Naturally occurring signaling molecules that modulate the expression of endogenous ES cell transcription factors are promising candidates for soluble factors that enhance reprogramming. The Wnt signaling pathway contributes to the maintenance of pluripotency in mouse and human ES cells (Sato et al., 2004; Ogawa et al., 2006; Singla et al., 2006; Cai et al., 2007), as well as the self-renewal of undifferentiated adult stem cells in multiple tissues (Reya and Clevers, 2005). In addition, initial studies of iPS cell generation suggested that constitutively active β-catenin, another downstream component of the Wnt pathway, might promote reprogramming of fibroblasts to pluripotency (Takahashi and Yamanaka, 2006). Studies of Tcf3, one of the key transcriptional regulators downstream of the Wnt pathway in embryonic stem cells, have revealed that this factor co-occupies almost all promoter regions occupied by ES cell specific transcription factors, including Oct4 and Nanog, and can regulate the expression of key ES cell transcription factors (Cole et al, 2008; Tam et al., 2008; Yi et al., 2008).

Since the Wnt pathway is intimately connected to the core circuitry of pluripotency, we hypothesized that the stimulation of the pathway using soluble factors could modulate the efficiency of inducing pluripotency in somatic cells. Here we examined the influence of Wnt stimulation on the reprogramming of murine fibroblasts in the absence of c-Myc retrovirus. For this, cells with GFP driven by the endogenous Oct4 promoter (Meissner et al., 2007) were infected with doxycycline (DOX)-inducible lentiviruses encoding Oct4, Sox2 and Klf4. Oct4/Sox2/Klf4 infected cells were either cultured in standard ES cell medium or in Wnt3a-conditioned medium (Wnt3a-CM) and were analyzed for GFP expression by flow cytometry at days 10, 15 and 20 after DOX induction. No GFP positive cells were present with or without Wnt3a-CM treatment on day 10 or day 15. By day 20 a small population of GFP expressing cells was detected only in the cells cultured in Wnt3a-CM (Figure 1a). The Wnt3a-CM exposed cultures formed GFP expressing colonies with morphology typical for ES or iPS cells (Figure 1b). However, unlike four factor transduced cells, which usually form a highly heterogeneous population of cells when propagated without selection, the Oct4/Sox2/Klf4/Wnt3a-CM colonies appeared homogenously ES-like, similar to previously reported Myc^[-] iPS clones (Nakagawa et al., 2008).

Several assays were performed to characterize the developmental potential of Myc^[-] iPS cells derived with Wnt3a-CM treatment. Immunocytochemistry confirmed the expression of markers of pluripotency, including the nuclear factor Nanog and the surface glycoprotein SSEA1 (Figure 1b). Functional assays confirmed that, like ES cells, these iPS cells were pluripotent. When injected into SCID mice subcutaneously, the Myc^[-] iPS cells gave rise to teratomas with histological evidence of cells differentiating into all three germ layers (Figure 1c). More importantly, Myc^[-] iPS cells derived with Wnt3a-CM treatment contributed to the formation of somatic tissues as well as the germline in chimeric mice (Figures 1d and 1e and Supplemental Table S1). These results indicate that Wnt3a-CM treated Myc^[-] clones are pluripotent cells that are morphologically and functionally indistinguishable from ES cells.

To quantify the effects of Wnt3a-CM, triplicate experiments were performed on Oct4/Sox2/Klf4-inducible MEFs carrying a G418 resistance cassette downstream of the Oct4 promoter (Figure 2a). G418 was added to the cultures at 15 days after infection to select for cells that had reactivated the Oct4 locus. When scored on day 28 after infection, only a few Myc^[-] G418 resistant colonies (between 0-3 colonies forming on each ten centimetre plate) were detected in standard ES cell culture conditions. In contrast, ~20 fold more drug resistant colonies formed when G418 selection was initiated on Wnt3a-CM-treated cells, consistent with the conclusion that activation of the Wnt pathway enhances reprogramming. It should be noted that conditioned medium from control fibroblasts (L cells, ATCC) lacking Wnt3a over-expression also caused a moderate increase in the number of G418-resistant colonies relative to standard ES medium, suggesting that normal fibroblasts may secrete factors, perhaps including Wnt3a, that promote reprogramming. So far, we have been unable to recapitulate the effects of Wnt3a-CM on reprogramming with small molecule inhibitors of glycogen synthase kinase-3 (Gsk3) that modulate the Wnt pathway and promote the self-renewal of ES cells (Sato et al., 2004; Ying et al., 2008) (Supplementary Figure S1). While there are many plausible explanations for this negative result, one possibility is that the half-life of these chemicals is too short to provide sustained stimulation of the Wnt-pathway required to enhance reprogramming.

To independently assess the effect of Wnt3a on reprogramming, we cultured cells in the presence of ICG-001 (Teo et al., 2005; McMillan and Kahn, 2005), an inhibitor of the Wnt/ß-catenin pathway. Figure 2a (right columns) shows that 4μM ICG-001 strongly inhibited the effect of Wnt3a-CM on Myc^[-] iPS formation. The effects of Wnt3a-CM and ICG-001 were also examined in MEFs over-expressing all four reprogramming factors, including c-Myc (Figure 2b). High numbers of G418 resistant colonies were observed in both standard ES cell media (66 colonies/32cm²) and Wnt3a-CM in four factor reprogrammed cells, with only a subtle increase in the number of colonies with Wnt3a-CM (80 colonies/32cm²). In contrast to the dramatic effect of ICG-001 on Myc^[-] cells, at the same dose, the compound had only a subtle effect on the number of G418 colonies in c-Myc transduced cells, and a relatively high number of resistant colonies were observed under these conditions (51 colonies/32cm²) (Figure 2b). When the experiment was repeated with higher doses of ICG-001, iPS colony numbers were further reduced, but even at 25μM multiple Oct4/Sox2/Klf4/c-Myc iPS colonies were observed (mean= 8 colonies/32cm²) (Supplemental Figure S2). These data suggest that Wnt signaling and c-Myc could have overlapping, but not completely redundant roles in the induction of pluripotency.

Recent reports have demonstrated that Myc is dispensable for inducing pluripotency in MEFs (Nakagawa et al., 2008; Wernig et al., 2008a), establishing Oct4, Sox2 and Klf4 as a core set of reprogramming factors, although a slightly different combination of factors also functions in human cells (Yu et al., 2007). Epigenetic reprogramming of MEFs with defined factors is a gradual process, where transcription factors progressively re-establish the core circuitry of pluripotency over the course of weeks (Brambrink et al, 2008; Stadtfeld et al., 2008). Myc accelerates this process (Nakagawa et al., 2008; Wernig et al., 2008a), conceivably because it enhances cell proliferation allowing the changes to unfold more rapidly. Additionally, Myc is suspected to have widespread effects on chromatin state (Knoepfler, 2008; Kim et al., 2008) and could facilitate the productive binding of the core reprogramming factors to their appropriate genomic targets (Jaenisch and Young, 2008).

The Wnt signaling pathway has been shown to connect directly to the core transcriptional regulatory circuitry of ES cells, suggesting a mechanism by which this pathway could directly promote the induction of pluripotency in the absence of c-Myc transduction (Figure 2c). In ES cells, Tcf3 occupies and regulates the promoters of Oct4, Sox2 and Nanog (Cole et al., 2008; Tam et al., 2008; Yi et al., 2008). In MEFs, these endogenous pluripotency transcription factors are silenced. During reprogramming, as exogenous Oct4, Sox2 and Klf4 contribute to the reactivation of the endogenous pluripotency factors (Jaenisch and Young, 2008), Wnt signaling could directly potentiate the effect of these transcription factors, as it does in ES cells (Cole et al., 2008).

Additionally, Wnt could serve to activate endogenous c-Myc directly, thereby substituting for exogenous c-Myc (Figure 2c). Indeed, c-Myc is a well-established target of the Wnt pathway in colorectal cancer cells (He et al., 1998). In ES cells, Tcf3 occupies the c-Myc promoter, and Wnt3a positively contributes to expression of the gene (Cole et al., 2008). However, quantitative PCR studies did not reveal Wnt3a-CM-dependent activation of c-Myc during the first 48 hours after induction of Oct4/Sox2/Klf4 in MEFs (Supplementary Figure S3). Although we cannot exclude that Wnt3a enhances c-Myc expression in the rare cells that will eventually become iPS cells, the expression data suggest that the role of Wnt in reprogramming could be independent of c-Myc induction, consistent with recent findings in ES cells (Ying et al., 2008). Wnt signaling and c-Myc could play partially redundant functional roles in iPS cell generation. For example, Wnt-induced effects on cell proliferation could help to accelerate the sequence of events that lead to the generation of Myc^[-] iPS colonies.

A major goal of current research is to identify transient cues that can reprogram somatic cells, eliminating the need for retroviruses. The studies described here establish that Wnt stimulation can be used to enhance the efficiency of reprogramming in combination with nuclear factors, Oct4, Sox2 and Klf4. By enhancing the efficiency of reprogramming in the absence of c-Myc retrovirus, soluble Wnt, or small molecules that modulate the Wnt signaling pathway, will likely prove useful in combination with other transient cues, yet to be identified, that can replace the remaining retroviruses.

01