Investigating the regulation of EML proteins through their N-terminal domain

The echinoderm microtubule-associated protein-like protein (EMLs) are a family of well conserved but poorly characterized microtubule-associated proteins (MAPs) that are also found in oncogenic fusions. One such fusion is EML4-ALK that is present in ~5% of non-small cell lung cancer patients and results from fusion of the N-terminal domain (NTD) of EML4 to the C-terminal tyrosine kinase domain of ALK. Many EML4-ALK patients show good initial response to ALK inhibitors. However, some patients do not respond and all eventually relapse highlighting the need for better understanding of how this fusion protein promotes cancer. Indeed, therapies that target pathways activated by the EML4 NTD would prove highly valuable.

The human genome encodes six EMLs, named EML1 to EML6. EML1 to EML4 have a similar organization consisting of an NTD comprising a trimerization motif and basic unstructured region, followed by a highly structured C-terminal tandem atypical β-propeller (TAPE) domain. The NTD is responsible for binding microtubules, while the TAPE domain binds soluble α/β-tubulin heterodimers. Interestingly, EML5 and EML6, as well as the founding member of this family, sea urchin echinoderm microtubule-associated protein (EMAP), lack a distinct NTD. Although EMAP is a MAP, it is not yet known whether EML5 and EML6 bind microtubules. From a functional perspective EML proteins control microtubule stability and there is growing evidence that this is regulated in a cell cycle-dependent manner through phosphorylation within the NTD.

In this study, we generated a purified version of the EML4 NTD. Using in vitro kinase assays and mass spectrometry we identified specific phosphorylation sites for the NEK9 and CDK1 mitotic kinases, while in vitro pull-down assays identified the domains of NEK9, and the related kinase NEK7, required for interaction with the EML4-NTD. These results add to the hypothesis that EML proteins are not only downstream substrates of NEK kinases but may also act as upstream regulators of their activity. Finally, we found no evidence for localisation of EML5 or EML6 to microtubules supporting the idea that, at least for human EMLs, the NTD is essential for microtubule association.