Selective dependency of CALR-mutant myeloproliferative neoplasms on TYK2 signaling.
Key claim: CALR-mutant and MPL-mutant MPNs selectively depend on TYK2 signaling, so TYK2 inhibition with deucravacitinib (alone or with fedratinib) suppresses these clones and ET/PMF colony growth while sparing JAK2V617F-driven disease.
Abstract
Myeloproliferative neoplasms (MPN) are driven by the oncoproteins JAK2V617F, mutant calreticulin (CALR), and mutant thrombopoietin receptor (TPOR; MPL), all of which activate JAK/STAT signaling. While JAK2 signaling is engaged in all MPNs, TYK2 is dispensable for JAK2V617F-driven disease. Here, we hypothesized a distinct role for TYK2 in CALR-mutant driven MPN. We found constitutive TYK2 phosphorylation in CALRdel52/ins5- and MPLW515K- but not JAK2V617F-expressing cells. Structural modeling predicted similar TPOR-binding affinities for JAK2 and TYK2, while micropatterning experiments demonstrated that both JAK2WT and JAK2V617F displace TYK2 from the receptor. The TYK2 inhibitor deucravacitinib reduced viability and STAT3/5 phosphorylation in CALRdel52/ins5- and MPLW515K- but not JAK2V617F-mutant cells, with enhanced efficacy when combined with the JAK2-selective inhibitor fedratinib. Cellular response correlated with JAK2 protein abundance, as CALRins5 JAK2clones outcompeted JAK2clones upon TYK2 inhibition. In primary samples, deucravacitinib significantly suppressed colony growth in ET and PMF but not PV, and selectively reduced CALR- but not JAK2V617F-mutant allele burden. Similarly, CALR-mutant patient-specific iPSC-derived CD34progenitors showed increased TYK2 phosphorylation and were more sensitive to TYK2 inhibition than their JAK2V617F counterparts. These findings identify TYK2 as a selective vulnerability in CALR-mutant MPN and support combined TYK2/JAK2 inhibition strategies to overcome JAK2-dependent resistance.
Why this matters
MPN subtype behavior has increasingly been explained by driver genotype rather than by clinical phenotype alone, and this finding sharpens that framing by identifying TYK2 as a signaling dependency unique to CALR- and MPL-mutant clones. For the CALR/MPL-enriched subsets of ET and PMF, it nominates a genotype-directed therapeutic axis (deucravacitinib, alone or with fedratinib) that is mechanistically distinct from pan-JAK inhibition, CALR-neoepitope antibodies, and LSD1/BET/XPO1 combinations. The explicit sparing of JAK2V617F disease also reinforces the emerging split between JAK2- and CALR/MPL-driven MPN biology that already appears in thrombosis risk, interferon response, and mutant-selective JAK2 programs.