The deoxycorticosterone acetate (DOCA)-salt model of hypertension has been used to study diastolic dysfunction and heart failure with preserved ejection fraction (HFpEF) in rats. Our goal was to validate DOCA-salt as a mouse model of HFpEF and to identify immune mediators of cardiac remodeling using CITE-Seq, a novel single cell technique that provides surface marker and transcriptomic data.
DOCA-salt mice underwent uninephrectomy, DOCA pellet implantation, and 1% NaCl supplementation of the drinking water for three weeks. Control mice underwent a sham procedure and received normal water. Echocardiography and invasive hemodynamics were performed to examine cardiac function, and urine was collected to examine renal function. CITE-seq was performed on sorted leukocytes from four sham and DOCA-salt left ventricles. To test whether Trem2 has a causal role in HFpEF, wild type and Trem2 deficient animals were subjected to DOCA-salt. To evaluate translational potential, soluble Trem2 was measured in plasma of patients with HFpEF and healthy controls.
Compared to control mice, DOCA-salt mice exhibit significantly increased systolic blood pressure, albuminuria, and cardiac hypertrophy. On echocardiography, DOCA-salt mice exhibited a preserved ejection fraction. Invasive hemodynamic studies revealed an increased tau constant (5.70 ± .24, 8.22 ± .45; n = 8,5) in DOCA-salt mice, indicating diastolic dysfunction. CITE-seq revealed major transcriptional changes in cardiac macrophages between groups with upregulation of genes known to be involved in HFpEF and cardiac remodeling, such as galectin-3 (Lgals3) and osteopontin (Spp1), as well as novel genes including Trem2. Using an independent cohort of Sham and DOCA-salt treated mice, we validated a two-fold increase in Trem2 gene expression in left ventricles by qRT-PCR and a significant increase in Trem2 mRNA in Cd68+ macrophages by RNAscope. After DOCA-salt treatment, Trem2-deficient mice displayed increased heart weight to tibia length ratios (9.39 ± 0.20, 11.21 ± 0.80; n = 6–7) and albumin to creatinine ratios (5.5 ± 1.2, 14.2 ± 2.9; n = 6–7) compared to wild-type controls. No difference in albuminuria was observed at baseline (0.36 ± 0.05, 0.46 ± 0.10; n = 10). Bone marrow-derived macrophages (BMDMs) from Trem2 deficient mice expressed significantly less Lgals3 and Spp1 compared to wild type BMDMs suggesting that Trem2 signaling is upstream of galectin 3 and osteopontin in HFpEF. Plasma levels of soluble Trem2 were elevated in patients with HFpEF compared to healthy controls (51 ± 9.5; 22.4 ± 2.4 ng/mL; n = 9,12).
The DOCA-salt mouse model induces key features of human HFpEF and identifies a Trem2-dependent gene program involved in cardiac remodeling. Loss of Trem2 resulted in increased cardiac hypertrophy and renal damage in the DOCA-salt model of HFpEF.