Question: Structural, functional and metabolic remodeling of the atria plays a central role in the development and, more importantly, perpetuation of atrial fibrillation (AF). We have shown that the histone deacetylase (HDAC) inhibitor valproic acid is able to delay the development of atrial remodeling and the onset of atrial fibrillation (AF) in CREM-IbΔC-X transgenic mice (TG), a well-characterized model that spontaneously develops AF along an extensive atrial remodeling. Since valproic acid at therapeutic doses inhibits HDAC isoforms 1, 2, 3 and 8 we started to investigate which aspects of atrial remodeling could be affected by targeting specific HDAC isoforms. Here, we investigated the structural and functional remodeling along the development of AF in TG mice with concomitant genetic inactivation of the HDAC isoform HDAC2.

Methods: Mice with cardiomyocyte-specific knockout of HDAC2 (KO) were generated by crossbreeding HDAC2^loxP/loxP and αMHC^Cre+/- mice and were subsequently mated with TG mice. In vivo ECG recordings, determination of atrial weight and structural analysis of atrial tissue was performed on CTR (FVB/N^Cre+/-), TG (CREM-IbΔC-X^Cre+/-) and TGxKO mice. Intracellular Ca²⁺ (Ca²⁺_i) transients were recorded in isolated atrial cardiomyocytes (ACMCs) with Indo1/AM and spontaneous intracellular Ca²⁺ releases with Fluo-4/AM using a myocyte calcium and contractility system (Ionoptix).

Results: ECG recordings revealed that TG mice reliably developed AF between 7^th and 12^th week of life. Already at 6 wks TG mice had developed a significant enlargement of the left atrium reflected by an increase of left atrial weight (LAW), which was markedly attenuated in TGxKO mice (LAW/body weight in mg/g; CTR: 0.15±0.01, TG: 0.24±0.03*, TGxKO: 0.19±0.04; n=5-14; *p<0.05 vs. CTR). At the same time the increase in average cell length and planimetric area of ACMCs were reduced in TGxKO mice vs. TG mice (cell length in µm±SD; CTR: 96±7, TG: 153±15, TGxKO: 132±10^#; cell area in µm²±SD; CTR: 1053±172, TG: 1527±240, TGxKO: 1228±162^#; *p<0.05 vs. CTR; ^#p<0.05 vs. TG; n=9-11, 25-52 cells per image). Ca²⁺_i transient recordings in ACMCs from 6-week-old mice, before onset of AF, revealed increased Ca²⁺_i transient amplitudes in TG vs. CTR but not in TGxKO (in R.U.±SD; CTR: 0.11±0.07, TG: 0.15±0.09*, TGxKO: 0.12±0.08^#; *p<0.05 vs. CTR, ^#p<0.05 vs.TG; n=64-86 ACMCs). At the same time the amount of ACMCs with provoked Ca²⁺_i oscillations increased from 21% in CTR to 41% in TG while only 11% of TGxKO ACMCs showed Ca²⁺_i oscillations (n=9-12 animals, 5-10 ACMCs/animal). Finally, the limitation of structural and functional changes in TG by concomitant inactivation of HDAC2 was accompanied by a significant delay in the development of AF in TGxKO vs. TG mice. 

Conclusion: Cardiomyocyte-specific inactivation of HDAC2 attenuated the structural and functional remodeling of atria from CREM-IbΔC-X transgenic mice and delayed the onset of AF. Our work demonstrates that targeting HDAC2 in this AF model can limit structural alterations as well as disturbed Ca²⁺ dynamics, both important aspects of atrial remodeling. Thus, our results suggest HDAC2 as a possible therapeutic target to attenuate atrial remodeling in AF.

(supported by the DFG)