Developmental disorders caused by haploinsufficiency of transcriptional regulators: a perspective based on cell fate determination

Abstract

Many human birth defects and neurodevelopmental disorders are caused by loss-of-function mutations in a single copy of transcription factor (TF) and chromatin regulator genes. Although this dosage sensitivity has long been known, how and why haploinsufficiency (HI) of transcriptional regulators leads to developmental disorders (DDs) is unclear. Here I propose the hypothesis that such DDs result from defects in cell fate determination that are based on disrupted bistability in the underlying gene regulatory network (GRN). Bistability, a crucial systems biology concept to model binary choices such as cell fate decisions, requires both positive feedback and ultrasensitivity, the latter often achieved through TF cooperativity. The hypothesis explains why dosage sensitivity of transcriptional regulators is an inherent property of fate decisions, and why disruption of either positive feedback or cooperativity in the underlying GRN is sufficient to cause disease. I present empirical and theoretical evidence in support of this hypothesis and discuss several issues for which it increases our understanding of disease, such as incomplete penetrance. The proposed framework provides a mechanistic, systems-level explanation of HI of transcriptional regulators, thus unifying existing theories, and offers new insights into outstanding issues of human disease. This article has an associated Future Leader to Watch interview with the author of the paper.

Keywords: Chromatin regulator; Cooperativity; Developmental disorder; Haploinsufficiency; Positive feedback; Transcription factor.

Box 2-Figure.

(A) Wild-type gene expression leads to the production of a functional protein (only one allele is shown throughout the figure). (B) Structural variation such as gene deletions leads to reduced gene dosage. (C) Nonsense and frameshift mutations produce premature stop codons, which results in nonsense-mediated mRNA decay (NMD), leading to reduced gene expression. (D) Missense mutations, splice-site mutations, and small in-frame indels that inactivate protein domains, rather than exert a dominant negative effect, lead to reduced protein activity. (E) Deletion, disruption, or disconnection of enhancers also leads to reduced gene expression (regulatory loss of function).

Fig. 1.

The significance of positive feedback and cooperativity for cell fate decisions. (A) The three phases of cell fate determination. Initially, master TFs from competing lineages are co-expressed in progenitor cells (coactivation phase). TF levels then change gradually in favor of one of the competing cell fate programs (gradual biasing phase). Finally, mutually exclusive gene expression programs commit each cell's fate (commitment phase). The process of fate determination is driven by master TFs both reinforcing their own expression program (auto-activation) and repressing competing fate programs (cross-antagonism). (B) At normal enhancers with a limited number of TF binding sites, TFs usually bind in a noncooperative manner, resulting in a gradually increasing transcriptional output. Conversely, super-enhancers with a large number of TF binding sites allow for cooperative TF binding, generating a sigmoidal (switch-like) transcriptional response. Part B adapted from Giorgetti et al. (2010).

Fig. 2.

A model of transcriptional regulation of cell fate and its misregulation due to HI. (A) Master TFs maintain their own expression through autoregulatory positive feedback by cooperatively binding to their own super-enhancers. Cooperative binding occurs through multivalent interactions between the IDRs of TFs and cofactors (inset) and is presumably associated with the formation of transcriptional condensates (grey circle). (B) When plotting the total TF concentration against the basal expression rate (which stems from a basal enhancer), biallelic (WT) expression generates a bistable domain, where two mutually exclusive cell fates are established by low versus high TF concentration. (C) HI (here caused by TF gene deletion) disrupts positive feedback and cooperativity, as well as condensate formation. (D) HI shifts and diminishes the bistable domain, thus interfering with cell fate determination.

Fig. 3.

Incomplete penetrance and variable expressivity during C. elegans cell fate determination. (A) Gene regulatory network (GRN) underlying C. elegans intestinal development. In the WT, end-1 is activated redundantly by SKN-1, MED-1/2 and END-3. This redundancy ensures that end-1 expression is sufficiently high to trigger elt-2 auto-activation, and thereby to induce intestinal development. skn-1 mutations remove this redundancy, resulting in highly variable end-1 expression, which frequently fails to trigger elt-2 auto-activation. Adapted from Raj et al. (2010). (B) GRN underlying C. elegans touch receptor neuron (TRN) differentiation. In the WT, UNC-86 induces mec-3 expression, and MEC-3 maintains its own expression in an autoregulatory loop. An additional positive feedback loop involving ALR-1 ensures TRN differentiation by reducing variability of mec-3 expression. alr-1 mutations remove the parallel feedback loop, thereby increasing variability in mec-3 expression. Adapted from Topalidou et al. (2011).