Induced proximity approaches for targeted protein dephosphorylation and degradation

Abstract

Post-translational modification (PTM) of proteins greatly expands the complexity encoded by the human genome since these chemical modifications enable regulation of virtually every aspect of protein function in cells, including localisation, stability and catalytic activity. Of the >600 reported types of PTMs, which provide precise control of cell signalling, around 90% involve protein phosphorylation, ubiquitination and acetylation. Consequently, dysregulation of these PTM processes has been implicated in a broad range of human pathologies, including cancers as well as neurodegenerative, metabolic and immune disorders. As a result, these PTM processes, particularly phosphorylation and ubiquitination, have provided an area of great interest for therapeutic intervention.

An emerging strategy to tackle dysregulation of PTMs involves the targeted manipulation of protein PTMs through induction of spatial proximity between a PTM-modifying enzyme and a protein of interest (POI). This has been demonstrated as a promising mechanism to alter POI function by various technologies, the majority of which focus on ubiquitination. For example, the affinity-directed protein missile (AdPROM) system and proteolysis-targeting chimera (PROTAC) compounds mediate proximity between a POI and an E3 ligase to induce targeted POI ubiquitination and proteasomal degradation. Such a targeted PTM-modifying approach could afford several advantages over current therapeutics.

Here, I explore the potential of targeting an alternative PTM, specifically phosphorylation. The ability to selectively modulate the phosphorylation status of a POI could overcome some of the current limitations associated with traditional small molecule kinase inhibitors, which can elicit off-target effects by impacting other substrates of the kinase in addition to the POI. Previously, the Sapkota lab developed the AdPROM system, which comprises a POI-binding polypeptide conjugated to an E3 ligase moiety. By exchanging the E3 ligase moiety of the AdPROM system for a phosphatase moiety, we develop the AdPhosphatase (affinity-directed phosphatase) system, which recruits a phosphatase to a POI to mediate targeted POI dephosphorylation. Focussing on GFP-tagged transcription factor SMAD3 (mothers against decapentaplegic homologue 3) as a target for this proof-of-concept work, I employed an AdPhosphatase composed of an anti-GFP nanobody conjugated to PPP2CA, the principle catalytic subunit of PP2A phosphatase, to successfully remove the TGFβ (transforming growth factor-β)-induced tail phosphorylation of GFP-SMAD3. I demonstrated this removal of GFP-SMAD3 phosphorylation to be dependent upon PPP2CA activity and uncovered that this AdPhosphatase-mediated targeted GFP-SMAD3 dephosphorylation resulted in impeded nuclear translocation of GFP-SMAD3.

Having demonstrated with nanobody-directed phosphatase recruitment that targeted dephosphorylation can impact POI function, I explored heterobifunctional small molecule-mediated induction of proximity between a phosphatase and a POI for targeted dephosphorylation. For this, I exploited a combination of protein tags, such as HaloTag, dTAG and bromoTAG, for which ligands have been established and utilised as warheads for PROTACs, and developed small bivalent molecules that would be predicted to induce proximity between two tags. A heterobifunctional molecule, HDPIC (HaloTag-dTAG proximity-inducing chimera), that I developed, as well as the reported compound PhosTAC7, appeared to cause the dephosphorylation of stably expressed Halo-SMAD3 by compound binding alone. Since neither HDPIC nor PhosTAC7 required the recruitment of dTAG-phosphatases to elicit the observed dephosphorylation of Halo-SMAD3, this suggested that that the Halo-SMAD3 binding to chloroalkane was causing the dephosphorylation. In contrast, another molecule that I developed, termed BDPIC (bromoTAG-dTAG proximity-inducing chimera), did not incur a similar dephosphorylation of dTAG-SMAD3 by binding alone. Excitingly, BDPIC facilitated robust targeted dephosphorylation of stably expressed dTAG-SMAD3 through recruitment of bromoTAG-PPM1H. Furthermore, at the endogenous level, BDPIC-mediated induction of proximity between bromoTAG-PPM1H and dTAG-SMAD3 in A549 ^{bromoTAG/bromoTAG}PPM1H/ ^dTAG/dTAGSMAD3 knock-in cells inhibited the transcription of SMAD3-dependent TGFβ-target gene SERPINE-1 (encodes PAI-1 (plasminogen activator-inhibitor 1)). These proof-of-concept studies confirm that bifunctional small molecules can successfully mediate targeted POI dephosphorylation, which can alter POI function and impact downstream cell signalling, potentially providing a novel therapeutic strategy.

A key limitation in the translation of targeted dephosphorylation efforts to therapeutically viable PhosTAC (phosphorylation-targeting chimera) molecules, which engage endogenous, untagged phosphatases and POIs, is the lack of available ligands to bind endogenous, untagged phosphatases. The lack of progress in developing phosphatase ligands can, in part, be attributed to a general lack of an advanced understanding of the physiological substrates of phosphatases. To address this, I harnessed the potential of PROTAC molecules to target dTAG-PPP2CA degradation to enable dissection of the substrate landscape of PP2A, which, together with PP1, is thought to conduct the vast majority of Ser/Thr dephosphorylation in cells. Indeed, targeted degradation of dTAG-PPP2CA in ^dTAG/dTAGPPP2CA knock-in HEK293 cells yielded a significant increase in the abundance of 6,280 phospho-peptides, implicating them as potential substrates of PPP2CA. Of these, some have been reported as PPP2CA substrates previously while most are novel. Bioinformatic analyses revealed involvement of these putative PPP2CA substrates in many cellular processes, including spliceosome function, the cell cycle, RNA transport and ubiquitin-mediated proteolysis. This work provides an in-depth atlas of potential PPP2CA substrates and demonstrates targeted protein degradation as a valuable tool to interrogate phosphatase substrate landscapes in cells.

Together, these findings exemplify the promise an induced proximity modality offers for achieving targeted POI dephosphorylation or degradation, both to interrogate and modulate POI function for research purposes and to explore for potential novel therapeutic applications.

Date of Award	2024
Original language	English
Awarding Institution	University of Dundee
Supervisor	Gopal Sapkota (Supervisor) & Alessio Ciulli (Supervisor)