A novel in-frame calreticulin mutation induces activation of JAK-STAT signaling and thrombocytosis Free

Mutations in the gene encoding calreticulin (CALR) represent the second most common genetic abnormality in BCR:ABL negative myeloproliferative neoplasms (MPNs). The CALR protein is an endoplasmic reticulum (ER)-resident calcium (Ca2+)-binding chaperone responsible for ER Ca2+ buffering and folding nascent glycoproteins. CALR mutations typically occur as heterozygous insertions and/or deletions in the last coding exon, resulting in a +1 base pair frame shift and generation of a neomorphic mutant C-terminus. This mutant C-terminus results in a gain of function (GOF) of binding to the thrombopoietin receptor (MPL) to activate JAK-STAT signaling to promote proliferation. The two most common CALR mutations are characterized as a 52 base pair deletion (CALRdel52) and a 5 base pair insertion (CALRins5).

We recently evaluated a 22 year old female patient with mild thrombocytosis (platelet count of 452 K/Cumm). Prior evaluation revealed platelet counts ranging from the high 400s to low 500s for at least the past four years. Next-generation sequencing revealed a deletion of base pairs 1214-1225 in the CALR gene, resulting in an in-frame deletion of four amino acids (p.Glu405_Asp408del) and loss of calcium (Ca2+) binding sites EEED (CALRdelEEED). Given the possibility that this may represent an atypical germline CALR variant, additional family members were tested for the mutation. The patient's identical twin sister tested positive for the same CALR mutation, and also was found to have mild thrombocytosis. The patient's father also tested positive for the mutation, consistent with germline transmission of the CALRdelEEED mutation.

Given these findings, we sought to determine if this atypical in-frame CALRdelEEED mutation has similar functionality to canonical CALR frameshift mutations.We first asked whether CALRdelEEED confers cytokine-independent growth. To do so, we generated Ba/F3-MPL-CALRdelEEED expressing cells and performed a proliferation assay in the presence or absence of interleukin 3 (IL-3). Upon IL-3 withdrawal, CALRdelEEED expressing cells indeed exhibited cytokine-independent growth. To determine how this cytokine-independent growth is driven, we next tested whether CALRdelEEED can bind to MPL and activate JAK-STAT signaling similar to canonical CALR frameshift mutations. We performed co-immunoprecipitation assays with FLAG-tagged CALR variants, and observed binding between FLAG-CALRdelEEED and MPL. Further, we observed phosphorylation of STAT5 in IL-3-starved Ba/F3-MPL-CALRdelEEED cells. We also performed colony forming assays on human CD34+ cells and found that CALRdelEEED confers a stem cell self-renewal advantage compared to control. Together, these data suggest that the CALRdelEEED mutation is capable of activating the MPL-JAK-STAT axis and promoting stem cell self-renewal.

We previously found that the CALRdel52 mutation, but not the CALRins5 mutation, loses Ca2+ binding function, resulting in chronic ER Ca2+ depletion. Because CALRdelEEED loses Ca2+ binding residues, we asked whether this is also the case for CALRdelEEED expressing cells. We performed a Stains-All assay and found that the CALRdelEEED protein exhibits decreased fluorescence, indicating a decreased ability to bind Ca2+. We also transfected HEK293T cells with CALR variants and R-CEPIA1er, a red-fluorescing ER Ca2+ sensor. Flow cytometry analysis revealed decreased ER Ca2+ levels in CALRdelEEED-expressing cells, although not to the extent of CALRdel52. Thus, the CALRdelEEED mutation functions similarly to, but not as potently as, CALRdel52.

Finally, we asked whether we could recapitulate the thrombocytosis observed in the aforementioned patients in an in vivo mouse model. We performed a retroviral bone marrow transplant in mouse cKit+ cells expressing GFP-tagged empty vector, CALR wildtype (CALRwt), CALRdel52, CALRins5, or CALRdelEEED. We found that CALRdelEEED-expressing mice exhibit elevated platelet levels similar to CALRdel52-expressing mice.In sum, we have identified a novel, in-frame CALR mutation that confers thrombocytosis and functions similarly to, although not as potently as, the canonical frameshift mutations. Our studies show that the neomorphic CALR mutant C-terminus may not be the only factor responsible for binding MPL and activating JAK-STAT signaling to drive disease. These findings also suggest that that atypical, in-frame CALR mutations may have clinical significance in individuals with thrombocytosis.