At present the molecular mechanisms involved in adult stem cell pluripotency and self renewal remain largely unknown. Furthermore, they appear to lack a blueprint of transcription factors proposed to be involved with 'stemness', as described for ES cells . It is not unreasonable to assume however, that similar gene patterns may exist for adult stem cells. The focus of this work was to ascertain whether human adult stem cells, specifically c-kit isolated cells from UCB were capable of expressing transcription factors described for ES cell pluripotency. The impetus for the study arose over controversy following the number of recent reports suggesting genes such as Oct4 play a role in regulation of adult stem cell multipotency [4–12] and the proposed pre-disposition of developmentally active genes to retro-transposition events .
From our initial results, it became clear that reliance on RT-PCR technology was highly dependent on the specificity of primer sequences and was an unreliable method of analysing ES cell pluripotency genes in adult stem cells. Amplification of 'Oct4' transcripts in differentiated cell types such as hepatocytes and various cell lines (together with c-kit+ progenitor cell populations), revealed the difficulty in designing primers capable of recognising only genuine Oct4. On examination of several published primer sequences, all aligned with Oct4 pseudogene sequences in addition to genuine Oct4, as determined using the NCBI BLAST alignment tool [10, 11, 21, 23, 24]. This raised the possibility that artefacts generated by Oct4 pseudogenes may have contributed to false identification of Oct4 and our current knowledge of adult stem cell gene expression. Recently, Liedtke et al approached this problem by alignment of Oct4 to all its known pseudogene sequences with design of exact primers incorporating a polymorphism unique to genuine Oct4. They demonstrated that their primers could discriminate between cDNA and genomic DNA. Furthermore, in accordance with our data, Liedtke et al  showed that Oct4 was not expressed in human cord blood mononuclear (MN) cells or peripheral blood MN cells, contradicting other recent reports [5, 7, 28].
In both ES and EC cell types, pseudogene expression appears less of a contaminating factor in such cell populations. This may be attributed to events which occur during early development, whereby genes which are not involved in embryonic patterning are silenced by promoter methylation. Genes such as Oct4 have an 'anti-silencing' promoter which prevents methylation during this stage of embryonic development . Hence, lack of pseudogene expression in these cells may likely be due to such generic gene silencing. Our sequencing results however, suggest that neither freshly isolated nor cultured c-kit+ progenitor cells (nor cell lines HepG2, Hek293T and HMC1) undergo such tight regulation of Oct4 pseudogene expression, whilst they also appear to be incapable of genuine Oct4 transcription. The reason why the frequency of retro-transposition events in ES cell-specific genes far exceeds those of non-ES cells is not known . Interestingly, this may have particular significance with regard to recent findings by Elliman et al  who have detected pseudogene expression of Dppa3/Stella in human adult tissue types including bone marrow, peripheral blood, pancreas, adrenal and thyroid gland. This supports the fact that developmental pseudogene expression in differentiated tissues is more widespread than previously thought and reiterates the need for caution regarding use of such genes as markers of adult stem cell pluripotency.
Our subsequent strategy was to select published primers designed to specifically amplify genuine Oct4 gene expression . Unfortunately, our sequencing results confirmed that in non ES cell samples, these too detected pseudogene expression. On inspection of where these primers aligned on genomic DNA from chromosome 6, the reason why they may have failed to preclude pseudogene amplification became apparent. The anti-sense primer sequence taken from Pickering et al had not incorporated an additional guanine base present in the Oct4 sequence obtained from NCBI, accession number NC_000006. Subsequent 'BLASTing' with the new primer sequence generated alignments to several pseudogene transcripts, thus providing an explanation for our unexpected RT-PCR false positives. (NB: the Oct4 DNA sequence used to identify primer sequences was NCBI version GI:51511722 which replaced version GI:42406225 on 24 august 2004). Therefore, it is possible that the previous DNA sequence had omitted the guanine base identified and may have influenced original primer design. Our sequencing results revealed that only the EC cell positive control expressed the genuine Oct4 transcript with 100% sequence homology to the 647 bp product. All the other cell types examined expressed a selection of the various Oct4 pseudogene mRNAs found on chromosomes 1,10 and 12, all which have >97% homology to genuine Oct4.
Recently, several reports have proposed that pseudogene transcripts may act as regulators of protein synthesis and mRNA stability, particularly during carcinogenesis [31, 18, 19]. Kandouz et al  revealed that a Connexin43 pseudogene [ψCx43] plays a functional role in some tumour cells. The protein Connexin43, involved in intercellular communication is often deregulated in many cancers. Interestingly, in some cancer cells ψCx43 becomes translated into a functional protein, playing a role in growth inhibition. This appears to contradict the common concept that pseudogenes represent a version of a gene mutated to inactivity. If however, the mutation is within a regulatory region which prevents transcription, perhaps under certain cellular conditions the pseudogene may be expressed. This putative role of Oct4 pseudogene transcripts in regulation of the parent gene has been proposed by Suo et al  who identified two pseudogenes transcribed in conjunction with the apparent genuine Oct4 in HepG2, MCF-7 and Hela cell lines. Nevertheless, this is speculative and recent work by Mueller et al , using a variety of techniques and in excess of 30 somatic tumour cell lines, have concluded that functional Oct4 is not expressed. As with our results, they also demonstrated that the only cells to exhibit genuine Oct4 expression were EC cell lines .
Although sequencing data in the present study revealed no genuine Oct4 expression in c-kit+ cells, we wished to investigate whether these cells were capable of utilising an exogenous Oct4 promoter construct. The reasoning was that if levels of pseudogene transcripts were relatively high in comparison to their genuine Oct4 counterparts, the pseudogenes may have been preferentially amplified. Our results corroborated the sequencing data in that the only cells able to express genuine Oct4 were the EC 2102Ep cells, with no GFP observed in c-kit+ HSCs or cell lines HepG2 and Hek203T. In summary our results suggest that Oct4 does not appear to be involved in adult stem cell multipotency, in human UCB progenitor cells. In support of our findings, work by Lengner et al have shown recently the same appears true for a variety of mouse somatic stem cells, using Oct4 gene ablation. They show that Oct4 is dispensable for self-renewal in intestinal epithelium, bone marrow mesenchymal stem cells and HSC, hair follicle, brain and liver in contrast with previous findings utilising mainly RT-PCR analysis.
In conclusion, together with a number of recent reports [27, 32, 34, 35] on the controversy surrounding the role of Oct4 in ASCs, our study has underlined the necessity of utilising more than one approach to identify embryonic genes involved in pluripotency before hypothesising their involvement in ASC multipotency. We were unable to substantiate recent reports that Oct4 is involved in the self-renewal of somatic stem cell populations, reiterating the need for caution in the interpretation of results, especially RNA derived when investigating genes prone to high level pseudogene expression.