Autophagy of ER-Retained Mpl Is Linked to Low Expression Levels of Jak2 and Provides an Unconventional Route to Cell Surface.

Mpl, the receptor for thrombopoietin (Tpo), activates Jak2. Mpl and Jak2 are expressed both on platelets and on hematopoietic progenitors, where they initiate critical cell survival and proliferation signals in response to Tpo. Thus, interfering with the “Mpl/Jak2 pair” is likely to alter myelopoiesis. Indeed, Bcr-Abl-negative myeloproliferative neoplasms (MPNs) are characterized by the presence of an activating mutation of the JAK2 gene (JAK2V617F) and platelets with low or absent expression of Mpl. However, aberrant Mpl expression (absence of the mature form, accumulation of the ER core-glycosylated form) in MPNs has been linked to low expression levels of wild type JAK2 (JAK2WT).

To determine how the presence of JAK2WT or JAK2V617F impacts Mpl trafficking to the cell surface in MPNs.

Transient transfections of HEL cells (JAK2 V617F^+/+) and K562 cells (JAK2 WT^+/+ and expressing the Bcr-Abl fusion protein characteristic of chronic myelogenous leukemia) were used to study the effects of Jak2WT and Jak2V617F on Mpl. We also studied the cellular localization of Jak2WT, Jak2V617F and Mpl, detected by immunofluorescence or expressed as chimeric proteins fused to Visible Fluorescent Proteins (VFP, derivatives of GFP). The sub-cellular compartment containing the core-glycosylated (immature) form of Mpl was identified by confocal co-localization studies with specific intracellular markers as well as by using a novel genetically encoded fluorescent protein (MiniSOG) recently developed for correlated light and electron microscopic analyses. Finally, Mpl trafficking was evaluated in HEL and K562 cells using a cell surface biotinylation assay.

In HEL cells, re-introduction of Jak2WT, alone or with MPL, favors an increase in the ratio of mature:immature Mpl. Transfected Mpl^mOrange2, in K562 cells, shows a significant co-localization with transfected and endogenous JAK2 WT, but not with transfected JAK2 V617F^-mCitrine. In addition, in HEL cells, transfected Mpl^mOrange2 does not show a significant co-localization with endogenous JAK2V617F. Both endogenous Mpl and transfected Mpl^mOrange2 were detected in punctate structures that stain positively with ER-Tracker and ER exit site (ERES) markers. These structures are negative for other common ER markers, as well as Golgi and endosomal proteins. Ultrastructural electron microscopy studies based on expression of Mpl^mKO2-MiniSOG revealed that Mpl accumulates in multivesicular bodies that bud from the ER. Consistent with an autophagic process, Mpl-containing structures stain positive for LC3. Thrombopoietin stimulus results in rapid delivery of ER core-glycosylated Mpl to the surface, likely through the “unconventional” autolysosome-mediated pathway.

Co-localization of Mpl and Jak2V617F is poor, whereas Mpl and Jak2WT are strongly co-localized. These results support the role of wild type Jak2 in the stabilization at the surface of Mpl and are indicating a role for Jak2WT as a chaperone for the exocytic or recycling pathways. Furthermore, our data obtained in transfected hematopoietic cells strongly suggest that aberrant Mpl trafficking occurs via a secretory autolysosome pathway. This provides a plausible route for limited amounts of ER-core glycosylated Mpl receptor to reach the cell surface. The emerging field of autophagy-based unconventional secretion, associated with such diverse processes as inflammation and tissue remodeling, may now be associated with both the causation and potential treatment of myeloproliferative diseases.

No relevant conflicts of interest to declare.