Structural models of full-length JAK2 kinase

Abstract

The protein JAK2 is a prototypical member of the Janus kinase family, and mediates signals from numerous cytokine receptors. The constitutively active V617F mutant of JAK2 is prevalent in many bone marrow disorders, blood cancers, and autoimmune diseases, and is an important drug target. Structures have been determined for each of the four individual domains making up JAK2, and for certain pairs of these domains, but no structure of full-length JAK2 is available, and thus the mechanisms underlying JAK2 regulation and the aberrant activity of the V617F mutant have been incompletely understood. Here we propose structural models of full-length JAK2 in both its active and inactive forms. Construction of these models was informed by long-timescale molecular dynamics simulations. Subsequent mutagenesis experiments showed that mutations at the putative interdomain interfaces modulated JAK2 activity. The models provide a structural basis for understanding JAK2 autoinhibition and activation, and suggest that the constitutive activity of the V617F mutant may arise from a dual effect of destabilizing the inactive conformation and stabilizing the active conformation.

Structural insights into the interaction of heme with protein tyrosine kinase JAK2

“…Peptide 1 might not be favored for heme binding, since heme binding to the catalytic loop could interfere with phosphorylation, for example by blocking access to the catalytic center. In contrast, heme binding to peptide 2 might disrupt the JH1-SH2L interaction and thus release the JH1 domain into the putative elongated active state, which would explain the observed hyperphosphorylation (34). Alternatively, one could speculate that the increased heme phosphorylation could stem from an indirect effect.…”

“…Parts of the motif are solvent-accessible, but heme binding might interfere with substrate binding at this site. Peptide 2, on the other hand, is located in a flexible loop on the surface of the JH1 domain, which has been hypothesized to be part of the JH1-Srchomology 2-like domain (SH2L) interface (34). Its surface accessibility would be beneficial for unhindered heme binding, as has been shown for other heme-regulated proteins (35).…”

“…The motifs KRYIH 974 RDLA (peptide 1) and RPDGC 1092 PDEI (peptide 2) are shown in yellow in the context of JH1 (light blue ribbons) and JH2 (dark blue ribbons). Peptide 1 is located in the catalytic loop (red ribbons), while peptide 2 is located in the putative JH1-SH2L interface (34). Visualization was performed using YASARA (39) and a model of JH1-JH2 (32).…”

Structural Insights into the Interaction of Heme with Protein Tyrosine Kinase JAK2**

“…Peptide 1 might not be favored for heme binding, since heme binding to the catalytic loop could interfere with phosphorylation, for example by blocking access to the catalytic center. In contrast, heme binding to peptide 2 might disrupt the JH1‐SH2 L interaction and thus release the JH1 domain into the putative elongated active state, which would explain the observed hyperphosphorylation [24] . Alternatively, one could speculate that the increased heme phosphorylation could stem from an indirect effect.…”

“… [23] Parts of the motif are solvent‐accessible, but heme binding might interfere with substrate binding at this site. Peptide 2, on the other hand, is located in a flexible loop on the surface of the JH1 domain, which has been hypothesized to be part of the JH1‐Src‐homology 2‐like domain (SH2L) interface [24] . Its surface accessibility would be beneficial for unhindered heme binding, as has been shown for other heme‐regulated proteins [25] .…”