Ataxia-Telangiectasia-Mutated deficiency is associated with postpartum cardiac dysfunction and hypertrophy

Background: Peripartum cardiomyopathy (PPCM) is defined by maternal systolic heart failure (HF) resulting from left ventricular (LV) dysfunction in previously heart healthy women, occurring between one month before, during, and the months after delivery. Recent data from the German PPCM registry indicated that gene variants associated with the DNA damage response (DDR) might promote PPCM onset. Here, we investigated the regulation of DDR associated genes in the postpartum mouse heart, the effect of impairment in this pathway during the peripartum phase using mice heterozygous for a major regulator of this pathway, Ataxia-Telangiectasia-Mutated (ATM), and determined the frequency of ATM gene variants  and plasma levels of ATM-targeting miRNAs in PPCM patients.

Methods and results: RNAseq analysis of LV tissue comparing healthy nulli pari (NP) and postpartum (PP) wildtype (WT) mice detected the upregulation of several DDR related genes. Whole exome sequencing in 80 PPCM patients from the German PPCM registry identified 7 carriers of heterozygous ATM gene variants classified as variants of uncertain significance (VUS, n=6) or pathogenic (P, n=1). MiRNA-screening by using human TaqMan Array Cards identified elevated plasma levels of 5 miRNAs (miR-18a-5p, miR-18b-5p, miR-203a-3p, miR-181a-5p, miR-181c-5p) potentially targeting ATM in PPCM patients compared to heart healthy postpartum-matched women. Overexpression of miRNA precursors of two of these miRNAs, miR-18a-5p and miR-18b-5p, resulted in the reduction of ATM protein levels in HL1 cardiomyocytes. Therefore, we investigated the role of ATM deficiency in the postpartum heart using a heterozygous ATM (ATM^+/-) mouse model generated by a targeted mutation (Atm^ins5790neo), which results in a truncated protein. These mice are born at the expected mendelian ratio, and reduction of ATM protein levels in ATM^+/- mice (-43%, P<0.01) was confirmed by western blotting. Cardiac function and heart weight (HW) were unchanged in ATM^+/- NP compared to WT NP mice. However, ATM^+/- mice who underwent 4 pregnancies and nursing periods (4xPP) displayed systolic dysfunction and cardiac hypertrophy unlike their WT 4xPP littermates (FAC (%): WT: 49±5 vs ATM^+/-: 37±9; n=11-12, P<0.001; HW/Tibia length (TL) (mg/mm): WT: 8.6±0.5 vs ATM^+/-: 9.4±1.0; n=11-12, P<0.05). In line with these findings, ATM^+/- 4xPP LV tissue revealed an increased myocyte cross sectional area (CSA (µm²): WT: 475±53 vs ATM^+/-: 572±81; n=6, P<0.01) and mRNA expression levels of hypertrophy markers ANKRD1 (+99%, n=7, P<0.01) and ANP (+52%, n=7, P<0.05) compared to WT 4xPP LVs. As a marker for DNA damage, we demonstrated that interferon type I related genes IRF7 (+109%, n=7, P<0.01), IRF9 (+76%, n=7, P<0.01) and STAT2 (+69%, n=7, P<0.01) were elevated exclusively in ATM^+/- 4xPP compared to WT 4xPP LVs without the induction of cardiac inflammation (total number of CD45⁺ cells).

In conclusion, upregulation of DDR genes in the postpartum maternal heart suggests an important role of these pathway for protection from pregnancy associated stress. The increased prevalence of ATM gene variants and plasma levels of ATM targeting miR-18a-5p and miR-18b-5p in a cohort of the German PPCM registry supports this notion and suggests a potential role of ATM deficiency in the pathophysiology of PPCM. In line with these findings, reduced ATM expression causes cardiac dysfunction and hypertrophy in the ATM^+/-PP mice after consecutive pregnancies and nursing periods