Synergistic activation of RARβ and RARγ nuclear receptors restores cell specialization during stem cell differentiation by hijacking RARα-controlled programs

Abstract

How cells respond to different external cues to develop along defined cell lineages to form complex tissues is a major question in systems biology. Here, we investigated the potential of retinoic acid receptor (RAR)–selective synthetic agonists to activate the gene regulatory programs driving cell specialization during nervous tissue formation from embryonic carcinoma (P19) and mouse embryonic (E14) stem cells. Specifically, we found that the synergistic activation of the RARβ and RARγ by selective ligands (BMS641 or BMS961) induces cell maturation to specialized neuronal subtypes, and to astrocytes and oligodendrocyte precursors. Using RAR isotype knockout lines exposed to RAR-specific agonists, interrogated by global transcriptome landscaping and in silico modeling of transcription regulatory signal propagation, revealed major RARα-driven gene programs essential for optimal neuronal cell specialization and hijacked by the synergistic activation of the RARβ and RARγ receptors. Overall, this study provides a systems biology view of the gene programs accounting for the previously observed redundancy between RARs, paving the way toward their potential use for directing cell specialization during nervous tissue formation.

Retinoids: Mechanisms of Action in Neuronal Cell Fate Acquisition

“…In contrast, RARβ and RARγ coactivation in either RARβ (−/−) or RARγ (−/−) cells presented a poor yield of upregulated genes associated with the studied neuronal subtypes or glial cells (Figure 2B), in agreement with the low cell differentiation performance. Finally, it is worth mentioning that these findings performed in P19 embryonal carcinoma cells were also confirmed in mouse embryonic stem cells, confirming the potency of the synergistic action of both RARβ and RARγ agonists to lead to neuronal cell specialization (Figure 2C) [25].…”

“…More recently, we found that, while RARβ or RARγ agonists are not able to induce neurogenesis, their synergistic action surprisingly allows them to do so, demonstrating that there exist redundant pathways driven by each RAR subtype, but also their distinctive influence [25]. In this study, in addition to the use of synthetic RAR agonists, we generated P19 CRISPR-Cas9 RAR knock-out lines, providing means to evaluate the redundancy between RARs (Figure 2A).…”

“…Interestingly, the capacity of each of the synthetic RAR agonists to regulate a specific transcriptional response has been described early on by Pierre Chambon's team [20][21][22], and more recently, we have disentangled such a specific response at the level of the various controlled gene programs [23][24][25].…”

“…While both studies provide evidence for the role of individual receptors distinctly influencing terminal cell fate [25,74], there are several key differences which might explain the discordant findings. For instance, different RARγ selective agonists were used in these studies (BMS961 versus CD666).…”

“…Specifically, what was relevant was the ability of RARβ and RARγ coactivation to lead to a rescued neuronal phenotype at the end of 10 days of cell differentiation in P19 embryonic carcinoma cells, while the individual activation of these receptors with their synthetic agonists (BMS641 and BMS961, respectively) did not do so. Taking a closer look at the global transcriptional profiles, within reconstructed gene regulatory networks, revealed that 45-60% of upregulated genes associated with markers for neurons, astrocytes, and oligodendrocyte precursor cells were recovered from the RARβ and RARγ coactivation in wildtype cells, but more than 70% recovery was seen with the same markers from RARβ and RARγ coactivation in RARα (−/−) cells [25]. In contrast, RARβ and RARγ coactivation in either RARβ (−/−) or RARγ (−/−) cells presented a poor yield of upregulated genes associated with the studied neuronal subtypes or glial cells (Figure 2B), in agreement with the low cell differentiation performance.…”