Integration of the results from systems-level determination of cis-acting elements into high-resolution maps of posttranscriptional regulatory events.

National Institute Of Arthritis And Musculoskeletal And Skin Diseases

Investigators

Linked publications & trials

Abstract

We recently identified a pair of intensely studied, unrelated RNA-binding protein families, Igf2bp and Lin28, with overlapping binding sites on target mRNAs. These families are attractive candidates for investigation of synergy or competition in posttranscriptional gene regulation. Mammalian LIN28A and B are members of an RBP family conserved in animals, containing two CCHC ZnF domains and one cold-shock domain (CSD). Initially, LIN28 was identified as a heterochronic gene in C. elegans development. In humans, LIN28 genes are also stage-specifically expressed in embryonic tissues and, to a lesser extent, in adult tissues. LIN28 proteins have been shown to regulate various aspects of mammalian development, mainly via direct repression of most members of the let-7 miRNA family with key functions in animal development. Our collaborator Dr. Stefan Muljo (Integrative Immunology Unit, NIAID) described reprogramming of adult hematopoietic stem cells (HSC) from mice into a fetal-like state by overexpression of LIN28B (Yuan, 2012). Interestingly, fetal HSCs exhibit high levels of LIN28 expression, while still maintaining some expression of the only LIN28-independent let-7 family member, suggesting additional roles for LIN28 in gene regulation. We confirmed reported findings that LIN28 proteins interact with thousands of mRNAs. The regulatory effect and molecular mechanism of ectopically expressed LIN28 on its mRNA targets remained elusive, possibly due to the lack of specific cofactors in the cell lines studied. We wanted to use hematopoietic cells as a model system to dissect the let-7 independent molecular mechanisms of this multifunctional RBP family and mapped LIN28A and B targets in a B-cell progenitor line (220-8). We found thousands of binding sites distributed over the CDS and 3UTR of 5,000 mRNAs and, as expected, LIN28 expression in this cell line also resulted in a repression of the miRNA let-7 family with a concomitant robust derepression of the 328 conserved let-7 target mRNAs expressed in 220-8. However, direct mRNA targets of LIN28 accumulated to an even higher level, with the top 1,403 LIN28 targets showing greater than 2-fold accumulation of mRNA levels, indicating that the bulk of the LIN28 effect was direct and independent of let-7 inhibition. Our observations were in sharp contrast to previous studies that found only a few direct regulatory effects on LIN28 targets, indicating that the pro-B-cell line represents a suitable system for the study of miRNA-independent regulation of LIN28 targets. Using unbiased proteomic approaches, we identified a direct interaction of the RBP IGF2BP3 with LIN28. This interaction was RNA independent and mapping of the IGF2BP3 binding sites in 220-8 cells revealed a significant overlap of LIN28 and IGF2BP3 bound sites. Intriguingly, the IGF2BP family of proteins shares LIN28s oncofetal expression pattern, which we have previously shown leads to modest stabilization of their target mRNAs. Most significantly, we found a synergistic stabilization of target mRNAs by binding of IGF2BP3 and LIN28 to overlapping sites, which constituted the first example of two RBPs from different families collaborating to amplify their respective effects on mRNA targets to our knowledge. In the future, together with Dr. Muljos group, we will perform gain-of-function retrogenic experiments to determine whether IGF2BP1-3 might collaborate with LIN28A/B in specifying the fetal HSC fate. If the preliminary results are promising, we will use CRISPR/Cas9 genome editing technology to generate triple mutant IGF2BP1-3 mice. Recently, we identified a second pair of unrelated RBPs that work together to destabilize target mRNAs, DND1 and NANOS3. DND1 and NANOS3 are RNA-binding proteins that are NANOS3 essential for primordial germ cell (PGC) survival. Their co-immunoprecipitation and overlapping loss-of-function phenotypes suggest joint function. Nevertheless, how they co-regulate target RNAs remains unclear. Here, we developed Tandem PAR-CLIP and identified a DND1–NANOS3 ribonucleoprotein (RNP) that specifically recognizes an AUGAAUU heptanucleotide on target mRNAs, termed the NANOS3-dependent DND1 Recognition Element (N3-DRE). mRNAs containing 3′UTR N3-DREs are aberrantly upregulated in DND1- or NANOS3-deficient germ cells and encode key cell-cycle and epigenome regulators, including the prominent target CDK1. Genome editing showed that the N3-DRE is essential for Cdk1 repression in mouse PGCs in vivo. Together with Dr. Traci Hall's group at the NIEHS IRP, we obtained a 1.7-Å crystal structure of a DND1–NANOS3–CDK1-N3-DRE RNA ternary complex, revealing a continuous RNA-binding surface and providing the collective sequence-specificity of the RNP. RNA-binding assays demonstrated that DND1–NANOS3 RNP formation is required for high-affinity binding. Together, these findings define the molecular and functional basis of N3-DRE-mediated mRNA regulation in germ cell development. Moreover, we provide a paradigm of two RBPs with low (DND1) or no (NANOS3) intrinsic sequence-specificity, jointly building a high-information content RNA sequence motif that is different from the sum of their individual preferences. Considering that RBP specificities are typically studied individually, rather than in the context of RNPs, this type of “2-factor authorization” may be an underappreciated mechanism to protect posttranscriptional gene regulatory networks from aberrant expression of an individual RNP component. Now that our group has identified two examples of unrelated RNA-binding proteins collaborating in larger RNPs, we are poised to systematically search for more examples, with the assumption that we observed an underappreciated mechanism of posttranscriptional gene regulation. We will use the binding profile data from all projects in the Hafner lab to identify pairs of RNA-binding proteins with overlapping or mutually exclusive binding profiles for further study of competition and synergy of RNA-binding proteins.