Abstract
It has been suggested that testicular germ stem cells represent the only adult body stem cells that dedifferentiate and reprogram into a pluripotent state without any genetic modification. Emerging debate about the authenticity of embryonic stem cell (ES)-like cells derived from adult testicular tissue has prompted us to put forth this letter. We wish to reinforce our findings that pluripotent very small embryonic-like stem cells (VSELs) exist as a small population in adult mammalian testis and may result in ES-like colonies. Because of their small size, it is felt that VSELs could be contaminating the initial cells used for seeding, although efforts were made to place a single germ cell per well in a 96-well plate for clonal expansion, or magnetic activated cell sorting (MACS)-sorted α6 integrin positive cells were used. On a similar note, it is felt that the presence of VSELs in various tissues along with mesenchymal stem cells (MSCs) may provide an alternative explanation to the transdifferentiation potential of MSCs. We conclude that like Oct-4 biology, presence of VSELs in adult body tissues has somewhat surprised stem cell biologists.
A
Ko et al. [10] and Tapia et al. [1] recently questioned the pluripotency of human testis–derived ES-like cells reported by Conrad et al. [6] on the basis that the microarray results extracted by them from Conrad's data and western blot results of Conrad's group do not match. Further, since Oct-4, Nanog, and Sox2 expression in the pluripotent stem cells cultured by Conrad's group were 100-, 10-, and 1,000-fold less, respectively, compared with ES cells; a need was felt to critically reevaluate claims made by Conrad's group. In reply to their comments, Conrad's group [10] argued that the culture procedures were different and the pluripotent stem cells reported by them were closer ES cells than are testicular fibroblasts, and that microarray data across studies cannot be compared.
However, the discrepancy between the microarray and western results could have occurred because Oct-4 exists as two transcripts, Oct-4A and Oct-4B, and the polyclonal antibody used did not differentiate between the 2. Oct-4 has indeed confused stem cell biologists, as previously suggested [11,12]. It is crucial to differentiate between Oct-4A and Oct-4B, since only Oct-4A is expressed in pluripotent cells and Oct-4B, though more abundant, has no role in maintaining pluripotent state of a cell. But what we found intriguing was the fact that various groups working on mouse [3 –6] and human [7 –9] testicular tissue speculated that testicular spermatogonial stem cells (SSCs) dedifferentiate and acquire pluripotent characteristics.
Isolation and culture of pluripotent stem cells from adult mice testicular stem cells was carried out by clonal derivation of GSCs [6,13]. Pluripotency of the stem cells grown in culture was confirmed by studying their differentiation potential into the three germ layers, germ cell contribution and chimera formation, teratoma development, etc. They used testes of Oct-4 green fluorescent protein (GFP) transgenic mice with GFP expression under the control of Oct-4 promoter for the derivation. Briefly, they cultured the testicular GSCs for 7 days in Stempro medium with varying growth factor combinations that resulted in grape-like colonies that were maintained for up to 37 passages. Seven days of culturing resulted in 15 Oct-4-GFP positive GSC colonies from 192 wells plated with single GSCs. These colonies were c-kit negative, indicating their primitive nature, and were further plated on mouse embryonic feeder (MEF) for expansion of pluripotent stem cells. They proved the unipotency and functionality of the cultured GSCs by transplanting them into the seminiferous tubules of germ cell-depleted w/w mutant mice, which resulted in restoration of spermatogenesis without any teratoma formation. One thousand GSCs were plated per well in a 24-well dish on mouse embryonic feeder layer in Dulbecco's modified Eagle medium with 15% fetal bovine serum. Interestingly, after 3–4 weeks, a few cells appeared, grew rapidly, and formed round ES-like colonies. Based on these results, the authors concluded that GSCs converted to pluripotent stem cells (PSCs) and deny the presence of a pluripotent subpopulation which could have given rise to ES-like colonies based on the following two arguments: (1) not even a single ES-like colony appeared when GSCs were cultured, and (2) if a subpopulation of PSCs exists, colonies should form in 2–3 days rather than after 3–4 weeks. Recently the group succeeded to achieve similar success using testicular biopsy rather than whole testis [3], thus highlighting the usefulness of the technology for regenerative medicine in humans.
But the conclusion that there was no subpopulation of pluripotent stem cells in the initial seeding of GSCs could have an alternative explanation. We have recently reported the presence of a distinct subpopulation of pluripotent stem cells [14] and stage-specific localization of c-kit shows in adult human testis, with Adark SSCs being negative for c-kit [15]. Presence of pluripotency network in adult testis was confirmed by using Oct-4 antibodies from 3 different sources, exon-specific primer sets for studying Oct-4A, and by in situ hybridization results. As expected, Oct-4A transcripts were detected in low abundance by quantitative polymerase chain reaction (Q-PCR) analysis. Oct-4A was immunolocalized in the nuclei of VSELs, whereas Oct-4B was in abundance and immunolocalized in the cytoplasm of slightly bigger Adark spermatogonial stem cells with dark stained nuclei, which underwent rapid clonal expansion and also exhibited the presence of cytoplasmic bridges. Based on the results we have proposed a revised model for premeiotic expansion of stem cells during spermatogenesis in humans. We propose that adult human testis has two distinct populations of stem cells: (1) quiescent and pluripotent VSELs with nuclear Oct-4A and (2) rapidly dividing progenitor stem cells with cytoplasmic Oct-4 (ie, Adark SSCs which undergo further differentiation and meiosis to differentiate into haploid sperm). We further discussed that the dark stained nuclei of Adark SSCs may reflect a simple stem cell phenomenon in vivo, wherein the open euchromatin of pluripotent VSELs gets compacted (hence appears dark), undergoes remodeling, gets heavily methylated, and is reprogrammed for differentiation into a particular lineage.
Presence of similar pluripotency network–comprising VSELs with nuclear Oct-4 and slightly bigger progenitor stem cells with cytoplasmic Oct-4B has been recently reported by our group in adult rabbit, sheep, monkey, and human perimenopausal ovary [16] and also in cord blood, bone marrow, and umbilical cord tissue [17] (Table 1). We have observed similar biology (ie, two distinct populations of stem cells in adult mice testicular tissue) by immunolocalization studies and Q-PCR (Fig. 1). Very small cells have nuclear Oct-4 and slightly larger cells have cytoplasmic Oct-4, and as expected, Q-PCR data shows abundance of Oct-4B.
Immunolocalization and relative expression of Oct-4 and Oct-4A transcripts in adult mice testis. Images from confocal microscopy on testicular smears show
HSCs, hematopoietic stem cells; MSCs, mesenchymal stem cells; OGSCs, ovarian germ stem cells; VSELs, very small embryonic-like stem cells.
The presence of two distinct stem cell populations in various tissues is in agreement with the results of Li and Clevers [18], who have also reported presence of a relatively quiescent and actively dividing stem cell populations in hair follicle, gut epithelium, and in bone marrow. Rataczak and his group were the first to report and publish extensively on VSELs [19,20]. VSELs exist in various body tissues of both mice and humans, including bone marrow and cord blood [21]; can be isolated by flow cytometry; and are pluripotent in nature, possibly arising very early during development at epiblast stage [19]. The epiblast-derived primordial germ cells (PGCs) during migration through the embryo proper on their way to the genital ridges are also understood to migrate and settle in various other body organs [22]. They have suggested that the VSELs could be the normal stem cells that get transformed into cancer stem cells resulting in cancer [23].
Various groups have failed to detect the presence of VSELs in normal testicular tissue because mostly germ cell markers have been used in various studies, whereas Oct-4 is a pluripotent stem cell marker, and because of shortcomings in immunolocalization protocols, including choice of fixative, antigen retrieval, et cetera.[14]. Similarly, Oosterhuis et al. [24] have reported that although Oct-4 is an extremely sensitive and specific marker for human malignant germ cell tumors, it may be false negative in biopsies fixed in Bouin's or Stieve's fixative. We propose that the VSELs present in adult testicular tissue of both mice and men may be responsible for ES-like colonies rather than dedifferentiation of GSCs. The VSELs, under yet not well-understood conditions, get transformed and multiply rapidly to give rise to germ cell tumors, as suggested earlier [23]. Similarly, emerging literature on the involvement of Oct-4–positive stem cells in ovarian cancers has been recently reviewed by Virant-Klun et al. [25].
Our reservations with the study reported by Conrad and group: As pointed out by Ko et al. [10] many of the discrepancies in the results are because Conrad's group did not use specific primer sets to amplify Oct-4A responsible for pluripotency and rather amplified both Oct-4A and Oct-4B [7]. Alpha6 integrin–sorted cells were used to initiate the cultures. The pluripotent stem cell profile reported in supplementary Fig 3 has several problems. If the cells are truly pluripotent they are not expected to exhibit DAZL and VASA (germ cell markers). Also, SSEA-4 in the figures is cytoplasmic and cells have fibroblast like morphology, whereas the pluripotent stem cells are expected to be spherical in shape with a high nucleo-cytoplasmic ratio and with a ring of SSEA-4 on the surface. In contrast, the nuclei are elliptical in some figures and round in some, and most importantly, the cells have abundant cytoplasm—which should actually be minimal in pluripotent stem cells. Thus, the colonies reported by Conrad's group are a heterogeneous mix of cells and not a pure population of true pluripotent stem cells. No explanation is offered by the group that no Oct-4 transcripts are detected in normal testicular tissue but isolated SSCs are Oct-4 positive by reverse transcriptase-polymerase chain reaction (RT-PCR). Getting both germ cell and PSC markers by RT-PCR suggests that the sample is actually a mix of PSCs and germ cells. However, presence of CD133 in all 3 samples is a very interesting result shown in their Fig 3a. It is a marker for VSELs raising a possibility that they had contaminating VSELs in the testicular biopsies which remained unnoticed because of their size.
Our reservations with the study reported by Ko and group: The unipotency of transplanted GSCs reported by Ko et al. [6] was expected, because in the testis, even though the VSELs may contaminate the GSCs at time of transplantation, it is not expected that being pluripotent they will differentiate into multiple lineages in the testicular microenvironment. They will most probably exist as a relatively quiescent reservoir of stem cells to maintain homeostasis and as a source to give rise to GSCs. Thus although they mention clonal expansion of cells to derive ES-like cultures, we feel that a high probability existed of VSELs contaminating their initial GSC cultures because of their very small size. Moreover, their reasoning that the pluripotent stem cells should have given rise to at least a few colonies in GS cell culture conditions may not be valid since it is possible that the culture conditions used by them may not be conducive for ES-like cultures. It is only when the cells get plated on mouse feeders in medium for culturing embryonic stem cells that a few cells in a clump start to expand and form ES-like colonies. Their argument that it takes 3 weeks for the colonies to appear and if a subpopulation of pluripotent stem cells is present, they should appear within 2–3 days of culture under ES-like conditions could possibly be because the VSELs are relatively quiescent and are different from embryonic stem cells in their epigenetic status [26]. They do not divide easily like embryonic stem cells, and attempts in our lab over the last year have failed to expand them in culture. It is most likely that these cells take time (about 3 weeks in culture under ES-like culture conditions) to get reprogrammed and then start behaving like ES-like cells in culture. If dedifferentiation and reprogramming of GSCs occurs to give rise to pluripotent stem cells, one could counterargue by observing the inefficiency in the process, as only 1–2 cells in a group give rise to the ES-like colony—in spite of all cells being exposed to same environment.
VSELs may be giving rise to ES-like colonies: The VSELs are very small in size, are not easily detected, and may “hide” or adhere with the bigger cells by a phenomenon termed “emperipolesis,” as reported earlier [19,27]. Thus, although Conrad's group used α6 integrin positive MACS sorted cells to initiate their cultures and Ko's group cultured a single germ stem cell per well, it is very likely that VSELs were contaminating their starting material and gave rise to ES-like colonies in culture. The inefficiency of the process is just because the VSELs are present in very few numbers. We reported similar alkaline phosphatase positive colonies using adult ovarian tissue culture [16], but the process is highly inefficient and the colonies could not be expanded further on feeder support. Furthermore, use of Oct-4 GFP mice with GFP expression under the control of Oct-4 promoter fails to discriminate between cells with nuclear and cytoplasmic Oct-4 (ie, the pluripotent and progenitor stem cells since both will express GFP and may result in more ambiguity).
VSELs, mesenchymal stem cells, and iPS cells: Our results suggest that the presence of a subpopulation of VSELs in various body tissues has somewhat surprised the stem cell biologists just like Oct-4 biology. Presence of VSELs with nuclear Oct-4 in umbilical cord tissue sections along with mesenchymal cells with cytoplasmic Oct-4 [17] can provide an explanation to a large body of existing “controversial” literature that MSCs are pluripotent and have the ability to transdifferentiate. The MSCs are not pluripotent but actually progenitor stem cells with cytoplasmic Oct-4. Transdifferentiation of MSCs (because of their pluripotent nature and give rise to all 3 germ layers) is controversial, and they may not necessarily differentiate, especially into ectoderm and endoderm. However, MSCs are multipotent, committed towards mesodermal lineage, and may give rise to various mesodermal lineages like osteoblast, adipocytes, and chondrocytes. Interestingly, reports are available that MSCs in undifferentiated state also express neural and endodermal transcripts (we believe that this is because of contaminating VSELs). Osonoi et al. [28] reported that human dermal fibroblasts are able to differentiate directly to all 3 germ layer derivatives [ie, neurons (ectodermal), skeletal myocytes (mesodermal) and insulin-producing cells (endodermal)]. They exhibit nestin, desmin, and insulin when exposed to specific cocktail of growth factors. Thus it is felt that achieving transdifferentiation on the basis of immunolocalization or presence of transcripts may not suffice. Rather, functional maturation needs to be demonstrated.
Our basic understanding of the presence of a pluripotent stem cell network in various adult body tissues forces us to conclude that the need to reprogram somatic cells to pluripotent state by induced pluripotent stem cells (iPS) technology requires further justification. The derivation process remains highly inefficient and there was a debate when this technology was first reported and still continues on whether it is really reprogramming of somatic cells or there exists a subpopulation of pluripotent stem/progenitor cells that grow in culture [29]. The somatic cells used to derive iPS cells may not be a good starting material since they accumulate mutations over time and have short telomeres. In contrast, VSELs are relatively healthy, pluripotent stem cells (which exist in adult body tissues) with intact genome and long telomeres (because of their quiescent nature). The 3 recent publications in Nature [30 –32] and commentaries on them [33 –35] do not surprise us at all. Further justification is required to reprogram somatic cells by iPS technology.
Indeed, VSELs may be the almighty stem cells for regenerative medicine, as suggested earlier [36], for the several reasons. Like embryonic stem cells, they are pluripotent with maximum regenerative potential; however, they do not form teratoma, can be isolated from an autologous source, and have no associated ethical issues. They are naturally occurring pluripotent stem cells in adult body tissues with healthy, intact genome, have long telomeres, get mobilized under disease conditions, and should exhibit maximum regenerative potential.
To conclude, whether the pluripotent stem cells during culture of mammalian testicular arise from unipotent GSCs or from a subpopulation of VSELs will get resolved in due time, but more important are the results showing that it is possible to derive and expand autologous pluripotent stem cells from testicular biopsies and that transplantation of cultured GSCs (which we believe are contaminated with VSELs) do not form teratoma. This holds lot of promise as an autologous source of pluripotent stem cells for regenerative medicine in future.
Footnotes
Author Disclosure
The authors declare no competing financial interests.