The “STK38-XPO1 axis”: its general relevance and mechanistic underpinnings remain to be further characterized

To the Editor:

In a recent review in Cellular and Molecular Life Sciences “The STK38-XPO1 axis, a new actor in physiology and cancer” [1], Martin et al. discuss, among other aspects of STK38 function, the newly uncovered role of this kinase as a regulator of the nuclear export receptor XPO1 (also called CRM1). Their discussion, in this specific section of the review, focuses largely on results previously reported by the same group showing that STK38 phosphorylates XPO1 Ser1055 residue, and thus promotes its own nuclear export, as well as the export of Beclin1, YAP1 and Centrin1 [2]. While I do not question that STK38 may play a significant role as a regulator of XPO1, I believe that this section of the review contains some statements that lend themselves to misinterpretation, and that do not accurately reflect what is supported by the experimental evidence currently available.

On one hand, the authors state “In addition to describe a new crucial role for STK38, we identified the first activator of the nuclear export receptor XPO1. Not only the first but also the ultimate one: without STK38 kinase activity, XPO1 remains stuck in an inactive form and cannot fulfill its physiological nuclear export duties like: 1) bringing Beclin1 into the cytoplasm upon starvation for the autophagy activation and 2) delivering YAP out of the nucleus upon cell confluence, limiting the risk of overgrowth.”. This statement seems to imply that phosphorylation by STK38 is a necessary event for XPO1 to act as a nuclear export receptor. However, it remains to be investigated to what extent the effect of STK38-mediated phosphorylation can be generalized to the export of XPO1 cargos other than the four indicated above. In this regard, it must be noted that XPO1 is an essential gene [3], while STK38 is not [4], a finding that seems to be difficult to reconcile with the notion that STK38 is an essential activator of general XPO1 function.

On the other hand, I think that the legend to Fig. 3 can be misleading. There, Martin et al. state “Activated STK38 (phosphorylated at least on Thr444) phosphorylates XPO1 on the Ser1055 embedded in XPO1s auto-inhibitory domain (AID). This phosphorylation induces a conformational change that reveals XPO1s cargo binding site. This results in the binding of the cargo to XPO1׳s nuclear export sequence binding domain (NES BD) and to the binding of RanGTP to XPO1´s Ran-binding domain (RAN BD). This newly quaternary STK38–XPO1–Cargo–RanGTP complex is then exported in the cytoplasm by GTP hydrolysis through the nuclear pore complex (NPC)”. To the best of my knowledge, this is just a hypothetical mechanism, originally proposed by the authors in their primary research report [2], but lacking any experimental support. The fact that Fig. 3 is just illustrating a hypothetical model should have been, in my opinion, clearly stated in the legend (or elsewhere in the review).

XPO1 is emerging as an important therapeutic target in cancer [5, 6]. With nearly 2000 papers on XPO1 or CRM1 available on PubMed (as of December 2020), many newly interested readers rely on reviews to enter this field. It is, therefore, crucial for reviews to provide an accurate and measured description of the function and regulation of this important nuclear export receptor.