Background: Cardiac remodeling can be driven by either septic or aseptic inflammation initialized by the activation of Toll-Like-Receptor 4 (TLR4). TLR4 activation by injection of Lipopolysaccharide (LPS) is known to decrease contractile function via nitric oxide signaling  and increase the probability of atrial and ventricular arrhythmia, however the exact mechanism is not fully explored. TLR4 can be found mainly on cells of innate immune system but also on highly specialized cells like cardiomyocytes. The difference in indirect effects mediated by TLR4 function on circulating cells or direct function on cardiomyocytes remains unclear. Therefore, in this study we aim to identify the cell type responsible for TLR4 dependent cardiac remodeling to suggest novel therapeutic principles.

Methods: A acute septic mouse model was generated by injecting the TLR4 activator LPS (2 mg/kg BW) or NaCl intraperitoneal and 3.5 hours later animals were sacrificed for further experiments. Wildtype C57/Bl6J mice and ubiquitous TLR4^-/- mice were compared in explanted Langendorff-perfused hearts by optical epicardial voltage-mapping using the voltage sensitive dye Di-4-ANEPPS and a highly light sensitive sCMOS camera. Conduction velocity (CV) as well as action potential shape were analyzed from atria and ventricles with up to 10.000 frames per second. In addition, single cardiomyocytes were dissociated from atria of untreated mice and exposed to LPS for 3 h in vitro before recording of action potentials by patchclamp analyses.

Results: In atria of LPS injected wildtype mice, voltage-mapping showed decreased conduction velocity (+LPS: 43.1 ± 3.1 cm/s, n = 5; +NaCl: 72.6 ± 9.8 cm/s, n = 10, p = 0.04). Analysis of syncytial action potential shape showed a tendency to decreased upstroke velocity (max. dV/dt: +LPS: 38.4 ± 2.9 mV/ms, n = 5; +NaCl: 54.8 ± 5.9 mV/ms, n = 10, p = 0.06) but action potential duration was unaltered (APD70: +LPS: 25.6 ± 4.1 ms, n = 5; +NaCl: 28.7 ± 3.2 ms, n = 10, p = 0.57). Decreased CV and max. dV/dt, both indicators of decreased Na⁺ channel availability, were not observed in TLR4^-/- mice. Interestingly, LPS stimulation of isolated atrial cardiomyocytes also lead to decreased max.dV/dt in patchclamp experiments (+LPS: 45.7 ± 4.1 mV/ms, n = 5; + NaCl: 156.3 ± 11.6 mV/ms, n = 4, p < 0.001), indicating a cardiomyocyte specific effect of TLR4 on Na⁺ channels. In addition we found proarrhythmogenic shortened action potential durations (APD70: +LPS: 10.3 ± 2.5 ms, n = 5; +NaCl: 19.1 ± 2.3 ms, n = 6, p = 0.003). In the left and right ventricle we observed significantly decreased CV (+LPS: 50.2 ± 2.2 cm/s, n = 6; + NaCl: 67.7 ± 5.0 cm/s, n = 10, p = 0.02), decreased upstroke velocity (max. dV/dt: +LPS: 32.8 ± 1.6 mV/ms, n = 6; +NaCl: 46.3 ± 0.9 mV/ms, n = 10, p < 0.0001) and shortened APD70 (+LPS: 41.5 ± 2.7 ms, n = 6; +NaCl: 67.7 ± 7.8 ms, n = 10, p = 0.02).