Small molecule-based reprogramming of cell differentiation


Michael Shen, Andrea Califano


In Progress


A major objective of regenerative medicine is the generation of any desired cell lineage through the generation of induced pluripotent stem cells (iPSC) followed by directed differentiation [1]. At present, such reprogramming methods generally require ectopic expression of Master Regulator (MR) gene cocktails [2-5]. The ability to replace these genetic manipulations with an equivalent combination of small molecules would constitute an important breakthrough for the field and would be highly desirable for the effective translation of reprogramming approaches to clinical applications. MAGNet investigators have developed novel algorithms to infer small molecule that can implement arbitrary genetic programs in the cell. We thus propose collaborating with MAGNet investigators to elucidate small molecule based cell reprogramming on a rational basis. 

We propose that the rational, systematic elucidation of druggable mechanisms presiding over cell reprogramming would greatly benefit from an accurate and comprehensive understanding of the regulatory networks that control these processes. We will thus extend systems biology approaches developed by MAGNet investigators for the comprehensive characterization and analysis of gene regulatory networks (interactomes) involved in cellular reprogramming. Specifically, we propose that interrogation of these models using reprogramming as well as small molecule perturbation signatures will help elucidate both reprogramming MRs as well as the small molecules that can substitute for their activities.

Preliminary data from our labs strongly support these goals. Specifically, we have used MAGNet reverse engineering algorithms to generate interactomes for two pluripotent cell types, mouse embryonic stem cells (ESC) and mouse epiblast stem cells (EpiSC). Importantly, ESC correspond to a naïve or ground state of pluripotency, while EpiSC represent a distinct downstream pluripotent state that is primed for lineage-specific differentiation [6-8]. Using large sets of gene expression profiles that we have generated using combinatorial differentiation treatments of these stem cells in culture, we have produced high-quality unbiased interactomes for ESC and EpiSC. Interrogation of the EpiSC interactome with expression signatures of lineage-specific differentiation has led to the identification of candidate MRs for pluripotency, and functional validation analyses have shown that a majority of inferred MRs are indeed required to maintain EpiSC pluripotency in culture. Finally, we have used computational approaches to predict small molecule compounds that can inhibit EpiSC differentiation in culture, and have performed initial validation studies to demonstrate their inhibitory properties.

Based on these results, our proposal will be articulated along the following two specific aims:

Aim 1: Reprogramming of primed pluripotent cells to ground state pluripotency by small molecule perturbations. We will interrogate our mouse EpiSC and ESC interactomes with signatures representing the ESC -> EpiSC transition to infer specific MRs whose activity is abrogated or activated during reprogramming. We will then analyze small molecule perturbation profiles to identify compounds that can invert the ESC->EpiSC MR-activity pattern, thus inducing de-differentiation of primed state EpiSC back to the ESC ground state. Compounds will be validated using cell culture and in vivo assays for ground state pluripotency.

Aim 2: Reprogramming of fibroblasts to pluripotent stem cells by small molecule perturbations. We will perform analysis of small molecule perturbation profiles to identify compounds and compound combinations that implement the pluripotency MR activity pattern identified in Aim 1. Combinations of ectopic MR expression and small molecule perturbations will be tested using cell culture and in vivo assays, with the ultimate goal of inducing reprogramming using only small molecule combinations. For both aims, we will develop new algorithms to identify synergistic MRs and small molecule combinations that can induce reprogramming. Using these approaches, we will infer small molecules that can mimic the activity of both known and newly discovered reprogramming factors, or that can induce their activity by modulating their upstream regulators.


Our initial studies have identified ten candidate small molecule compounds that were predicted to mimic the activity of key MRs of pluripotency. Of these ten compounds, we found that eight could inhibit the differentiation of EpiSCs, using an assay in which EpiSCs cultures were transferred to media that would induce their differentiation. Since the EpiSC line used contains an Oct4-GFP knock-in allele, we could then monitor the expression of endogenous Oct4 by measuring GFP fluorescence. With this assay, we could show that several predicted compounds could successfully inhibited differentiation. Importantly, the computational prediction identified ascorbic acid (vitamin C) as a candidate pluripotency-inducing compound, which was notable since ascorbic acid was previously known to improve the efficiency of iPSC formation. 

In recent studies, we have continued to improve our assay for monitoring the effects of small molecules upon EpiSC differentiation, in order to improve its sensitivity, and have validated compound combinations that can inhibit EpiSC differentiation. We are now using Oct4 immunostaining in combination with the InCell-2000 scanner to provide quantitative assessments of EpiSC differentiation, which will allow the accurate assessment of potential synergistic combinations of small molecules. With this improved assay, we are have now investigated the effects of several predicted compound combinations with synergistic activity, and have validated their activity as compound combinations . Interestingly, this analysis has not only identified combinations including compounds that were previously predicted (and validated) in the single compound analysis, but has also found combinations of new molecules that were not previously implicated. 

In ongoing work to validate candidate MRs of pluripotency, we have performed preliminary studies that have identified a novel gene that can reprogram EpiSC to ESC following its shRNA mediated silencing. This is particularly notable since no other genes have been previously shown to have such property. Since this gene was identified as a candidate negative regulator of pluripotency, we therefore tested the effects of its knock-down to induce reprogramming to pluripotency. For this purpose, we used lentiviral shRNA-mediated knock-down in EpiSC, and then shifted the cultures to ESC conditions (2i+LIF), which are not permissive to EpiSC growth. We found that the knock-down cells, but not the control cells infected with a Scramble virus, were able to re-express Oct4 and displayed rounded colony morphology in 2i+LIF medium. Passaging of the knock-down cells resulted in cultures that were morphologically indistinguishable from ESC and expressed ESC-specific markers such as Rex1

Based on these observations, we have computationally identified several small molecules that can mimic or antagonize the activity of this novel gene. We are now testing these compounds for their activity in EpiSC reprogramming assay.


  1. Robinton DA, Daley GQ. The promise of induced pluripotent stem cells in research and therapy. Nature. 2012 Jan 18;481(7381):295-305. 
  2. Pereira CF, Lemischka IR, Moore K. Reprogramming cell fates: insights from combinatorial approaches. Ann N Y Acad Sci. 2012 Aug;1266:7-17. 
  3. Adachi K, Scholer HR. Directing reprogramming to pluripotency by transcription factors. Curr Opin Genet Dev. 2012 Oct;22(5):416-22.
  4. Ng HH, Surani MA. The transcriptional and signalling networks of pluripotency. Nat Cell Biol. 2011 May;13(5):490-6.
  5. Young RA. Control of the embryonic stem cell state. Cell. 2011 Mar 18;144(6):940-54.
  6. Nichols J, Smith A. Pluripotency in the embryo and in culture. Cold Spring Harb Perspect Biol. 2012 Aug 1;4(8):a008128.
  7. Silva J, Smith A. Capturing pluripotency. Cell. 2008 Feb 22;132(4):532-6. 
  8. Nichols J, Smith A. Naive and primed pluripotent states. Cell Stem Cell. 2009 Jun 5;4(6):487-92.