Gut Microbiota Promote Chronic Inflammation in Jak2V617F Mice

Chronic inflammation is regularly seen in patients with myeloproliferative neoplasms (MPN). Increased levels of pro-inflammatory cytokines and elevated white blood cell counts (WBC) are positively correlated with increased thrombotic risk and disease progression in MPN patients (Gerds A, Blood, 2024; Ronner L, Blood, 2020). The underlying cause of this chronic inflammation is multifaceted, likely involving factors such as the presence of JAK2 mutated clones leading to abnormal secretion of inflammatory cytokines. Previously, chronic inflammation has been shown to result from dysbiosis (i.e., proportional dysregulation of gut microbiome composition compared to healthy controls), and a weakened intestinal barrier (Potrykus M, Nutrients, 2021). Our group has already demonstrated that microbiota suppression using broad-spectrum antibiotics or an iron-deficient diet reduced chronic inflammation in sickle cell disease mice (Zhang D, Nature, 2015; Li H, Blood, 2023). Recent studies indicate that gut microbiota in MPN patients differs from healthy controls and exhibits increased pro-inflammatory signatures(Eickhardt-Dalboge C, Eur J Haematol, 2024; Oliver A, Microbiol Spectr, 2022). We hypothesize that alterations in gut microbiota contribute to inflammation in MPN patients and aim to better understand the relationship between the gut microbiota and chronic inflammation using Jak2^V617F mice.

First, we evaluated the degree of intestinal inflammation in Jak2^V617F mice and identified significantly increased leukocyte infiltration in the intestinal villi of Jak2^V617F mice as compared to the healthy controls (6±0.7 vs 2.1±0.2 % area, P<0.001). Second, Jak2^V617F mice exhibited dysbiosis with increased Firmicutes (112±37 vs 8±4x10⁷ copies/mg, P=0.02) and Actinobacteria (12±44 vs 2±0.6x10⁵ copies/mg, P=0.05), 2 common gram-positive bacterial phyla, compared to healthy controls. No difference was found in Bacteroidetes, a common gram-negative bacterial phylum. To further assess the significance of this dysbiosis, we treated Jak2^V617F mice with vancomycin (1 g/L) in the drinking water for 12 weeks to both target gram-positive intestinal bacteria and avoid confounding effects of antibiotic absorption. Our results demonstrate that vancomycin treatment leads to the following consequences: (1) decreased circulating WBC counts (7.4±0.6 vs 15.1±1.7x10³ cells/ul, P<0.001); (2) decreased the percentage (5±0.5 vs 7±0.5%, P=0.03) and the absolute number (0.4±0.05 vs 1.0±0.2 x10³ cells/ul, P<0.001) of circulating monocytes; (3) decreased the numbers of megakaryocytes in the spleen (173±27 vs 425±58 cells/mm², P=0.002) without impacting the degree of splenomegaly; and (4) reduced inflammatory cytokine transcripts both in the liver, e.g. transforming growth factor-β (TGF-β) (7±1 vs 18±4x10^-3 /GAPDH, P=0.03) and Vcam-1 (34±5 vs 62±7x10^-3 /GAPDH, P=0.01) and in the bone marrow, e.g. TGF-β (8±0.3 vs 16±3x10^-3 /GAPDH, P=0.004) and Vcam-1 (1.2±0.1 vs 1.9±0.3 /GAPDH, P=0.02).

Our results confirm the presence of microbial dysbiosis and inflammation in the intestine of Jak2^V617 mice and demonstrate that chronic inflammation is reversible with oral nonabsorbable vancomycin. Changes observed in vancomycin treated Jak2^V617 mice are more likely due to altered gut microbiota than the systemic effects of vancomycin. We thus speculate that these data provide a strong rationale for exploring cost-effective microbiota-targeted therapies for MPN patients to potentially alleviate systemic symptom burdens, improve patient quality of life, and delay disease progression.