It has been almost two years since I left the lab, a move precipitated largely by the increasing unlikelihood of further funding to sustain an attritional research battle, wherein I’d spent the majority of the previous decade chewing my own ass ragged. Letting go was in some ways a relief; but also difficult, in the realisation that I had accumulated (incomplete) data that I failed to publish. This is not easy for a scientist to take. And it still bothers me. I worked on a problem that, judging by the sparseness of directly relevant literature, few others were (productively) progressing. Which can mean all sorts of things: ethical constraints governing access to source tissue; a consequence of culture intractability; others having the sense to discontinue wasting valuable resources. So, although I’ve left research, and no longer have cause to regularly trawl PubMed, I do still find myself occasionally eyeing out for any update on human embryonic germ cells (hEGCs). And last week I came across this recent paper:

Genome-Wide Profiling of Pluripotent Cells Reveals a Unique Molecular Signature of Human Embryonic Germ Cells

Aside from a(nother derivation method) paper from the same lead researchers last year, this is (as far as I can ascertain) the first to be published on the properties of this recalcitrant in vitro cell species since (another method paper in) 2009. Oh, in case you’re not with me and are wont to exclaim ‘So what?!’, allow me to make clear that I’m discussing the culture of an elusive human cell type derived from primary cultured primordial germ cells (PGCs; or gonocytes, depending on your preferred terminology); and further consider the dearth of available information on these cells relative to the copious literature pertaining to (the routinely maintained) human embryonic stem cells (hESCs). And please also note the ‘direction’ in culture plasticity: the derivation of a mooted pluripotent cell type (EGC) from a primary cultured, unipotent germ cell; not the directed differentiation of an established pluripotent cell line (ESC) into cells with germ-like characteristics.

Since the original report of hEGC derivation (in 1998, coincident with the landmark first report on maintained culture of hESCs), a limited number of other research groups have necessarily (though often derivatively) reported reproducing this (no mean) feat, a few of which are cited by Pashai et al (unfortunately including a 2004 report subsequently retracted in 2006). It is good to see the reporting of new data on these cells, and headway being made. Clearly, a sterling group tour de force has yielded a highly informative evaluation of the gene expression profile of cultured hEGCs, and discusses the significance of the similarities and differences in comparison to their antecedent PGCs, and to (other) pluripotent cells (ie those with the potential to differentiate into any cell type): ESCs, induced pluripotent stem cells (iPSCs), and embryonal carcinoma cells (ECCs). I’m not going to discuss in detail here the new data, as I have been out the game a while and, frankly, am no longer ‘qualified.’ But there are still points that perhaps need to be raised, in order to clarify for other groups still (considering) investing venturesome effort and resources in culturing hPGCs – hEGCs.

Analogous to the in vitro ‘reprogramming’ of blastocoelic inner cell mass cells to ESCs, EGCs arise from a poorly understood conversion event (upon which this new report sheds transcriptional light). The authors reiterate the unipotency of PGCs, which in situ are fated to form germ cells (either sperm or oocyte). Away from their natural nurturing environment, however, PGCs are highly vulnerable and typically apoptose rapidly in vitro. Unless, that is, a culture regimen employing specific medium formulation is adhered to, wherein they can overcome this limitation and acquire strong proliferative properties – a criterion that, coupled with their marker profile and in vitro differentiation potential, has been taken to define hEGCs as akin to hESCs, under whose pluripotent umbrella they have consequently been lumped (– including by yours truly; at least some years back before I became sceptical towards that designation).

So, are hEGCs ‘pluripotent’? Their designation as such has relied on the ensuing in vitro detection of (markers indicative of) cells representative of the three germ lineages (yielding some encouraging reports of the regenerative promise of these differentiated derivatives). However, (in lieu of ethically prohibited human-mouse chimeras) the traditional ‘gold standard’ criterion for assigning the label ‘pluripotent’ – in vivo teratomata formation following engraftment into immunocompromised mice – has not been met by hEGCs. It is pleasing (to one having repeatedly attempted these protracted experiments) that the authors remind us of this, with the seeming disregard of the 2009 paper’s outlying report to the contrary. But it seems that the term has become a fluid one: it is not an either/or state; but a multi-valent property, with the elucidation of a unique hEGC molecular signature here assigning a novel version of pluripotency, overlapping – yet distinct from – that of the unipotent PGC at one end of a spectrum, and pluripotent ESCs/iPSCs/ECCs at the other.

Of interesting relevance here is the expression of SOX2, one of a cohort of transcription factors (including most familiarly POU5F1/OCT4 and NANOG) characteristic of pluripotent cell types. A 2008 study (to which I contributed) conducted a detailed comparison of SOX2/SOX2 expression between various hESC and hECC lines, and human germ cells in situ/vivo and in vitro. Neither transcript nor protein was convincingly detectable in hPGCs, and in cultures of proliferating (POU5F1/POU5F1-positive) germ cells. This is, at first glance, seemingly at odds with the new report, which identifies an up-regulation of SOX2 gene expression in cultured hEGCs, relative to hPGCs (thus comprising part of the distinct hEGC molecular signature). The authors postulate transcriptional level regulation as explaining the absence of SOX2 protein – in hPGCs. But its presence or absence in their SOX2-expressing hEGCs has not been confirmed; thus (being pedantic) it cannot be ruled out that regulation might also apply at the post-transcriptional/translational level. Certainly, it is not unreasonable to predict a role for microRNAs somewhere in this molecular program. However, perhaps consideration of SOX2 is not so relevant: mouse PGCs are SOX2-positive, but are nullipotent in vivo, in that they are not (unlike pluripotent mouse EGCs) tumourigenic. And SOX2 protein is detectable in the nucleus of both pluripotent and nullipotent hECC lines.

Primary culture of hPGCs obtained from aborted foetal tissue is complicated by intrinsic and extrinsic heterogeneity with inevitable inconsistency within and between laboratories. Consequently, despite similar methodology, reports include discrepancies, such as hEGC derivation efficiency, and culture duration and characteristics. We might reasonably predict that SOX2 protein was present in these newly reported hEGCs. Of note is their passage number (15-18) versus the early passage SOX2-negative cultures tested in the 2008 study. It would seem that very different cell entities are apparent. Personally, I became very reluctant to label cultures ‘hEGC’, preferring to fall-back on ‘vigorously proliferative’ (‘VP’; clearly no longer PGC, but… ?). Whether ‘pluripotent hEGCs’ can accurately refer to cells continuing to proliferate in vitro beyond the time their cognate PGCs/gonocytes normally cease to divide in situ is uncertain. Although this group recently reported the benefit to cultures of exogenous BMP4, the medium formulation has changed little (since 1998), with the qualitative and quantitative effects of the other added factors (LIF, forskolin and FGF2) on human PGCs/EGCs remaining poorly defined. Indeed, the difficulties in deriving, maintaining, cryopreserving, and preventing differentiation of, these cells have drawn frequent comment – including by the authors of the latest study. Unlike hESCs, hEGCs have not, despite these commendably high passage numbers, demonstrated persistence, let alone ‘immortality’, in culture. Hence, to date there remain (to the best of my knowledge) no banked hEGC ‘lines.’

Clearly, however, this new report defines a distinct cell type that (contrary to some speculation) apparently exists and, moreover, that its culture is not practically impossible (as has also been suggested). Hopefully, the new information here will now enable others to make progress. However, before they can do so they will still need to become adept at deriving and culturing the darn things, and not succumb to the frustration of a priapic dachshund attempting to climb Escher’s stairs.

• Pashai N, Hao H, All A, Gupta S, Chaerkady R, De Los Angeles A, Gearhart JD, & Kerr CL (2012). Genome-wide profiling of pluripotent cells reveals a unique molecular signature of human embryonic germ cells. PloS one, 7 (6) PMID: 22737227