007 (x2)-3-3: 14-3-3 targeted compounds as double agents

Abstract

14-3-3 proteins enact a wide range of cellular functions through protein-protein interactions with client proteins. Kaplan and colleagues [1] recently demonstrated that a semi-synthetic compound was able to selectively stabilize or disrupt specific interactions, depending on the binding partner. This finding presents an exciting possibility of designing other 14–3-3 compounds to regulate critical 14-3-3 interactions.

The inability of neurons in the central nervous system (CNS) to repair themselves leads to permanent functional deficits following injury. Despite intensive research on the subject, no therapies currently restore the reparative function to neurons. Stimulating axonal growth could aid in restoring lost interneural connections but finding targets to go after pharmacologically remains a challenge. Recently, Kaplan and colleagues [1] offered one such target in the 14-3-3 proteins.

14-3-3s are a highly conserved family of proteins with high CNS expression that play important roles in functions crucial to neurons, including cell survival, protein folding, and axonal/dendritic growth [2]. Their cellular effects are mediated via interactions with a wide range of client proteins through a conserved binding pocket [3]. Because of their potential interactions with a few hundred proteins, traditional therapeutic approaches have avoided 14-3-3s due to a concern of off-target effects. However, recent work by Kaplan et al. [1] demonstrates that targeting hub proteins like 14-3-3s that regulate multiple cellular pathways may lead to new therapeutic approaches for treating complex neurological disorders.

Previously, Kaplan et al. [4] identified a small molecule 14-3-3 interactor, fusicoccin-A (FC-A), which stabilizes 14-3-3 protein-protein interactions (PPIs) and stimulates axonal growth. FC-A binds within a hydrophobic pocket created by the binding of certain protein interactors to 14-3-3s [5]. FC-A increases neurite growth and axonal regeneration in vitro [4]. FC-A also reduces corticospinal dieback after spinal cord injury and stimulates optic nerve regeneration in vivo [4]. Using mass spectrometry, Kaplan et al. identified FC-A binding partners, including the stress response protein, General control non-derepressible 1 (GCN-1). Stabilization of the GCN-1:14-3-3 complex increased GCN-1 turnover through proteasome degradation, and loss of GCN-1 function increased neurite outgrowth. These findings not only suggested a novel target for axonal regrowth following injury, but also raised the exciting possibility of creating modifications of the FC-A molecule that could target specific 14-3-3 binding partners [4].

In a recent work, Kaplan et al. [1] looked deeper into 14-3-3 PPI modulators. The authors created and screened derivatives of FC-A to identify nine compounds with higher potency than FC-A in a neurite outgrowth assay. The top three hits all contained an amide substitution that allows for an additional hydrogen bond between the compound and 14-3-3s. All three showed improved neurite outgrowth with a much lower EC₅₀ than the parent FC-A molecule. Treatment with one of these compounds, FC-NCPC, significantly induced axon regrowth in an axonal injury model.

The authors next looked at how FC-A and FC-NCPC impacted 14-3-3 PPIs by proteomic analysis of GST-14-3-3ε pulldowns from cortical neuron lysates. This analysis revealed 265 unique interacting partners in total, which covered a wide range of cellular functions. While many binding partners showed no change in the presence of either FC-A or FC-NCPC, several showed either stabilization or inhibition of 14-3-3 binding. Further, as predicted, FC-NCPC showed a higher-fold impact on these interactions than FC-A, indicating that the FC-NCPC was better at stabilization and inhibition than the parent molecule.

KEGG pathwayⁱ analysis pointed to the Rap1 signaling pathway among the pathways that were enriched with FC-A or FC-NCPC treatment. Rap1 is a Ras-related small GTPase involved in cell adhesion and cell junction formation [6]. Several proteins in this pathway were either stabilized or inhibited by FC-NCPC. Using small molecule inhibitors to target different proteins in the Rap1 pathway, the authors found that inhibition of four proteins in this pathway significantly reduced neurite outgrowth. The authors then looked more closely at two proteins within the Rap1 pathway, Pak6 and Rapgef2, by creating peptides flanking putative phosphorylation sites that likely mediated 14-3-3 binding. Per the mass spectrometry analysis, Pak6 showed enhanced binding in the presence of FC-NCPC, while Rapgef2’s binding was inhibited. The crystal structures of the 14-3-3:Pak6 and the 14-3-3:Rapgef2 peptide interactions in the presence and absence of FC-NCPC confirmed expectations: FC-NCPC enhanced the binding of Pak6 by binding one of the residues of the interacting molecule, while it competed out binding of Rapgef2. Therefore, slight structural shifts in the 14-3-3 binding pocket by FC compounds had opposing effects to alternatively stabilize or destabilize specific 14-3-3 PPIs (Figure 1). Such structural shifts could thus help select for activation of those proteins that promote neurite outgrowth over those that may impair outgrowth.

Figure 1. 14-3-3 agents with dual function.

Binding of FC-A compounds into the hydrophobic pocket of 14-3-3 proteins promote a slight structural shift in the 14-3-3 binding pocket that can either disrupt (A) or stabilize (B) binding to other client proteins.

A key question raised by this work includes the isoform specific effects of FC compounds. The 14-3-3 family includes seven, highly homologous members that are expressed differentially in the central and peripheral nervous systems. In their recent work, Kaplan et al. [1] focused on 14-3-3ε in their proteomic analysis. How much the 14-3-3 interactome differs between each isoform is not well understood, and how FC-A and related compounds may impact each 14-3-3 isoform’s PPIs is an important question for future studies.

Kaplan et al.’s work demonstrate that a single small molecule can differentially impact 14-3-3s’ interactions with many different proteins involving different cellular pathways. Surprisingly, by binding to the same hydrophobic pocket of 14-3-3s, FC compounds have dual functions to stabilize or disrupt specific 14-3-3 PPIs (Figure 1). The success of FC-based compounds in treating axonal injury in mammalian models points to a new polypharmacological approach in the management of complex disorders that lack a single genetic cause and involve a broad array of signaling pathways. As 14-3-3 alterations have been reported in numerous neurologic disorders, including Alzheimer’s disease (AD) and Parkinson’s disease (PD) [7], 14-3-3-based dual agents could be beneficial in the treatment of these disorders. Indeed, Pak6 has been shown to regulate 14-3-3s’ interaction with LRRK2, mutations of which are the most common genetic cause for PD [8]. Modifying the molecular structure of the parent compound could allow for increased selectivity in those PPIs that promote the desired phenotype. Examination of the crystal structure of 14-3-3s bound to different binding partners could be important for designing other small molecules that could be used to promote or inhibit specific 14-3-3 interactions. However, caution is still required with neurological disorders that require chronic treatment; alterations of 14-3-3 PPIs could have long-term consequences not evident in short-term laboratory studies. Yet, the success of Kaplan et al.’s studies in targeting 14-3-3 PPIs by FC compounds to promote axonal regeneration opens up a new therapeutic avenue for complex disorders and highlights the importance of future studies to test similar compounds in AD, PD and other disorders in which genetic manipulation of 14-3-3s have shown promise [9, 10].