Preprint highlights, selected by the biological community

Functional diversification of Ser-Arg rich protein kinases to control ubiquitin-dependent neurodevelopmental signalling

Francisco Bustos, Anna Segarra-Fas, Gino Nardocci, Andrew Cassidy, Odetta Antico, Lennart Brandenburg, Thomas Macartney, Rachel Toth, C. James Hastie, Robert Gourlay, Joby Vargese, Renata Soares, Martin Montecino, Greg M. Findlay

Preprint posted on April 02, 2020 https://www.biorxiv.org/content/10.1101/2020.04.02.005041v1

Article now published in Developmental Cell at https://www.cell.com/developmental-cell/fulltext/S1534-5807(20)30757-7

SR protein kinases go beyond splicing.

Selected by Ram

Categories: biochemistry, cell biology, developmental biology, genetics, molecular biology, neuroscience

Context

Serine/arginine-rich proteins (SR proteins), a group of proteins that harbor an arginine/serine (RS) domain either at their N’ or C’ terminus, play crucial roles in a plethora of biological processes such as cell cycle and signaling, developmental pathways, DNA replication and repair, transcription and mRNA splicing. Of note, phosphorylation of SR proteins mediated by Serine/arginine protein kinases (SRPKs) drive their functionality¹.

Three mammalian SRPKs – SRPK1, SRPK2, and SRPK3 – relay information between environmental cues and gene expression by regulating SR protein phosphorylation^1,2. These kinases have similar nucleotide-binding and active sites (fig.a) and share an overlapping substrate and functional landscape³; intriguingly they also carry out distinct biological functions². For example, down-regulation or chemical inhibition of SRPK1 effectively blocks angiogenesis by changing the RNA splicing patterns⁴. Additionally, the C. elegans orthologue of SRPK1 is essential for embryogenesis⁵, and human SRPK1 is essential for spermatogenesis⁶. On the other hand, SRPK2 plays a pivotal role in regulating RNA splicing and DNA damage response in neuronal and cancer cells^7,8. SPRK3 (predominantly expressed in muscle cells) drives muscle development by regulating muscle-specific mRNA splicing⁹. However, the authors of the current preprint set out to investigate the non-splicing functions of SRPKs and their potential role in metazoan developmental pathways.

Key outcomes

The authors identified E3 ubiquitin ligase RNF12 (also known as RLIM) as a potential SRPK substrate using an in silico motif search analysis in the mouse proteome database. The possible reason to choose RNF12 among other candidates could be based on its role in stem cell biology and neural development (the host lab’s previous work¹⁰). In this work, their analysis found four serines (S212/S214/S227/S229) adjacent to arginine residues, reminiscent of a classic RS domain (fig.b). Guided by a range of in vitro, in vivo, and mass spectrometry-based studies, they report that these residues are phosphorylated by SRPKs. Interestingly, their mutational analysis demonstrates processive phosphorylation of RNF12 from C’ to N’ direction, a mechanism characteristic of SRPKs¹¹. Moreover, phosphorylation of all four serines – harbored in the nuclear localization signal – supports RNF12 nuclear retention in mouse embryonic stem cells (mESCs).
They then focused on REX1 (or mouse Zfp42) transcription factor, a canonical downstream substrate of RNF12 that plays a crucial role in X-chromosome inactivation¹². RNF12 mediated ubiquitylation of REX1 leads to REX1 protein degradation. Additionally, REX1 levels increased in ubiquitinylation and phosphorylation impaired RNF12 mutant mESCs. Furthermore, SRPK1 and SRPK2 mediated phosphorylation of RNF12 stimulates its ubiquitylation activity in vitro.
Previous data from the host lab report that functional RNF12 restricts mESC differentiation to neurons¹⁰, a finding reinforced in the current study as revealed by gene expression data (RNA-seq and/or qPCR) collected from RNF12 wildtype, phosphorylation, and ubiquitylation impaired mutants. Moreover, RNF12 and SRPK2 are robustly expressed in the adult mouse brain during in vitro neuronal maturation. Additionally, gene expression data from RNF12-REX1 double knockouts reveal REX1 as a key downstream substrate of the SRPK-RNF12 axis to control the neuro-developmental gene expression program.
Lastly, the authors extended their previously published study of RNF12 mutations that cause Tonne-Kalscheuer Syndrome (TOKAS), a neurodevelopmental disorder, and an X-linked intellectual disability¹⁰. In the current work, they found SRPK1 & 2 deletions and SRPK3 variants in patients suffering from intellectual disabilities. They also demonstrate that the same SRPK3 mutations dampen RNF12 phosphorylation reiterating SRPK-RNF12 signaling anomalies in intellectual disability disorders.

Conclusion and Perspective

The current work elegantly reports a novel role of SRPKs in metazoan developmental pathways revealing a novel SRPK-RNF12-REX1 signaling axis in neurodevelopmental pathways (fig.c). This is particularly exciting as SRPK2 seems to play a meaningful role in neurodevelopmental pathways that could be relevant for future studies to unravel the etiology of complex neurological diseases¹³.

Acknowledgments

I am grateful to all the authors for their support, especially Dr. Greg Findlay and Dr. Francisco Bustos for being open to discuss the work and replying promptly.

References:

Tags: mesc, neurodevelopment, srpks

Posted on: 22nd September 2020 , updated on: 20th October 2020

doi: https://doi.org/10.1242/prelights.24850

Read preprint

(No Ratings Yet)

Issue: *
Your Name: *
Your Email: *

Details: *

Author's response

Francisco Bustos (FB) and Greg Findaly (GF) shared

1.The initial part of the study gauges different CMGC kinases but focuses on SRPKs. What inspired the authors to focus on SRPKs when other CMGC kinases (like CLK2) robustly phosphorylate SR proteins and RNF12 (Fig 1A and 1E)?

FB and GF: Our in vitro data do indeed indicate that several CMGC kinases can efficiently phosphorylate the RNF12 SR motifs. However, once we interrogated each of these potential kinases using potent and selective kinase inhibitors in vivo, it became clear that the major kinase activity required for RNF12 SR motif phosphorylation in ES cells is that of SRPK. The take-home message is that to identify a kinase you have to combine in vitro and in vivo approaches.

2. The authors investigated the role of all three SRPKs in mESCs, but they interchanged them in different experimental settings. It is possible SRPKs have redundant functions and could functionally compensate one another. However, the data cumulatively seem to support more in favor of SRPK2 (like Figs 2G, 2H, 4, and 6). It would be interesting to hear the authors comment on this.

FB and GF: We hypothesize that SRPKs do indeed function redundantly – in fact, our data suggest that SRPK1 and SRPK2 redundantly catalyze RNF12 SR-motif phosphorylation in ES cells. However, it is conceivable that in different tissues, the relative contribution of family members may be different, for example in tissues where SRPK3 is expressed.

3. The authors report gene expression anomalies in mESCs with dysfunctional RNF12 and/or REX1 (Fig 5 and Fig6A). However, considering the proposed model where SRPK drives RNF12 function and thereby REX1, would the authors investigate if genetic perturbation or chemical inhibition of SRPKs leads to similar changes in gene expression?

FB and GF: Yes, this is an important prediction of our model and an interesting idea that we are currently pursuing! We are particularly excited about the possibility of chemically manipulating RNF12 signaling using SRPK inhibitors.

4. The authors show that phosphorylation of RNF12 supports its nuclear retention, but not cytoplasm-nuclear localization. Although ectopically expressed SRPKs are predominantly cytoplasmic, SRPK2 has some nuclear localization (current study). However, it might be informative to look into endogenous SRPK localization⁸. This may be relevant as nuclear SRPKs could locally phosphorylate and regulate RNF12 function that in turn impact REX1 protein levels and drive neuro-developmental gene expression. It will be interesting to hear what the authors think.

FB and GF: SRPKs are thought to regulate nuclear translocation of SR proteins by phosphorylating them in the cytosol, but it is of course possible that nuclear SRPKs may contribute to regulation. As we find that RNF12 phosphorylation drives activity the experiment you suggest would provide great insight into how RNF12 regulates transcription factor substrates and gene transcription.

5. Following the host lab’s past work¹⁰, the current study reinforces the role of RNF12 in neurodevelopmental pathways and proposes a novel SRPK-RNF12-REX1 signaling axis. The authors report the deregulation of these factors in patients suffering from intellectual disabilities. In light of their evidence, how do the authors reconcile their model to mitigate this challenge?

FB and GF: This is an important issue that remains to be addressed. It is clear from the study of intellectual disability mutations that either accelerated or delayed neuronal development has the potential to cause intellectual disability. As you might imagine, it is critical for intellectual functioning that the nervous system develops at the correct time. In the case of the SRPK-RNF12 pathway, REX1 accumulates in engineered cells carrying Tonne Kalscheuer syndrome mutations, which drives aberrant induction of neural genes in pluripotent stem cells. Specially targeted protein degradation approaches as PROTACs or AdPROMs could be employed to induce degradation of REX1 and normalize signaling and gene expression in these cells. However, as REX1 is transcriptionally repressed following exit from pluripotency, identification of additional RNF12 substrates that are deregulated in intellectual disability patients will be key for therapeutic intervention.