Embryonic Stem Cell Transplantation into Seminiferous Tubules: A Model for the Study of Invasive Germ Cell Tumors of the Testis

Abstract

Over the last 15 years, cell transplantation into seminiferous tubules has become a valuable tool to study germinal cell biology and related matters. This is particularly so, because the blood–testis permeability barrier establishes a sealed compartment which protect against certain influences such as immunological rejection. In the light of the functional and genetic similarities between carcinoma in situ (CIS) of the testis and embryonic stem (ES) cells, our laboratory has developed a tumor assay to study cancer invasion processes in testicular germ cell tumors (TGCT) based on the transplantation of ES cells into the seminiferous tubules. Here, we describe this new tumor assay and provide additional information regarding the transplantation techniques used and their application for the study of TGCTs. Finally, we discuss the practical implications of our experimental approach and its potential application for the understanding of TGCT invasive processes and the development of new antineoplastic strategies.

Developmental Pathology of Testicular Germ Cell Tumors

Germ cell tumors represent a heterogeneous group of neoplasias generally found in the gonads, although they can also be found in extragonadal areas. The incidence of these malignancies has increased during recent years and nowadays they represent one of the most frequent tumors among adolescents and young adults (37). The common precursor of testicular germ cell tumors (TGCTs) appears to be the so-called carcinoma in situ (CIS) of the testis, also named intratubular germ cell neoplasia or in situ gonocitoma. It was first described in 1972 when atypical spermatogonia were found in patients who later developed TGCTs (31). CIS cells are probably present, in a latent state, at the moment of birth, but it is not until puberty or later on when they start to proliferate as pluripotent stem cells, invade the stroma of the testis (9), and disseminate throughout the body.

The germinal or embryonic origin of CIS of the testis is still a matter of debate. The fact that germ cell tumors appear mostly in the gonads, together with the finding that transplantation of mouse embryo genital ridges into testis originates tumors (33), point to a germ cell origin. However, an embryonic stem (ES) cell origin is supported by the fact that these tumors can also be generated by the transplantation of early embryos (34) and that the gene expression pattern and the differentiation potential of the tumor cells are similar to those of ES cells (2). The existence of morphologically identical tumors arising from extragonadal locations also suggests their nongerminal origin. However, these extragonadal germ cell tumors may also be derived from primordial germ cells (PGCs), retained during organogenesis in their migration to the gonads. This hypothesis is supported by the fact that these tumors appear in the body midline, which coincides with the path that PGCs follow in their migration (24,26). Interestingly enough, mouse PGCs cultured under specific conditions turn into immortal cell lines named embryonic germ (EG) cells that have the ability, just like ES cells, to contribute to chimeras when injected into host blastocysts (13). Although there are significant differences between human and mouse germinal cells, human PGCs have as well proven to be a source of pluripotent cells (27). The ability of human EG cells to differentiate into derivatives of the three germ layers has been demonstrated; however, the specific developmental potency of these cells still remains unknown. It is likely that the same molecular mechanisms involved in the conversion of PGCs into totipotent EG cells might be regulating the acquisition of the malignant phenotype in protumorigenic PGCs.

TGCTs appear to be genetically regulated because relatives of patients with this tumor have a higher risk of developing this pathology (as much as double in the case of brothers); in addition, specific genes related to the development of this kind of tumor have been identified. For example, a link between modification of the q27 band in the X chromosome and the development of bilateral germ cell tumors has been established (17). Other studies have revealed the importance of epigenetic or environmental factors during embryonic development for the induction of TGCTs. Thus, in vivo experiments or exposure to pollution toxics have revealed that hormonal deregulation, such as increased estrogen levels during pregnancy or toxicity due to external endocrine disruptors, can result in higher TGCT incidence in descendents (4,32).

TGCTs can be classified according to several criteria (18), but histology is the tool most used by pathologists (21). This approach classifies germinal tumors in terms of the presence of one or more morphological patterns. Thus, tumors with only one histological pattern are less frequent and include seminoma, spermatocytic seminoma, yolk sac carcinoma, embryonal carcinoma (EC), polyembrioma, choriocarcinoma, and benign teratoma. Those with more than one histological pattern include combinations of the previous ones, with EC and partial teratoma differentiation (called teratocarcinoma or malignant teratoma) being the most frequent (8).

Teratocarcinoma belongs to the group, which is nowadays called nonseminoma neoplasms, and it is characterized by the presence of embryonic and extra embryonic tissues, together with a population of pluripotent stem cells. These cells, called EC, were the first cancer stem cells to be experimentally isolated (12) and represent the most aggressive cell population of the tumor, although they do exhibit the ability to spontaneously differentiate into teratoma. This property, which is also found in other cancer types, opens the possibility to induce the differentiation of the tumor as an antineoplastic treatment (3). EC cells have also been shown to be responsible for the transplantability of tumors between immunocompatible animals (10). Furthermore, these cells are considered the pathological counterpart of ES cells, because when they are microinjected into blastocysts, they lose their malignancy and participate in the development of normal tissues (5). Thus, in the resulting chimeric mice, the transplanted EC cells give rise to tissues derived from the three germ layers, including germinal cells (20).

As previously mentioned, it is believed that the common precursor of most TGCTs is the CIS of the testis. Several studies have demonstrated genetic and phenotypic similarities between CIS and ES cells (1,2,16,25). On the basis of these observations, we have developed an in vivo model for nonseminoma germ cell tumors based on the transplantation of ES cells into seminiferous tubules (29). In this regard, we believe that the tumorigenic events that take place after transplantation faithfully reproduce the invasive process and the transformation of CIS cells into EC cells in the spontaneous TGCTs.

Cell Transplantation Into seminiferous Tubules

Brinster and Zimmermann devised the original technique of cell transplantation, which involves the microinjection of germinal cell suspensions into single seminiferous tubules (7). Initial experiments using this technique showed the ability of transplanted spermatogonia to reach the basal membrane of the seminiferous tubules, repopulate the testis in azoospermic animals, and differentiate into mature and fertilization-competent spermatozoa (6). Later, it was described that this capacity was a characteristic not only of adult germ cells, but also of transplanted PGCs (11). However, in this particular case, the transplanted seminiferous tubules exhibited intraluminal segments of germinal epithelium, with their basement membranes dividing the tubule into two or more segments. Although the authors did not unequivocally demonstrate that luminal spermatogenesis was derived from the transplanted PGCs, such structures were only observed after transplantation.

Besides germinal cells, somatic cells have also been injected into the seminiferous tubules. For example, Sertoli cells have been transplanted into seminiferous tubules and it was shown they were capable of accomplishing their normal function in the host testis as well (28). Other groups have proposed bone marrow stem (BMS) cell transplantation into the seminiferous tubules as a possible approach to recover fertility in men with testicular failure. Experiments carried out in mice have demonstrated the capacity of the seminiferous microenvironment to drive the transdifferentiation of BMS cells into germ cells (22). Furthermore, experiments using c-kit mutant mice proved that BMS cells are also able to give rise to somatic cell lineages of the testis, namely Sertoli and Leydig cells, when transplanted into the seminiferous tubules or in the testicular interstitial space, respectively (19).

The early approach of Brinster's group to perform the transplantation was to inject cell suspensions into individual seminiferous tubules (Fig. 1a–d). To access them, a small incision in the tunica albuginea had to be made and then some seminiferous tubules were exposed by gently pressing the testis. In order to facilitate the visualization of the individual seminiferous tubules, a small volume of a sterile staining solution (e.g., trypan blue or bromophenol blue) can be dropped onto the incision. Once one of the tubules is cannulated, injection pressure is applied and partial filling of the seminiferous tubules with the cell suspension is achieved. Because seminiferous tubules are interconnected through the rete testis, injection offers some resistance. So, this technique is only partially efficient because the filling of the complete tubular network of the testis is often not possible.

Figure 1.

Three different methods of cell transplantation into the seminiferous tubules. The trypan blue or bromophenol blue-colored cell suspensions can be injected directly into single seminiferous tubules (a–d), through the rete testis (e–h), or through the efferent ducts of the testis (i–l). Note that with the first method the seminiferous tubules network is only partially filled (d). Scale bars: 1 mm.

Later on, two additional technical approaches for cell transplantation into the seminiferous tubules were successfully developed by one of us (23). The first consists in the injection of cell suspensions by piercing directly the rete testis with the micropipette (Fig. 1e–h). In this case, the injection needle is inserted nearly parallel to the testis surface, in the region close to the vascular pediculum of the testis. Although this method easily allows a faster and complete filling of the whole seminiferous network it is possible to rip the rete testis epithelium and extravasate liquid into the testicular stroma, when applying high pressure to the cell suspension. It is important also to note that due to the thinness of the rete testis structure and the hardness of the tunica albuginea, it is easy to accidentally traverse it.

An alternative and even more efficient method is the injection through the efferent ducts of the testis (Fig. 1i–l). Each mammalian testis has between three and five efferent ducts, all of them surrounded by a dense collagen layer. To pierce this coat it is important to use a short and sharp-pointed needle. For this purpose, the testis must be placed with the efferent ducts running parallel to the injection needle, which then is inserted halfway between the rete testis and the epididymis, taking care not to damage any blood vessel. This method allows a faster and more complete filling of the whole seminiferous tubule network without liquid extravasations.

An Experimental Model to Study Germinal Cancer Stem Cell Behavior

TGCTs are very uncommon among nonhuman mammals and thus have been no suitable animal models for their study available until now. In this regard, Stevens (33) described an inbred subline of mice, named 129-terSv, with a high incidence of spontaneous teratomas. However, most of these tumors are fully differentiated (benign teratomas) and, for that reason, are not adequate for the study of human testis teratomas, which are usually malignant. Recently, it has been reported that TGCT-derived cell lines transplanted into the seminiferous tubules of germline-depleted mice are able to populate the seminiferous tubules and progress into tumors (14). This group transplanted two human cell lines, the JKT-1 seminoma cell line and the 833K EC cell line, with both lines giving rise to tumors that expressed TGCT markers.

As previously mentioned, CIS cells share many functional and genetic similarities with ES cells (1). For example, it has been estimated that nearly 50% of the genes upregulated in human ES cells are also expressed in CIS cells, including pluripotentiality genes (NANOG, Oct4, KIT, SFRP1, TlA-2, TFAP2C, etc.) and undifferentiation-related genes (AP-2γ). Brinster and Avarbock (6) have previously reported the inability of ES cells to differentiate into functional germ cells when injected into the seminiferous tubules. On the contrary, these cells gave rise to tumors in both syngenic and non-syngenic mice.

Taken together, these observations led us to establish a TGCT model based in the transplantation of ES cells into the seminiferous tubules. We believe that this transplantation procedure mimics more accurately than others the early stages of TGCT development (29). In order to monitor specifically the fate of the transplanted cells, we generated a stable GFP-transfected ES cell line, coded as AB1^GFP.

At the onset of the microinjection procedure, the injected fluorescent cells could be found in the lumen of the seminiferous tubules (Fig. 2a, b) and 36 h after transplantation, some of the cells were already integrated into the seminiferous epithelium, close to its basal membrane, a localization similar to that of the spontaneous CIS of the testis (Fig. 2c). Five weeks after transplantation, a tumor was formed (Figs. 2d, e). Microscopic analysis of cryostat sections, in which the autofluorescence of the transplanted cells is not lost, showed that most of the structures of the tumor where fluorescent and thus, were derived from the transplanted cells (Fig. 2f). However, the autofluorescence intensity of different cells in the tumors was very heterogeneous. One possible reason for this observation could be the variable oxygen concentration in the different regions of the tumors caused by the chaotic vascular system (30), because a direct relationship between oxygen tension and the fluorescent emission of GFP has been shown (35).

Figure 2.

Different time points of the process of invasion of experimental teratocarcinomas, after seminiferous tubule transplantation of ES cells. (a) Histological section of a mouse testis a few minutes after transplantation of AB1^GFP cells (arrow) into the seminiferous tubules. (b) Immunochemistry with an anti-GFP HRP-labeled antibody of an identical sample shows the injected GFP cells in the lumen of a seminiferous tubule (arrow). (c) Thirty-six hours after transplantation, some ES cells remained in the lumen of the seminiferous tubules, while most were found integrated into the seminiferous epithelium, as revealed with an anti-GFP HRP-labeled antibody. Five weeks after transplantation, regions with proliferative embryonal carcinoma cells could be found (d). In the formed tumors also well-differentiated areas could be found, showing, among others, neural differentiation (rosettes, e). (f) Most of the tumor tissue was found to be fluorescent, indicating its origin from transplanted ES cells. Scale bars: 50 μm.

The experimental tumors formed highly resembled the spontaneous testicular teratocarcinomas at the histological level. Regions with structures derived from the three germ layers, including neural (Fig. 2d), osteoid matrix, muscular, and epidermoid differentiation were observed. In addition, tumors presented regions with an undifferentiated appearance, which would correspond to the EC component of spontaneous teratocarcinomas (Fig. 2e).

In summary, ES cell transplantation into mouse seminiferous tubules represents a model with very valuable potential applications, because it mimics in many ways the preinvasive state of TGCTs. The resulting tumors recapitulate the phenotypic and molecular features of spontaneous human teratocarcinomas. Furthermore, donor cells can be transfected with different transgenes before transplantation, to evaluate the effects of certain genes during the invasive process, such as growth, neo-vascularization and metastasis and tumorigenesis-related factors (15,36). This approach is also potentially useful for the screening of novel therapeutic drugs, including inhibitiors of angiogenesis and metastasis.