EHA Library - The official digital education library of European Hematology Association (EHA)

Abstract: S883

Type: Oral Presentation

Presentation during EHA23: On Saturday, June 16, 2018 from 16:30 - 16:45

Location: Room A8

Background
Recurrent somatic mutations in exon 9 of the calreticulin (CALR) gene are found in a majority of JAK2-unmutated myeloproliferative neoplasms (MPNs). All known mutations result in a +1-frameshift in the CALR reading frame that creates a novel C-terminal domain encoded by an alternative reading frame. We and others have previously demonstrated that mutant CALR binds to and activates the thrombopoietin receptor MPL and confers cytokine-independent growth. Moreover, we showed that these oncogenic activities of mutant CALR are dependent on a threshold of basic amino acid residues in the mutant C-terminus. However, relatively less is known regarding the role of the N-domain of mutant CALR in oncogenic transformation. 

Aims
We aim to identify key features of the N-domain of mutant CALR which are required for its oncogenic activity.

Methods
Mutations in the FLAG-tagged human CALR cDNA harbouring a “type I” del52 mutation (CALR^MUT) were introduced by site-directed mutagenesis. Transforming activity of CALR^MUT variants was assessed based on their ability to confer cytokine-independence to Ba/F3 cell line expressing human MPL (Ba/F3-MPL). Interaction between CALR^MUT variants and MPL was assessed by FLAG-pulldown experiments. 

Results
We previously demonstrated that the lectin binding motif of CALR^MUT was essential for oncogenic activity. The crystal structure of the CALR globular domain reveals that the carbohydrate binding pocket is proximal to a patch of histidine (His) residues which are required to coordinate Zn²⁺ and mediate regional domain structure of wild-type CALR. Here, we undertook a systematic mutagenesis screen to test the role of 3 His residues (His⁸², His¹²⁸ and His¹⁵³) and their effects on Zn²⁺ homeostasis in CALR^MUT oncogenic activity. We observed that CALR^MUT harbouring of any single His residue retained the ability to confer cytokine-independence, bind to MPL and activate Stat3/5. In contrast, CALR^MUT harbouring loss of His⁸² in combination with either His¹²⁸ or His¹⁵³ (but not His²⁵ which resides on the opposite face of the CALR N-domain) were compromised in their capacity to confer cytokine independence and bind MPL, suggesting that His⁸² needs to cooperate with other nearby Zn²⁺-coordinating His residues on the carbohydrate binding face of CALR^MUT to facilitate oncogenic activity. The stability of all CALR^MUT His variants exhibited comparable stability and folding in vitro, suggesting these changes in CALR^MUT activity likely reflect a disruption of a critical domain within CALR rather than affecting protein stability. Finally, given the role of the His residues in wild-type CALR in regulating Zn²⁺ binding, we tested that CALR^MUT activity may be affected by intracellular Zn²⁺ levels. We used CRISPRi to repress expression of metallothionein 1a (MT1a) which is required for Zn²⁺ uptake, and observed that MT1a depletion rendered CALR mutant cells more sensitive to the JAK inhibitor ruxolitinib. These data suggest that Zn²⁺ homeostasis is able to regulate MPN pathogenesis. 

Conclusion
We show that Zn²⁺ binding by histidine residues are essential for CALR^MUT to bind to MPL and engender transformation, and that disrupting intracellular Zn²⁺ levels has the potential to sensitise CALR^MUT-expressing cells to existing MPN therapies such as ruxolitinib.

Session topic: 15. Myeloproliferative neoplasms – Biology & Translational Research

Keyword(s): mutation analysis, Myeloproliferative disorder, Receptor tyrosine kinase