- Email: [email protected]
Transcriptional Intricacies of Stem Cells
Cell Systems
Previews Transcriptional Intricacies of Stem Cells Andrew P. Hutchins1,2,* and Paul Robson3,* 1Key Laboratory of Regenerative Biology, South China Institute for Stem Cell Biology and Regenerative Medicine, Guangzhou Institutes of Biomedicine and Health, Chinese Academy of Sciences, Guangzhou 510530, China 2Guangdong Provincial Key Laboratory of Stem Cell and Regenerative Medicine, South China Institute for Stem Cell Biology and Regenerative Medicine, Guangzhou Institutes of Biomedicine and Health, Chinese Academy of Sciences, Guangzhou 510530, China 3The Jackson Laboratory for Genomic Medicine, Farmington, CT, USA 06030 *Correspondence: [email protected] (A.P.H.), [email protected] (P.R.) http://dx.doi.org/10.1016/j.cels.2015.08.007
Two studies shed light on Oct4 behavior under different conditions to illuminate how cells maintain the pluripotent state and exit from it. The complex interplay of transcription factors is critical to controlling cell fate decisions. In this issue, two studies of stem cell behavior highlight contrasting approaches to understanding cell type stability and fate changes. Both center on the key transcription factor, Oct4. Ding et al. (2015) take a systematic, genome-wide approach to understanding the genomic control of the steady state epiblast stem cells (EpiSCs). They characterize the layers of regulation pluriopotent cells place on Oct4 expression. Meanwhile, Sokolik et al. (2015) suggest that commitment to a neurectoderm cell fate is critically dependent on a two-state switch between the activity of Oct4 and a second transcription factor Brn2. Their work demonstrates how competition between transcription factors executes the switching behavior required to exit a stable cell fate. Overall, both studies illustrate that characterizing the dynamics of transcriptional regulation will be essential to mechanistically understand these and other developmental systems (Figure 1). Genome-wide screens, like that conducted by Ding et al., typically end with the characterization of a few specific knockdowns and the description of a putative protein complex, already a substantial amount of work. However, Ding et al. step things up: not only do they characterize the knockdowns genome-wide, they illustrate pair-wise interactions between genes and move beyond gene expression to define ‘‘protein-level dependency’’ within a phenotypically important subset of genes. Protein-level dependency uses GFP reporters to measure how knockdown of one factor affects the protein levels of all the other members of the analyzed gene set.
Naturally, these later stages of analysis lose the genome-wide power of the shRNA knockdown screen as only a restricted set of genes can feasibly be analyzed, but they provide critical information about how manipulation of each factor affects others, both transcriptionally and post-transcriptionally. Ding et al. characterize the regulatory network underlying the EpiSC cell state, a cell type thought to be related to the late embryonic epiblast and primitive streak stage of gastrulating embryos (Kojima et al., 2014). These observations culminate in the startling observation that, even in unstimulated EpiSCs, Oct4, despite being functionally required (DeVeale et al., 2013), is under significant transcriptional repression. This contrasts with embryonic stem cells (ESCs), which are derived from the earlier inner cell mass of the pre-implantation embryo and see no such wide-scale repression of Oct4. The work by Ding et al. provides a tantalizing hint about a crucial point in development, the exit from pluripotency in the gastrulating embryo, a critical but poorly understood process at the molecular level. Their work suggests that the active suppression of Oct4 is a requirement for epiblast-like cells to leave the gastrulating primitive streak and so commit to a differentiated cell type. This implies that the EpiSCs are a barely stable cell type, precariously balanced on the exit from pluripotency, while the developmentally earlier ESCs are more robust to perturbation. This hypothesis rests on two caveats. First, the exact relationship of the well-characterized ESCs to the cells of the pluripotent inner cell mass remains unclear (Boroviak et al., 2014), much less the closest in vivo cell type for the less
100 Cell Systems 1, August 26, 2015 ª2015 Elsevier Inc.
well-characterized EpiSCs. This is an important consideration because the ability to indefinitely self-renew and maintain pluripotency is not a feature of in vivo epiblast cells which only exist transiently in the embryo. Instead, the default state is to exit pluripotency and rapidly commit to a differentiated cell type. The long term in vitro culture of embryonic-like cells remains a technical artifact, albeit one that is incredibly useful. In a related study also published in this issue, Sokolik et al. explore the mechanisms by which ESCs maintain a stable state when confronted with a differentiation-inducing transcription factors. Specifically, they study the ectopic expression of a light-inducible Brn2 (also called Pou3f2), a transcription factor known to drive mESCs to neurectoderm (Lodato et al., 2013). This elaborate experimental set up allows Sokolik et al. to modulate the level of Brn2 expression across time, so the authors can tease apart the precise thresholds required for exit from pluripotency, as measured by fluorescence of a Nanog-GFP reporter. Their mathematical models indicate that Oct4 and Brn2 act as a ‘‘two-state switch’’ to activate or deactivate Nanog expression, subject to two critical thresholds of Brn2 activity— magnitude and duration. When both thresholds are crossed, cells decrease Nanog levels and exit pluripotency. Mechanistically, Sokolik et al. characterize the switch-like exit from pluripotency as a (relatively) simple competition between Oct4 and Brn2 for binding to Sox2: Sox2’s recruitment into a Sox2Brn2 complex effectively reducing Nanog transcription by depleting the Sox2-Oct4 complex from the Nanog promoter. However, recent work indicates that Brn2, along with Oct6 and Brn1, has additional
Cell Systems
Previews example, Sokolik et al. measure Nanog at the protein level (i.e. not solely a transcriptional read out), thus the complete characterization of Nanog regulation likely also needs to invoke post-transcriptional controls, for instance microRNAs associating with Nanog mRNA (Tay et al., 2008). Independent of these mechanistic considerations, Sokolik et al. provide an important experimental and conceptual model for how cells deal with stochastic inputs to exit (or not) from pluripotency. It will be extremely valuable to extend this experimental technique to other transcription factors and other cell fate commitment systems. REFERENCES Boroviak, T., Loos, R., Bertone, P., Smith, A., and Nichols, J. (2014). Nat. Cell Biol. 16, 516–528.
Figure 1. Decision Making at the Exit of Pluripotency Embryonic stem cells can interconvert to EpiSCs and can terminally differentiate to neurectoderm. Ding et al. describe the role of TOX4 and the Paf1C complex in maintaining OCT4 and describe the curious observation that in EpiSCs OCT4 is under tremendous negative repression, despite being essential for the EpiSC state. Sokolik et al. describes how the interplay of OCT4, SOX2, and BRN2 establish a ‘‘twostate switch’’ to control noise inputs in determining the exit from pluripotency and the entrance to a neurectoderm cell fate.
roles within the cell and may directly regulate transcription by predominantly forming homodimers on palindromic octamer motifs (specifically ‘‘more palindromic octamer recognition elements’’), cis elements associated with neurectoderm lineage genes (Mistri et al., 2015). When taken in this light, the two-state switch described by Sokolik et al. reflects larger biological processes that build on, or augment, Oct4/Brn2 competition for binding to Sox2. This feature is not a generality for all differentiation-inducing transcription factors as the authors show a gradient rather than a two-state switch induced by MyoD. Perhaps Brn2 acts as
a pioneering factor to open up chromatin in regulatory regions of the neurectoderm lineage (Lodato et al., 2013), exposing new cis elements for Sox2 to bind. Such new sites could act as a Sox2 sink reducing the availability of Sox2 to interact with Oct4 and the Nanog promoter. The informative experimental data of Sokolik et al. is a fruitful place to begin dissecting the molecular mechanisms of cell-fate decision making in mESCs. Their proposed computational model is purposefully coarse grained—additional layers of regulation likely contribute to making this cell-fate decision. For
DeVeale, B., Brokhman, I., Mohseni, P., Babak, T., Yoon, C., Lin, A., Onishi, K., Tomilin, A., Pevny, L., Zandstra, P.W., et al. (2013). PLoS Genet. 9, e1003957. Ding, L., Paszkowski-Rogacz, M., Winzi, M., Chakraborty, D., Theis, M., Singh, S., Ciotta, G., Poser, I., Roguev, A., Chu, W.K., et al. (2015). Cell Sys 1, this issue, 141–151. Kojima, Y., Kaufman-Francis, K., Studdert, J.B., Steiner, K.A., Power, M.D., Loebel, D.A., Jones, V., Hor, A., de Alencastro, G., Logan, G.J., et al. (2014). Cell Stem Cell 14, 107–120. Lodato, M.A., Ng, C.W., Wamstad, J.A., Cheng, A.W., Thai, K.K., Fraenkel, E., Jaenisch, R., and Boyer, L.A. (2013). PLoS Genet. 9, e1003288. Mistri, T.M., Devasia, A.G., Chu, L.T., Ng, W.P., Halbritter, F., Colby, D., Martynoga, B., Tomlinson, S.R., Chambers, I., Robson, P., et al. (2015). EMBO Rep. Published online on August 11, 2015. http:// dx.doi.org/10.15252/embr.201540467. Sokolik, C., Liu, Y., Bauer, D., McPherson, J., Broeker, M., Heimberg, G., Qi, L.S., Sivak, D.A., and Thomson, M. (2015). Cell Sys 1, this issue, 117–129. Tay, Y.M., Tam, W.L., Ang, Y.S., Gaughwin, P.M., Yang, H., Wang, W., Liu, R., George, J., Ng, H.H., Perera, R.J., et al. (2008). Stem Cells 26, 17–29.
Cell Systems 1, August 26, 2015 ª2015 Elsevier Inc. 101
Previews Transcriptional Intricacies of Stem Cells Andrew P. Hutchins1,2,* and Paul Robson3,* 1Key Laboratory of Regenerative Biology, South China Institute for Stem Cell Biology and Regenerative Medicine, Guangzhou Institutes of Biomedicine and Health, Chinese Academy of Sciences, Guangzhou 510530, China 2Guangdong Provincial Key Laboratory of Stem Cell and Regenerative Medicine, South China Institute for Stem Cell Biology and Regenerative Medicine, Guangzhou Institutes of Biomedicine and Health, Chinese Academy of Sciences, Guangzhou 510530, China 3The Jackson Laboratory for Genomic Medicine, Farmington, CT, USA 06030 *Correspondence: [email protected] (A.P.H.), [email protected] (P.R.) http://dx.doi.org/10.1016/j.cels.2015.08.007
Two studies shed light on Oct4 behavior under different conditions to illuminate how cells maintain the pluripotent state and exit from it. The complex interplay of transcription factors is critical to controlling cell fate decisions. In this issue, two studies of stem cell behavior highlight contrasting approaches to understanding cell type stability and fate changes. Both center on the key transcription factor, Oct4. Ding et al. (2015) take a systematic, genome-wide approach to understanding the genomic control of the steady state epiblast stem cells (EpiSCs). They characterize the layers of regulation pluriopotent cells place on Oct4 expression. Meanwhile, Sokolik et al. (2015) suggest that commitment to a neurectoderm cell fate is critically dependent on a two-state switch between the activity of Oct4 and a second transcription factor Brn2. Their work demonstrates how competition between transcription factors executes the switching behavior required to exit a stable cell fate. Overall, both studies illustrate that characterizing the dynamics of transcriptional regulation will be essential to mechanistically understand these and other developmental systems (Figure 1). Genome-wide screens, like that conducted by Ding et al., typically end with the characterization of a few specific knockdowns and the description of a putative protein complex, already a substantial amount of work. However, Ding et al. step things up: not only do they characterize the knockdowns genome-wide, they illustrate pair-wise interactions between genes and move beyond gene expression to define ‘‘protein-level dependency’’ within a phenotypically important subset of genes. Protein-level dependency uses GFP reporters to measure how knockdown of one factor affects the protein levels of all the other members of the analyzed gene set.
Naturally, these later stages of analysis lose the genome-wide power of the shRNA knockdown screen as only a restricted set of genes can feasibly be analyzed, but they provide critical information about how manipulation of each factor affects others, both transcriptionally and post-transcriptionally. Ding et al. characterize the regulatory network underlying the EpiSC cell state, a cell type thought to be related to the late embryonic epiblast and primitive streak stage of gastrulating embryos (Kojima et al., 2014). These observations culminate in the startling observation that, even in unstimulated EpiSCs, Oct4, despite being functionally required (DeVeale et al., 2013), is under significant transcriptional repression. This contrasts with embryonic stem cells (ESCs), which are derived from the earlier inner cell mass of the pre-implantation embryo and see no such wide-scale repression of Oct4. The work by Ding et al. provides a tantalizing hint about a crucial point in development, the exit from pluripotency in the gastrulating embryo, a critical but poorly understood process at the molecular level. Their work suggests that the active suppression of Oct4 is a requirement for epiblast-like cells to leave the gastrulating primitive streak and so commit to a differentiated cell type. This implies that the EpiSCs are a barely stable cell type, precariously balanced on the exit from pluripotency, while the developmentally earlier ESCs are more robust to perturbation. This hypothesis rests on two caveats. First, the exact relationship of the well-characterized ESCs to the cells of the pluripotent inner cell mass remains unclear (Boroviak et al., 2014), much less the closest in vivo cell type for the less
100 Cell Systems 1, August 26, 2015 ª2015 Elsevier Inc.
well-characterized EpiSCs. This is an important consideration because the ability to indefinitely self-renew and maintain pluripotency is not a feature of in vivo epiblast cells which only exist transiently in the embryo. Instead, the default state is to exit pluripotency and rapidly commit to a differentiated cell type. The long term in vitro culture of embryonic-like cells remains a technical artifact, albeit one that is incredibly useful. In a related study also published in this issue, Sokolik et al. explore the mechanisms by which ESCs maintain a stable state when confronted with a differentiation-inducing transcription factors. Specifically, they study the ectopic expression of a light-inducible Brn2 (also called Pou3f2), a transcription factor known to drive mESCs to neurectoderm (Lodato et al., 2013). This elaborate experimental set up allows Sokolik et al. to modulate the level of Brn2 expression across time, so the authors can tease apart the precise thresholds required for exit from pluripotency, as measured by fluorescence of a Nanog-GFP reporter. Their mathematical models indicate that Oct4 and Brn2 act as a ‘‘two-state switch’’ to activate or deactivate Nanog expression, subject to two critical thresholds of Brn2 activity— magnitude and duration. When both thresholds are crossed, cells decrease Nanog levels and exit pluripotency. Mechanistically, Sokolik et al. characterize the switch-like exit from pluripotency as a (relatively) simple competition between Oct4 and Brn2 for binding to Sox2: Sox2’s recruitment into a Sox2Brn2 complex effectively reducing Nanog transcription by depleting the Sox2-Oct4 complex from the Nanog promoter. However, recent work indicates that Brn2, along with Oct6 and Brn1, has additional
Cell Systems
Previews example, Sokolik et al. measure Nanog at the protein level (i.e. not solely a transcriptional read out), thus the complete characterization of Nanog regulation likely also needs to invoke post-transcriptional controls, for instance microRNAs associating with Nanog mRNA (Tay et al., 2008). Independent of these mechanistic considerations, Sokolik et al. provide an important experimental and conceptual model for how cells deal with stochastic inputs to exit (or not) from pluripotency. It will be extremely valuable to extend this experimental technique to other transcription factors and other cell fate commitment systems. REFERENCES Boroviak, T., Loos, R., Bertone, P., Smith, A., and Nichols, J. (2014). Nat. Cell Biol. 16, 516–528.
Figure 1. Decision Making at the Exit of Pluripotency Embryonic stem cells can interconvert to EpiSCs and can terminally differentiate to neurectoderm. Ding et al. describe the role of TOX4 and the Paf1C complex in maintaining OCT4 and describe the curious observation that in EpiSCs OCT4 is under tremendous negative repression, despite being essential for the EpiSC state. Sokolik et al. describes how the interplay of OCT4, SOX2, and BRN2 establish a ‘‘twostate switch’’ to control noise inputs in determining the exit from pluripotency and the entrance to a neurectoderm cell fate.
roles within the cell and may directly regulate transcription by predominantly forming homodimers on palindromic octamer motifs (specifically ‘‘more palindromic octamer recognition elements’’), cis elements associated with neurectoderm lineage genes (Mistri et al., 2015). When taken in this light, the two-state switch described by Sokolik et al. reflects larger biological processes that build on, or augment, Oct4/Brn2 competition for binding to Sox2. This feature is not a generality for all differentiation-inducing transcription factors as the authors show a gradient rather than a two-state switch induced by MyoD. Perhaps Brn2 acts as
a pioneering factor to open up chromatin in regulatory regions of the neurectoderm lineage (Lodato et al., 2013), exposing new cis elements for Sox2 to bind. Such new sites could act as a Sox2 sink reducing the availability of Sox2 to interact with Oct4 and the Nanog promoter. The informative experimental data of Sokolik et al. is a fruitful place to begin dissecting the molecular mechanisms of cell-fate decision making in mESCs. Their proposed computational model is purposefully coarse grained—additional layers of regulation likely contribute to making this cell-fate decision. For
DeVeale, B., Brokhman, I., Mohseni, P., Babak, T., Yoon, C., Lin, A., Onishi, K., Tomilin, A., Pevny, L., Zandstra, P.W., et al. (2013). PLoS Genet. 9, e1003957. Ding, L., Paszkowski-Rogacz, M., Winzi, M., Chakraborty, D., Theis, M., Singh, S., Ciotta, G., Poser, I., Roguev, A., Chu, W.K., et al. (2015). Cell Sys 1, this issue, 141–151. Kojima, Y., Kaufman-Francis, K., Studdert, J.B., Steiner, K.A., Power, M.D., Loebel, D.A., Jones, V., Hor, A., de Alencastro, G., Logan, G.J., et al. (2014). Cell Stem Cell 14, 107–120. Lodato, M.A., Ng, C.W., Wamstad, J.A., Cheng, A.W., Thai, K.K., Fraenkel, E., Jaenisch, R., and Boyer, L.A. (2013). PLoS Genet. 9, e1003288. Mistri, T.M., Devasia, A.G., Chu, L.T., Ng, W.P., Halbritter, F., Colby, D., Martynoga, B., Tomlinson, S.R., Chambers, I., Robson, P., et al. (2015). EMBO Rep. Published online on August 11, 2015. http:// dx.doi.org/10.15252/embr.201540467. Sokolik, C., Liu, Y., Bauer, D., McPherson, J., Broeker, M., Heimberg, G., Qi, L.S., Sivak, D.A., and Thomson, M. (2015). Cell Sys 1, this issue, 117–129. Tay, Y.M., Tam, W.L., Ang, Y.S., Gaughwin, P.M., Yang, H., Wang, W., Liu, R., George, J., Ng, H.H., Perera, R.J., et al. (2008). Stem Cells 26, 17–29.
Cell Systems 1, August 26, 2015 ª2015 Elsevier Inc. 101