Acquired Resistance to JAK Inhibitors in Calr-Mutated Myeloproliferative Neoplasms

Background: Calreticulin (CALR) mutations are one of the major driver mutations in BCL-ABL1-negative myeloproliferative neoplasm (MPN) and are frequently detected in JAK2/MPL-unmutated essential thrombocythemia and primary myelofibrosis. Mutant CALR activates JAK-STAT signaling through an MPL-dependent mechanism to mediate pathogenic thrombopoiesis in MPNs. Although JAK inhibitors such as ruxolitinib can provide important clinical benefits to MPN patients including those harboring CALR mutations, JAK inhibition does not preferentially target the MPN clone and acquired resistance develops over time. We aimed to characterize the mechanisms of acquired resistance to JAK inhibitors in CALR-mutated hematopoietic cells and to screen for novel therapeutic approaches specifically target CALR-mutant cells in this study.

Methods: UT-7/TPO-derived cell lines expressing wild-type and type 1 and type 2 mutant CALR (CALRdel52 and CALRins5) were kindly provided by Drs. Komatsu and Araki. JAK2-inhibitor-resistant cells were generated by co-cultured with ruxolitinib and fedratinib (TG101348, a JAK2-selective inhibitor). JAK-STAT signaling was evaluated by Western blot on CALR-wild-type and mutated cells exposed to JAK2 inhibitor compared to untreated cells. For the detection of acquired secondary mutations in CALR-mutated cells exposed to JAK2 inhibitor, whole exome sequencing (WES) was performed using the BGISEQ-500 Sequencing platform (BGI, Shenzhen, China) with the 2 x 100 bp paired-end protocol. Genome Analysis Toolkit was used for variation detection. Reads were aligned to human reference genome hg19 using BWA version 0.7.15. Targeted resequencing was performed on leukocytes from patients with MPN who had been treated with ruxolitinib. Screening with chemical libraries/novel compounds will be conducted on UT7/TPO-CALR cell lines.

Results: Compared to the parental cells, ruxolitinib-resistant UT7/TPO-CALR mutant cell lines have developed significant cross resistance to other JAK inhibitor as shown in the cell viability study. Signalling downstream of JAK2 in all 3 inhibitor-naïve UT-7/TPO/CALR parental cell lines was inhibited by acute treatment of ruxolitinib as shown on Western blot. Whereas, constitutive JAK2 activation was observed in all 3 inhibitor-resistant UT-7/TPO/CALR cell lines. No change in the expression of Epo and MPL receptors in these cell lines was found. Interestingly, constitutive JAK3 activation was also seen in inhibitor-resistant UT-7/TPO/CALR cells in comparison with parental cells. These findings indicated the presence of transphosphorylation by JAK3 in inhibitor-resistant UT-7/TPO/CALR cell lines. In addition, the results of WES identified several acquired secondary mutations in 3 inhibitor-resistant UT-7/TPO/CALR cell lines including SH2B1, ABCC1, HOXB3 and KRTAP4-5. No acquired secondary mutation was identified in CALR and other genes involved in JAK-STAT signaling. Acquired secondary mutation will be screened in primary MPN patients' samples treated with JAK inhibitor. Type II JAK inhibitor such as BBT-594 has been shown to inhibit JAK activation and signaling in JAK-persistent/resistant cells.

Conclusions: Our results confirmed that the in vitro efficacy of JAK2 inhibition on CALR-mutant cells. Our data also suggested that JAK2 transphosphorylation and acquired secondary mutations could be underlying mechanisms for acquired resistance to JAK inhibitors in CALR-mutated cells. Novel therapeutics approaches should be developed to overcome acquired resistance in CALR-mutated cells.