Hypoxia Promotes Atrial Tachyarrhythmias via Opening of ATP-Sensitive Potassium Channels

BACKGROUND: Hypoxia-ischemia predisposes to atrial arrhythmia. Atrial ATP-sensitive potassium channel (KATP) modulation during hypoxia has not been explored. We investigated the effects of hypoxia on atrial electrophysiology in mice with global deletion of KATP pore-forming subunits. METHODS: Whole heart KATP RNA expression was probed. Whole-cell KATP current and action potentials were recorded in isolated wild-type (WT), Kir6.1 global knockout (6.1-gKO), and Kir6.2 global knockout (6.2-gKO) murine atrial myocytes. Langendorff-perfused hearts were assessed for atrial effective refractory period (ERP), conduction velocity, wavefront path length (WFPL), and arrhymogenicity under normoxia/hypoxia using a microelectrode array and programmed electrical stimulation. Heart histology was assessed. RESULTS: Expression patterns were essentially identical for all KATP subunit RNA across human heart, whereas in mouse, Kir6.1 and SUR2 (sulphonylurea receptor subunit) were higher in ventricle than atrium, and Kir6.2 and SUR1 were higher in atrium. Compared with WT, 6.2-gKO atrial myocytes had reduced tolbutamide-sensitive current and action potentials were more depolarized with slower upstroke and reduced peak amplitude. Action potential duration was prolonged in 6.1-gKO atrial myocytes, absent of changes in other ion channel gene expression or atrial myocyte hypertrophy. In Langendorff-perfused hearts, baseline atrial ERP was prolonged and conduction velocity reduced in both KATP knockout mice compared with WT, without histological fibrosis. Compared with baseline, hypoxia led to conduction velocity slowing, stable ERP, and WFPL shortening in WT and 6.1-gKO hearts, whereas WFPL was stable in 6.2-gKO hearts due to ERP prolongation with conduction velocity slowing. Tolbutamide reversed hypoxia-induced WFPL shortening in WT and 6.1-gKO hearts through ERP prolongation. Atrial tachyarrhythmias inducible with programmed electrical stimulation during hypoxia in WT and 6.1-gKO mice correlated with WFPL shortening. Spontaneous arrhythmia was not seen. CONCLUSIONS: KATP block/absence leads to cellular and tissue level atrial electrophysiological modification. Kir6.2 global knockout prevents hypoxia-induced atrial WFPL shortening and atrial arrhythmogenicity to programmed electrical stimulation. This mechanism could be explored translationally to treat ischemically driven atrial arrhythmia.

were backcrossed onto a C57Bl/6 background for at least six generations.Homozygous Kir6.1 global KO (Kir6.1(-/-),6.1-gKO) mice and littermate controls were generated by crossbreeding of the Kir6.1(+/-)heterozygous mice.Global Kir6.2 KO mice (6.2-gKO) were produced with the help of The Medical Research Council Centre for Mouse Genetics, Harwell Campus, Oxfordshire, UK (MRC Harwell) in collaboration with the International Mouse Phenotyping Consortium (IMPC), the European Conditional Mouse Mutagenesis Program (EUCOMM) and the International Knockout Mouse Consortium (IKMC). 47The process has been described previously by the IMPC. 47Briefly, a construct containing a lacZ sequence followed by a neomycin-resistance cassette was integrated via homologous recombination into the Kcnj11 gene upstream of its sole exon.The lacZ sequence together with the neomycin-resistance cassette were flanked by FRT sites and the neomycin-resistance cassette together with the Kcnj11 exon flanked by loxP sites producing a construct termed tm1a (Figure S8).The construct was transfected into embryonic stem cells (ESCs) harvested from a developing blastocyst creating the ES cell clone (clone id EPD0974_3_G05).ESCs exhibiting the transfected sequence were selectively cultured in antibiotic containing media.The cultured ESCs were then injected back into a blastocyst and implanted into a pseudo-pregnant female.Chimeric tm1a pups were born and crossed with C57Bl/6 wild-type mice to enable germ line transmission.Thereafter tm1a sperm were isolated and in-vitro fertilisation (IVF) performed in the presence of soluble Cre thus mediating excision of the loxP flanked neomycin-resistance cassette and Kcnj11 exon and generating pups heterozygous for the knockout tm1b allele (Kir6.1 (+/-)).When crossed these mice gave rise to litters containing pups with homozygous global knockout of Kir6.2 (Kir6.2(-/-),6.2-gKO) and wild-type litter`mate controls.Mice were bred to order generating stock totalling wild-type 61 and gKO 58.Mice were housed at MRC facilities until 6 weeks of age before being shipped and then acclimatised at our facilities until experimental age of 3-6 months.

Genotyping
DNA was extracted from mouse ear samples by proteinase K digestion.PCR to confirm Kir6.1 gene knock-out used the BioMix™ Red stable Taq DNA polymerase system (Bioline Reagents Ltd, UK) and the following primers: For the wild-type allele -sense 5' The presence of a tm1b knockout allele is driven only by a positive copy number from the LacZ assay and no copy number from any other assay.The presence of the WT allele is driven only by a positive copy number from the WT breakpoint sequence and WT critical sequence assays and no copy number from the other assays.

Quantitative RT-PCR
Total RNA was extracted from mouse tissues using the RNAeasy kit from Qiagen.
Commercially available isolated human total RNA from all four heart chambers (right atrium, left atrium, right ventricle and left ventricle) from three separate donors and said to be normal with no past medical history and normal cardiac structure and function, was purchased from AMS Biotechnology (Europe) Limited.Donors were a 49 year old male, 69 year old male and 65 year old male.
Total RNA was DNase I-treated and reverse-transcribed using the high-capacity cDNA reverse transcription kit (Applied Biosystems).
Quantitative RT-PCR was performed with 90 ng of murine cDNA and 50 ng of human cDNA using customized TaqMan gene expression assays (Applied Biosystems).We used the commercially available probes for all the genes as listed below: Mouse -Mm00434620_m1 for

Histology and immunohistochemistry
Mouse hearts collected from 10-12 week old animals were rinsed thoroughly in PBS to remove excess blood and fixed in 10 % formalin for at least 24 hours.After the fixation process, they were washed twice in PBS and stored in 70 % ethanol before paraffin embedding.Paraffinembedded myocardium were cut at 5 µm thick sections and mounted on clear, plus microscope slides.For histological analysis, sections were stained for haematoxylin and eosin with automated Leica autostainer XL system (Leica Biosystems, UK) and Trichrome stain kit Atrial cells were dissociated manually using a fire-polished Pasteur pipette in Kraft-Brühe solution and allowed to rest for 5 min prior to re-adaptation.The cells were allowed to re-adapt to physiological levels of Na + and Ca 2+ by incremental addition of a solution containing 10 mmol/L NaCl and 1.8 mmol/L CaCl2 and finally normal Tyrode's solution with 1 mg/ml BSA.
Cells were then plated on laminin-coated 10 mm coverslips and left for 30 minutes at 37 o C in Tyrode's solution with 1 mg/ml BSA prior to patch clamp recordings.

Supplemental Tables
Supplemental Table I Parameter . PCR cycle conditions were denaturation at 94 o C for 2 minutes, 35 cycles of 94 o C for 30 secs, 65 o C for 30 secs, 68 o C for 40 secs, extension at 68 o C for 8 minutes.This yielded a WT allele band of 474 bp and a knock-out allele band of 724 bp.Failure of knock-out leads to a product between the two primers too large to PCR under standard conditions (Figure S9).Confirmation of Kir6.2 gene knockout was performed by The Medical Research Council Centre for Mouse Genetics, Harwell Campus, Oxfordshire, UK (MRC Harwell) in collaboration with the International Mouse Phenotyping Consortium (IMPC) utilising qPCR of gDNA. 47All qPCR assays were FAM labelled using GTX Taqman master mix (Applied Biosystems) and run in duplex with a VIC labelled internal control, Dot1l (Primer 1 = GCCCCAGCACGACCATT, Primer 2 = TAGTTGGCATCCTTATGCTTCATC, Probe = CCAGCTCTCAAGTCG).Signal is calibrated against controls with known allele copy number.A LacZ assay designed around the sequence of the LacZ reporter (Primer 1 = CTCGCCACTTCAACATCAAC, Primer 2 = TTATCAGCCGGAAAACCTACC, Probe = TCGCCATTTGACCACTACCATCAATCC) is used in conjunction with a neomycin assay designed around the sequence of the neomycin resistance cassette (Primer 1 = GGTGGAGAGGCTATTCGGC, Primer 2 = GAACACGGCGGCATCAG, Probe = TGGGCACAACAGACAATCGGCTG) and 2 WT assays designed around the WT breakpoint sequence that is lost around the loxP sites of the synthetic cassette (Primer 1 = GGGCACGTGGAAAGTGAAG, Primer 2 = AGCCGAGCAGGGTCTTGTC, Probe = TAGAGTGGTGGGGTGCGAGC), and an assay designed around the critical sequence flanked by the two loxP sites (Primer 1 = ACGACCTGGCTCCTAGTGA, Primer2 = ACCACGCCTTCCAAGATGAC, Probe = CTGCACCACCACCAGGACCTG) (Figure S8).