Abstract 4134148: Reduced BMPR2 in Monocytes De-represses HERV-K and Sustains Inflammation in Pulmonary Arterial Hypertension

Abstract

Background&Rationale: Human endogenous retroviral (HERV) elements are embedded in 8% of our genome. They remain quiescent but can be transiently activated by viral infection to induce an interferon (IFN) response. In pulmonary arterial hypertension (PAH), HERV-K expression is elevated in myeloid cells, and thus can promote inflammation observed in pulmonary arteries (PAs) with progressive obstructive remodeling. PAH is linked to a mutation or a decrease in BMPR2 expression which our laboratory has recently shown in monocytes (MONO). HERVs are silenced by SUMOylated (S) TRIM28 (KAP1, a methylase) and SPEN (a histone deacetylase) factors that are also sequestered by the lnc RNA XIST which is increased in females to inactivate the extra X-chromosome. When TRIM28 is phosphorylated (P) by DNA-PK, it loses methylase activity and can act as a transcription factor. Hypothesis: Reduced BMPR2 in MONO results in DNA-PK mediated P-TRIM28 causing demethylation of XIST targets (X-related genes) and HERV-K thus increasing their expression. Competitive SPEN binding to the increased XIST also de-represses HERV-K . Approach: We used THP-1 MONOs with BMPR2 shRNA and induced pluripotent stem cell (i) MONOs derived from BMPR2 mutant PAH patients and assessed HERV-K , XIST , IFNs and related our findings to DNA-PK–P-TRIM28 and sequestration of SPEN. Results: XIST and HERV-K mRNA were significantly elevated in THP-1 MONOs with shBMPR2 vs. shControl and in PAH-iMONO vs. control iMONO based on gender. Reduced BMPR2 also increased IFN genes. These findings were consistent with nuclear DNA-PK and P-TRIM28 shown to bind in affinity purification mass spectrometry analysis. Bisulfite conversion assay demonstrated decreased methylation of HERV-K in BMPR2 -deficient MONOs attributed to a reduction in the methylase S-TRIM28. RNA-IP in THP-1 MONOs with shBMPR2 vs. shControl indicated increased binding of XIST with TRIM28 and SPEN, suggesting a competitive reduction in binding to HERV-K. Conclusions: Loss of BMPR2 increases XIST and HERV-K in MONO related to DNA-PK induced P-TRIM28 and reduced methylase S-TRIM28. Increased HERV-K may also be a function of the increase in XIST sequestering its deacetylase SPEN. These molecular features underlie a chronic HERV-K -IFN inflammatory response in PAH. They may explain the female predisposition to PAH, because the increase in XIST in females heightens the propensity to activate HERV-K -IFN signaling and chronic inflammation in PAs infiltrated by MONO.