of the patient’s spermatogenic state and reflected the cell types of the testicular origin. One exception to the rule consisted of a patient presenting FS where no cluster formation occurred and where SC contained many lipid droplets and lysosomes. Although a single case, this observation might emphasize the role of morphologically normal SC for cluster formation. We conclude that the in vitro reorganization of low amounts of human testicular cells obtained from normogonadotropic men is feasible as well as highly reproducible under 2D culture conditions. The tested culture system is suited to maintain clusters and SC for up to 3 months, but in its current state, far from optimal for the survival of germ cells. Because SPG were only detected on the surface or outside of clusters, it is not unlikely that the compact three-dimensional structure is insufficient to serve as a niche for long-term germ cell maintenance, propagation and differentiation. Hence, three-dimensional culture systems may provide more favourable conditions for the generation of human testicular organoids as a consequence of providing compartimentalization and concentration gradients (Alves-Lopes et al., 2017). The development of an optimal system for testicular organoids could be useful for in vitro spermatogenesis or for the support of spg stem cell proliferation, and therefore for infertility treatment in daily clinical practice. Supplementary data Supplementary data are available at Molecular Human Reproduction online. Acknowledgements We would like to thank Gabriele Hahn and Barbara Holstermann for technical assistance, Prof. Martin Bergmann (JLU, Giessen), Prof. Andreas Meinhardt (JLU, Giessen), Prof. Davor Ježek (University of Zagreb) and Mina Mincheva (CERA, Münster) for fruitful discussions, and Dr Rudolph Reimer (Heinrich Pette Institute, Hamburg), Dr Sabine Windhorst and Dr Markus Geissen as well as the UKE Microscopy Imaging Facility for supplying technical equipment. We are grateful to Prof. Stefan W. Schneider for his support. Authors’ roles W.S. and A.S. delivered biopsy samples from a therapeutic testicular sperm extraction and classified the samples via light microscopy. K.v.K., M.S., E.S. and B.R. carried out the experiments. K.v.K., M.S., E.S. and A.N.S. analysed the data. K.v.K., W.S., A.N.S. and S.S. interpreted the results. K.v.K. and A.N.S. wrote the article. Funding Deutsche Forschungsgemeinschaft (DFG) grant to K.v.K. (KO 4769/2-1). Conflicts of interest None declared. References Alves-Lopes JP, Söder O, Stukenborg JB. Testicular organoid generation by a novel in vitro three-layer gradient system. Biomaterials  2017; 130: 76– 89. Google Scholar CrossRef Search ADS PubMed  Baert Y, De Kock J, Alves-Lopes JP, Söder O, Stukenborg JB, Goossens E. Primary human testicular cells self-organize into organoids with testicular properties. Stem Cell Rep  2017; 8( 1): 30– 38. Google Scholar CrossRef Search ADS   Benton L, Shan LX, Hardy MP. Differentiation of adult Leydig cells. J Steroid Biochem Mol Biol  1995; 53( 1−6): 61– 68. Google Scholar CrossRef Search ADS PubMed  Cappallo-Obermann H, Feig C, Schulze W, Spiess AN. Fold-change correction values for testicular somatic transcripts in gene expression studies of human spermatogenesis. Hum Reprod  2013; 28( 3): 590– 598. Google Scholar CrossRef Search ADS PubMed  Chakravarty G, Rider B, Mondal D. Cytoplasmic compartimentalization of SOX9 abrogates the growth arrest response of breast cancer cells that can be rescued by trichostatin A treatment. Cancer Biol Ther  2011; 11( 1): 71– 83. Google Scholar CrossRef Search ADS PubMed  Chávez-Galán L, Olleros ML, Vesin D, Garcia I. Much more than M1 and M2 macrophages, there are also CD169+ and TCR+ macrophages. Front Immunol  2015; 6: 263. Google Scholar PubMed  Clermont Y. Contractile elements in the limiting membrane of the seminiferous tubules of the rat. Exp Cell Res  1958; 15: 438– 440. Google Scholar CrossRef Search ADS PubMed  Clermont Y. The cycle of the seminiferous epithelium in man. Am J Anat  1963; 112: 35– 51. Google Scholar CrossRef Search ADS PubMed  Clermont Y. Renewal of spermatogonia in man. Am J Anat  1966; 118: 509– 524. Google Scholar CrossRef Search ADS PubMed  Clermont Y. Kinetics of spermatogenesis in mammals: seminiferous epithelium cycle and spermatogonial renewal. Physiol Rev  1972; 52( 1): 198– 236. Google Scholar CrossRef Search ADS PubMed  De Santa Barbara P, Moniot B, Poulat F, Berta P. Expression and subcellular localization of SF-1, SOX9, WT1, and AMH proteins during early human testicular development. Dev Dyn  2000; 217: 293– 298. Google Scholar CrossRef Search ADS PubMed  Dempster AG, Molteno ACB, Bevin TH, Thompson AM. Otago glaucoma surgery outcome study: electron microscopy of capsules around Molteno implants. Glaucoma  2011; 52: 8300– 8309. Ehmcke J, Schlatt S. Identification and characterization of spermatogonial subtypes and their expansion in whole mounts and tissue sections from primate testes. Methods Mol Biol  2008; 450: 109– 118. Google Scholar CrossRef Search ADS PubMed  Ezeasor DN. Light and electron microscopical observations on the Leydig cells of the scrotal and abdominal testes of naturally unilateral cryptorchid West African dwarf goats. J Anat  1985; 141: 27– 40. Google Scholar PubMed  Feig C, Kirchhoff C, Ivell R, Naether O, Schulze W, Spiess AN. A new paradigm for profiling testicular gene expression during normal and disturbed human spermatogenesis. Mol Hum Reprod  2007; 13( 1): 33– 43. Google Scholar CrossRef Search ADS PubMed  Gassei K, Ehmcke J, Schlatt S. Initiation of testicular tubulogenesis is controlled by neurotrophic tyrosine receptor kinases in a three-dimensional Sertoli cell aggregation assay. Reprod  2008; 136( 4): 459– 469. Google Scholar CrossRef Search ADS   Goluža T, Boscanin A, Cvetko J, Kozina V, Bernat MM, Kasum M, Kaštelan Z, Ježek D. Macrophages and Leydig cells in testicular biopsies of azoospermic men. Biomed Res Int  2014; 2014: 828697. Google Scholar CrossRef Search ADS PubMed  Gregory CW, DePhilip RM. Detection of steroidogenic acute regulatory protein (StAR) in mitochondria of cultured rat Sertoli cells incubated with follicle-stimulating hormone. Biol Reprod  1998; 58( 2): 470– 474. Google Scholar CrossRef Search ADS PubMed  Holstein AF, Roosen-Runge EC Atlas of Human Spermatogenesis . Berlin: Grosse Verlag, 1981. Holstein AF, Wulfhekel U. Die Semidünnschnitt-Technik als Grundlage für eine cytologische Beurteilung der Spermatogenese des Menschen. Andrologia  1971; 3: 65– 69. Google Scholar CrossRef Search ADS   Ishikawa T, Hwang K, Lazzarino D, Morris PL. Sertoli cell expression of steroidogenic acute regulatory protein-related lipid transfer 1 and 5 domain-containing proteins and sterol regulatory element binding protein-1 are interleukin-1beta regulated by activation of c-Jun N-terminal kinase and cyclooxygenase-2 and cytokine induction. Endocrinology  2005; 146( 12): 5100– 5111. Google Scholar CrossRef Search ADS PubMed  Ježek D, Knuth UA, Schulze W. Successful testicular sperm extraction (TESE) in spite of high serum follicle stimulating hormone and azoospermia: correlation between testicular morphology, TESE results, semen analysis and serum hormone values in 103 infertile men. Hum Reprod  1998; 13( 5): 1230– 1234. Google Scholar CrossRef Search ADS PubMed  Jiang S, Fu Y, Williams J, Wood JA, Pandarinathan L, Avraham S, Makriyannis A, Avraham S, Avraham HK. Expression and function of cannabinoid receptors CB1 and CB2 and their cognate cannabinoid ligands in murine embryonic stem cells. PLoS One  2007; 7: e641. Google Scholar CrossRef Search ADS   Johnson L, Welsh TH, Curley KO, Johnston CE. Anatomy and physiology of the male reproductive system and potential targets of toxicants. Compr Toxicol (Second Edition), Elsevier Ltd  2010; 11: 5– 59. Lee NPY, Cheng CY. Ectoplasmatic specialization, a testis-specific cell-cell actin-based adherens junction type: is this a potential target for male contraceptive development? Hum Reprod Update  2004; 10( 4): 349– 369. Google Scholar CrossRef Search ADS PubMed  Lucket-Chastain L, Calhoun K, Schartz T, Gallucci RM. IL-6 influences the balance between M1 and M2 macrophages in a mouse model of irritant contact dermatitis. J Immunol  2016; 196( 1): 196.17. Maekawa M, Kamimura K, Nagano T. Peritubular myoid cells in the testis: their structure and function. Arch Histol Cytol  1996; 59( 1): 1– 13. Google Scholar CrossRef Search ADS PubMed  Maekawa M, Kazama H, Kamimura K, Nagano T. Changes in the arrangement of actin filaments in myoid cells and Sertoli cells of rat testes during postnatal development and after experimental cryptorchidism. Anat Rec  1995; 241: 59– 69. Google Scholar CrossRef Search ADS PubMed  Malki S, Berta P, Poulat F, Boizet-Bonhoure B. Cytoplasmic retention of the sex-determining factor SOX9 via the microtubule network. Exp Cell Res  2005; 309: 468– 475. Google Scholar CrossRef Search ADS PubMed  Martinez FO, Gordon S, Locati M, Mantovani A. Transcriptional profiling of the human monocyte-to-macrophage differentiation and polarization: new molecules and patterns of gene expression. J Immunol  2006; 177: 7303– 7311. Google Scholar CrossRef Search ADS PubMed  Medrano JV, Rombaut C, Simon C, Pellicer A, Goossens E. Human spermatogonial stem cells display limited proliferation in vitro under mouse spermatogonial stem cell culture conditions. Fertil Steril  2016; 106( 6): 1539– 1549.e8. Google Scholar CrossRef Search ADS PubMed  Morais da Silva S, Hacker A, Harley V, Goodfellow P, Swain A, Lovell-Badge R. Sox9 expression during gonadal development implies a conserved role for the gene in testis differentiation in mammals and birds. Nat Genet  1996; 14: 62– 68. Google Scholar CrossRef Search ADS PubMed  Mori H, Christensen AK. Morphometric analysis of Leydig cells in the normal rat testis. J Cell Biol  1980; 84: 340– 354. Google Scholar CrossRef Search ADS PubMed  Mosser DM, Edwards JP. Exploring the full spectrum of macrophage activation. Nat Rev Immunol  2008; 8( 12): 958– 959. Google Scholar CrossRef Search ADS PubMed  Mulder CL, Zheng Y, Jan SZ, Struijk RB, Repping S, Hamer G, van Pelt AM. Spermatogonial stem cell autotransplantation and germline genomic editing: a future cure for spermatogenic failure and prevention of transmission of genomic diseases. Hum Reprod Update  2016; 22( 5): 561– 573. Google Scholar CrossRef Search ADS PubMed  Mäkelä JA, Toppari J, Rivero-Müller A, Ventelä S. Reconstruction of mouse cellular microenvironments in long-term seminiferous tubule culture. PLoS One  2014; 9( 3): e90088. Google Scholar CrossRef Search ADS PubMed  Pan F, Lifeng C, Schlatt S. Effects of nanostructures and mouse embryonic stem cells on in vitro morphogenesis of rat testicular cords. PLoS One  2013; 8( 3): e60054. Google Scholar CrossRef Search ADS PubMed  Pendergraft SS, Sadri-Ardekani H, Atala A, Bishop CE. Three-dimensional testicular organoid: a novel tool for the study of human spermatogenesis and gonadotoxicity in vitro. Biol Reprod  2017; 96( 3): 720– 732. Google Scholar CrossRef Search ADS PubMed  Raucci F, D’Aniello A, Di Fiore MM. Stimulation of androgen production by D-aspartate through the enhancement of StAR, P450scc and 3β-HSD mRNA levels in vivo rat testis and in culture of immature rat Leydig cells. Steroids  2014; 84: 103– 110. Google Scholar CrossRef Search ADS PubMed  Reuter K, Ehmcke J, Stukenborg JB, Simoni M, Damm OS, Redmann K, Schlatt S, Wistuba J. Reassembly of somatic cells and testicular organogenesis in vitro. Tissue Cell  2014; 46( 1): 86– 96. Google Scholar CrossRef Search ADS PubMed  Ross MH. The fine structure and development of the peritubular contractile cell component in the seminiferous tubules of the mouse. Am J Anat  1967; 121: 523– 558. Google Scholar CrossRef Search ADS PubMed  Sato T, Katagiri K, Gohbara A, Inoue K, Ogonuki N, Ogura A, Kubota Y, Ogawa T. In vitro production of functional sperm in cultured neonatal mouse testes. Natural  2011; 471: 504– 507. Google Scholar CrossRef Search ADS   Sawazaki R, Ishihara T, Usui S, Hayashi E, Tahara K, Hoshino T, Higuchi A, Nakamura S, Tsubota K, Mizushima T. Diclofenac protects cultured human corneal epithelial cells against hyperosmolarity and ameliorates corneal surface damagein a rat model of dry eye. Invest Ophthalmol Vis Sci  2014; 55( 4): 2547– 2556. Google Scholar CrossRef Search ADS PubMed  Schulze C. Sertoli cells and Leydig cells in man. Adv Anat Embryol Cell Biol  1984; 88: 1– 104. Google Scholar CrossRef Search ADS PubMed  Schulze W, Thoms F, Knuth UA. Testicular sperm extraction: comprehensive analysis with simultaneously performed histology in 1418 biopsies from 766 subfertile men. Hum Reprod  1999; 14( Suppl 1): 82– 96. Google Scholar CrossRef Search ADS PubMed  Sidney LE, Branch MJ, Dunphy SE, Dua HS, Hopkinson A. Concise review: evidence for CD34 as a common marker for diverse progenitors. Stem cells  2014; 32: 1380– 1389. Google Scholar CrossRef Search ADS PubMed  Tapley P, Lamballe F, Barbacid M. K252a is a selective inhibitor of the tyrosine protein kinase activity of the trk family of oncogenes and neurotrophon receptors. Oncogene  1992; 7( 2): 371– 381. Google Scholar PubMed  Thorrez L, Van Deun K, Tranchevent LC, Van Lommel L, Engelen K, Marchal K, Moreau Y, Van Mechelen I, Schuit F. Using ribosomal protein genes as reference: a tale of caution. PLoS One  2008; 3: e1854. Google Scholar CrossRef Search ADS PubMed  Tomokiyo A, Maeda H, Fujii S, Monnouchi S, Wada N, Kono K, Yamamoto N, Koori K, Teramatsu Y, Akamine A. A multipotent clonal human periodontal ligament cell line with neural crest cell phenotypes promotes neurocytic differentiation, migration, and survival. Cell Physiol  2012; 227( 5): 2040– 2050. Google Scholar CrossRef Search ADS   von Kopylow K, Schulze W, Salzbrunn A, Spiess AN. Isolation and gene expression analysis of single potential human spermatogonial stem cells. Mol Hum Reprod  2016; 22( 4): 229– 239. Google Scholar CrossRef Search ADS PubMed  Yin X, Mead BE, Safaee H, Langer R, Karp JM, Levy O. Engineering stem cell organoids. Cell Stem Cell  2016; 18( 1): 25– 38. Google Scholar CrossRef Search ADS PubMed  Yokonishi T, Sato T, Katagiri K, Komeya M, Kubota Y, Ogawa T. In vitro reconstruction of mouse seminiferous tubules supporting germ cell differentiation. Biol Reprod  2013; 89( 1): 15. Google Scholar CrossRef Search ADS PubMed  Zhang J, Hatakeyama J, Eto K, Abe S. Reconstruction of a seminiferous tubule-like structure in a 3 dimensional culture system of re-aggregated mouse neonatal testicular cells within a collagen matrix. Gen Comp Endocrinol  2014; 205: 121– 132. Google Scholar CrossRef Search ADS PubMed  © The Author(s) 2018. 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@oup.com http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Molecular Human Reproduction Oxford University Press

Dynamics, ultrastructure and gene expression of human in vitro organized testis cells from testicular sperm extraction biopsies

Loading next page...
 
/lp/ou_press/dynamics-ultrastructure-and-gene-expression-of-human-in-vitro-YWyqQkLXz0
Publisher
Oxford University Press
Copyright
© The Author(s) 2018. 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@oup.com
ISSN
1360-9947
eISSN
1460-2407
D.O.I.
10.1093/molehr/gax070
Publisher site
See Article on Publisher Site

Abstract

Abstract STUDY QUESTION Is it possible to induce in vitro reorganization of primary human testis cells from testicular sperm extraction (TESE) biopsies, maintain their long-term cultivation in a 2D system and identify cellular compositions? SUMMARY ANSWER In vitro reorganization of primary human testis cells from TESE biopsies and their long-term cultivation on uncoated cell culture dishes is feasible and the cellular compositions can be uncovered through gene expression and microscopic analyses. WHAT IS KNOWN ALREADY It has been shown in the rodent model that mixtures of testicular cell types are able to reassemble into clusters when cultivated on different kinds of surfaces or three-dimensional matrices. Two recent publications demonstrated the ability of primary human testicular cells to assemble into testicular organoids and their cultivation for a period of 3–4 weeks. STUDY DESIGN SIZE, DURATION Primary human testis cells from TESE biopsies from 16 patients were reorganized in vitro and the clusters were cultivated long term on uncoated cell culture dishes, providing a solid ground for in vitro spermatogenesis. Gene expression analysis as well as fluorescence/transmission electron microscopy (TEM) were employed to uncover the cellular composition of the clusters. PARTICIPANTS/MATERIALS, SETTING, METHODS Testis biopsies from adult, normogonadotropic patients displaying full spermatogenesis (n = 11), hypospermatogenesis (n = 2), predominantly full spermatogenesis with some hypospermatogenic tubules (n = 1), meiotic arrest (n = 1) or mixed atrophy (n = 1) were enzymatically digested and dispersed cells were cultivated on 96-well plates or chamber dishes as aggregate-free cell suspensions. Time-lapse imaging of cluster formation was performed over a period of 48 h. For receptor tyrosine kinase inhibition of cluster formation, cells were treated twice with K252a within 2–3 days. Immunofluorescence staining and confocal microscopy was carried out on clusters after 1–3 weeks of cultivation to identify the presence of Sertoli cells (SC) (SOX9), peritubular myoid cells (SMA), Leydig cells (LC) (STAR), undifferentiated spermatogonia (FGFR3), differentiating spermatogonia/spermatocytes (DDX4) and postmeiotic germ cells (PRM1). Single clusters from four patients and a pool of eight larger clusters from another patient were manually picked and subjected to quantitative real-time PCR to evaluate the presence of SC (SOX9, AR), LC (INSL3, STAR, HSD3B1), peritubular myoid cells (ACTA2), fibroblasts (FSP1), endothelial cells (CD34), macrophages (CD68), undifferentiated spermatogonia (FGFR3), differentiating spermatogonia/spermatocytes (DDX4) and postmeiotic germ cells (PRM1). Finally, an ultrastructural investigation was conducted based on TEM of clusters from six different patients, among them 3-month cultivated large clusters from two patients. MAIN RESULTS AND THE ROLE OF CHANCE Quantitative PCR-based analysis of single-picked testicular cell clusters identified SC, peritubular myoid cells, endothelial cells, fibroblasts, macrophages, spermatids and LC after 1, 2 or 3 weeks or 3 months of cultivation. Immunofluorescence positivity for SC and peritubular myoid cells corroborated the presence of these two kinds of testis niche cells. In addition, round as well as elongated spermatids were frequently encountered in 1 and 2 weeks old clusters. Transmission electron microscopical classification confirmed all these cell types together with a few spermatogonia. Macrophages were found to be of the proinflammatory M1 subtype, as revealed by CD68+/CD163−/IL6+ expression. Time-lapse imaging uncovered the specific dynamics of cluster fusion and enlargement, which could be prevented by addition of protein kinase inhibitor K252a. LARGE SCALE DATA N/A. LIMITATIONS REASON FOR CAUTION Cell composition of the clusters varied based on the spermatogenic state of the TESE patient. Although spermatids could be observed with all applied methods, spermatogonia were only detected by TEM in single cases. Hence, a direct maintenance of these germ cell types by our system in its current state cannot be postulated. Moreover, putative dedifferentiation and malignant degeneration of cells in long-term cluster cultivation needs to be investigated in the future. WIDER IMPLICATIONS OF THE FINDINGS This work demonstrates that the reorganization of testicular cells can be achieved with TESE biopsies obtained from men enroled in a standard clinical assisted reproduction program. The formed clusters can be cultivated for at least 3 months and are composed, to a large extent, of the most important somatic cell types that are essential to support spermatogenesis. These findings may provide the cellular basis for advances in human in vitro spermatogenesis and/or the possibility for propagation of spermatogonia within a natural stem cell niche-like environment. STUDY FUNDING AND COMPETING INTERESTS The project was funded by a DFG grant to K.v.K. (KO 4769/2-1). The authors declare they have no conflicts of interest. testis, cell cluster formation, organoid, time-lapse, in vitro spermatogenesis Introduction Testicular cells reorganizing under cell culture conditions may constitute a viable alternative to organ culture and could form the basis of in vitro spermatogenesis. Morphogenetic dynamics of in vitro cultivated testis cells are known from rodent studies (Gassei et al., 2008; Pan et al., 2013; Yokonishi et al., 2013; Mäkelä et al., 2014; Reuter et al., 2014; Zhang et al., 2014; Alves-Lopes et al., 2017), where mixtures of testicular cell types are able to reassemble into clusters when cultivated on different kinds of surfaces or three-dimensional matrices (matrigel, nanogrids, collagen sponges). Within these de-novo formed aggregates, Sertoli cells (SC), peritubular myoid cells, fibroblasts and germ cells exist up to the meiotic stage. Baert et al. (2017) demonstrated the ability of primary human testicular cells, derived from whole testes of bilateral orchidectomized ± 75-year-old men and biopsies of a pubertal boy, to repopulate a human decellularized cadaveric testicular matrix. In a parallel setup, multi-layered spheroid structures were formed in a scaffold-free, agarose-layer containing system. Moreover, Leydig cells (LC) and SC constitutively secreted testosterone and Inhibin B, respectively, which could not be stimulated by hCG or FSH. In their study, testis somatic cells with preserved functionality as well as proliferating spermatogonia (SPG) were still detectable after a cell culture period of 4 weeks. Also recently, Pendergraft et al. (2017) used a hanging drop cell culture system for human testicular organoid formation to test in vitro gonadotoxicity. For this purpose, spermatogonial (spg) stem cells, as well as SC and LC from testes of three 56–61-year-old patients were isolated, and SC and LC were lentivirally immortalized and subsequently cultured under addition of solubilized extracellular matrix components. After initial formation on cell culture Day 2, organoids were maintained in 96-well ultra-low attachment U-bottom plates. By employing qPCR and immunofluorescent staining on organoid culture Day 23, markers for undifferentiated and differentiating SPG and postmeiotic germ cells as well as somatic cells were detected. In this work, we demonstrate that cluster formation of primary human testis cells can be achieved on simple uncoated cell culture dishes using small biopsies (~2 × 30 mg) obtained from adult (28–58 years), normogonadotropic patients presenting with full spermatogenesis (FS), hypospermatogenesis (HS), meiotic arrest (MA) or mixed atrophy (FS plus Sertoli cell-only (SCO) tubules). The clusters behave extremely dynamically by constantly merging to form new and larger clusters, as revealed by time-lapse imaging, and this formation can be prevented by addition of protein kinase inhibitor K252a. To identify the cellular composition of the clusters, we applied semithin-sectioning, transmission electron microscopy (TEM) and three-dimensional-construction of cluster models from immunofluorescence (IF) laser-scanning confocal microscopy. To validate the identified cell types, a panel of somatic- and germ-cell specific transcripts served for qPCR-based quantitation. Materials and Methods Patients, testicular biopsies and ethical approval Testis biopsies were surgically extracted parallel to therapeutic testicular sperm extraction (TESE) and diagnostic purposes (Ježek et al., 1998; Schulze et al. 1999; Feig et al., 2007) from 16 adult (28–58 years) normogonadotropic (FSH: 2.2–5.8 IU/ml; LH: 2.24–6.49 IU/ml; T: 2.43–13.39 ng/ml; volume per testis: 15–28 ml) patients. Specifically these were (i) 11 patients presenting with FS (clinical indication: postvasectomized, obstructive azoospermia or CBAVD), (ii) two patients with HS, (iii) one patient with predominantly FS and some HS tubules, (iv) one patient with MA and (v) one patient with mixed atrophy (FS and SCO tubules). Spermatogenic stages of the patients were classified microscopically during test-TESE as follows: FS = > 10 spermatozoa/field of view, normal thickness of tubular epithelium and open tubulus lumen; HS = complete spermatogenesis, with only few mature spermatids (1–10 spermatozoa/field of view); MA = no spermatids, many spermatocytes; mixed atrophy = only one patient with tubules presenting FS and SCO (only SC, no germ cells). Informed and written consent as well as Ethical Committee Approval (WF-007/11 and WF-005/13) were obtained, and the study was performed accordant to the Declaration of Helsinki ethical principles. Preparation of cell suspensions from testis biopsies Two 30 mg fragments of testis tissue per patient were collected separately in 1 ml DMEM (Life Technologies Gibco, Paisley, UK) and stored at 37°C, before enzymatic digestion with Collagenase D (1 mg/ml; Roche, Mannheim, Germany) was performed at 37°C for 30 min with shaking every 10 min. The reaction was stopped through the removal of collagenase by pipetting, addition of 600 μl 37°C prewarmed HBSS (Life Technologies Gibco, Paisley, UK) and three 6 min centrifugations at 300g and 37°C. The tissue was transferred to a 1 ml 37°C preheated DMEM-containing glass Petri dish and minced by means of a microscopy scissor. Subsequent filtration through Falcon® 70 μm-cell strainers (BD Biosciences, Franklin Lakes, USA) generated aggregate-free cell suspensions. For volume restriction, cells were centrifuged at 300g for 10 min. Cell culture For cell culture experiments, biopsies were processed fresh directly after surgery. For each of the 16 patients, aggregate-free cell suspensions were obtained by pooling the two biopsies (from one testis or both testes of the same patient) and these were subsequently split into 18 aliquots for cultivation in uncoated wells of 96-plastic-well-plates (F-bottom). Alternatively, eight aliquots were employed in case of cultivation in Permanox® plastic chambers. For both, incubation was performed at 35°C under 5% CO2 in a total volume of 200 or 500 μl DMEM, respectively. Cell culture media were supplemented with 15% Knockout serum replacement (KO SR; Gibco Invitrogen, Carlsbad, CA, USA), 0.1 mM 2-Mercaptoethanol, 2 mM l-Glutamin (both from Gibco Life Technologies, Carlsbad, CA, USA), 0.1 mM 100× MEM non-essential amino acids (Gibco Invitrogen, Carlsbad, CA, USA), 1% Penicillin/Streptamycin (PAA, Pasching, Austria), 40 ng/ml EGF (Life Technologies, Carlsbad, CA, USA), 20 ng/ml FGFa (Peprotech, Rocky Hill, NJ, USA), 20 ng/ml FGFb (Gibco, Invitrogen, Carlsbad, CA, USA), 20 ng/ml FGF9, 100 ng/ml GDNF and 10 ng/ml IGF (all from Peprotech, Rocky Hill, NJ, USA). Cell feeding was carried out under complete removal of the cell culture medium including the non-adherent cells (with exception of qPCR patient 3, TEM for the same patient and qPCR patient 5, where the cell sediment was re-transferred at each time-point of feeding) every second or third day. Photo documentation was performed in the same time interval prior to feeding. Time-lapse imaging Cells were cultured in Lab-Tek® eight well Permanox® chambers (# 177 445, Thermo Fisher Scientific Inc., Waltham, MA, USA) for the first experiment or in Cellstar 35 mm-cell culture dishes (# 627 160; Greiner Bio-One GmbH, Frickenhausen, Germany) for the second experiment. Photos were taken every 10 min over a period of 48 h with a Keyence BZ-9000 microscope using ×40 magnification. During the experiment, samples were maintained in an incubator at 35°C and 5% CO2 under water-saturated atmosphere. Kinase inhibition Cells were cultured on 96-well-plates as described above. For receptor tyrosine kinase inhibition, cells were treated twice with 5 nM–5 μM K252a (# K1639, Sigma, St. Louis, MO, USA), initially at time of plating and then at first feeding 2–3 days later. Finally, cells were cultured without feeding for another 4 days. As a solvent control, cell culture medium without kinase inhibitors but DMSO in the same concentration (0.5%) was used. The experiment was carried out for four different patients with two to four technical replicates. IF staining IF staining was carried out on clusters cultivated in Permanox® chambers after 1 or 2 weeks of cell culture and on cytospin preparations of 1, 2 and 3 weeks 96-well plates-cultivated clusters. Specimens were Bouin-fixed (1 h, 4°C) and the first staining steps were performed employing the whole mount staining protocol of Ehmcke and Schlatt (2008) with some modifications. In short, samples were rinsed with 1× TBS buffer, treated with 1 M HCl for HCl retrieval (15 min, room temperature (RT) each) and washed with distilled water (3 min). To permeabilize cellular membranes, complete dehydration (1 h 85%, 1 h 95%, overnight 100% ethanol) was followed by complete rehydration (1 h 95%, 1 h 85% and overnight or 3–6 h 70% ethanol). Permeabilization (20 min, 1% Triton X-100 in 1× TBS, RT) and rinsing (20 min, 1× TBS, RT) were continued by the blocking of unspecific antibody binding (1 h, 1× TBS plus 2% normal goat serum, humid chamber, RT). Primary antibody staining for DDX4, FGFR3, PRM1, SMA, SOX9 and STAR with antibodies as listed in Supplemental Table S1 was conducted for 60–72 h (4°C, humid chamber), and further permeabilization and rinsing (see above) were carried out. Secondary antibody staining (Alexa Fluorochromes 488 and 555, dilution 1:200) was achieved by incubation for 4 h (RT, humid and dark chamber) prior to another permeabilization/rinsing step. Cell nuclei were displayed by using DNA dye DRAQ5 (#4084 S, Cell Signaling Technology, Danvers, MA, USA; dilution 1:200). Clusters incubated under omission of the primary antibodies served as negative controls. Samples were mounted and stored at 4°C until evaluation on a Zeiss Confocor 2 confocal scanning system consisting of an Axiovert 200 M inverted microscope equipped with LSM/FCS software (Carl Zeiss, Jena, Germany) and a Zeiss Axiocam digital camera. Alternatively, a Leica TCS SP5 confocal scanning system consisting of a Leica DMI 6000 inverted microscope equipped with LEICA LAS AF software was employed. Acquired confocal fluorescence signals were confirmed to be bleed-through-free between channels. Quantitative PCR Single, 1 and 2 weeks cultivated clusters (n = 8–10) from each of four patients (patient 1, HS; patient 2, FS; patient 3, FS in left testis and FS mixed with SCO tubules in right testis; patient 5, FS with some HS tubules) were micromanipulator-picked, transferred into a 0.5 ml Eppendorf DNA LoBind Tube® (Eppendorf AG, Hamburg, Germany) and immediately frozen at minus 20°C. A pool of eight larger clusters (3 weeks culture time on 96-well plates) from another patient (patient 4, FS) was manually picked using a 1000 μl-pipettor attached to a heat-extruded Pasteur pipette and frozen as above. After thawing, 10 μl DNA-free water (Qiagen, Venlo, Netherlands) were added before the clusters were heat-lysed 4 min at 90°. Second strand synthesis was performed using SuperScript™ III First-Strand Synthesis SuperMix for quantitative real-time PCR (qRT-PCR) (Invitrogen, Carlsbad, CA, USA) according to the manufacturer’s protocol. Samples of 2 μl undiluted cDNA were used for each qRT-PCR reaction, which was performed on a LightCycler™ 1.0 (Roche, Basel, Switzerland) employing SYBR® Premix DimerEraser™ (Takara Bio Europe, Saint-Germain-en-Laye, France). Working procedures were conducted according to manufacturer’s protocols, but with primer concentrations of 0.5 μM in a total volume of 20 μl. Primers (Eurofins Genomics, Ebersberg, Germany) were designed using Primer3 (http://www.ncbi.nlm.nih.gov/tools/primer-blast/), intron-overlapping, and captured all reference nucleotide sequences listed on NCBI with the exception of insulin like 3 (INSL3; only for amplification of transcript variant 1). Primer sequences and amplicon sizes are summarized in Supplemental Table 2. PCR cycling conditions were: 60 cycles of 10 s at 95°C, 20 s at 60°C and 30 s at 72°C. As a control transcript, ribosomal protein S27 (RPS27) was employed (Thorrez et al., 2008; von Kopylow et al., 2016) to test (i) whether clusters were correctly deposited in the tube, (ii) if the reverse transcription reaction was successful and (iii) for normalization to relative expression values. Equal-volume amounts of cluster-free cell culture medium were also tested for background transcript contamination. Specificity evaluation of amplicons was carried out by melting point determination, sizing by capillary electrophoresis (Bioanalyzer; Agilent Technologies, Santa Clara, CA, USA) and sequencing (Eurofins Genomics, Ebersberg, Germany). Differential expression between 1 and 2 weeks of culture and corresponding P-values was calculated by unequal variance T-tests on the log2-transformed and normal distributed relative expression values. Epon embedding, semithin section technique and TEM Clusters from Permanox chambers after 1 or 2 weeks culture time were obtained by up-and-down pipetting of the cell culture medium and 15 min centrifugation (300g, 37°C) for cluster sedimentation. Two individual 3-month cultivated 600–700 μm large clusters from two different patients were picked with a pair of tweezers after a cell culture period of 3 months in 96-well plates. The sediment or single large clusters were fixed in 1 ml phosphate-buffered 5.5% glutaraldehyde (pH 7.4) for 1 h at 4°C, centrifuged again (15 min, 300g) and further processed as described in Holstein and Wulfhekel (1971). In short, samples were postfixed in 1% osmium tetroxide in 0.1 m PBS, dehydrated, included in a drop of 2% liquid gelatine before embedding in Epon 812, then cut in semithin sections (1 μm), and stained with toluidine blue and pyronine (9 + 1). For TEM, 600–800 Å thin sections were prepared and double-contrasted (1% uranyl acetate, 20 min; 1% Reynolds lead citrate, 5 min). Specimens were photographed using a Philips CM100 transmission electron microscope (Philips, Amsterdam, Netherlands) equipped with a Quemesa Olympus TEM CCD camera (Olympus, Tokyo, Japan). Results Cluster formation of human primary testis cells and cluster development as investigated by light microscopy and time-lapse imaging After the first cell culture days under the described conditions, signs of cluster formation became apparent, with initiation time points depending on the kind of cell culture dish (e.g. 5 days in uncoated plastic wells of 96-well plates; Fig. 1C). From this time point, the formation of distinct round clusters (Fig. 1D) with subsequent cluster enlargement (Fig. 1E) was observed. To uncover the enlargement of clusters in more detail, time-lapse experiments were performed from cell culture Day 11 to cell culture Day 13 or 14 (Fig. 1, F1–F8). Here, rolling motion dynamics of the clusters became evident, in which clusters in near vicinity fused to larger, but again round shaped, clusters. Sometimes, adherent and extending cells captured clusters in a protruding, lasso-like manner, thereby causing cluster fusion with fast, only minute-elapsing constriction movements (Fig. 1, F6–F7; also compare Supplemental Video 1). Figure 1 View large Download slide Cluster formation chronology of cultured human primary testis cells. (A, B) Starting point (Day 0) of cell culture displaying homogeneous dispersed cells. (C) First signs of cluster formation on Day 5. (D) Appearance of distinct round clusters (arrow) on Day 7. (E) Enlarged clusters (arrows) are visible on Day 11. At this time point, fast cluster fusion could be recorded by time-lapse imaging after 0 (F1), 11 (F2), 16 (F3), 18 (F4), 21.5 (F5), 24.5 (F6), 24.7 (F7) and 31.5 (F8) hours (compare Supplemental Video 1). (A–E) Patient with MA (96-well plate). F1–F8: Patient with FS (8-well Permanox chamber). Figure 1 View large Download slide Cluster formation chronology of cultured human primary testis cells. (A, B) Starting point (Day 0) of cell culture displaying homogeneous dispersed cells. (C) First signs of cluster formation on Day 5. (D) Appearance of distinct round clusters (arrow) on Day 7. (E) Enlarged clusters (arrows) are visible on Day 11. At this time point, fast cluster fusion could be recorded by time-lapse imaging after 0 (F1), 11 (F2), 16 (F3), 18 (F4), 21.5 (F5), 24.5 (F6), 24.7 (F7) and 31.5 (F8) hours (compare Supplemental Video 1). (A–E) Patient with MA (96-well plate). F1–F8: Patient with FS (8-well Permanox chamber). Three-dimensional-cellular organization, transcript- and protein-based analysis and ultrastructure of single testicular clusters Epon-embedded semithin sections of clusters were morphologically analysed to disclose the types and arrangement of cluster-composing cells. Light-microscopic evaluation showed elongated peritubular-like cells at the outer cluster borders, whereas cell nuclei in the cluster interior often resembled those of SC (Fig. 2A, B). For an initial protein-based identification of these two cell types, antibodies against the SC marker SOX9 and the peritubular myoid cell marker smooth-muscle actin (SMA) were employed in IF whole cluster staining. Laser-scanning confocal fluorescence microscopy and Z-stack imaging demonstrated both markers in 1 (Fig. 2C–E), 2 (Fig. 2F, G; also compare Supplemental Video 2) and 3 (data not shown) week old clusters. Based on these results, SC as well as peritubular myoid cells are likely to contribute to cluster formation in samples from patients with FS (Fig. 2C–E), as well as HS (Fig. 2F, G). Clusters from another set of two patients (FS and FS with some HS tubules) were employed for further evaluation with the spermatid marker PRM1, undifferentiated SPG marker FGFR3, differentiating SPG and spermatocyte marker DDX4 and LC marker STAR. These were used in conjunction with SOX9 staining to investigate the localization of these cell types relative to SC. Whereas no FGFR3, DDX4 and STAR protein was detected at the protein level in cluster-included cells (Supplemental Fig. S1A–F), round and/or elongated spermatids were identified by means of PRM1 expression in 1 and 2 weeks old clusters of both patients (Fig. 3). Nevertheless, culture dishes after 1 week of cluster culture contained some DDX4-positive 2D but not three-dimensional germ cell aggregates (Supplemental Fig. S1G, H). IF staining of testis sections for the germ cell markers DDX4, FGFR3 and PRM1, as well as for the LC marker STAR and the SC marker SOX9, showed expected and specific signals in the corresponding cell type, i.e. differentiated germ cells, SC, LC, spermatids and undifferentiated SPG (Supplemental Fig. S2A–D). Figure 2 View largeDownload slide Semithin sections and immunofluorescence staining of whole clusters for smooth-muscle actin (SMA) and SOX9. (A, B) Semithin sections of two different glutaraldehyde-fixed, toluidine blue/pyronine-stained clusters from a patient with MA, cultivated for 11 days on 96-well plates. Note the different morphologies of the cluster-composing cells. Arrows denote elongated peritubular-like cells at the outer border of the clusters. (C) One of the upper layers from a 1-week-old cluster, and (D) the bottom layer of adherent cells (optical sectioning) from the same patient presenting FS. (E) Middle layer of a 1-week-old cluster from another patient with FS. (F, G) One of the upper and lower layers of a 2-week-old cluster from a patient with HS. (H) Negative control. (C–H) DNA staining by DRAQ5. Figure 2 View largeDownload slide Semithin sections and immunofluorescence staining of whole clusters for smooth-muscle actin (SMA) and SOX9. (A, B) Semithin sections of two different glutaraldehyde-fixed, toluidine blue/pyronine-stained clusters from a patient with MA, cultivated for 11 days on 96-well plates. Note the different morphologies of the cluster-composing cells. Arrows denote elongated peritubular-like cells at the outer border of the clusters. (C) One of the upper layers from a 1-week-old cluster, and (D) the bottom layer of adherent cells (optical sectioning) from the same patient presenting FS. (E) Middle layer of a 1-week-old cluster from another patient with FS. (F, G) One of the upper and lower layers of a 2-week-old cluster from a patient with HS. (H) Negative control. (C–H) DNA staining by DRAQ5. Figure 3 View largeDownload slide Immunofluorescence double staining of whole clusters for PRM1 and SOX9. (A, B) Two different 1-week-old clusters from a patient with FS. (C, D) Two different 2-week-old clusters from a patient presenting predominantly FS but also some tubules with HS. DNA co-staining by DRAQ5. Arrows point to two round spermatids (A), longitudinal optical sections of testicular spermatozoa (B, C), and an elongated spermatid head (D). Figure 3 View largeDownload slide Immunofluorescence double staining of whole clusters for PRM1 and SOX9. (A, B) Two different 1-week-old clusters from a patient with FS. (C, D) Two different 2-week-old clusters from a patient presenting predominantly FS but also some tubules with HS. DNA co-staining by DRAQ5. Arrows point to two round spermatids (A), longitudinal optical sections of testicular spermatozoa (B, C), and an elongated spermatid head (D). For a further elucidation of other potential cluster composing cell types, a transcript-based analysis of single-picked clusters was performed. In an initial step, qPCR reactions were conducted for the control transcript RPS27 to check for successful reverse transcription and for normalization. A cell type-specific analysis on 8–10 single clusters per patient/time-point then uncovered the presence of SC from positivity of AR (androgen receptor) and SOX9, peritubular myoid cells by ACTA2, fibroblasts by FSP1 (fibroblast-specific protein-1; official gene name: S100 calcium binding protein A4, S100A4; Mäkelä et al., 2014), endothelial cells by CD34 (Sidney et al., 2014), postmeiotic germ cells by protamine 1 (PRM1), and macrophages by CD68/CD163/IL6 (Martinez et al., 2006; Goluža et al., 2014; Chávez-Galán et al., 2015; Lucket-Chastain et al., 2016) (Fig. 4A–D). Contrasting this, no LC-specific INSL3 expression was detected in single clusters after 1 and 2 weeks cell culture (Fig. 4A–C). In the 3-week cultured eight-cluster pool, INSL3 was detectable only in traces (data not shown). The expression of the two other Leydig cell markers steroidogenic acute regulatory protein (STAR) and hydroxy-delta-5-steroid dehydrogenase, 3 beta- and steroid delta-isomerase 1 (HSD3B1; Fig. 4C, E) indicated the presence of dedifferentiated/immature LC in the clusters (Benton et al., 1995; Raucci et al., 2014), however, the STAR results must be viewed with caution because of possible co-expression in SC (Gregory and DePhilip, 1998; Ishikawa et al., 2005). Spermatogonial and spermatocytic marker DDX4 (Fig. 4A–C) and undifferentiated SPG / potential spg stem cell marker FGFR3 (tested on five 1-week-old clusters; data not shown) were not detectable. SOX9 and AR were present in single clusters at 25–70 and 25–50%, respectively, without (Fig. 4A, B) and at 9–25 vs 50–73% with (Fig. 4C) re-centrifugation. The latter high proportion of SC is explicable by cell-type enrichment in cases of many tubules with the SCO phenotype (mixed atrophy) (Cappallo-Obermann et al., 2013). The lower frequency of SOX9-positive qPCR reactions compared to more positive AR qPCR reactions may partially be explained by the high GC-content of SOX9 (locally up to 95%) that renders cDNA synthesis problematic. ACTA2 expression increased from week 1 to week 2, indicating a proliferative effect on myoid cells. As there was no PRM1, ACTA2 and FSP1 detectable in the three weeks cultivated eight-cluster pool (Fig. 4D), we do not rule out a potential PCR-inhibiting effect of larger cell culture medium remnants from the pooling procedure. In this pool, CD34 was detected, indicating the existence of a few endothelial cells whose transcripts are otherwise under the detection limit. This assumption was supported by endothelial cells in TEM images (see below). PRM1 expression in samples of patient 1 and 2 declined after week 1, pointing to the maintenance of postmeiotic germ cells in the cell culture dish. A longer detectable PRM1 expression (2 weeks) in the two re-centrifuged samples (patients 3 and 5; Fig. 4C, E) correlates well with the presence of sperm heads found in the TEM preparations of the same patient (Supplemental Fig. S4) and the IF stainings (Fig. 3). A more detailed quantitative qPCR analysis of the single clusters from the four patients (Fig. 5) revealed unaltered cell proportions between 1 and 2 weeks of culture for SC (SOX9, AR), fibroblasts (FSP1), LC (STAR) and spermatids (PRM1) (under the assumption that the RPS27-normalized expression values correlate to some extent to the cell numbers). In contrast, peritubular myoid cells (ACTA2) displayed a significant (P < 0.05) increase either in cellular proportions or from differential regulation. Hence, for this transcript, the quantitative data confirm the qualitative observations in Fig. 4, although a rise in cellular proportions need not necessarily coincide with increased overall transcript levels, as is the case for FSP1. Figure 4 View largeDownload slide Cellular composition of 1, 2 and 3 weeks old clusters as revealed by cell-specific mRNA expression. (A) Clusters from patient 1 (HS), (B) patient 2 (FS) and (C) patient 3 (FS, mixed atrophy) were cultured for 1 (left bars) and 2 (right bars) weeks in Permanox chambers. (D) Clusters from patient 4 (FS) were cultured for 3 weeks in 96-well plates. (E) Clusters from patient 5 (FS, HS) were cultured for 1 (left bars) and 2 (right bars) weeks in 96-well plates. Single clusters (n = 8–10) from patients 1–3 and 5 as well as a pool of eight clusters from patient 4 were interrogated by qPCR for cell-specific transcripts (Sertoli cells: SOX9, AR; postmeiotic germ cells: PRM1; peritubular cells: ACTA2; fibroblasts: FSP1; macrophages: CD68; endothelial cells: CD34; spermatogonia / spermatocytes: DDX4; Leydig cells: INSL3, STAR, HSD3B1; housekeeping gene: RPS27). Abbreviations: n.d., not determined; negative, no expression detected. Figure 4 View largeDownload slide Cellular composition of 1, 2 and 3 weeks old clusters as revealed by cell-specific mRNA expression. (A) Clusters from patient 1 (HS), (B) patient 2 (FS) and (C) patient 3 (FS, mixed atrophy) were cultured for 1 (left bars) and 2 (right bars) weeks in Permanox chambers. (D) Clusters from patient 4 (FS) were cultured for 3 weeks in 96-well plates. (E) Clusters from patient 5 (FS, HS) were cultured for 1 (left bars) and 2 (right bars) weeks in 96-well plates. Single clusters (n = 8–10) from patients 1–3 and 5 as well as a pool of eight clusters from patient 4 were interrogated by qPCR for cell-specific transcripts (Sertoli cells: SOX9, AR; postmeiotic germ cells: PRM1; peritubular cells: ACTA2; fibroblasts: FSP1; macrophages: CD68; endothelial cells: CD34; spermatogonia / spermatocytes: DDX4; Leydig cells: INSL3, STAR, HSD3B1; housekeeping gene: RPS27). Abbreviations: n.d., not determined; negative, no expression detected. Figure 5 View largeDownload slide qPCR-based relative expression values of SOX9, AR, PRM1, ACTA2, FSP1 and STAR from 1 and 2 weeks cultured testis clusters. Relative expression was calculated by the delta-Ct method using RPS27 as the normalizing gene to correct for cDNA synthesis bias. P-values were calculated by unequal variance T-tests on the log2-transformed data (significant in case of asterisks) between the two time points. Numbers (x/y) reflect the number of positive qPCR reactions (dots) in relation to all qPCR reactions. These figures include all Cq values obtained from patients 1 to 3 and 5 (compare Fig. 4). Figure 5 View largeDownload slide qPCR-based relative expression values of SOX9, AR, PRM1, ACTA2, FSP1 and STAR from 1 and 2 weeks cultured testis clusters. Relative expression was calculated by the delta-Ct method using RPS27 as the normalizing gene to correct for cDNA synthesis bias. P-values were calculated by unequal variance T-tests on the log2-transformed data (significant in case of asterisks) between the two time points. Numbers (x/y) reflect the number of positive qPCR reactions (dots) in relation to all qPCR reactions. These figures include all Cq values obtained from patients 1 to 3 and 5 (compare Fig. 4). The presence of the macrophage marker CD68 in 25–73% of single clusters (Fig. 4A–C) indicated a possible role of this cell type for either positive or negative cluster integrity. To shed light on the involved macrophage subtypes, we tested for the proinflammatory M1 macrophage cytokine IL6 and the anti-inflammatory M2 macrophage marker CD163 on 1 and 2-week-old clusters of two patients. Here, we found IL6 expression in 16/19 (Cq values: 31.23–38.53) and CD163 expression in 1/16 clusters (Cq value: 37.66), pointing to macrophages of a predominantly M1 subtype. Currently we cannot distinguish between native M1 macrophages or those that arise from the M2 to M1 subtype reprogramming (Mosser and Edwards, 2008), where the latter would imply a severe effect of the mechanical disruption and/or cluster formation. Here, the application of a macrophage marker panel may offer a more robust discrimination of involved subtypes, however, the amount of RNA isolated from single picked clusters does not suffice for this approach. Quantitative PCR analysis of cluster-free cell culture medium was always negative, demonstrating the absence of cell-released transcripts in the cell culture dishes. The identification of the clusters’ cell types via TEM based on their ultrastructure in overview (Fig. 6) and in more detail (Fig. 7, Supplemental Figs S3 and S4) is founded on TEM images depicted in mostly older literature. It is somewhat complicated by putative dedifferentiation of cultured cells after removal from their natural context, e.g. LC that barely preserve their typical lipid droplets. Contrasting this, SC (Holstein and Roosen-Runge, 1981; Schulze, 1984; Fig. 7G, Supplemental Fig. S3) were clearly identifiable in almost all cases based on their typical morphology. They were present in clusters of all patients independent of spermatogenic state, cell culture duration and type of cell culture dish, hence, appearing as the most dominating and stable cell type in the clusters. This outcome fits well to the importance of SC for the formation of rat testicular organoids within a three-layer gradient system (Alves-Lopes et al., 2017). Figure 6 View largeDownload slide TEM depiction of cluster cell typology of three different patients. (A, B) Clusters cultivated for 11 days on 96-well plates from a patient with MA. (C, D) Clusters from two different patients with FS cultivated for 7 days in Permanox chambers. Clusters in (D) were re-transfered to the same well at each time-point of feeding. Abbreviations: LC, Leydig cell; EC, endothelial cell; PC, peritubular cell; SC, Sertoli cell; FB, fibroblast; SPG, spermatogonium; in this case Apale, higher magnification in Fig. 7A. Asterisks denote layers of collagen fibres. Figure 6 View largeDownload slide TEM depiction of cluster cell typology of three different patients. (A, B) Clusters cultivated for 11 days on 96-well plates from a patient with MA. (C, D) Clusters from two different patients with FS cultivated for 7 days in Permanox chambers. Clusters in (D) were re-transfered to the same well at each time-point of feeding. Abbreviations: LC, Leydig cell; EC, endothelial cell; PC, peritubular cell; SC, Sertoli cell; FB, fibroblast; SPG, spermatogonium; in this case Apale, higher magnification in Fig. 7A. Asterisks denote layers of collagen fibres. Figure 7 View largeDownload slide Higher magnification TEM micrographs of cluster cell types from three different patients. (A) Apale SPG, and (B–D) spermatids from 1 week cultivated clusters from a patient with FS (96-well plate). (E) Peritubular cell, (F) endothelial cell, (G) Sertoli cell, (H) Leydig cell from 11 days cultivated clusters from a patient with MA (96-well plate). (I) Macrophage from a 11-week cultivated cluster from a patient with FS (Permanox chamber). White arrow in (A) points to a typical Apale nucleolus, peripherally located to the nuclear membrane. The nucleus of the cell appears homogeneous pale and shows no rarefaction-zone in contrast to Adark SPG (Clermont, 1963, 1966, 1972). Black arrow in (I) points to a primary lysosome as a macrophage-typical cellular structure. Figure 7 View largeDownload slide Higher magnification TEM micrographs of cluster cell types from three different patients. (A) Apale SPG, and (B–D) spermatids from 1 week cultivated clusters from a patient with FS (96-well plate). (E) Peritubular cell, (F) endothelial cell, (G) Sertoli cell, (H) Leydig cell from 11 days cultivated clusters from a patient with MA (96-well plate). (I) Macrophage from a 11-week cultivated cluster from a patient with FS (Permanox chamber). White arrow in (A) points to a typical Apale nucleolus, peripherally located to the nuclear membrane. The nucleus of the cell appears homogeneous pale and shows no rarefaction-zone in contrast to Adark SPG (Clermont, 1963, 1966, 1972). Black arrow in (I) points to a primary lysosome as a macrophage-typical cellular structure. In addition, Fig. 7 and Supplemental Fig. S4 exemplify different cluster-composing cell types, i.e. peritubular myoid cells (PC; Ross, 1967; Maekawa et al., 1995, 1996; Johnson et al., 2010), endothelial cells (EC; Maekawa et al., 1996; Lee and Cheng, 2004; ; Johnson et al., 2010), fibroblasts (FB; Ross, 1967), macrophages (MP; Ross, 1967; Johnson et al., 2010; Goluža et al., 2014; Dempster et al., 2011), and spermatozoa (Holstein and Roosen-Runge, 1981). Infrequently, LC (Mori and Christensen, 1980; Schulze, 1984; Ezeasor, 1985; Johnson et al., 2010), and in the case of a re-centrifuged sample, single SPG (Holstein and Roosen-Runge, 1981) were encountered. It cannot be excluded that these SPG, only observed in the periphery of the cluster cells (Fig. 6D), are a consequence of the experimental setup. Extracellular matrix, as presented by collagen fibrils (Clermont, 1958; Ross, 1967; Lee and Cheng, 2004; Fig. 6B, Supplemental Fig. S5A–D) and typical actin filaments in the cytoplasm of a peritubular myoid cell (Supplemental Fig. S5D–F) were also identified. Prevention of cluster formation by kinase inhibition K252a is an alkaloid from the prokaryote Nonomuraea longicatena that is often used as a selective inhibitor of tyrosine protein kinase activity, e.g. of neurotrophin receptors (Tapley et al., 1992). Here, in analogy to the induced disturbance of rat in vitro SC aggregation (Gassei et al., 2008), we were able to prevent cluster formation of primary adult testis cells with K252a concentrations of 500 nM and higher (Fig. 8A–F). At these concentrations, cluster formation and length/number of protrusions are significantly reduced. The corresponding solvent controls (0.5% DMSO; Fig. 8G) and no-solvent controls (Fig. 8H) displayed normal cluster formation. Figure 8 View largeDownload slide Disruption of cluster formation by protein kinase inhibition (6 days after plating). (A–F) Addition of kinase inhibitor K252a in concentrations ranging from 5 nM to 1 μM to the culture medium. (G) Solvent control (0.5% DMSO), (H) no solvent control (0% DMSO). Cells from the testis of a patient displaying FS. Figure 8 View largeDownload slide Disruption of cluster formation by protein kinase inhibition (6 days after plating). (A–F) Addition of kinase inhibitor K252a in concentrations ranging from 5 nM to 1 μM to the culture medium. (G) Solvent control (0.5% DMSO), (H) no solvent control (0% DMSO). Cells from the testis of a patient displaying FS. Discussion In this work, we describe a stable experimental protocol for in vitro reorganization of human testicular cells obtained from rice grain-sized TESE biopsies. The aggregate-free cell suspensions were cultured in simple cell culture dishes without matrix support, similar to the mouse study of Mäkelä et al. (2014), but in contrast to the agarose-containing cultivation system of a recent human study (Baert et al., 2017). The observed structures display a distinct and intrinsic pattern of cluster formation and fusion that possibly recapitulates the cellular morphogenetic recognition program, a well-known process when organ models consist of multiple cell types (Yin et al., 2016), and which occurs in embryonic testicular cord development (Gassei et al., 2008). Our time-lapse-based investigation, in extension to previous findings in the rodents (Pan et al., 2013; Mäkelä et al., 2014), reveals a transition from an aggregating phase to a fusing phase when clusters acquire a size of ~60–80 μm. This is accompanied by a highly dynamic process where large cellular protrusions capture clusters in near vicinity, suggesting the involvement of chemotactic compounds as observed in embryoid body formation (Jiang et al., 2007). The impairment of cluster formation and absence of protrusions after protein kinase inhibition may therefore (among other potential effects) be attributed to the modulation of the chemotactic machinery. To date, we cannot pinpoint the affected cell type, but the observed protrusions are to some extent analogous to the NGF-induced, K252a-inhibitable neurite-like processes of the PC12 cell line (Tomokiyo et al., 2012). Our results indicate putative dedifferentiation of SC and LC entrapped in the clusters. Along these lines, while in adult SC SOX9-protein localization pertains mostly to nuclei (Morais da Silva et al., 1996), in the investigated clusters, it always appeared cytoplasmic. Expression of SOX9 during cell proliferation (Chakravarty et al., 2011) does not normally occur in the male adult, but can be encountered in the reversion to an immature embryonal-like state of male gonadal somatic cells (De Santa Barbara et al., 2000) or in a disrupted microtubular network (Malki et al., 2005). The latter may be caused from experimental procedures such as the in vitro culture conditions. A recent study, in which mouse spg stem cell culture conditions were tested on human testicular cells, resulted partly in testicular clusters and a similar cytoplasmatic SOX9 localization in SC (Medrano et al., 2016) as described in this work. Dedifferentiation of supporting LC is indicated by the absence of classical lipid droplets, absence of INSL3 mRNA expression, downregulation of HSD3B1 mRNA and the presence of STAR (Medrano et al., 2016), where in our case the latter was only significantly expressed at the mRNA level. A possible cause for the dedifferentiation of these two somatic cell types might be the absence of vital stimulatory factors such as hCG and FSH, and indeed, preliminary cluster stimulations with 50 ng/ml FSH (Gassei et al., 2008) resulted in up to 50% increase in the cluster diameter (data not shown). However, other works revealed an overall hCG stimulatability only until cell culture Day 2 (Pendergraft et al., 2017), or not at all, even under optimized three-dimensional culture conditions (Baert et al., 2017). Here, the addition of spg survival factors may provide more favourable conditions for in vitro spermatogenesis (Sato et al., 2011) or the proliferation of spg stem cells for male infertiliy treatment (Mulder et al., 2016). Equally important as the lack of essential stimulatory factors is an impairment of cluster or organoid function by the presence of proinflammatory molecules. The large proportion of M1 macrophages and corresponding CD68/IL6 expression in our clusters, with the absence of CD163, indicated a proinflammatory situation in the culture system. Similar elevated IL6 secretion was also encountered in one and four week old human testicular organoids (Baert et al., 2017). Moreover, testicular organoid formation in a three-dimensional model in rat was disrupted by addition of proinflammatory interleukin 1 alpha and tumour necrosis factor alpha (Alves-Lopes et al., 2017). These observations suggest that future research on testicular clusters or organoids should focus at least partly on the suppression of proinflammatory signalling (e.g. by non-steroidal anti-inflammatory drugs), as this may promote normal cell function (Sawazaki et al., 2014). Our employed methodology on single picked clusters enables a highly sensitive detection and classification of the clusters’ cellular composition, in which qPCR-based classification with cell-specific transcripts is supplemented by IF stainings and confirmed by the ultrastructure-based classification of single cells by TEM. With respect to germ cells, a few SPG were encountered in border regions of re-centrifuged clusters earlier than 2 weeks, as opposed to a considerably higher quantity found elsewhere (Baert et al., 2017; Pendergraft et al., 2017). Also, some remnant spermatozoa entrapped in the clusters could be identified by PRM1-positive qPCR, IF and TEM, demonstrating, to some extent, the potential of human testis-derived clusters to propagate postmeiotic germ cells. It is worth mentioning that the ability of cluster to form per se was independent of the patient’s spermatogenic state and reflected the cell types of the testicular origin. One exception to the rule consisted of a patient presenting FS where no cluster formation occurred and where SC contained many lipid droplets and lysosomes. Although a single case, this observation might emphasize the role of morphologically normal SC for cluster formation. We conclude that the in vitro reorganization of low amounts of human testicular cells obtained from normogonadotropic men is feasible as well as highly reproducible under 2D culture conditions. The tested culture system is suited to maintain clusters and SC for up to 3 months, but in its current state, far from optimal for the survival of germ cells. Because SPG were only detected on the surface or outside of clusters, it is not unlikely that the compact three-dimensional structure is insufficient to serve as a niche for long-term germ cell maintenance, propagation and differentiation. Hence, three-dimensional culture systems may provide more favourable conditions for the generation of human testicular organoids as a consequence of providing compartimentalization and concentration gradients (Alves-Lopes et al., 2017). The development of an optimal system for testicular organoids could be useful for in vitro spermatogenesis or for the support of spg stem cell proliferation, and therefore for infertility treatment in daily clinical practice. Supplementary data Supplementary data are available at Molecular Human Reproduction online. Acknowledgements We would like to thank Gabriele Hahn and Barbara Holstermann for technical assistance, Prof. Martin Bergmann (JLU, Giessen), Prof. Andreas Meinhardt (JLU, Giessen), Prof. Davor Ježek (University of Zagreb) and Mina Mincheva (CERA, Münster) for fruitful discussions, and Dr Rudolph Reimer (Heinrich Pette Institute, Hamburg), Dr Sabine Windhorst and Dr Markus Geissen as well as the UKE Microscopy Imaging Facility for supplying technical equipment. We are grateful to Prof. Stefan W. Schneider for his support. Authors’ roles W.S. and A.S. delivered biopsy samples from a therapeutic testicular sperm extraction and classified the samples via light microscopy. K.v.K., M.S., E.S. and B.R. carried out the experiments. K.v.K., M.S., E.S. and A.N.S. analysed the data. K.v.K., W.S., A.N.S. and S.S. interpreted the results. K.v.K. and A.N.S. wrote the article. Funding Deutsche Forschungsgemeinschaft (DFG) grant to K.v.K. (KO 4769/2-1). Conflicts of interest None declared. References Alves-Lopes JP, Söder O, Stukenborg JB. Testicular organoid generation by a novel in vitro three-layer gradient system. Biomaterials  2017; 130: 76– 89. Google Scholar CrossRef Search ADS PubMed  Baert Y, De Kock J, Alves-Lopes JP, Söder O, Stukenborg JB, Goossens E. Primary human testicular cells self-organize into organoids with testicular properties. Stem Cell Rep  2017; 8( 1): 30– 38. Google Scholar CrossRef Search ADS   Benton L, Shan LX, Hardy MP. Differentiation of adult Leydig cells. J Steroid Biochem Mol Biol  1995; 53( 1−6): 61– 68. Google Scholar CrossRef Search ADS PubMed  Cappallo-Obermann H, Feig C, Schulze W, Spiess AN. Fold-change correction values for testicular somatic transcripts in gene expression studies of human spermatogenesis. Hum Reprod  2013; 28( 3): 590– 598. Google Scholar CrossRef Search ADS PubMed  Chakravarty G, Rider B, Mondal D. Cytoplasmic compartimentalization of SOX9 abrogates the growth arrest response of breast cancer cells that can be rescued by trichostatin A treatment. Cancer Biol Ther  2011; 11( 1): 71– 83. Google Scholar CrossRef Search ADS PubMed  Chávez-Galán L, Olleros ML, Vesin D, Garcia I. Much more than M1 and M2 macrophages, there are also CD169+ and TCR+ macrophages. Front Immunol  2015; 6: 263. Google Scholar PubMed  Clermont Y. Contractile elements in the limiting membrane of the seminiferous tubules of the rat. Exp Cell Res  1958; 15: 438– 440. Google Scholar CrossRef Search ADS PubMed  Clermont Y. The cycle of the seminiferous epithelium in man. Am J Anat  1963; 112: 35– 51. Google Scholar CrossRef Search ADS PubMed  Clermont Y. Renewal of spermatogonia in man. Am J Anat  1966; 118: 509– 524. Google Scholar CrossRef Search ADS PubMed  Clermont Y. Kinetics of spermatogenesis in mammals: seminiferous epithelium cycle and spermatogonial renewal. Physiol Rev  1972; 52( 1): 198– 236. Google Scholar CrossRef Search ADS PubMed  De Santa Barbara P, Moniot B, Poulat F, Berta P. Expression and subcellular localization of SF-1, SOX9, WT1, and AMH proteins during early human testicular development. Dev Dyn  2000; 217: 293– 298. Google Scholar CrossRef Search ADS PubMed  Dempster AG, Molteno ACB, Bevin TH, Thompson AM. Otago glaucoma surgery outcome study: electron microscopy of capsules around Molteno implants. Glaucoma  2011; 52: 8300– 8309. Ehmcke J, Schlatt S. Identification and characterization of spermatogonial subtypes and their expansion in whole mounts and tissue sections from primate testes. Methods Mol Biol  2008; 450: 109– 118. Google Scholar CrossRef Search ADS PubMed  Ezeasor DN. Light and electron microscopical observations on the Leydig cells of the scrotal and abdominal testes of naturally unilateral cryptorchid West African dwarf goats. J Anat  1985; 141: 27– 40. Google Scholar PubMed  Feig C, Kirchhoff C, Ivell R, Naether O, Schulze W, Spiess AN. A new paradigm for profiling testicular gene expression during normal and disturbed human spermatogenesis. Mol Hum Reprod  2007; 13( 1): 33– 43. Google Scholar CrossRef Search ADS PubMed  Gassei K, Ehmcke J, Schlatt S. Initiation of testicular tubulogenesis is controlled by neurotrophic tyrosine receptor kinases in a three-dimensional Sertoli cell aggregation assay. Reprod  2008; 136( 4): 459– 469. Google Scholar CrossRef Search ADS   Goluža T, Boscanin A, Cvetko J, Kozina V, Bernat MM, Kasum M, Kaštelan Z, Ježek D. Macrophages and Leydig cells in testicular biopsies of azoospermic men. Biomed Res Int  2014; 2014: 828697. Google Scholar CrossRef Search ADS PubMed  Gregory CW, DePhilip RM. Detection of steroidogenic acute regulatory protein (StAR) in mitochondria of cultured rat Sertoli cells incubated with follicle-stimulating hormone. Biol Reprod  1998; 58( 2): 470– 474. Google Scholar CrossRef Search ADS PubMed  Holstein AF, Roosen-Runge EC Atlas of Human Spermatogenesis . Berlin: Grosse Verlag, 1981. Holstein AF, Wulfhekel U. Die Semidünnschnitt-Technik als Grundlage für eine cytologische Beurteilung der Spermatogenese des Menschen. Andrologia  1971; 3: 65– 69. Google Scholar CrossRef Search ADS   Ishikawa T, Hwang K, Lazzarino D, Morris PL. Sertoli cell expression of steroidogenic acute regulatory protein-related lipid transfer 1 and 5 domain-containing proteins and sterol regulatory element binding protein-1 are interleukin-1beta regulated by activation of c-Jun N-terminal kinase and cyclooxygenase-2 and cytokine induction. Endocrinology  2005; 146( 12): 5100– 5111. Google Scholar CrossRef Search ADS PubMed  Ježek D, Knuth UA, Schulze W. Successful testicular sperm extraction (TESE) in spite of high serum follicle stimulating hormone and azoospermia: correlation between testicular morphology, TESE results, semen analysis and serum hormone values in 103 infertile men. Hum Reprod  1998; 13( 5): 1230– 1234. Google Scholar CrossRef Search ADS PubMed  Jiang S, Fu Y, Williams J, Wood JA, Pandarinathan L, Avraham S, Makriyannis A, Avraham S, Avraham HK. Expression and function of cannabinoid receptors CB1 and CB2 and their cognate cannabinoid ligands in murine embryonic stem cells. PLoS One  2007; 7: e641. Google Scholar CrossRef Search ADS   Johnson L, Welsh TH, Curley KO, Johnston CE. Anatomy and physiology of the male reproductive system and potential targets of toxicants. Compr Toxicol (Second Edition), Elsevier Ltd  2010; 11: 5– 59. Lee NPY, Cheng CY. Ectoplasmatic specialization, a testis-specific cell-cell actin-based adherens junction type: is this a potential target for male contraceptive development? Hum Reprod Update  2004; 10( 4): 349– 369. Google Scholar CrossRef Search ADS PubMed  Lucket-Chastain L, Calhoun K, Schartz T, Gallucci RM. IL-6 influences the balance between M1 and M2 macrophages in a mouse model of irritant contact dermatitis. J Immunol  2016; 196( 1): 196.17. Maekawa M, Kamimura K, Nagano T. Peritubular myoid cells in the testis: their structure and function. Arch Histol Cytol  1996; 59( 1): 1– 13. Google Scholar CrossRef Search ADS PubMed  Maekawa M, Kazama H, Kamimura K, Nagano T. Changes in the arrangement of actin filaments in myoid cells and Sertoli cells of rat testes during postnatal development and after experimental cryptorchidism. Anat Rec  1995; 241: 59– 69. Google Scholar CrossRef Search ADS PubMed  Malki S, Berta P, Poulat F, Boizet-Bonhoure B. Cytoplasmic retention of the sex-determining factor SOX9 via the microtubule network. Exp Cell Res  2005; 309: 468– 475. Google Scholar CrossRef Search ADS PubMed  Martinez FO, Gordon S, Locati M, Mantovani A. Transcriptional profiling of the human monocyte-to-macrophage differentiation and polarization: new molecules and patterns of gene expression. J Immunol  2006; 177: 7303– 7311. Google Scholar CrossRef Search ADS PubMed  Medrano JV, Rombaut C, Simon C, Pellicer A, Goossens E. Human spermatogonial stem cells display limited proliferation in vitro under mouse spermatogonial stem cell culture conditions. Fertil Steril  2016; 106( 6): 1539– 1549.e8. Google Scholar CrossRef Search ADS PubMed  Morais da Silva S, Hacker A, Harley V, Goodfellow P, Swain A, Lovell-Badge R. Sox9 expression during gonadal development implies a conserved role for the gene in testis differentiation in mammals and birds. Nat Genet  1996; 14: 62– 68. Google Scholar CrossRef Search ADS PubMed  Mori H, Christensen AK. Morphometric analysis of Leydig cells in the normal rat testis. J Cell Biol  1980; 84: 340– 354. Google Scholar CrossRef Search ADS PubMed  Mosser DM, Edwards JP. Exploring the full spectrum of macrophage activation. Nat Rev Immunol  2008; 8( 12): 958– 959. Google Scholar CrossRef Search ADS PubMed  Mulder CL, Zheng Y, Jan SZ, Struijk RB, Repping S, Hamer G, van Pelt AM. Spermatogonial stem cell autotransplantation and germline genomic editing: a future cure for spermatogenic failure and prevention of transmission of genomic diseases. Hum Reprod Update  2016; 22( 5): 561– 573. Google Scholar CrossRef Search ADS PubMed  Mäkelä JA, Toppari J, Rivero-Müller A, Ventelä S. Reconstruction of mouse cellular microenvironments in long-term seminiferous tubule culture. PLoS One  2014; 9( 3): e90088. Google Scholar CrossRef Search ADS PubMed  Pan F, Lifeng C, Schlatt S. Effects of nanostructures and mouse embryonic stem cells on in vitro morphogenesis of rat testicular cords. PLoS One  2013; 8( 3): e60054. Google Scholar CrossRef Search ADS PubMed  Pendergraft SS, Sadri-Ardekani H, Atala A, Bishop CE. Three-dimensional testicular organoid: a novel tool for the study of human spermatogenesis and gonadotoxicity in vitro. Biol Reprod  2017; 96( 3): 720– 732. Google Scholar CrossRef Search ADS PubMed  Raucci F, D’Aniello A, Di Fiore MM. Stimulation of androgen production by D-aspartate through the enhancement of StAR, P450scc and 3β-HSD mRNA levels in vivo rat testis and in culture of immature rat Leydig cells. Steroids  2014; 84: 103– 110. Google Scholar CrossRef Search ADS PubMed  Reuter K, Ehmcke J, Stukenborg JB, Simoni M, Damm OS, Redmann K, Schlatt S, Wistuba J. Reassembly of somatic cells and testicular organogenesis in vitro. Tissue Cell  2014; 46( 1): 86– 96. Google Scholar CrossRef Search ADS PubMed  Ross MH. The fine structure and development of the peritubular contractile cell component in the seminiferous tubules of the mouse. Am J Anat  1967; 121: 523– 558. Google Scholar CrossRef Search ADS PubMed  Sato T, Katagiri K, Gohbara A, Inoue K, Ogonuki N, Ogura A, Kubota Y, Ogawa T. In vitro production of functional sperm in cultured neonatal mouse testes. Natural  2011; 471: 504– 507. Google Scholar CrossRef Search ADS   Sawazaki R, Ishihara T, Usui S, Hayashi E, Tahara K, Hoshino T, Higuchi A, Nakamura S, Tsubota K, Mizushima T. Diclofenac protects cultured human corneal epithelial cells against hyperosmolarity and ameliorates corneal surface damagein a rat model of dry eye. Invest Ophthalmol Vis Sci  2014; 55( 4): 2547– 2556. Google Scholar CrossRef Search ADS PubMed  Schulze C. Sertoli cells and Leydig cells in man. Adv Anat Embryol Cell Biol  1984; 88: 1– 104. Google Scholar CrossRef Search ADS PubMed  Schulze W, Thoms F, Knuth UA. Testicular sperm extraction: comprehensive analysis with simultaneously performed histology in 1418 biopsies from 766 subfertile men. Hum Reprod  1999; 14( Suppl 1): 82– 96. Google Scholar CrossRef Search ADS PubMed  Sidney LE, Branch MJ, Dunphy SE, Dua HS, Hopkinson A. Concise review: evidence for CD34 as a common marker for diverse progenitors. Stem cells  2014; 32: 1380– 1389. Google Scholar CrossRef Search ADS PubMed  Tapley P, Lamballe F, Barbacid M. K252a is a selective inhibitor of the tyrosine protein kinase activity of the trk family of oncogenes and neurotrophon receptors. Oncogene  1992; 7( 2): 371– 381. Google Scholar PubMed  Thorrez L, Van Deun K, Tranchevent LC, Van Lommel L, Engelen K, Marchal K, Moreau Y, Van Mechelen I, Schuit F. Using ribosomal protein genes as reference: a tale of caution. PLoS One  2008; 3: e1854. Google Scholar CrossRef Search ADS PubMed  Tomokiyo A, Maeda H, Fujii S, Monnouchi S, Wada N, Kono K, Yamamoto N, Koori K, Teramatsu Y, Akamine A. A multipotent clonal human periodontal ligament cell line with neural crest cell phenotypes promotes neurocytic differentiation, migration, and survival. Cell Physiol  2012; 227( 5): 2040– 2050. Google Scholar CrossRef Search ADS   von Kopylow K, Schulze W, Salzbrunn A, Spiess AN. Isolation and gene expression analysis of single potential human spermatogonial stem cells. Mol Hum Reprod  2016; 22( 4): 229– 239. Google Scholar CrossRef Search ADS PubMed  Yin X, Mead BE, Safaee H, Langer R, Karp JM, Levy O. Engineering stem cell organoids. Cell Stem Cell  2016; 18( 1): 25– 38. Google Scholar CrossRef Search ADS PubMed  Yokonishi T, Sato T, Katagiri K, Komeya M, Kubota Y, Ogawa T. In vitro reconstruction of mouse seminiferous tubules supporting germ cell differentiation. Biol Reprod  2013; 89( 1): 15. Google Scholar CrossRef Search ADS PubMed  Zhang J, Hatakeyama J, Eto K, Abe S. Reconstruction of a seminiferous tubule-like structure in a 3 dimensional culture system of re-aggregated mouse neonatal testicular cells within a collagen matrix. Gen Comp Endocrinol  2014; 205: 121– 132. Google Scholar CrossRef Search ADS PubMed  © The Author(s) 2018. 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@oup.com

Journal

Molecular Human ReproductionOxford University Press

Published: Mar 1, 2018

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 12 million articles from more than
10,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Unlimited reading

Read as many articles as you need. Full articles with original layout, charts and figures. Read online, from anywhere.

Stay up to date

Keep up with your field with Personalized Recommendations and Follow Journals to get automatic updates.

Organize your research

It’s easy to organize your research with our built-in tools.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

Monthly Plan

  • Read unlimited articles
  • Personalized recommendations
  • No expiration
  • Print 20 pages per month
  • 20% off on PDF purchases
  • Organize your research
  • Get updates on your journals and topic searches

$49/month

Start Free Trial

14-day Free Trial

Best Deal — 39% off

Annual Plan

  • All the features of the Professional Plan, but for 39% off!
  • Billed annually
  • No expiration
  • For the normal price of 10 articles elsewhere, you get one full year of unlimited access to articles.

$588

$360/year

billed annually
Start Free Trial

14-day Free Trial