Heterogeneous cardiac sympathetic innervation gradients promote arrhythmogenesis in murine dilated cardiomyopathy

Ventricular arrhythmias (VAs) in heart failure are enhanced by sympathoexcitation. However, radiotracer studies of catecholamine uptake in failing human hearts demonstrate a proclivity for VAs in patients with reduced cardiac sympathetic innervation. We hypothesized that this counterintuitive finding is explained by heterogeneous loss of sympathetic nerves in the failing heart. In a murine model of dilated cardiomyopathy (DCM), delayed PET imaging of sympathetic nerve density using the catecholamine analog [11C]meta-Hydroxyephedrine demonstrated global hypoinnervation in ventricular myocardium. Although reduced, sympathetic innervation in 2 distinct DCM models invariably exhibited transmural (epicardial to endocardial) gradients, with the endocardium being devoid of sympathetic nerve fibers versus controls. Further, the severity of transmural innervation gradients was correlated with VAs. Transmural innervation gradients were also identified in human left ventricular free wall samples from DCM versus controls. We investigated mechanisms underlying this relationship by in silico studies in 1D, 2D, and 3D models of failing and normal human hearts, finding that arrhythmogenesis increased as heterogeneity in sympathetic innervation worsened. Specifically, both DCM-induced myocyte electrical remodeling and spatially inhomogeneous innervation gradients synergistically worsened arrhythmogenesis. Thus, heterogeneous innervation gradients in DCM promoted arrhythmogenesis. Restoration of homogeneous sympathetic innervation in the failing heart may reduce VAs.


m-HED Synthesis Protocol
All commercially available reagents and materials were used as received.Dimethyl sulfoxide (DMSO, anhydrous, 99.9 %) and N,N-Dimethylformamide (DMF, anhydrous, 99.8 %) were obtained from Sigma-Aldrich.Ethanol (EtOH, 200 proof, anhydrous) was obtained from Decon.All water used was purified to 18 MΩ and passed through a 0.1-mm filter.Metaraminol (free base) and meta-Hydroxyephedrine (m-HED) hydrochloride were obtained from ABX. Sodium chloride (NaCl, USP) and ammonium acetate (NH4OAc, HPLC grade) was purchased from Fisher Chemical.Millex-GV sterile filters with a 0.22μm pore size were purchased from Millipore Sigma.
Analytical Methods -Radio high-performance liquid chromatography (radio-HPLC) chromatograms for quality control were registered using an 1100 Series HPLC system (Agilent Technologies) equipped with a GabiStar flow-through gamma detector (Raytest) and a 4.6 x 250 mm Aqua 5u C18 125Å column (Phenomenex).Data acquisition and processing was performed using Gina Star 6 software (Raytest).Elution was performed at a constant flow rate of 1 mL/min with 10% EtOH in NH4OAc solution.Detection wavelength was 280 nm.
Semi-preparative HPLC was performed with the TRACERlab FX2 C (GE Healthcare) equipped with a S 1122 HPLC pump (Sykam) and a VP 250/10 NUCLEOSIL 100-5 C18 Nautilus column (Macherey-Nagel).Elution was performed at a constant flow rate of 4 mL/min with 5% EtOH in isotonic 25 mM NH4OAc (NaCl added for isotonicity).Storage solution for the column was 70% EtOH in water.
Radiosynthesizer Setup -Before each run, the TRACERlab FX2 C synthesizer was checked for leaks in the gas reaction circuit and the reaction vessel.The HPLC purification column and the fraction collection lines were rinsed with 5% EtOH in isotonic 25 mM NH4OAc.The reaction vessel was charged with metaraminol (0.7 mg, 4.2 µmol) dissolved in DMF:DMSO = 3:1 (320 µL).Vessel 3 was charged with 5% EtOH in isotonic 25 mM NH4OAc (1.78 mL).Before delivery of the activity, the methane oven was conditioned with a flow of H2 (50 mL/min) for 20 min at 350 °C.
Synthesis Protocol -[ 11 C]CO2 was produced by the 14 N(p,α) 11 C nuclear reaction in an 11 MeV RDS-112 cyclotron (Siemens) at 35 µA using an aluminum target body.The activity was unloaded from the target and pushed through the methane oven at room temperature.Trapped [ 11 C]CO2 was converted into [ 11 C]CH4 with a stream of H2 (50 mL/min) at 350 °C.[ 11 C]CH4 was trapped in the methane trap under liquid N2 cooling at -75 °C and purified.Conversion to [ 11 C]CH3I was achieved by reaction with I2 at 740 °C in a gas circulating process system.Accumulated [ 11 C]CH3I was released from the trap upon heating to 190 °C and pushed into the reaction vessel containing the precursor solution at -20 °C.The mixture was reacted for 5 min at 100 °C, cooled to room temperature and diluted with 5% EtOH in isotonic 25 mM NH4OAc in vessel 3. The mixture was transferred into the HPLC injection loop and purified by HPLC.The product fraction was typically collected from 11 min -14 min and sterile filtered directly.The final formulation was taken out of the hot cell for quality control and shipment.

Evaluation of noradrenergic nerves in heart sections of mice
Transmural mouse heart sections (100 m) were stained using primary and secondary fluorescent antibodies.Each section was placed in 400 μl of blocking solution containing 10% Horse Serum (HS), 0.2% Triton X-100, and 0.01 M PBS + 0.02% NaAz solvent for overnight incubation with slight agitation.Next, blocking solution was replaced with a primary antibody solution, containing the same blocking solution from the night before, 1:500 Rabbit PGP 9.5 (Protein Gene Product 9.5), 1:200 Sheep TH (Tyrosine Hydroxylase), and 1:400 Goat NPY (Neuropeptide-Y).After 3 nights of incubation with slight agitation, the sections were washed with 0.01 M PBS twice every hour for a minimum of 5 hours.A secondary antibody solution, containing the same initial blocking solution and 1:200 Anti-Rabbit 488, 1:200 Anti-Sheep Cy3, and 1:200 Anti-Goat 647, was added.Sections were covered to prevent light exposure and incubated for 3 nights with slight agitation.Next, sections were washed with 0.01 M PBS twice every hour for at least 5 hours, all while avoiding light exposure.Under dimly lit conditions soft paint brushes were used to place each section in the middle of a 75 mm long, 25 mm wide, and 1 mm thick glass microscope slide.Tissue mounting media -RIMS (refractive index matching solution) -was added in drops until the sections were covered entirely without excess media and an 18 mm x 18 mm x 0.25 mm coverslip was placed over the tissue mounting media covering the tissue.Slides were labeled, and taken for immediate imaging after a 20-minute drying period (in the dark).
A Zeiss LSM-880 confocal laser microscope (Oberkochen, Germany) was used to image the murine cardiac tissues.The following laser settings were used: 488 -PGP 9.5 (Green), 562 (Cy3) -TH (Red), 647 -NPY (Blue).Whole tissues were imaged at 10x and stitched together using 'tile scan'.A representative area of ~ 20 μm thickness was designated to the Zstack.All images were exported as .CZI files and converted to Maximum Intensity Projections (MIP) for subsequent processing.

Evaluation of noradrenergic nerves in sections of human left ventricle
Paraffin sections (5 m) through left ventricular free wall were obtained from archived tissue collected at autopsy from normal hearts and hearts with DCM.Slides were coded to maintain investigator blinding during tissue processing and microscopic evaluation.Sections were deparaffinized, hydrated, treated for antigen retrieval (1 mM EDTA, pH 8, 40 min at 92-94 C), and stained for the noradrenergic marker, tyrosine hydroxylase (TH) at room temperature using the ABC immunohistochemistry method (Rabbit ABC-HRP Kit, PK-4001, Vector Labs).Briefly, slides were rinsed with PBS (pH 7.3), incubated for 10 min in PBS containing 0.4% Triton X-100 and 0.5% bovine serum albumin (BSA), treated for 15 min with 1.0% H2O2 in PBS, rinsed an additional time with PBS and incubated 10 min in PBS containing 0.4% Triton X-100 and 0.5% BSA.Slides were then placed in an incubation box and covered with blocking buffer (PBS containing 1% BSA, 0.4% Triton X-100, and normal goat serum).After 2 hours, the blocking buffer was replaced with fresh blocking buffer containing the primary antibody (rabbit anti-TH, Pel-Freez, Cat.No. P40104-150, 1:500 dilution) and incubated overnight.Sections were washed with PBS and PBS containing 0.5% BSA followed by a two-hour incubation in biotinylated secondary antibody (1:200 dilution) from the kit.Slides were washed again, before a 1.5-hour incubation with the ABC reagent from the kit.Slides were next washed for 20 min in 50 mM Tris buffer (pH 7.6) before treatment for 1-10 min with the chromogen (Vector ImmPACT VIP Kit, SK4605) to visualize localization of TH (purple reaction product).Slides were washed, dehydrated and cover glasses attached using Cytoseal XYL (Thermo Scientific Cat.No. 8312-4).Labeled sections were viewed and digital images collected with an Olympus BX41 microscope equipped with an Olympus DP74 digital camera and cellSens software (Olympus America Inc., Center Valley, PA).Two sections from each sample were viewed to determine the presence of noradrenergic nerves in the subepicardium, midwall, and subendocardium, and samples were assigned to the normal or DCM group based on observed innervation pattern.Representative images for each group were collected for figure construction.

Primary Antibody Concentration Source
Protein Gene Product 9.5 (PGP 9.5)

Long-Axis Heterogeneity Score
Ratio Score Region with ratio of transmural gradients between 1.5-1.99 Region with ratio of transmural gradients between 2.0-2.99

points / region
Region with ratio of transmural gradients 3 and above

points / region
The human ventricular action potential model by O'Hara et al was used (1).For normal electrophysiology, control values for the O'Hara model parameters were used with GKs = 0.01 mS/µF given the small control value.Endocardial and epicardial differences in conductance parameters were taken from the original O'Hara model.For heart failure electrophysiology, changes in ionic currents and SERCA uptake were made according to the experimental ranges detailed in Elshrif et al (2), including specific heart failure remodeling for endocardial versus epicardial cells.(See table below for specific parameters.)A representative action potential for both normal control electrophysiology and heart failure electrophysiology is shown in Figure 7.

Cardiac tissue and anatomical ventricle modeling
The 1D cable was modeled with the following partial differential equation for voltage (V) as (1) and the isotropic 2D tissue as where Cm is the membrane capacitance and was set to Cm=1 µF/cm 2 , and D is called the diffusion constant which is proportional to the gap junction conductance.D=0.0005 cm 2 /ms was used.A time-adaptive Euler method with Dx=Dy=0.015cm and Dt=0.01-0.10ms was used for numerical integration of Eqs. 1 and 2.
The human anatomical ventricle model was adapted from our previous study Liu et al. (34), originally based on a model developed by Ten Tusscher et al (3,4).In brief, the governing equation for V is (3) where is the diffusion tensor describing the fiber directions in the ventricles (3).Endocardial and epicardial layers were created by defining an equidistant boundary between the epicardial and endocardial surfaces.This boundary was calculated using a fast-marching method as implemented in the scikit-fmm Python module.M-cells were not simulated for simplicity and due to the ongoing debate on their existence.The Purkinje network model was unchanged from our previous study (34) and based on the action potential model by Stewart et al. (5).Pacing was initiated from the AV-node through the His-Purkinje system.
Pseudo-ECGs were calculated for 1D and 2D tissue and the human whole-ventricle model using the following formula: (4) where and (x0, y0, z0) is the position of the ECG electrode.Note that for the anatomical ventricle simulations, P-waves are absent since the atria are not modeled.

Supplemental Figure 3 :
Cholinergic staining of the ventricular myocardium is primarily restricted to the epicardium in DCM and control mice.Confocal microscopy (10x) of IHC with VAChT in control (left) and DCM mice (right) shows staining being localized in the epicardium across all four walls of the left cardiac ventricle (n = 4 for control, n = 4 for DCM).Image scale bars are 100 μm.