Concise Review: Spermatogenesis in an Artificial Three-Dimensional System§


  • Huleihel Mahmoud

    Corresponding author
    1. The Shraga Segal Department of Microbiology and Immunology, Faculty of Health Sciences, Ben-Gurion University of the Negev, Beer-Sheva, Israel
    • The Shraga Segal Department of Microbiology and Immunology, Faculty of Health Sciences, Ben-Gurion University of the Negev, Beer-Sheva 84105, Israel

    Search for more papers by this author

  • Author contribution: M.H.: wrote the review.

  • Disclosure of potential conflicts of interest is found at the end of this article.

  • §

    First published online in STEM CELLSEXPRESS August 7, 2012.


Culture of spermatogonial stem cells has been performed under a variety of conditions. Most featured two-dimensional systems, with different types of sera, conditioned media, feeder layers, and growth factors. Some have used three-dimensional (3D) matrices produced from gelatin, collagen, or other material. In spite of their increasingly sophisticated composition, however, complete spermatogenesis in vitro has not yet been achieved. In the seminiferous tubules, spermatogenesis occurs in an environment where cells are embedded in a 3D structure with specific niches regulating each stage of germ cell maturation mediated by hormones and paracrine/autocrine factors. We have recently reported achievement of complete in vitro spermatogenesis of mouse testicular germ cells in a 3D culture system featuring a soft agar matrix. This review discusses the advantages of the 3D culture system for studying the spermatogenic process in its entirety. Also discussed are the steps necessary to expand the applicability of the 3D culture system to human germ cell development and determine the functionality of culture-produced spermatozoa for generating offspring. STEM CELLS2012;30:2355–2360



Spermatogenesis is the process of male germ cell proliferation and differentiation. It begins with a differentiating cell division of diploid spermatogonial stem cells (SSCs) and continues with sequential cell divisions of spermatogonia and meiosis of spermatocytes to form round spermatids [1, 2]. Successful differentiation of round spermatids into the complex structure of the spermatozoon is called spermiogenesis [1, 3].

Tissue-specific stem cells in mammals are defined by their capacity for both self-renewal and differentiation. In most stem cell systems, the stem cells do not derive differentiated cells directly, but rather through intermediate cells that while unable to generate new stem cells can self-renew and differentiate. These progenitor cells play a homeostatic role in maintaining a specific and necessary ratio between proliferating and differentiating cells [4, 5].

The process of SSC development is distinct in different species of mammals. In human and non-human primates, two morphologically distinguishable type of undifferentiated spermatogonia exist, called A(dark) and A(pale). A(dark) exhibits characteristics consistent with their identification as testicular stem cells, while, the A(pale) exhibits characteristics of a progenitor cell [3, 5–9]. In rodents, such as rats and mice, the spermatogonia are grouped into type A. In the mouse, seven types of A spermatogonia have been described [A(single), A(pair), A(aligned), A1–A4] [3, 5, 10, 11]. A(single) are considered to be the SSCs [3, 5]. The A(pair) and A(aligned) spermatogonia are further clonally expanded colonies not synchronized with the seminiferous epithelial cycle [3, 5]. The A1–A4 spermatogonia are usually considered a further expansion of these spermatogonial clones, which are now synchronized with the seminiferous epithelial cycle [3, 5]. B and intermediate spermatogonia are observed at defined spermatogenic stages [3, 5].

In all species, a small population of testicular stem cells functions as a regenerative reserve which is distinguished by an enormous capacity for recolonization of the seminiferous epithelium [3, 5, 12]. Stem cells are active in each cycle of the seminiferous epithelium. SSCs were initially recognized as undifferentiated spermatogonia, present as individual cells on the basement membrane of the seminiferous tubules until type B spermatogonia divide to produce preleptotene spermatocytes [3, 5].

Microenvironment for SSC Development

In mammalian species, spermatogenesis occurs in the seminiferous tubules of the testis and relies on the appropriate expansion of undifferentiated and differentiated spermatogonia prior to the entry of germ cells into meiosis and subsequent spermiogenesis [3, 5]. The process of spermatogenesis is under the general control of the endocrine system, and locally by a variety of direct and indirect interactions and signals between the developing germ cells and the surrounding microenvironment via autocrine/paracrine factors [13–16]. Critical in this regard are cell-cell interactions between the developing germ cells and surrounding sustentacular (i.e., Sertoli) cells. In addition, peritubular cells and various cells in the interstitium (i.e., Leydig cells, macrophages, and endothelial cells) regulate germ cell development by secretion of different factors. These factors might be part of the basement membrane and/or might directly or indirectly affect germ cells and Sertoli cells within the seminiferous tubule [17–20].

The cellular compartment and the basement membrane biomolecules of the seminiferous tubule generate different niches and microenvironments that are specific and distinct for the different stages of SSCs throughout the epithelial cycle. These as yet unidentified niches control SSC self-renewal and/or their differentiation. They may be composed of variable associated and/or soluble factors within the different compartments of the seminiferous tubule (i.e., basal, intraepithelial, and adluminal). Thus, the type of cell, stage of differentiation, and the molecular signals available in the specific microenvironment/niche are key factors in the regulation of SSC development [21]. Current knowledge suggests that the Asingle/Apair/Aaligned population is localized to an area adjacent to the vasculature and interstitium surrounding the seminiferous tubules. From there, upon initiation of differentiation, they migrate out to a niche for A1 spermatogonia, a region likely specialized to provide a microenvironmental niche for Asingle/Apair/Aaligned spermatogonia as well as for stem cells [22–25].


It has been suggested that fluid flow within the seminiferous tubules creates specific local gradients (or compartmentalization) of paracrine/autocrine factors dynamically formed in basal, adluminal, and luminal compartments [26]. Within the three-dimensional (3D) tissue structure, concentration gradients might exist for any soluble culture-medium component consumed or produced by endogenous cells. Diffusion within the culture is affected by a number of factors, including tissue thickness, cell density, and the concentration of the substances on the surface of the tissue [27]. In 3D cultures, static flow may lead to generation of passive gradation of substances that will, over time, be rapidly equilibrated. A forced but controlled perfusion of culture media will, therefore, be crucial [26]. Practical and theoretical aspects of the generation of active diffusion of such signaling agents have previously been considered for 3D culture technology [27]. The loss of autocrine signaling has been cited as one element explaining observed high flow rates through 3D tissues. While posited to improve nutrient transfer, these can paradoxically diminish cell survival and function [27]. Indeed, vascularized tissues exhibit slow interstitial flow, stimulated by small mechanical stresses on the tissue that result in a convection influence on the extracellular transport of growth factors and other large molecules. In this way, interstitial flow is an important coupling factor between mechanical stress and signaling in the 3D matrix [27]. Thus, the distribution regulatory factors must be considered when working with 3D culture systems in order to optimize conditions for cell behavior to most closely approximate that in vivo condition.

Development of Spermatogenesis In Vitro

Culture of isolated SSCs has become a popular approach to study the influence of milieu and identify biomolecular factors involved in the regulation of their proliferation and the differentiation of their progeny [3, 5]. It is known, however, that the cellular composition and the signals produced in two-dimensional (2D) cultures of SSC differ significantly from the in vivo condition, in which conditions necessary for survival, proliferation, differentiation, and recapitulation of SSCs are optimally provided. Numerous 2D culture studies have been performed to explore the optimal conditions for complete spermatogenesis. These studies had to overcome daunting obstacles including: (a) Identification of SSCs. Until very recently, there were no biochemical markers for unambiguous identification of SSCs, which largely precluded assessment of their role as starting cells in culture. The subsequent identification of specific markers for testicular germ cells, some of whom double as markers for SSCs, for example, GFR-α1, CD9, and Thy-1, lead to isolation in this tissue of stem cells and made their use in vitro possible [28]. (b) The very low number of SSCs in the testis (i.e., only two to three stem cells exist per 104 germ cells in the adult mouse). This obstacle remains, although different factors and conditions have been shown to induce proliferation of these cells in vitro, for example, glial cell line-derived neurotrophic factor (GDNF), leukemia inhibitory factor (LIF), and contributions from supporting cells [3]. (c) Conditions promoting differentiation of SSCs in vitro remain elusive [2, 29].

Approaches to Inducing Spermatogenesis In Vivo and In Vitro

Enhancing Recovery of SSCs

To overcome the above-mentioned limitations, several strategies have been adopted. Germ cell transplantation was initially undertaken to identify bioactive SSCs in culture or in suspensions [30]. Isolation/enrichment of SSCs was performed with antibodies specific to germ cell markers using a MiniMACS apparatus. SSC cultures were prepared from neonatal or prepubertal testis to provide a large number of undifferentiated SSCs at less differentiated stages [31]. In addition, induction of pathological conditions to reduce or eliminate differentiated SSCs has been used [3, 32]. Additionally, SSC immortalized lines were established to overcome some of the above limitations. Culture on adherent cell feeder layer (e.g., Vero cells [an epithelial cell line of monkey kidney origin], Sertoli and Leydig cells), serum-free or feeder layer-free culture conditions, conditioned media of different cells, or recombinant growth factors (e.g., addition of specific diffusible factors, e.g., LIF, GDNF, basic fibroblast growth factor, and stem cell factor) has been used to induce SSC survival and proliferation in vitro [3, 32].

None of the 2D conditions mentioned above have succeeded in inducing SSC generation of mature spermatozoa, although differentiation of male germ cells in vitro from earlier (albeit postspermatagonial) developmental stages has been described [3, 33–36]. It appears that both the initiation of spermatogonial differentiation from stem cells and the entry of differentiating spermatogonia into meiosis are blocked in vitro, despite the fact that SSCs can be maintained for up to several months in culture and their transplantation reinitiates spermatogenesis [29, 37]. The passage of spermatogonia into meiosis (and hence, migration through the seminiferous epithelium) depends upon the structural support of the seminiferous epithelium, interaction with and signaling through the extracellular matrix (ECM), and the availability of specific factors in the testicular microenvironment that in as yet poorly understood ways participate in mediating this event. It is interesting to speculate that these microenvironmental structures form a system of hierarchical niches within the seminiferous tubules that are essential for the exquisite temporal and spatial-regulation of SSC proliferation and differentiation events during spermatogenesis.

Exploitation of the In Vivo Somatic Environment to Experimentally Induce Spermatogenesis

Organ culture

In the past, the study of spermatogenesis in vitro began by establishing organ cultures from neonatal rodents under varying conditions. The main advantage of this approach is that germ cells maintain their spatial arrangement and their normal cellular and microenvironmental composition when grown in vitro. These studies largely failed to demonstrate spermatogenesis beyond the meiotic stages [38–40]. Recently, however, Sato et al. have reported the generation of fertile sperm from ex vivo culture of isolated seminiferous tubules obtained from immature mice using a unique methodology [41].

Germ cell transplantation

This approach was used by Brinster and coworkers to confirm the presence of biologically active SSCs [42]. Injection of these cells to the rete testes of mice treated with the chemotherapeutic agent, Busulfan, demonstrated their migration into the seminiferous tubules. Subsequently, these cells participated in full spermatogenesis, including the generation of fertile sperm. Production of mature sperm by transplanted SSCs illustrates the importance of the testicular microenvironment in this process; conditions that are inevitably absent in 2D cell cultures.


Testicular fragments are injected under the skin of sexually mature, castrated, immune-deficient mice. Generation of fertile sperm from testicular allografts and xenografts was previously demonstrated and is age and species dependent [43, 44].

All the three above-mentioned techniques support the argument that SSCs require a precise cellular composition and exposure to the microenvironment of the seminiferous tubule, including specific hormones and paracrine factors to engage in full spermatogenesis. As stated previously, these elements are typically absent in existing 2D culture systems. It follows that improved culture conditions that provide a microenvironment similar to the in situ organization of the seminiferous tubules should permit generation of fertile sperm from SSCs [20, 36, 45–47].

Development of 3D In Vitro Culture Systems

3D culture systems were originally established for clonogenic assays to explore the complex mechanisms involved in multipotent hematopoietic cell proliferation and differentiation [48]. Adaptation of this approach to the male reproductive system provided unequivocal evidence that germ cells can be routinely developed outside the body to the stage of elongating spermatids [20, 36, 45–47].

Today, 3D culture of testicular germ cells is generally considered to most closely mimic the microenvironment of the in situ seminiferous epithelium. This system profoundly deepened our understanding of the interactions between Sertoli cells and germ cells and between germ cells and the ECM, and the influence of those interactions in the process of spermatogenesis. The essential role of Sertoli cells and ECM on cell survival, differentiation, and remodeling of Sertoli cells and on germ cell development is well-established [20, 49–53] and presupposes 3D relations. Reproducing the third dimension in vitro is achieved by embedding the various (dissociated) cell types of the seminiferous tubule in a collagen gel matrix. In this way, a suitable support is provided for isolated spermatocytes to interact with Sertoli and other structural and hormone-producing elements. Under such conditions, the SSCs can be induced to produce daughter cells that self-renew and others that differentiate into spermatids [46]. This approach also yields enhanced viability, germ cell meiosis, and postmeiotic differentiation into differentiating spermatid-like cells [20]. Demonstration of spermatogenesis in 3D cultures has thus far been reported in a variety of species. For example, testicular cells from neonatal bulls cultured in a 3D scaffold produced by sodium alginate, a polysaccharide that forms a hydrogel when complexes with calcium, have produced (haploid) germ cells [46, 54].

These 3D cultures, which more closely reproduce the local (in situ) environment (i.e., niches) within the seminiferous tubule, must be presumed to provide more efficacious cellular/molecular support related to the more natural architectural arrangement of the cellular elements with respect to one another [47, 55]. Despite the observed advantages, 3D cultures have also failed to generate fertile sperm.

Embryonic stem cells (ESCs) have also been used to generate sperm. One approach to induce differentiation of mouse ESCs features a monolayer culture [56], while another is based on a (3D) embryoid body (EB) differentiation strategy [57, 58]. The generation of mouse vasa homolog-positive germ cells was found to depend on EB formation, and their proliferation was induced within the EBs when cocultured with somatic cells expressing bone morphogenic protein 4. These cells could participate in spermatogenesis when transplanted into reconstituted seminiferous tubules [57]. Geijsen et al. reported a spontaneous generation of male primordial germ cells from in vitro-derived mouse EBs [58]. These cells differentiated in the EBs to haploid cells with the capacity to induce oocytes to form blastocyst-like structures [58]. By establishing SSC lines from ESCs, Nayernia et al. succeeded in inducing proliferation and differentiation of SSCs in vitro in the presence of growth factors and retinoic acid to generate functional haploid male gametes with sperm-like morphology [56].

Recently, we described a novel 3D soft agar culture system (SACS) as well as a methyl cellulose culture system (Fig. 1) that promote mouse testicular germ cell development in vitro. We have hypothesized that these culture systems more closely approximate conditions in vivo, particularly in regard to the microenvironment existing within the seminiferous tubules [26, 36, 47, 59–61]; Figure 2.

Figure 1.

Scheme of the SACS. The SACS was composed of two layers: The solid, lower layer (0.5% [wt/vol] agar) and the soft upper layer (0.37% [wt/vol] agar) in 24-well plates. Testes from immature mice were removed aseptically and decapsulated and then mechanically separated into interstitium and seminiferous tubules. The tubules were enzymatically digested, and the isolated tubular cells were seeded (106 cells per well per 200 μl) into the upper layer of the soft agar medium. Cultures were incubated in 5% CO2 at 37°C [61]. Abbreviations: FCS, fetal calf serum; SACS, soft agar culture system.

Figure 2.

Schematic illustration of the microenvironment of developed tubular cells in the upper layer of SACS. Isolated tubular cells cultured in the upper layer of SACS give rise to colonies of germ cell origin in addition to SC and PTC. Our results [61] indicate, however, that the colonies are not connected to SC and PTC. Thus, the upper layer of SACS provides a three-dimensional (3D) structure for the germ cells (similar to that in the seminiferous tubule); cell-cell interactions (through secretion of paracrine/autocrine factors) between all the cells present and developed in the SACS (SC/PT and cells present in the developed colonies, developed cells in the same colony, and also between cells from different colonies). These cell-cell interactions and the 3D structure occur in and define the specific niches for the germ cell to develop full spermatogenesis, including spermatozoa. Abbreviations: PTC, peritubular cell; SACS, soft agar culture system; SC, sertoli cell.

SACS, first established to characterize clonal expansion of bone marrow cells and identify factors involved in their proliferation and differentiation [62–65]. We used it as two layers of different agar concentrations forming, respectively, a gel and a solid phase (Fig. 1) [61, 64]. This arrangement allows addition of supplemental factors and/or supporter cells (e.g., Sertoli cells) to the solid agar phase without contaminating the gel phase containing the enriched SSCs [36, 47, 61]. In contrast to conventional cell cultures where the dish is coated with gelatin, collagen, matrigel, and so forth, the 3D matrix provided in SACS exists as a thick (several millimeters to several centimeters) layer in which the germ and supporting cells are embedded (Fig. 1).

This approach has allowed us to replicate male germ cell development (proliferation and differentiation to sperm) simply and reliably. Briefly, seminiferous tubules from the testes of 7-day-old mice were enzymatically dissociated, and intratubular cells were embedded in the upper (gel) layer of the SACS in Roswell Park Memorial Institute (RPMI) medium supplemented with fetal calf serum (FCS). The lower layer of the SACS contained only RPMI medium supplemented with FCS. Our results showed development of germ cell “colonies” in the upper layer. Immunofluorescence and qPCR analysis of the developed colonies after 4 weeks in SACS with specific markers [(premeiotic: Vasa, Dazl, OCT-4, C-Kit, GFRα-1, CD9, and α-6-integrin), (meiotic: lactate dehydrogenase, Crem-1, and Boule), and (postmeiotic: Protamine-1, Acrosin, and SP-10)] demonstrated the presence of premeiotic, meiotic, and postmeiotic cells, respectively. Significantly, spermatozoa exhibiting normal morphology, including acrosome, were observed in the fixed cultures [61].

In our previous studies, SACS was also used to examine supporting and limiting effects on enriched SSCs in the gel phase of somatic testicular cells cocultured in the solid phase of the system [47]. We concluded that our approach is highly advantageous for exploration of SSC expansion. Our latest results now indicate that the system supports/promotes differentiation up to the level of postmeiotic spermatozoa. Thus, we have concluded that SACS can be used as a novel in vitro system for the maturation of premeiotic mouse germ cells to postmeiotic stages and also to morphologically normal spermatozoa [36, 61]. Thus far, our system has succeeded in generating a relatively low number of sperms (approximately 16 per well per 106 cells seeded), which makes their identification in SACS somewhat challenging. Indeed, we were able to identify sperms only after fixation on slides. For this reason, we did not assess their motility or their capacity to fertilize an oocyte. However, identified sperms exhibited an acrosome, and the acrosome reaction was induced in the presence of ionophore, suggesting their potential for fertility [61]. Our efforts are now focused on determining the conditions necessary to increase the number of sperms generated and for their live analysis in culture.

The advantage of SACS lays in its provision of a complete microenvironment that result in establishment of niches in a spatial arrangement that closely resembles the in vivo condition (Fig. 2). We hypothesize that it also provides an improved architecture that promotes the cell-cell interactions essential for clonal expansion and differentiation of germ cells. In addition, we believe that SACS will facilitate a comprehensive understanding and practical definition of the optimal temporal and spatial conditions required for maintenance of the self-renewal capacity of SSCs. Finally, this system should prove useful in avoiding the ischemia that hampers long-term organ culture of testis and maintaining the normal organization of germ cells in densely packed clusters—the loss of which in traditional culture paradigms impedes germ cell–germ cell contacts necessary for their efficient differentiation [36].

Previously published studies have demonstrated the importance of a 3D structure for the differentiation of mouse and human testicular cells and the support of in vitro spermatogenesis [20, 46, 47]. In this regard, our 3D system could be used to bypass some of the limitations inherent in the use of testicular biopsies for studies of human spermatogenesis (Fig. 3). Most importantly, it has the potential to be applied in future therapeutic strategies for male infertility, particularly in men with azoospermic syndrome who cannot generate sperm but who have SSCs. Similarly, it could benefit individuals undergoing cancer treatment, in particular, prepubertal patients, since they cannot generate sperm to be cryopreserved before chemotherapy/radiotherapy. A cryopreserved biopsy from prepubertal cancer patients before chemotherapy/radiotherapy treatment may be used, in conjunction with SACS, to restore their fertility. Indeed, recent studies demonstrated in vitro propagation of SSC isolated from normal (adult) human testes and prepubertal cancer patients over several weeks [66, 67]. This was achieved in the presence of different growth factors on laminin-coated dishes. Using the same system, Koruji et al., reported induction of SSC proliferation from cells isolated from azoospermic patients [68]. A similar finding in isolated SSC from azoospermic patients was demonstrated in vitro in the presence of Sertoli cells with or without growth factors [69]. Conversely, Lim et al. reported success in inducing proliferation of SSCs isolated from obstructive and nonobstructive azoospermic patients when cultured in the presence of growth factors alone [70].

Figure 3.

Proposed scheme for application of SACS to promote in vitro spermatogenesis from human testicular biopsies. One of the main limitations of using human testicular cells in vitro is the very limited number of viable cells obtainable in testicular biopsy. Hence, the approach described above involves the isolation of the testicular cells and then maintaining them in SACS (A) or, alternately, to induce their proliferation in two-dimensional culture in the presence of conditioned media and growth factors (A-1) [3, 71]. After their proliferation, the cells will be transferred to SACS (B). After developing colonies in SACS (C), sperm generation will be assessed. If sperm is generated and isolated (D), they can be used in ICSI technology to fertilize oocyte (E). Abbreviations: ICSI, intracytoplasmic sperm injection; PTC, peritubular cell; SACS, soft agar culture system; SC, sertoli cell.


SACS provides a novel, 3D structure-supportive microenvironment with specific niches required for initiation of spermatogenesis and progression to the final stages, ending in morphologically normal spermatozoa. This system also offers various options for manipulations through the addition of factors, cells, or other changes that may facilitate mechanistic studies of spermatogenesis and genetic modification of the male germ line. The fertility of sperm generated in this system will be addressed in future studies. Finally, the novel method described here must be assessed in human germ cells to determine its potential utility in addressing male infertility.


I'am grateful to Prof. Tony plant, the University of Pittsburgh, PA, USA and Prof. Ze'ev Silverman and Dr. Lesli Lobel, Faculty of Health Sciences, Ben-Gurion University for comments on the manuscript and language editing. I thank German-Israel Foundation (GIF), Israel Ministry of Sciences and Technology, Israel Ministry of Health.


The author indicates no potential conflicts of interest.