SAS Sprague Dawley rats (Strain: code 400, Charles River), aged P4-P5, were euthanized and testes were removed. All animal procedures were performed as approved by the Animal Care Committee, University of Calgary. Using a pair of forceps, the tunica albuginea was removed to release the tubules, which were then washed in Hank’s Balanced Salt Solution (HBSS, Gibco, cat# 14025092) containing 1% penicillin-streptomycin (ThermoFisher Scientific, cat# 15070063). The tubules were digested using collagenase type IV (Worthington-Biochem, cat# LS004189) in HBSS (2 mg/mL) for 25 min at 37°C. The tubules were then sedimented by centrifugation at 90x g for 1.5 min and washed with 5 mL HBSS thrice. Finally, the tissue was digested with 0.25% trypsin–EDTA (Sigma-Aldrich, cat# T4049) and DNase I (Sigma-Aldrich, cat# DN25) in HBSS (10 mg/mL) for 5 min to obtain the starting cell population (11). All experiments were replicated using a minimum of three independently prepared cell suspensions.

AggreWell 400 plates (STEMCELL Technologies Inc, Vancouver, Canada, cat# 34450) were treated with Anti-adherence Rinsing Solution (STEMCELL Technologies Inc, Vancouver, Canada, cat# 07010) according to the manufacturer’s instructions. Each well was filled with 500 μL of organoid formation medium (OFM) (Dulbecco Modified Eagle Medium F/12 (Gibco, cat# 11330-032) supplemented with insulin 10 μg/mL, transferrin 5.5 μg/mL, selenium 6.7 ng/mL (Gibco, cat#41400-045); 20 ng/mL epidermal growth factor (R & D Systems, cat# 236-EG); 1% Penicillin-Streptomycin) and the plates were then centrifuged at 2000x g for 2 minutes to release any trapped air bubbles (11). Each well was then seeded with 6 × 10⁵ rat testicular cells suspended in 500 μL OFM. Finally, the plates were centrifuged at 500x g for 5 minutes to sediment the cells in the microwells. 500 μL of media were removed from each well and fresh media supplemented with Corning Matrigel Growth Factor Reduced (GFR) Basement Membrane Matrix (1:100 dilution; Life Sciences, cat# 354230) was used to replenish each well. Microwell plates were divided into three groups and placed in 3 incubators setup for three conditions: (i) ambient O₂ tension (18.4% or 6.34 mM O₂, 5% CO₂, 37°C), (ii) 15% O₂ tension at 37°C (15% or 5.18 mM O₂, 5% CO₂) and (iii) 15% O₂ tension at 34°C (15% or 5.22 mM O₂, 5% CO₂). Culture in OFM was carried out for 3-5 days with 50% media changes every second day.

After culturing the organoids for 3 days, which were designated as day 0 undifferentiated organoids, the OFM media was completely removed, and the culture was continued in organoid differentiation medium (ODM). ODM was composed of Minimum Essential Medium α (ThermoFisher Scientific, cat# 12571063) supplemented with 10% KnockOut Serum Replacement (ThermoFisher Scientific, cat# 10828028), hepatocyte growth factor (5 ng/mL) (R&D Systems, cat# 294-HG), activin A (100 ng/mL) (Sigma-Aldrich, cat# a4941), follicle stimulating hormone (1 ng/mL) (Sigma-Aldrich, cat# F4021), luteinizing hormone (1 ng/mL) (Sigma-Aldrich, cat# L5259), testosterone (1 μM) (Steraloids, cat# A6950-000), recombinant human BMP-4 (20 ng/mL) (R&D Systems, cat# 314-BP), recombinant human BMP-7 (20 ng/mL) (R&D Systems, cat# 354-BP), 3,3’,5-triodo-L-thyronine sodium (2 ng/mL) (Sigma-Aldrich, cat# T6397), l-ascorbic acid-2-glucoside (1 mM) (Matrix Scientific, cat# 092375) and 1% Penicillin-Streptomycin (22, 23). Culture was carried out for an additional 28 days, with full media changes every second day. The organoids were sampled every 7 days, including day 0, for analysis.

Testes tissue from 43-day old rats were fixed in 4% paraformaldehyde, dehydrated with a gradient series of ethanol, and embedded in paraffin wax to prepare sections of 5 µm thickness. The rat starting cell populations and organoids were fixed using 2% paraformaldehyde and spun down on slides using cytospin centrifugation (1000 rpm for cells and 500 rpm for organoids) (Cytospin 4, Thermo Scientific). The samples were then permeabilized using a gradient series of methanol (24) and blocked with 10% donkey serum. The testes tissue was incubated with anti-γ-H2AX (Gamma H2A Histone Family X) (25) and anti-SYCP3 (Synaptonemal Complex Protein 3) (26) ( Supplementary Table 1 ). For rat testicular cells, the slides were incubated with anti-GATA4 (GATA Binding Protein 4) (27), anti-VASA (DEAD-Box Helicase 4) (28), anti-α-SMA(alpha-Smooth Muscle Actin) (29), anti-3β-HSD (3 Beta-Hydroxysteroid Dehydrogenase) (28) ( Supplementary Table 1 ). In addition to the antibodies mentioned above, rat organoids were also incubated with anti-Collagen IV (30), anti-Laminin (31), anti-Fibronectin (32), anti-Claudin 11 (33), anti-UCHL1 (Ubiquitin C-terminal Hydrolase L1), anti-TNP1 (Transition Protein 1) (34), anti-PRM1 (Protamine 1) (35), anti-ACR (Acrosin) (36) and anti-AR (Androgen Receptor) (37) antibodies overnight at 4°C ( Supplementary Table 1 ). Fluorescence labelling was done with secondary antibodies conjugated with Alexa Fluor 488 and 555 ( Supplementary Table 1 ). DAPI (4′,6-diamidino-2-phenylindole) (Vector, cat# H1200) was used for labelling the nuclei. The cells were analyzed using Zeiss Imager.M2 fluorescence microscopy and the percentages of testicular cell types were determined by counting the cells with ImageJ software. The organoids were analyzed using a Leica TCS-SP8 confocal laser scanning microscope with the Leica Las X software.

Immunohistochemistry for VASA was performed on day 7 and 28 organoids, while SYCP3 staining was performed on day 28 organoids. Using confocal microscopy, organoids were selected blindly based on only DAPI and scanned across the z-axis to quantify the number of VASA^+ve and SYCP3^+ve cells in each organoid. From each of the three independent experiments (n = 3) performed, a total of 30 organoids were analyzed for VASA^+ve and 10 organoids were analyzed for SYCP3^+ve cell counts.

RNA was isolated from 1200 organoids using RNeasy Micro Kit (QIAGEN, cat # 74004) and then reverse transcription was performed using SuperScript™ IV VILO™ Master Mix (Thermo Fisher Scientific, cat# 11756050). RT-qPCR with the primers listed in Supplementary Table 2 was performed with a 7.500 Fast Real-Time PCR System (Applied Biosystems) using SsoFast Eva Green Supermix with Low ROX (Bio-Rad laboratories, cat# 1725211). The expression levels were presented relative to Gapdh. Statistical analysis was performed on the mean of ΔΔCt.

Day 19 organoids were treated with 1 μM MEHP and 0.25 μM CdCl₂. Controls were treated with equivalent volumes of DMSO. After 48 hours of treatment, at day 21, the organoids were harvested and analyzed with immunofluorescence and RT-qPCR.

Day 5 organoids were treated with 0.5, 1 and 1.5 μM of MEHP (Sigma, Cat# 796832) and 0.01, 0.05, 0.25 and 1.25 μM of CdCl₂ (Sigma, Cat# 202908). The control groups for MEHP and CdCl₂ were treated with equivalent volumes of DMSO for 1.5 μM of MEHP and 1.25 μM of CdCl_2, respectively. Controls of 1.5 μM and 1.25 μM DMSO were treated with equivalent volumes of phosphate buffered saline (Thermo Fisher Scientific, cat# 14190144). At day 7 (48 hours after treatment), the organoids were harvested and approximately 60 organoids suspended in 50 μL ODM were seeded in each well of a 96-well plate as duplicates to perform the MTT assay (Abcam, cat# ab211091) according to manufacturer’s instructions. Absorbance at OD 590 nm was measured using SpectraMax i3x plate reader (Molecular Devices). MTT assay was used to measure cellular metabolic activity as an indicator of cell viability.

All the results described here are from at least three independent experiments performed with three separately prepared rat starting cell population (n = 3). Data were analyzed using the GraphPad Prism 8 software. Unpaired two-tailed t-tests were performed for single comparisons between two groups. For more than two groups, one-way ANOVA with Tukey’s multiple comparison tests were performed. A value of p < 0.05 was set as the limit of statistical significance.

A rat (P4-P5) testicular starting cell population which contained 92.2 ± 1.28% GATA4^+ve Sertoli cells (38), 1.7 ± 0.3% VASA^+ve germ cells (a marker for spermatogonia, spermatocytes and round spermatids) (39), 2.6 ± 0.7% 3β-HSD^+ve Leydig cells (40) and 19.97 ± 2.8% α-SMA^+ve peritubular myoid cells (41) (n = 3) was used to generate rat testicular organoids with organoid formation media (OFM). Since lower O₂ tension is known to support higher differentiation potential of rat testicular cells (22), initial cultures were carried out in incubators set up for ambient and 15% O₂ tensions (37°C). Both culture conditions supported the initial generation of testicular organoids (72 hours), with organotypic morphology similar to our previously published porcine and murine model (11). However, unlike the porcine or murine models (11), the rat organoids cultured in ambient O₂ tension underwent a loss of testis-specific tissue architecture at day 6 of culture while day 6 organoids cultured at 15% O₂ tension had distinct internal-interstitial and external-seminiferous epithelial compartments. The two compartments were separated by a collagen IV^+ve, fibronectin^+ve and laminin^+ve basement membrane ( Figure 1A ). The external compartment was composed of VASA^+ve germ cells and GATA4^+ve Sertoli cells. α-SMA^+ve peritubular myoid cells were located lining the basement membrane in the interior compartment, while 3β-HSD^+ve Leydig cells were distributed throughout the interior compartment ( Figure 1A ). In contrast, the organoids cultured at ambient O₂ tension showed increased Sertoli cell numbers, generation of large Sertoli cell clusters and complete or partial separation of internal and external compartments ( Figure 1B ). Thus, subsequent experiments were carried out at 15% O₂ tension.

Matsumura et al. (22) and Sato et al. (23) reported efficient spermatogenic differentiation of rodent testicular organ cultures at 34°C. To evaluate the effect of temperature on somatic cell and spermatogonial maturation, rat organoids were generated with OFM (72 hours: day 0 undifferentiated organoid) and then cultured with organoid differentiation medium (ODM) at 37°C and 34°C (22, 42) for up to 28 days. Immunofluorescence analysis revealed no morphological differences between the two groups at day 28 ( Figure 2A ) and both cultures showed expression of Claudin 11, a component of Sertoli cell tight junctions ( Figure 2B ) (43). Except for expression of Shbg (sex hormone binding globulin) (Sertoli and Leydig cells), which was increased 5.7-fold (n = 3, p < 0.05) in 34°C cultures, no significant differences in transcription levels were observed between the two groups for the somatic cell markers Fshr (follicle stimulating hormone receptor) (Sertoli cells), Star (steroidogenic acute regulatory protein), Cyp17a1 (cytochrome P450 family 17 subfamily a member 1) (Leydig cells) and Hsd17b3 (hydroxysteroid 17-beta dehydrogenase 3) (Sertoli and Leydig cells) ( Supplementary Figure 1A ) (n = 3, p > 0.05) (44–47). At day 7, rat testicular organoids contained UCHL1^+ve undifferentiated spermatogonia at both temperatures ( Supplementary Figure 1B ). Both at day 7 and 28, there was no difference in the number of VASA^+ve germ cells between 37°C and 34°C cultures (n = 3, p > 0.05) ( Figures 2C, D ). However, the number of VASA^+ve germ cells was lower at day 28 compared to day 7 at both temperature conditions (n = 3, p < 0.05) ( Figure 2D ). SYCP3^+ve spermatogenic cells were observed in both conditions, with a staining pattern similar to spermatocytes in 43-day old rat testes ( Supplementary Figure 1E ), starting from day 21. γ-H2AX, which is induced by the DNA double stranded break in leptotene and early zygotene, was also observed in both culture conditions with a staining pattern similar to 43-day old rat testes ( Supplementary Figures 1D, E ) (25). The number of SYCP3^+ve was quantified and their number was found to be higher at day 28 in the organoids cultured at 34°C compared to 37°C (n = 3, p < 0.05) ( Figures 2E, F ) (48, 49). Therefore, the 37°C cultures were excluded from further analysis.

To characterize maturation during the 28-day long culture, transcription levels of the immature Sertoli cell marker Amh (anti-mullerian hormone) and Sertoli and Leydig cell markers Fshr, Shbg, Star, Hsd17b3 and Kitlg were analyzed for day 0, 7, 14, 21, 28 organoids. The transcription levels of Amh were undetectable within a week (n = 3, p < 0.05) while expression of Cyp17a1 showed no significant changes over the duration of the culture (n = 3, p > 0.05). Fshr, Shbg and Kitlg were upregulated by 2.7-, 17.9- and 3.3-fold at day 21 (n = 3, p < 0.05) ( Figure 2G ). Transcription levels of Star and Hsd17b3 increased 26.9- and 10-fold, respectively, by day 28 of culture (n = 3, p < 0.05) ( Figure 2G ). In addition, an increasing number of Sertoli cells in the organoids started to express AR with subsequent weeks of culture, indicating cell maturation ( Figure 2H ). Along with punctate SYCP3 staining (potentially leptotene spermatocytes) ( Figure 2E ), an elongated pattern of SYCP3 staining (as observed in early zygotene spermatocytes) (50) ( Supplementary Figure 1C ) was also detected at day 28. However, cells found with elongated SYCP3 expression were no longer adhered to the organoids at the time of analysis ( Supplementary Figure 1C ). In contrast, no PRM1^+ve, TNP1^+ve or ACR^+ve cells were observed in the cultures.

Initial dose-response experiments on monolayers of rat testicular cells (P4-P5) were used to select the dosages of 0.5, 1 and 1.5 μM for phthalic acid mono-2-ethylhexyl ester (MEHP); and 0.01, 0.05, 0.25 and 1.25 μM for the heavy metal cadmium chloride (CdCl₂) to be tested on organoids. Then, relative cell viability assessments were performed on day 7 (treatment began on day 5) organoids treated with the aforementioned dosages with a 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl-2H-tetrazolium bromide (MTT) assay, revealing that 1 μM MEHP and 0.25 μM CdCl₂ were the highest doses without adverse effects on viability (n = 3, p < 0.05) ( Figures 3A, B ). Thus, all other dosages were excluded from further analyses. Since most of the maturation markers showed robust upregulation at day 21, toxicological effects of MEHP and CdCl₂ were evaluated at day 21 by treating the organoids with MEHP and CdCl₂ at day 19 and then harvesting and analyzing 48 hours later. MEHP treatment was associated with upregulation of Fshr and Star and downregulation of the expression of Cyp17a1 (n = 3, p < 0.05) ( Figure 3D ). Expression of Fshr, Shbg, Hsd17b3 and Cyp17a1 was drastically downregulated upon exposure to CdCl₂ (n = 3, p < 0.05) ( Figure 3D ). Transcription of Star, in contrast, was upregulated 7-fold compared to DMSO controls (n = 3, p < 0.05) ( Figure 3D ). In contrast to MEHP which led to partial loss of tight-junction protein Claudin11, CdCl₂ treatment caused a total loss of Claudin 11 ( Figure 3C ).