Pitx2 prevents susceptibility to atrial arrhythmias by inhibiting left-sided pacemaker specification

Abstract

Atrial fibrillation (AF), the most prevalent sustained cardiac arrhythmia, often coexists with the related arrhythmia atrial flutter (AFL). Limitations in effectiveness and safety of current therapies make an understanding of the molecular mechanism underlying AF more urgent. Genome-wide association studies implicated a region of human chromosome 4q25 in familial AF and AFL, approximately 150 kb distal to the Pitx2 homeobox gene, a developmental left-right asymmetry (LRA) gene. To investigate the significance of the 4q25 variants, we used mouse models to investigate Pitx2 in atrial arrhythmogenesis directly. When challenged by programmed stimulation, Pitx2(null+/-) adult mice had atrial arrhythmias, including AFL and atrial tachycardia, indicating that Pitx2 haploinsufficiency predisposes to atrial arrhythmias. Microarray and in situ studies indicated that Pitx2 suppresses sinoatrial node (SAN)-specific gene expression, including Shox2, in the left atrium of embryos and young adults. In vivo ChIP and transfection experiments indicated that Pitx2 directly bound Shox2 in vivo, supporting the notion that Pitx2 directly inhibits the SAN-specific genetic program in left atrium. Our findings implicate Pitx2 and Pitx2-mediated LRA-signaling pathways in prevention of atrial arrhythmias.

Fig. 1.

Pitx2 is expressed in the left atrium and pulmonary vein of postnatal mice. (A–G) Whole-mount LacZ staining and sagittal sections of Pitx2 ^LacZ allele at P3. Pitx2 is expressed in the left atrium, right ventricle, and pulmonary vein (arrows). Boxed areas in D and F are shown at higher magnification in E and G, respectively. (H–J) LacZ staining (arrows) on sagittal sections of Pitx2 ^LacZ allele on P42 shows LacZ activity in the left atrium. Boxed areas in H are shown at higher magnification in I and J. (K–N) LacZ staining (arrows) on sagittal sections of 1-year-old Pitx2 ^LacZ allele. Pitx2 is expressed in rare left atrial myocardial cells in left atrium. Boxed areas in K are shown at higher magnification in L, M, and N. (Magnification in E, G, I, J and L-N: ×200.) (O) RT-PCR of Pitx2c in WT left atrium and right atrium shows Pitx2c is expressed in left atrium. (P) Quantitative RT-PCR analysis of Pitx2c comparing embryonic and adult stages. Arrows indicate Lac staining. *, statistically significant difference (P < 0.05). LA, left atrium; LV, left ventricle; PV, pulmonary vein; RA, right atrium; RV, right ventricle.

Fig. 2.

Surface ECG (A) and intracardiac electrograms (B and C) revealing an episode of pacing-induced atrial tachycardia in a Pitx2^{null +/−} mouse. (A) Surface ECG showing the last three paced beats of the burst protocol. The tracing reveals the absence of P waves and irregular RR intervals, suggestive of atrial arrhythmia. (B) Atrial electrogram shows the presence of rapid and regular a waves, suggestive of AFL/tachycardia. (C) Ventricular electrograms reveals irregular intervals between v waves.

Fig. 3.

Pitx2 inhibits the SAN program. (A). Heat map of microarray data from Pitx2^{null −/−}-mutant embryo hearts compared with WT hearts at 13.5 dpc. All pacemaker genes and ion-channel genes designated to be present at significant levels were identified in the normalized data, followed by log2 transformation, subjected to hierarchical clustering and heat-map generating. (B) Quantitative RT-PCR validation of microarray analysis of Pitx2-mutant hearts compared with WT control hearts at 13.5 dpc. The quantitative real-time RT-PCR analysis showed that Kcnq1, ANF, Hcn4, Shox2, and Tbx3 were up-regulated in Pitx2-mutant hearts. (C). Whole-mount in situ analysis at 12.5 dpc and 13.5 dpc with indicated probes. Genotypes and probes are labeled. Arrows designate areas of relevant gene expression in control and mutant embryos. (D) Heat map of microarray data and (E) quantitative RT-PCR validation of microarray on WT and Pitx2^{null +/−} mouse hearts at P42. Potassium-channel genes including Kcnq1, as well as the SAN genes Shox2 and Hcn4, were up-regulated in the Pitx2^{null +/−}-mutant hearts. In B and E, * indicates statistically significant differences (P < 0.05).

Fig. 4.

Pitx2 directly represses Shox2. (A) Summary of exon usage by Pitx2 isoforms. AUG and UGA: start and stop codons; E1->E6, exons 1->6. (B) Targeting strategy of generating the Pitx2^Flag allele. (C) Southern blots for the Pitx2^Flag allele showing the WT 7-kb and mutant 4-kb bands. (D) Western blot showed Pitx2^Flag-directed Pitx2 expression in embryoid bodies and P0 heart. D7-D10, culture day 7–10. (E) Alignment of Pitx2 recognition elements in Shox2. (F) In vivo ChIP indicated that Shox2 was bound by Pitx2 in embryoid bodies (EB) and adult left atrium (H). (G) Luciferase activity assay experiments indicated that Pitx2 repressed Shox2 in P19 cells and that this repression was lost when the Pitx2 binding site was mutated. At top, highlight in red indicates point mutation inserted to disrupt Pitx2 binding. * indicates statistically significant differences (P < 0.05).