AUTHORS
Maihi Skeen, BS, Nicole S. Arellano, MS2, Juan Li, PhD3, Anton Rets, MD, PhD4, Shannon E Elf, PhD5
AFFILIATIONS
1 Huntsman Cancer Institute - Cancer Research North, Salt Lake City, Utah, United States
2 Huntsman Cancer Institute - Cancer Research North, Salt Lake City, Utah, United States
3 Jeffrey Cheah Biomedical Centre, Cambridge, United Kingdom
4 ARUP Laboratories, Salt Lake City, Utah, United States
5 Huntsman Cancer Institute - Cancer Research North, Salt Lake City, Utah, United States
ABSTRACT
Introduction:
Calreticulin (CALR) mutations are the second most common driver of BCR::ABL1-negative myeloproliferative neoplasms (MPNs). The widely used CALRdel52 knock-in mouse model employs Mx1-Cre-mediated recombination to induce mutant CALR expression; however, induction requires polyinosinic:polycytidylic acid (poly I:C), which elicits systemic interferon signaling, causing acute hematopoietic suppression, weight loss, and potentially confounding disease phenotypes. We therefore compared disease progression in CALRdel52 mice crossed with inducible Mx1-Cre or constitutive Vav-Cre, which eliminates the need for poly I:C-mediated induction. We also evaluated three poly I:C dosing regimens to determine whether reduced inflammatory signaling could improve the Mx1-Cre model while preserving recombination efficiency.
Additionally, although the CALRdel52/Mx1-Cre model has been widely used to study ET and MF, it remains unknown whether these mice progress through a bona fide pre-fibrotic MF stage, representing a major gap in current MPN mouse models and limiting mechanistic studies of disease progression. We therefore asked whether either the Mx1-Cre or Vav-Cre model recapitulates this stage. Identification of a pre-fibrotic MF stage would establish a window to investigate the mechanisms driving progression from ET to pre-fibrotic MF and ultimately overt MF, as well as to evaluate therapeutic strategies that prevent disease progression.
Methods:
CALRdel52 knock-in mice were crossed with Mx1-Cre or Vav-Cre mice. Homozygous, heterozygous, and wild-type littermates were analyzed. Mx1-Cre mice received one of three poly I:C dosing regimens. Body weight and complete blood counts were serially monitored. Genomic recombination was assessed by PCR, and spleen, liver, and femurs were analyzed histologically.
Results:
In the Mx1-Cre cohort, poly I:C induction significantly influenced body weight and hematologic parameters in CALRdel52 mice, with effects proportional to dose. Mice receiving 10 mg/kg poly I:C lost 15-20% of body weight following the initial injection, whereas 1 mg/kg caused minimal weight loss. Although transient leukopenia was anticipated after poly I:C administration, white blood cell counts showed no consistent decline across dosing groups. Platelet counts demonstrated the greatest phenotypic differences. Lower poly I:C doses produced a more rapid rise in platelet counts, whereas higher doses attenuated thrombocytosis, suggesting that interferon-driven inflammation suppresses the CALRdel52 disease phenotype. Genomic PCR demonstrated that there is no significant in difference in recombination across dosing groups. This suggests that the 1 mg/kg dose of poly I:C is optimal to balance sufficient recombination with suppression of hematopoiesis and weight loss.
In the Mx1-Cre cohort, thrombocytosis developed as early as 6 weeks after 1 mg/kg poly I:C induction in homozygous mice and 7 weeks in heterozygous mice. Platelet counts in heterozygous mice plateaued at approximately 2,000 ± 500 × 10³/µL by 12 weeks, whereas homozygous mice exhibited progressive thrombocytosis throughout the study. Histopathologic analysis identified a pre-fibrotic stage by 3 months in homozygous mice receiving 1 mg/kg poly I:C. In contrast, disease progression was delayed by approximately 1-2 months in mice receiving 10 mg/kg poly I:C, consistent with higher dose poly I:C suppressing the CALRdel52 disease phenotype.
By comparison, Vav-Cre mice developed thrombocytosis by 5 weeks in both homozygous and heterozygous animals without poly I:C-induced toxicity. Platelet counts in heterozygous mice plateaued at approximately 2,000 ± 200 × 10³/µL by 7 weeks, whereas homozygous mice exhibited progressive thrombocytosis and developed histologic features consistent with pre-fibrotic myelofibrosis by 2 months, approximately one month earlier than observed in the Mx1-Cre model.
Conclusions:
Poly(I:C) substantially alters the hematologic phenotype of CALRdel52-driven MPN, potentially confounding studies using the Mx1-Cre model. In contrast, CALRdel52/Vav-Cre mice develop disease without inflammatory induction, providing a more physiologically relevant and reproducible model of CALR-mutant MPN. Importantly, we identified a bona fide pre-fibrotic MF stage, establishing a platform for mechanistic studies of disease progression and evaluation of therapies that prevent progression to overt MF.
DISCLOSURES
Maihi Skeen, BS: No relevant conflicts of interest to declare.