Innovative Approach for Spermatogonial Stem Cell Proliferation

Germ Cell to Somatic Cell Ratio is Critical for Efficient Expansion of Spermatogonial Stem Cells In Vitro – a Proof of Concept

Running head: Innovative approach for SSC proliferation

Key words: Fertility preservation, prebubertal boys, gonadotoxic treatments, spermatogonial stem cells, differential plating, GFRα1.

List of abbreviations

SSC – Spermatogonial Stem Cells.

DP – Differential plating.

NOA – Non-obstructive azoospermia.

MACS – Magnetic-activated cells sorting.

FACS- Fluorescence-activated cells sorting.

Abstract

Spermatogonial Stem Cell (SSC) expansion in vitro remains a major challenge in efforts to preserve fertility among pubertal boys. The current study focused on innovative approaches to optimize SSC expansion. Six-eight week old CD-1 murine testicular samples were harvested by mechanical and enzymatic digestion. Cell suspensions were incubated for differential plating (DP). After DP, we established two experiments comparing single vs. repetitive DP (S-DP and R-DP, respectively) until passage 2 (P2). Each experiment included a set of cultures consisting of 5 floating-to-attached cell ratios (5, 10, 15, 20 and 25) and control cultures containing floating cells only. We found similar cell and colony count drops during P0 in both S- and R-DP. During P2, counts increased in S-DP in middle ratios (10, 15, and especially 20) relative to low and high ratios (5 and 25, respectively). Counts dropped extensively in R-DP after passage 2. The superiority of intermediate ratios was demonstrated by enrichment of GFRα1 by qPCR. The optimal ratio of 20 in S-DP contained significantly increased proportions of GFRα1-positive cells (25.8±5.8%) as measured by flow cytometry compared to after DP (1.9±0.7%, p<0.0001), as well as positive immunostaining for GFRα1 and UTF1, with rare Sox9-positive cells. This is the first report of the impact of initial floating-to-attached cell ratios on SSC proliferation in culture.

Introduction

Spermatogonial stem cells (SSCs) are unipotent cells localized to the seminiferous basement membrane which are responsible for sperm production during the process of spermatogenesis (Gassei, Valli, & Orwig, 2014). These cells have been the focus of many studies worldwide due to their potential to restore fertility among sterile patients treated by gonadotoxic regimens during childhood or men with non-obstructive azoospermia (NOA), in whom sperm is absent or sparse in the testes (Koruji et al., 2012; Livne-Segev, Forbes, & Lo, 2016).

Despite intensive research, SSCs remain enigmatic due to their unique features. Firstly, they exist at very low frequencies, estimated at 1 in 3,000-4,000 cells in the mouse (Tegelenbosch & de Rooij, 1993).  For clinical applications, ex-vivo expansion of human SSCs will likely be required. Secondly, SSCs are difficult to identify and purify due to a lack of SSC-specific markers and possibly due to the existence of subpopulations of SSCs in various species (Albert et al., 2012; Izadyar et al., 2011; Kossack et al., 2013). Thirdly, current strategies focusing on ex-vivo SSC  expansion are complex, time- and resource-consuming and often result in low cell purity and vitality (He, Lu, Zhang, Hao, & Yang, 2015).

The conventional approach for SSC expansion ex-vivo includes sorting followed by expansion in vitro (Conrad et al., 2014). A basic sorting method is differential plating (DP), which includes the incubation of a testicular cell suspension over gelatin overnight (Steinberger & Steinberger, 1966). During DP, somatic cells tend to attach to gelatin while the germ cells remain floating, which makes their separation easy and simple. However, DP by itself cannot separate germ cells in various stages of differentiation, and complete separation from somatic cells is not 100% efficient (Langenstroth et al., 2014). To this effect, more advanced methodologies such as magnetic-activated and fluorescence-activated cells sorting (MACS and FACS, respectively) using SSC-specific markers have been assessed (Gonzalez & Dobrinski, 2015). However, these sophisticated manipulations often result in substantial cell death and lack of available SSCs for further culture (Valli, Sukhwani, et al., 2014), and rely on markers that are not fully elucidated with regards to their specificity for SSCs. In addition, various culture conditions have been shown to promote SSC expansion including feeder vs. non-feeder cultures and use of several medias amongst others (He et al., 2015).

Although long term ex-vivo SSC expansion has been reported from pre-pubertal as well as adult human testicular tissue samples (Conrad et al., 2008; Sadri-Ardekani et al., 2009), these reports were questioned due to the persistence of somatic cell marker expression in cultured cells, raising uncertainty regarding the purity of these SSC cultures (Ko et al., 2010). Consequently, adaptations are required to establish reliable in vitro SSC expansion systems. Some recent studies focused on short term cultures (Langenstroth et al., 2014; J. F. Smith et al., 2014), trying to improve early stage culture conditions for successful long term proliferation. However, a major obstacle is testicular somatic cell overgrowth which appears to inhibit SSC proliferation. As demonstrated by Smith et al, 2014, human SSC markers UTF1, FGFR3, and DAZL expression progressively declined throughout culture while somatic cell markers VIM, ACTA2 and GATA4 increased (J. F. Smith et al., 2014).

Our knowledge of SSC biology arises mostly from studies in rodents (Boitani, Di Persio, Esposito, & Vicini, 2016). While long term SSC proliferation among human samples is debatable, murine  SSC expansion over months in vitro has been reported by multiple researchers (Kanatsu-Shinohara et al., 2005; Kanatsu-Shinohara et al., 2003; Seandel et al., 2007) and murine models have been used extensively for SSC culture studies which focused on novel approaches and techniques for in vitro SSC propagation (He et al., 2015; Kanatsu-Shinohara et al., 2003) and differentiation (Abu Elhija, Lunenfeld, Schlatt, & Huleihel, 2012). It should be noted that most murine SSC culture experiments included juvenile mice whose testes have a relatively high SSC concentration and were harvested prior to initiation of spermatogenesis (Abu Elhija et al., 2012; He et al., 2015). Therefore the relevance of this model for adult pathologies, such as NOA, is uncertain. The aim of the current study was to perform preliminary assessments of innovative approaches to optimize SSC proliferation in vitro using an adult mouse model. The first approach focused on establishing an optimal ratio of testicular floating cells (presumably enriched for germ cells)-to-attached cells (presumably enriched for somatic cells) following DP. We hypothesized that effective SSC proliferation would be achieved by the addition of somatic cells in an optimal ratio which provides vital support, while minimizing somatic cell overgrowth. The second approach included the use of repeated DP at every passage in order to maintain a low number of somatic cells. We hypothesized that repeated DP might increase SSC proliferation by preventing somatic cell overgrowth.

Methods

Tissue harvesting, mechanical and enzymatic digestion

All work was performed in accordance with CACC guidelines and with approval from the Toronto Centre for Phenogenomics animal care committee (#17-0245). 6-8 week old CD-1 murine testicular samples were harvested as described previously (Kanatsu-Shinohara et al., 2003). Following euthanasia by CO2 inhalation, tissue harvest was performed from mice by midline incision followed by extraction of the testis and removal of the tunica. Mechanical digestion was performed by cutting testes into small pieces (1-2 mm each) followed by mincing in HBSS (Life technologies, # 14175-095) using fine needles. Tissue pieces were transferred to 15mL tubes and washed three times, with 5 minute spins at 1,200rpm. Enzymatic digestion was performed in two steps. First, 2mL of collagenase solution (Sigma-Aldrich) was added for 15 minutes at 37^oC with repetitive shaking followed by double washing with HBSS. Second, 1.6mL 0.25% trypsin (Thermo Fischer Scientific, Waltham, MA, USA) and 0.4mL DNase (Sigma-Aldrich, St Luis, Missouri, USA) were added for 10 minutes at 37^oC with constant shaking. Thereafter, 5mL of Stempro34 (ThermoFischer Scientific, Waltham, MA, USA, #10640) with 2% Stempro34 supplements (Invitrogen Cat#10639) were added to the single cell suspensions. Samples were centrifuged at 1,200 rpm for 5 minutes, the supernatant was removed and cells were counted using the automated Cell Countess (Life Technologies).

Differential plating (DP)

Cell suspensions were  incubated overnight in basal SSC media as described previously (Kanatsu-Shinohara & Shinohara, 2010) including 2% FBS (Hyclone, Lot # AWK24007), 1% Penicillin/ Streptomycin and 1L/mL gentamicin (Sigma-Aldrich, #G1397-10MI) in 10cm dishes (BD Biosciences) coated with 0.2% gelatin (Sigma-Aldrich, Cat. G1393) in 5%  CO₂ at 37^oC at a density of 150,000 cells per cm². The next morning, floating cells were collected and transferred for the second DP in identical conditions. Attached cells in the first DP dish were hydrated with PBS (Millipore, #6508). Floating cells were collected after 4-5 hours. Attached cells were combined from both first (overnight) and second (4-5 hours) DP dishes by removing PBS and adding TrypleE (Life technologies, #12605028) to all first and second DP for 3-5 minutes. All attached cells were separated from the floating cells and combined. All tubes were centrifuged at 1,200rpm for 5 minutes and the supernatant was aspirated. Both floating and attached cell populations were counted using the automated Cell Countess (Life Technologies) prior to establishing cultures.

Experiment 1: culture in various floating/germ to attached/somatic cells ratios

Floating cells were co-cultured with attached cells after DP in 0.2% laminin- (VWR International Co) coated 12 well plates (BD Biosciences) with supplemented StemPro34 media. The following growth factors and reagents were added the media: GDNF (10ng/mL, R&D system, Minneapolis, USA, Cat212-GD-010), LIF (103142 U/mL, EMD Millipore, 143 Darmstadt, Germany; Cat. GF342,), EGF (20ng/mL, Peprotech, Rocky Hill, NJ, USA; Cat. AF-144 100-15), bFGF (10ng/mL, Peprotech, Rocky Hill, NJ, USA; Cat. Af-100-18B) (Kanatsu-Shinohara & Shinohara, 2010).

Six cultures were established (P0) as followed: 5 study cultures with specific 5 floating-to-attached cells ratios: 5, 10, 15, 20 and 25. Each culture ratio contained 10,000 attached cells and corresponding number of floating cells – 50,000, 100,000, 150,000, 200,000 and 250,000, respectively. An additional control culture condition consisted of  100,000 floating cells only to mimic previous reports of SSC expansion in a murine model without the addition of attached/somatic cells after DP (Kanatsu-Shinohara et al., 2003).

Cells were passaged from P0 to P1 and from P1 to P2 every 14-18 days according to cell confluence. After P0, 40,000 cells in all ratios (except cultures which contained lower numbers of cells) were passaged to P1 in order to prevent quantity-related bias. All cells were passaged from P1 to P2. As opposed to P0 establishment, in which cell number was calculated separately for the floating/germ vs. attached/somatic cells, combined cell populations were passaged to P1 and P2. The experiment was ended upon the loss of the majority of cells in ratios 5 and 25 after P2 (after 45-50 days in culture) and was replicated 6 times.

Experiment 2: Use of repeated DP at each passage

P0 culture conditions included identical ratios as described in experiment 1. However, in these experiments we repeated the DP at every passage to maintain a low density of attached cells (<10,000 per culture). DP was performed by overnight incubation on 0.2% gelatin-coated 12 well plates using basal SSC media with 2% FBS – 1% Pen Strep + 1µl/ml gentamicin as described above. The next morning floating and attached cell fractions were separated and counted. Up to 10,000 attached cells were passaged. With regards to floating cells, a constant number (40,000) between ratios was passaged from P0 to P1, and all cells were passaged from P1 and to P2. Therefore, similarly to experiment 1, passage to P1 included a constant number of cells in order to focus on the ratio between cells fractions. In contrast to experiment 1, experiment 2 included separate floating and attached cell quantities based on calculation at each passage. This experiment was replicated 3 times.

Quantification of cell numbers and colonies

During both experiments, we assessed cell quantities using two methods. First, we counted the cells at every passage using an automated cell counter (Life Technologies). While in experiment 1 we counted and passaged combined germ and somatic cells fractions, in experiment 2 we separated floating and attached cell fractions and each of them was counted separately. Colonies, defined as clusters of more than 10 small round cells, were counted by two independent researchers, blinded to the cell ratio in each culture dish.

Q-PCR

Cells were resuspended in 350L of Buffer RL + 0.1% β-mercaptoethanol (Norgen Biotek Corp., Thorold, ON, CAN) and lysed. Genomic DNA was sheared by passing the lysate through a 25G needle. RNA was purified as per manufacturer’s instructions. Briefly, prior to loading the column, 100% ethanol was added to the lysate to precipitate the RNA.. The RNA-bound column was washed, the column was dried to remove residual ethanol, and the RNA was eluted in 25L  of elution buffer. RNA concentration was assessed by NanoVue (GE Healthcare, Little Chalfont, UK).

100ng of total RNA was used for preamplification and cDNA synthesis using the RT² PreAMP cDNA Synthesis Kit (Qiagen, Toronto, ON, CAN) as per manufacturer’s instructions. Briefly, genomic DNA was eliminated and first strand cDNA was synthesized. The cDNA was then amplified using RT² q-PCR Primer Assays for SSC and somatic cells markers GFRα1 and GATA4, respectively (Qiagen, Toronto, ON, CAN).

Q-PCR was performed using 1L  of preamplified cDNA, 2.5µM of primers and 12.5µl of 2X RT² SYBR Green FAST Mastermix (Qiagen, Toronto, ON, CAN). Cycling conditions were 95°C for 10 minutes to activate the polymerase, and 40 cycles at 95°C for 15 seconds and 60°C for 1 minute. A negative control (without template) was used for each primer pair and all samples were analyzed in triplicate. Melting curves were generated for each reaction to ensure primer specificity. Relative fold change was calculated using the 2^–ΔΔCt method with GAPDH as a reference.

Flow cytometry

SSC enrichment was assessed using flow cytometry for GFRα1, a well-known mouse SSC marker (Abu Elhija et al., 2012; Huleihel, Nourashrafeddin, & Plant, 2015; Khajavi et al., 2014; Li et al., 2016; Stukenborg et al., 2008). We compared the percentage of GFRα1-positive cells after differential plating (prior to P0) and in the optimal ratio 20 after P2. Cells were collected into flow cytometry tubes with addition of 3% FBS–PBS (FACS buffer). Primary rabbit anti-GFR1α antibody (1:100, Abcam, AB80126) was added followed by 30 minutes incubation at 4°C. Following a single wash, cells were resuspended in FACS buffer containing AlexaFluor-conjugated -goat anti-rabbit Alexa Fluor 488 (1:100, Cell Signaling Technologies, #4412S) incubated for 30 minutes. After one wash, cell suspensions were stained with Propidium Iodide (PI; Sigma-Aldrich, St. Louis, Missouri, United States). Unstained cells and cells stained only with secondary antibody acted as negative controls. Analysis was performed on FACS Aria II-SC BRV (Becton Dickinson, Mississauga, Ontario) at the Sickids-UHN Flow Cytometry Facility (Toronto, Ontario).

Immunocytochemistry (ICC):

Cells were transferred to slides and left to be dried in laminar flow, fixed in 4% paraformaldehyde for 20 minutes followed by single washing with PBS, and stored in 4°C for up to 2 weeks. Slides were rinsed 3 times with PBS and then permeabilized for 15 minutes in 0.1% Triton X (Millipore, #9410) and 1% Bovine Serum Albumin (MedStore, #323208) diluted in PBS. Blocking was performed in 5% normal goat serum (NGS, Thermo Fischer, 01-6201) in 1% BSA – PBS for 1 hour at room temperature. Hereafter, primary antibodies including rabbit anti-GFRα1 (1:200, Abcam, Ab8026), mouse anti-UTF1 (1:500, Millipore MAB4337) and rabbit anti-sox9 (1:200, Abcam Ab185230) diluted in 5% NGS- 1% BSA – PBS were added and incubated at 4°C overnight. Controls were carried out under identical conditions except that the primary antibodies were replaced with 1% BSA – PBS. After washing three times with 1% BSA – PBS, cells were incubated with secondary Alexa Fluor555-conjugated goat anti-rabbit antibodies (Cell Signaling Technologies #4413) and Alexa Fluor488-conjugated anti-mouse antibodies (Cell Signaling Technologies #4408) diluted 1:500 in 5% NGS – 1% BSA – PBS. Three additional washes were performed with 1% BSA – PBS followed by addition of Hoechst (Thermo Fischer, H1399) diluted 1:2000 in 1% BSA – PBS for 3-5 minutes. After three washes with 1% BSA – PBS, slides were mounted with Permafluor  (Thermo Fisher Scientific, TA030FM) and a cover slip. Slides were evaluated using a fluorescence microscope (Nikon 80i, Japan). The microscopic appearance of the cell culture was captured using a color SPOT RT camera (Diagnostic Instruments, USA).

Statistical analysis

Cell and colony counts, presented as mean ± standard errors, were compared between ratios in each experiment as well as between similar ratios in experiments 1 and 2. ANOVA was performed and where significant differences were observed, a post hoc Tukey analysis was performed. Changes in gene and protein expression were analyzed using a student t test or one-way ANOVA and Bonferroni’s multiple comparison tests. P values <0.05 were considered significant throughout analysis.

Results

Cell and colony counts during culture

Experiment 1: Single DP (S-DP):

Six cultures with specific floating/attached cells ratios were established after DP (P0). During P0, cell counts declined by 40-60% compared to the plated cell counts. Cell and colony counts at the end of P0 correlated with the initial number of cells in each ratio, as ratios 5 and 25 contained the lowest and highest average numbers of cells and colonies, respectively (Figures 1A, 1B). In order to focus on the impact of cell ratios and to avoid biases related to cell quantities, a constant number of cells (40, 000) was passaged from each culture from P0 to P1 in all ratios except ratio 5 which contained a slightly reduced number of cells. At P1, we noted the highest proliferation in ratio 20.  At P2, the ratios 15 and 20 exhibited the highest proliferation rates as reflected in both cell numbers (338,000 ± 165,000 and 336, 700 ± 31,800, respectively) and colony numbers (418.3 ± 165.2 and 581.7 ± 133.8, respectively). Cell and colony counts in ratio 20 were significantly higher compared to ratios 5, 10, 25 and control (p<0.05) (Figures 1A, 1B). Similar significant differences were observed in ratio 15 compared to ratios 5, 10, 25 and control (p<0.05), except that the increase in colony counts did not reach statistical significance compared to ratio 10. Colonies were observed not only in high numbers but also had homogeneous appearances (Figure 1E). The loss of cells in the extreme ratio 5, 25 and control cultures led us to stop the experiment after P2 (Figures 1A, 1B).

Experiment 2: Repeated DP during each passage (R-DP):

Cells counts after P0 declined in correlation to initial amount of cultured cells (30,700 ± 2,600 floating cells in ratio 5, 36,000 ± 15,800 in 10, 74,700 ± 1,500 in ratio 15, 75,700 ± 18,700 in ratio 20, 99,000 ± 23,000 in ratio 25 and 104,700 ± 37,900 in control cultures). Colony counts correlated with cell counts and increased continuously from ratio 5 to 25. An identical number of cells were transferred after performing R-DP to P1 in all ratios to prevent quantity-related bias, including up to 10,000 attached cells and 40,000 floating cells. At the end of P1, ratio 15 included the highest number of cells and colonies (97,000 ± 800 and 41 ± 7.5, respectively) followed by ratio 20 (72,700 ± 6,800 and 35.3 ± 10.7, respectively) compared to lower and higher ratios, without reaching statistical significance. At P2, the overall cell and colony counts declined in all ratios, while maintaining the trend of increased counts in middle compared to extreme ratios (Figure 1C, 1D).

Comparison between S- and R-DP

Both experiments led to similar findings. First, as equal culture conditions were established in P0, it is not surprising that comparable drops in cell counts and colonies counts were found in correlation with initial amount of cultured cells at the end of P0. Second, an initial pattern of increased counts in the middle ratios 15 and 20 when compared to extreme ratios evolved during P1, and became more prominent in P2. While counts in P0 and P1 were similar, major differences between the S- and R-DP evolved during P2. In S-DP cultures, cell and colony counts increased dramatically in middle ratios (especially ratio 20 but also 15 and 10) while low and high ratio (5 and 25, respectively) counts remained relatively constant as in P1. On the contrary, cells and colony counts in R-DP cultures dropped during P2, resulting with dramatic gaps between the experiments (Figure 1A-D). Therefore, the higher proliferation of intermediate ratios compared to extreme ratios during P2 was observed in both experiments but reached statistical significant only in S-DP cultures.

Cell characterization

In order to investigate SSC enrichment in the culture, we performed q-PCR and flow cytometry for GFRα1, one of the most commonly used murine SSC marker. We focused on the S-DP culture conditions since it gave us the highest numbers of colonies. As expected, GFRα1 expression after DP was significantly higher in the floating cell fraction compared to attached cells (Figure 2A). Moreover, the expression of GFRα1 after P2 was significantly increased in the ratios of 15 and 20 (p<0.01) and 10 (p<0.05) but there was no difference between the extreme ratios of 5 and 25 when compared to control (Figure 2B). The expression of somatic cell-associated marker GATA4  was similar after DP in floating and attached cell populations (Figure 2C) while, after P2, all ratios showed significantly increased GATA4 expression (p<0.05) compared to control, which did not include attached cells (Figure 2D). Thereafter, we compared the proportion of GFRα1-positive cells at P2 in the ratio 20 condition of the S-DP experiment (which showed optimal results) to the floating cell population after DP at the start of P0. The percentage of GFRα1-positive cells in ratio 20 was 25.8 ± 5.8% compared to 1.9 ± 0.7% in the floating cell fraction after DP (p<0.0001), representing an enrichment of 13.6 fold (Figure 3A, 3B). In absolute numbers, while the initial ratio 20 culture after DP included an average of 3,771 GFRα1-positive cells, after P2 it contained an average of 87,204 GFRα1-positive cells which is a 23.1 fold increase (Figure 3C).  This is in spite of intentionally excluding the majority of cells between P0 to P1 to match the number of passaged cells between ratios.

GFRα1 and UTF1 were used as SSC markers in addition to Sertoli cell-associated marker Sox9 were used to further characterize the cells by immunocytochemistry. At P2, cell clusters from ratio 20 were positive for both SSC markers. A few cells stained positive for Sox9, confirming the existence of supporting cells within the expansion of cell and colonies (Figure 4 and supplementary Figure 1).

Discussion

Spermatogenesis depends on cell-cell interactions between somatic and germ cell populations. The complex interplay within the testicular niche includes growth factors provided by Sertoli and interstitial cells in addition to vascular network stimuli targeted to SSCs (Franca, Hess, Dufour, Hofmann, & Griswold, 2016). Unique junctional complexes between SSCs and Sertoli cells are formed over the basement membrane through active cytoskeleton, enabling metabolic and immunologic germ cells support (L. B. Smith et al., 2012). SSC expansion in vitro is crucial before further clinical fertility preservation applications such as auto-transplantation or in vitro spermatogenesis followed by IVF-ICSI can become a reality. While SSC can be greatly expanded in culture and even restore fertility upon transplantation in rodents and large animals (Valli, Phillips, et al., 2014), long term SSC proliferation using human cells remains challenging.

Previous studies have suggested various culture systems for long term murine SSC culture, mostly by DP followed by culturing the floating fraction (enriched for germ cells) alone (He et al., 2015; Kanatsu-Shinohara et al., 2011; Kanatsu-Shinohara et al., 2005; Kanatsu-Shinohara, Mori, & Shinohara, 2013; Kanatsu-Shinohara, Morimoto, & Shinohara, 2012; Kanatsu-Shinohara et al., 2003). Since murine models are well-accepted tools to investigate novel concepts and methodologies, the current research examined innovative interventions to prevent somatic cell overgrowth which is a major hindrance to ex-vivo SSC propagation. We found that a delicate numerical balance between somatic and germ cells provides crucial somatic cell support for germ cells while avoiding their overgrowth. The importance of somatic cell support is emphasized while comparing the optimal ratios 15 and 20 to control culture, which included only floating cells as described previously (Kanatsu-Shinohara et al., 2003). Utilizing DP, a simple cell enrichment methodology, both cell fractions can be easily separated (although sub-optimally purified), followed by seeding the cells in culture-specific ratios. An optimal expansion of cells and colonies was achieved in the intermediate floating-to-attached cells ratios (10, 15 and 20) after DP. On the other hand, too low (5) or too high (25) ratios resulted in a loss of cells after culture for 45-50 days in both S- and R-DP experiments. The superiority of intermediate compared to extreme ratios was demonstrated not only quantitatively but also by increased GFRα1 expression and GRF1a-positive cells. Specifically, ratio 20 in the S-DP experiment demonstrated the optimal cell proliferation and colony counts after P2, resulting in significantly increased proliferation rate compared to control culture conditions which contained floating cells only (Kanatsu-Shinohara et al., 2005; Kanatsu-Shinohara et al., 2003). Moreover, while previous studies used prepubertal samples (He et al., 2015; Huleihel et al., 2015; Valli, Phillips, et al., 2014), we used adult mice, whose testes contain a lower SSC concentration, emphasizing the importance of the demonstrated SSC proliferation and purification.

The second approach involved repeated DP at every passage in order to control and maintain a low number of somatic cells. While P0 and P1 resulted with similar cell and colony counts compared to the S-DP set of cultures, P2 was characterized by opposing trends of decreased vs. increased cell and colony counts in the rec- and non-DP cultures, respectively. Such a difference can be explained by the excessive cell manipulations performed during the R-DP methodology. Interestingly, R-DP experiment suggested a survival advantage in the intermediate ratios (to a lesser extent than S-DP cultures), despite impaired cell survival.

Mammalian spermatogenesis shows several common features between species: a) the importance of SSC self-renewal to preserve the SSC pool; b) spermatogenesis is divided into a mitotic phase, meiotic stage and spermiogenesis and; c) complex cell-cell interactions within the testicular niche (Chalmel et al., 2014; L. B. Smith et al., 2012). However, fundamental differences between mouse and primate spermatogenesis should be noted such as spermatogenesis duration (Gonzalez & Dobrinski, 2015), number of mitotic divisions prior to initiating meiosis (Gassei et al., 2014), and SSC markers (Abu Elhija et al., 2012; Kokkinaki, Djourabtchi, & Golestaneh, 2011; Nickkholgh et al., 2014; Sadri-Ardekani et al., 2009). Using human samples, our initial results are very similar and promising. We found significantly higher SSC-like aggregates after 12 days in specific floating/attached cell ratio co-cultures, compared to floating cells alone (Gat et al., 2016). These findings emphasize the importance of somatic support for germ cell survival in vitro (J. F. Smith et al., 2014). However, in contrast to the mouse model, DP followed by floating-germ cell only cultures with human cells eventually resulted in somatic cell over growth (Langenstroth et al., 2014). This suggests that adding a low number of attached cells to the floating cells at P0 may improve initial germ cell survival, but will likely eventually lead to somatic cell overgrowth. Therefore, additional methods need to be devised to limit the detrimental impact of somatic cell overgrowth. In our opinion, the unsatisfactory outcome of R-DP compared to S-DP in the mouse model should not preclude attempting to study this in human samples. However, the excessive manipulations and time involved in R-DP are a potential drawback of this intervention.

The current study has some limitations. First, since we stopped the cultures after P2 due to loss of cells in ratios 5 and 25, we could not assess the impact of cell ratios in longer term culture. Second, the current study did not include a germ cell transplantation assay to demonstrate a true SSC phenotype in the expanded colonies (Gassei et al., 2014). However, the increase in GFRα1 (a well-accepted murine SSC marker (Li et al., 2016)) RNA and protein expression after P2, suggests that we achieved successful enrichment including a relatively high proportion of SSC in our colonies. While the current study introduces promising results, we intend to conduct transplantation assays in future experiments and also longer term cultures of cells at the optimal cell ratios of 15 and 20, as demonstrated in these experiments. Third, since we have performed ICC to colonies after P2 to demonstrate their SSC enrichment, cells are crowded over each other, therefore optimal distinction between nuclear and membrane localization is extremely challenging. However, we believe that the current ICC results, accompanied by QPCR and flow cytometry supply comprehensive picture regarding the SSC enrichment with our specific culture conditions.

To the best of our knowledge this is the first study of the effect of floating to attached cell ratios after DP on SSC proliferation and colony formation. R-DP seems to be less effective compared to S-DP in this CD-1 adult mouse model. While a ratio of 20 for S-DP appears to be the optimal conditions for in vitro generation of SSC-like cells, further studies are needed in order to evaluate the precise ratio in longer culture duration as well as performing xenotransplanation. Translating these methodologies to human samples may help optimize in vitro SSC expansion and thereby lead to novel fertility preservation strategies for young boys prior to cancer treatment and to potential therapeutic interventions for some men with severely abnormal spermatogenesis.

Acknowledgements: The authors would like to thank Ekaterina Shlush for general contribution to the study, Lianet Lopez for technical assistance and Kevin Quach for REB submission and statistical analysis.

The authors declare no conflict of interest

Author contribution statement:

IG designed the study, performed cells culture, collected the data, participated in data analysis and wrote the manuscript. LM and MF performed cells cultures. SH and BW performed QPCR. KZ contributed to manuscript editing. PZ analysed FACS data. AGF contributed to data analysis and manuscript editing. CL contributed to manuscript editing and approved the final version.

References

Abu Elhija, M., Lunenfeld, E., Schlatt, S., & Huleihel, M. (2012). Differentiation of murine male germ cells to spermatozoa in a soft agar culture system. Asian J Androl, 14(2), 285-293. doi: 10.1038/aja.2011.112

Albert, S., Wistuba, J., Eildermann, K., Ehmcke, J., Schlatt, S., Gromoll, J., & Kossack, N. (2012). Comparative marker analysis after isolation and culture of testicular cells from the immature marmoset. Cells Tissues Organs, 196(6), 543-554. doi: 10.1159/000339010

Boitani, C., Di Persio, S., Esposito, V., & Vicini, E. (2016). Spermatogonial cells: mouse, monkey and man comparison. Semin Cell Dev Biol. doi: 10.1016/j.semcdb.2016.03.002

Chalmel, F., Com, E., Lavigne, R., Hernio, N., Teixeira-Gomes, A. P., Dacheux, J. L., & Pineau, C. (2014). An integrative omics strategy to assess the germ cell secretome and to decipher sertoli-germ cell crosstalk in the Mammalian testis. PLoS One, 9(8), e104418. doi: 10.1371/journal.pone.0104418

Conrad, S., Azizi, H., Hatami, M., Kubista, M., Bonin, M., Hennenlotter, J., . . . Skutella, T. (2014). Differential gene expression profiling of enriched human spermatogonia after short- and long-term culture. Biomed Res Int, 2014, 138350. doi: 10.1155/2014/138350

Conrad, S., Renninger, M., Hennenlotter, J., Wiesner, T., Just, L., Bonin, M., . . . Skutella, T. (2008). Generation of pluripotent stem cells from adult human testis. Nature, 456(7220), 344-349. doi: 10.1038/nature07404

Franca, L. R., Hess, R. A., Dufour, J. M., Hofmann, M. C., & Griswold, M. D. (2016). The Sertoli cell: one hundred fifty years of beauty and plasticity. Andrology, 4(2), 189-212. doi: 10.1111/andr.12165

Gassei, K., Valli, H., & Orwig, K. E. (2014). Whole-mount immunohistochemistry to study spermatogonial stem cells and spermatogenic lineage development in mice, monkeys, and humans. Methods Mol Biol, 1210, 193-202. doi: 10.1007/978-1-4939-1435-7_15

Gat, I.,  Maghen, L.,  Felice, M.,  Wyse, B., Zohni, K.,  Jarvi, K., Lo, C., Gauthier-Fisher, A.,  LibrachC.

Optimal Culture Conditions are Critical for Efficient Expansion of Human Testicular Somatic and

Germ Cells In Vitro. Fertility and Sterility. Accepted.

Gonzalez, R., & Dobrinski, I. (2015). Beyond the mouse monopoly: studying the male germ line in domestic animal models. ILAR J, 56(1), 83-98. doi: 10.1093/ilar/ilv004

He, B. R., Lu, F., Zhang, L., Hao, D. J., & Yang, H. (2015). An alternative long-term culture system for highly-pure mouse spermatogonial stem cells. J Cell Physiol, 230(6), 1365-1375. doi: 10.1002/jcp.24880

Huleihel, M., Nourashrafeddin, S., & Plant, T. M. (2015). Application of three-dimensional culture systems to study mammalian spermatogenesis, with an emphasis on the rhesus monkey (Macaca mulatta). Asian J Androl, 17(6), 972-980. doi: 10.4103/1008-682X.154994

Izadyar, F., Wong, J., Maki, C., Pacchiarotti, J., Ramos, T., Howerton, K., . . . Copperman, A. (2011). Identification and characterization of repopulating spermatogonial stem cells from the adult human testis. Hum Reprod, 26(6), 1296-1306. doi: 10.1093/humrep/der026

Kanatsu-Shinohara, M., Inoue, K., Ogonuki, N., Morimoto, H., Ogura, A., & Shinohara, T. (2011). Serum- and feeder-free culture of mouse germline stem cells. Biol Reprod, 84(1), 97-105. doi: 10.1095/biolreprod.110.086462

Kanatsu-Shinohara, M., Miki, H., Inoue, K., Ogonuki, N., Toyokuni, S., Ogura, A., & Shinohara, T. (2005). Long-term culture of mouse male germline stem cells under serum-or feeder-free conditions. Biol Reprod, 72(4), 985-991. doi: 10.1095/biolreprod.104.036400

Kanatsu-Shinohara, M., Mori, Y., & Shinohara, T. (2013). Enrichment of mouse spermatogonial stem cells based on aldehyde dehydrogenase activity. Biol Reprod, 89(6), 140. doi: 10.1095/biolreprod.113.114629

Kanatsu-Shinohara, M., Morimoto, H., & Shinohara, T. (2012). Enrichment of mouse spermatogonial stem cells by melanoma cell adhesion molecule expression. Biol Reprod, 87(6), 139. doi: 10.1095/biolreprod.112.103861

Kanatsu-Shinohara, M., Ogonuki, N., Inoue, K., Miki, H., Ogura, A., Toyokuni, S., & Shinohara, T. (2003). Long-term proliferation in culture and germline transmission of mouse male germline stem cells. Biol Reprod, 69(2), 612-616. doi: 10.1095/biolreprod.103.017012

Kanatsu-Shinohara, M., & Shinohara, T. (2010). Germline modification using mouse spermatogonial stem cells. Methods Enzymol, 477, 17-36. doi: 10.1016/S0076-6879(10)77002-6

Khajavi, N., Akbari, M., Abdolsamadi, H. R., Abolhassani, F., Dehpour, A. R., Koruji, M., & Habibi Roudkenar, M. (2014). Role of Somatic Testicular Cells during Mouse Spermatogenesis in Three-Dimensional Collagen Gel Culture System. Cell J, 16(1), 79-90.

Ko, K., Arauzo-Bravo, M. J., Tapia, N., Kim, J., Lin, Q., Bernemann, C., . . . Scholer, H. R. (2010). Human adult germline stem cells in question. Nature, 465(7301), E1; discussion E3. doi: 10.1038/nature09089

Kokkinaki, M., Djourabtchi, A., & Golestaneh, N. (2011). Long-term Culture of Human SSEA-4 Positive Spermatogonial Stem Cells (SSCs). J Stem Cell Res Ther, 2(2). doi: 10.4172/2157-7633.S2-003

Koruji, M., Shahverdi, A., Janan, A., Piryaei, A., Lakpour, M. R., & Gilani Sedighi, M. A. (2012). Proliferation of small number of human spermatogonial stem cells obtained from azoospermic patients. J Assist Reprod Genet, 29(9), 957-967. doi: 10.1007/s10815-012-9817-8

Kossack, N., Terwort, N., Wistuba, J., Ehmcke, J., Schlatt, S., Scholer, H., . . . Gromoll, J. (2013). A combined approach facilitates the reliable detection of human spermatogonia in vitro. Hum Reprod, 28(11), 3012-3025. doi: 10.1093/humrep/det336

Langenstroth, D., Kossack, N., Westernstroer, B., Wistuba, J., Behr, R., Gromoll, J., & Schlatt, S. (2014). Separation of somatic and germ cells is required to establish primate spermatogonial cultures. Hum Reprod, 29(9), 2018-2031. doi: 10.1093/humrep/deu157

Li, L., Wang, M., Wu, X., Geng, L., Xue, Y., Wei, X., & Jia, Y. (2016). A long non-coding RNA interacts with Gfra1 and maintains survival of mouse spermatogonial stem cells. Cell Death Dis, 7, e2140. doi: 10.1038/cddis.2016.24

Livne-Segev, D., Forbes, E. C., & Lo, K. C. (2016). State-of-the art advances in fertility preservation for the male cancer patient. Future Oncol, 12(14), 1691-1694. doi: 10.2217/fon-2016-0185

Nagano, M. C. (2003). Homing efficiency and proliferation kinetics of male germ line stem cells following transplantation in mice. Biol Reprod, 69(2), 701-707. doi: 10.1095/biolreprod.103.016352

Nickkholgh, B., Mizrak, S. C., Korver, C. M., van Daalen, S. K., Meissner, A., Repping, S., & van Pelt, A. M. (2014). Enrichment of spermatogonial stem cells from long-term cultured human testicular cells. Fertil Steril, 102(2), 558-565 e555. doi: 10.1016/j.fertnstert.2014.04.022

Sadri-Ardekani, H., Mizrak, S. C., van Daalen, S. K., Korver, C. M., Roepers-Gajadien, H. L., Koruji, M., . . . van Pelt, A. M. (2009). Propagation of human spermatogonial stem cells in vitro. JAMA, 302(19), 2127-2134. doi: 10.1001/jama.2009.1689

Seandel, M., James, D., Shmelkov, S. V., Falciatori, I., Kim, J., Chavala, S., . . . Rafii, S. (2007). Generation of functional multipotent adult stem cells from GPR125+ germline progenitors. Nature, 449(7160), 346-350. doi: 10.1038/nature06129

Smith, J. F., Yango, P., Altman, E., Choudhry, S., Poelzl, A., Zamah, A. M., . . . Tran, N. D. (2014). Testicular niche required for human spermatogonial stem cell expansion. Stem Cells Transl Med, 3(9), 1043-1054. doi: 10.5966/sctm.2014-0045

Smith, L. B., Milne, L., Nelson, N., Eddie, S., Brown, P., Atanassova, N., . . . Peters, J. (2012). KATNAL1 regulation of sertoli cell microtubule dynamics is essential for spermiogenesis and male fertility. PLoS Genet, 8(5), e1002697. doi: 10.1371/journal.pgen.1002697

Steinberger, A., & Steinberger, E. (1966). In vitro culture of rat testicular cells. Exp Cell Res, 44(2), 443-452.

Stukenborg, J. B., Wistuba, J., Luetjens, C. M., Elhija, M. A., Huleihel, M., Lunenfeld, E., . . . Schlatt, S. (2008). Coculture of spermatogonia with somatic cells in a novel three-dimensional soft-agar-culture-system. J Androl, 29(3), 312-329. doi: 10.2164/jandrol.107.002857

Tegelenbosch, R. A., & de Rooij, D. G. (1993). A quantitative study of spermatogonial multiplication and stem cell renewal in the C3H/101 F1 hybrid mouse. Mutat Res, 290(2), 193-200.

Valli, H., Phillips, B. T., Shetty, G., Byrne, J. A., Clark, A. T., Meistrich, M. L., & Orwig, K. E. (2014). Germline stem cells: toward the regeneration of spermatogenesis. Fertil Steril, 101(1), 3-13. doi: 10.1016/j.fertnstert.2013.10.052

Valli, H., Sukhwani, M., Dovey, S. L., Peters, K. A., Donohue, J., Castro, C. A., . . . Orwig, K. E. (2014). Fluorescence- and magnetic-activated cell sorting strategies to isolate and enrich human spermatogonial stem cells. Fertil Steril, 102(2), 566-580 e567. doi: 10.1016/j.fertnstert.2014.04.036

Figure capitations

Fig 1: Cell and colony counts throughout culture in S- and R-DP at various adherent to floating cell ratios. 40%-60% drop in cell counts were observed during P0 throughout culture conditions. In order to focus on the cell ratio impact, a constant number of cells were passaged in all cultures from P0 to P1 except cultures in ratio 5 which contained a reduced number of cells. All cells were passaged from P1 to P2 in all cultures. After P2, significantly higher cell and colonies counts were observed in the S-DP experiment, with an  impressive rise specifically during P2. R-DP had inferior outcomes compared to S-DP in both cell and colony counts throughout the experiment. Horizontal lines represent significant difference (p<0.05) between optimal and other ratios.

Fig 2: Q-PCR after DP, presenting significantly higher GFRα1 expression in the floating cells compared to attached cells (A). After P2, GFRα1 expression was significantly higher in the intermediate ratios (especially in ratios 15 and 20, p<0.01) without differences between extreme ratios 5 and 25 and control (B). Somatic cell marker GATA4 expression was similar after DP on floating and attached cell populations (C), while after P2 all ratios had significantly higher GATA4 expression (p<0.05) compared to control, which did not include attached cells (D).

Fig 3: Comparison of GFRα1 positive cells in flow cytometry between floating-germ cell population after DP and ratio 20 in S-DP experiment after P2, demonstrating enrichment from 1.9±0.7% positive cells to 25.8±5.8% (A and B, respectively). In absolute numbers, we observed a 23.1 fold increase in GFRα1 positive cells in ratio 20 of S-DP experiment, despite passaging less than 40% of the cells from P0 to P1 (C).

Fig 4: Immunocytochemistry of colonies in ratio 20 in the S-DP experiment after P2. Cells within colonies were positive for both SSC markers GFRα1 and UTF1 (A). Very few cells within the colonies stained positive for Sox9 (B).