MOD_LOK_YxT_1

Lymphocyte-oriented kinase (LOK, gene name STK10) is a serine/threonine kinase that is a member of the STE20 subfamily (defined first in yeast). Many of the STE20 group have a preference to phosphorylate threonine residues (Miller,2019; Johnson,2023) and LOK has this Thr preference. It is mainly expressed in lymphocytes and other lymphoid organs such as spleen, thymus, and bone marrow. It is a large nearly 1000 residue protein. Outside its ~250 residue catalytic domain it does not show any sequence similarity to other members of the STE20 family (Kuramochi,1997). It has extensive regions of predicted intrinsic disorder and long coiled-coils. It is primarily a peripheral membrane-associated protein but does not insert into the membrane. Like many other kinases LOK undergoes dimerization and autophosphorylation as part of its activation.

ERM proteins (ezrin-radixin-moesin) are important for cell shape and motility in almost all vertebrate cell types by serving as a phospho-regulated linkage between the plasma membrane and the actin cytoskeleton. Other animal lineages mostly have a single ERM gene/protein. They have an inactive closed state, but phosphorylation of a single, C-terminal, ultraconserved threonine residue leads to opening and activation, enabling them to form a bridge between the plasma membrane and actin filaments. LOK (O94804) is a basophilic kinase whose knockout phenotype, localization and optimal substrate sequence preference all imply that it is a specialized kinase responsible for placing this activating phosphorylation onto ERM proteins (Belkina,2009). The LOK and SLK kinases were demonstrated to be required for the phosphorylation and proper functioning of ERM proteins (SLK can substitute LOK functionality to a certain extent due to similar phosphosite preferences (Belkina,2009; Garland,2021) even though it is not associated to the membrane) through studying the phenotypes of knockout cell lines and rescuing the phenotype by adding back LOK or phosphorylated ezrin (Zaman,2021). It is not clear whether LOK has other substrate proteins in the cell besides ERM proteins.

LOK showed characteristic plasma membrane association and co-localization with ERM proteins. When assessing the ability of LOK to phosphorylate a panel of 90 peptides from the human proteome that have been specifically selected as candidate substrates for basophilic kinases due to harbouring 3–5 basic residues near the S/T residue, LOK only phosphorylated ~10% of the peptides, which indicates a very narrow specificity and also distinct from that of other basophilic kinases, such as ROK and PKC (Belkina,2009). Based on a powerful approach utilizing degenerate peptide libraries, besides a general preference for basic residues, LOK turned out to have an unusual, strong preference for Y at P–2 (2 amino acids before the phosphorylation site). LOK also showed a preference for the C-terminal moesin peptide and did not modify other optimal basophilic kinase substrates. Furthermore, LOK showed a largely reduced phosphorylation efficiency when the Y in the -2 position was changed to R in the moesin peptide, while other tested basophilic kinases phosphorylated the mutated moesin peptide with increased catalytic efficiency (Belkina,2009). Interestingly, the phosphorylation site responsible for the activation of ERM proteins has a Y exactly at the -2 position which is also absolutely conserved among animal ERM proteins from worms to human (Belkina,2009). In vivo studies confirmed that overexpression of GFP-tagged LOK kinase domain in Jurkat cells led to increased ERM phosphorylation and impairment of cell migration.

The LOK and SLK mammalian kinases have a single homolog in D. melanogaster called SLIK (Q8MLP1). The SLIK kinase was demonstrated to phosphorylate the same conserved C-terminal threonine site within D. melanogaster moesin (P46150), that is the only ERM protein in D. melanogaster (Hipfner,2004; Hughes,2006; Kunda,2008; Carreno,2008). The residues around the C-terminal phosphorylation site of D. melanogaster moesin are fully conserved compared to the human homologues, therefore fly SLIK and mammalian LOK show conserved specificities. Furthermore, SLIK was demonstrated to phosphorylate Merlin in D. melanogaster, a more distant homolog of ERM proteins (Hughes,2006). Interestingly, in D. melanogaster and other insects Merlin has an optimal C-terminal phosphorylation site with Y in the -2 position, however, in vertebrates, although the phosphorylated threonine is conserved, the Y residue has been changed to serine and the preceding positive charge was also lost. Accordingly, merlin is not phosphorylated by LOK in vertebrates, but by its homologue, SLK.

Microvilli are specifically localized to the apical surface of epithelial cells and ezrin is a key protein in the assembly of microvilli by serving as a regulated linkage between the apical membrane and F-actin bundles. In a very elegant study, Pelaseyed and colleagues showed that the phosphorylation of T657 of human ezrin that leads to its activation is mediated by LOK in a multistep mechanism wherein the two proteins establish an intricate relationship depending on several different interaction surfaces (Pelaseyed,2017). First, the FERM domain of ezrin binds to PIP2 lipids in the plasma membrane that induces a conformational change whereby the C-terminal, actin-binding domain of ezrin is released from FERM domain binding. This opening up of ezrin enables the insertion of the LOK C-terminal domain between the membrane and F-actin-binding domains of ezrin in a wedge-like arrangement. Subsequently, the N-terminal kinase domain of LOK can dock to a site within the ezrin CTD that is approximately 40 residues distal from the target phosphorylation site, which allows direct phosphorylation of the appropriate threonine residue, T657 (Pelaseyed,2017).
These findings suggest that LOK phosphorylation also requires a specific docking site within the target proteins, however, the sequence and structural requirements of this docking mechanism are yet to be elucidated.