Right versus left ventricular remodeling in heart failure due to chronic volume overload

Mechanisms of right ventricular (RV) dysfunction in heart failure (HF) are poorly understood. RV response to volume overload (VO), a common contributing factor to HF, is rarely studied. The goal was to identify interventricular differences in response to chronic VO. Rats underwent aorto-caval fistula (ACF)/sham operation to induce VO. After 24 weeks, RV and left ventricular (LV) functions, gene expression and proteomics were studied. ACF led to biventricular dilatation, systolic dysfunction and hypertrophy affecting relatively more RV. Increased RV afterload contributed to larger RV stroke work increment compared to LV. Both ACF ventricles displayed upregulation of genes of myocardial stress and metabolism. Most proteins reacted to VO in a similar direction in both ventricles, yet the expression changes were more pronounced in RV (pslope: < 0.001). The most upregulated were extracellular matrix (POSTN, NRAP, TGM2, CKAP4), cell adhesion (NCAM, NRAP, XIRP2) and cytoskeletal proteins (FHL1, CSRP3) and enzymes of carbohydrate (PKM) or norepinephrine (MAOA) metabolism. Downregulated were MYH6 and FAO enzymes. Therefore, when exposed to identical VO, both ventricles display similar upregulation of stress and metabolic markers. Relatively larger response of ACF RV compared to the LV may be caused by concomitant pulmonary hypertension. No evidence supports RV chamber-specific regulation of protein expression in response to VO.


Study Design
Eight-week old male Sprague Dawley rats (from on-site certified breeding colony at IKEM) weighing 280-320g were randomly distributed into two groups and underwent needle ACF (n=46)/sham (n=20) operation as described previously 1 . Briefly, under general anesthesia induced by intraperitoneal (ip) injection of ketamine/midazolam mixture (50 mg and 5 mg/kg of body weight, respectively), a midline laparotomy was performed and an 18-gauge needle (outer diameter 1.2 mm) was inserted into the infrarenal abdominal aorta and advanced through the vessel wall into the inferior vena cava to create ACF.
The needle was removed after temporary clamping the aorta above the site of puncture, which was then sealed with acrylamide tissue glue. The shunt creation was verified by observing pulsatile, bright flow in the inferior vena cava. Sham-operated (control) animals underwent the same procedure without puncture of the vessels.
The rats were kept in air-conditioned animal facility on a 12-/12-h light/dark cycle, were fed standard salt/protein diet (0.45% NaCl, 19-21% protein, SEMED, CR) and had free access to tap water.
The rats were weighed weekly and the onset of HF was monitored by scoring as before 1 . Monitored HF signs included piloerection, lethargy, peripheral cyanosis, dyspnea and abdominal swelling (ascites).
These parameters were weekly assessed by a single experienced technician and scored 0-3 points (0=absence, 3=maximal presence). HF score was calculated as the sum of points.
All animals that survived till 24 week after surgery were examined (65% survival in ACF group, 95% survival in control group) by at the end of protocol as described below. The presence of ACF was verified from laparotomy (animals with failed ACF would be excluded, but such situation did not happen) and the animals were exsanguinated. The coronary tree of the excised heart was rapidly flushed with cardioplegic solution. The organs were weighed and normalized to body weight. The investigation conformed to the NIH Guide for the care and use of laboratory animals (NIH Publication No. 85-23, 1996), Animal protection laws of the Czech Republic (311/1997) and was approved by the Animal Ethic Committee of IKEM (#16600/2014-OVZ-30.0-14.3.14). The study was carried out in compliance with the ARRIVE guidelines, if not explicitly stated otherwise.

Echocardiography and Hemodynamics
Echocardiography (available from all animals) was performed under general anesthesia (ketamine/midazolam mixture i.p. as described above) with 10 MHz transducer (Vivid System 7, GE, USA) 24th week post-ACF creation. End-systolic (ESV) and end-diastolic (EDV) LV volumes were derived by cubic equation and stroke volume (SV) as their difference. Cardiac output (CO) was calculated as product of stroke volume times heart rate (HR). Relative LV wall thickness was defined as sum of enddiastolic interventricular septum and posterior wall thickness, divided by end-diastolic LV diameter. LV fractional shortening (FS) was calculated as difference of end-diastolic and end-systolic LV diameter, divided by end-diastolic diameter. RV fractional area change (FAC) was defined as difference of end-diastolic and end-systolic RV area, divided by end-diastolic area. RV volumes were calculated using monoplane ellipsoid approximation method 2 . 30 ACF animals and 19 control animals were analyzed by echocardiography.
Subsequently, rats were intubated with a plastic cannula, relaxed with pancuronium (Pavulon, 0.16 mg/kg, N.V. Organon, Oss, Netherlands) and artificially ventilated (rodent ventilator; Ugo Basile, Gemonio VA, Italy, FiO2 = 21%). Vagal blockade (atropine 0.10 mg/kg) was administered to prevent interfering reflexes. LV function was invasively assessed by 2F Pressure-Volume (P-V) micromanometer-tip catheter (Millar Instruments, Houston, TX, USA) introduced into the LV cavity via the right carotid artery. Simultaneously, another 2F PV catheter was introduced into the right ventricle via the internal jugular vein to study RV function. The volume signals were calibrated by end-diastolic and end-systolic volumes obtained by echocardiography shortly before invasive recordings as done before 3 .
Data were acquired using an 8-channel Power Lab recorder and were analysed by LABCHART PRO software (ADInstruments, Bella Vista, NSW, Australia). Chamber wall stress was calculated as (peak chamber pressure x chamber end-diastolic volume)/chamber mass obtained by direct weighting 4 . Usable hemodynamic pressure volume recordings were obtained from 26 ACF animals and from 16 shamoperated control animals.

Contractility and action potential duration measurements
The studies were conducted as described elsewhere 5 . The papillary muscles were dissected from both ventricles and placed into an experimental chamber. The preparation was perfused with 36°C warm, oxygenated Tyrode solution at a constant flow rate (6-10 mL/min). After a stabilization period (30 min), the preparation was stimulated at various frequencies (0.5, 1, 2, 3, and 5 Hz; Pulsemaster Multi-Channel Stimulator A300, World Precision Instruments, Inc., FL, USA). Contraction force was measured by an isometric force transducer (model F30; Hugo Sachs Electronik -Harvard Apparatus, GmBH, Germany) and membrane potential was acquired using glass microelectrodes filled with 3M KCl (resistance >20 MΩ; Microelectrode Puller P-1000, Sutter Instrument, CA, USA). APD was measured at 50% and 90% levels of repolarization (APD50, APD90). Data were recorded and analyzed using the National Instruments data acquisition hardware and software (National Instruments, Austin, TX, USA).

Gene expression analysis
Samples were taken from RV and LV free wall and placed into RNAlater. Total RNA was isolated by RNeasy Micro Kit (Qiagen) according to the procedure for fibrous tissues (cardiomyocytes). All extracts were treated by DNase I (Qiagen) to remove contaminating genomic DNA. The quantity of the RNA was measured on a NanoDrop ND-1000 (NanoDrop Technologies LLC). RNA integrity was assessed on Agilent 2100 Bioanalyser (Agilent Technologies). All RNA samples had RNA integrity number RIN > 7. The RNA (1000 ng) was reverse transcribed using QuantiTect Reverse Transcription Kit (Qiagen). Either 1.5 ng or 5 ng of cDNA was used in the second step of RT-qPCR using RealTime  (Hprt1, Sdha, Tbp). Each assay included primers and a short FAMlabelled hydrolysis probe containing locked nucleic acid. The kit consisted of ten 384 well plates that contained inter-plate calibrator samples. The protocol was performed on a LightCycler LC480 (Roche) instrument according to manufacturer's protocol.
The resulting data were analyzed using the ∆C_p method 6 within the R/Bioconductor statistical environment 7 . The inter-plate calibration was performed as follows: within each plate and for each target, C_p of calibrator samples were subtracted from C_p of biological samples and mean C_p of calibrator samples from all plates was added. Next, missing C_p values (less than 1 % of all values) were imputed using the non-detects package 8 . Finally, the mean of three stable reference genes (Hprt1, Tbp and Sdha) was subtracted to obtain 〖∆C〗p values.

Proteomic analysis and Western blotting
Myocardial sample preparation: All chemicals were from Sigma-Aldrich, unless stated otherwise.
Myocardial samples from both ventricles were pulverized using mortar and pestle under liquid nitrogen.  Unreviewed -29,953 entries). Percolator was used for FDR estimation, and 1 % FDRs limits for peptides and proteins were used. Quantification data were normalized on total peptide amount. Unique and razor peptides were used for quantification.
Western blot analysis: Pulverized pooled heart samples were lysed in lysis buffer (RIPA buffer, Sigma-Aldrich) for 30 minutes in 4 °C and sedimented at 20,000-× g for 20 minutes at 4 °C. The supernatants were collected, and the total protein concentration were determined by BCA assay (Bicinchoninic Acid Kit, Sigma-Aldrich) using Nanodrop One (ThermoFisher). Protein samples (40 µg) were denatured at 100 °C for 10 minutes using sample buffer (60 mM Tris-HCL pH 6.8, 2 % SDS, 10 % glycerol, 0.02 % bromphenol blue) and separated by SDS-PAGE. Proteins were transferred to PVDF membranes and blocked at room temperature for 30 minutes in 5 % nonfat dry milk in PBS-T buffer

Statistics
Data were assembled and statistically analyzed using JMP 14 software package (SAS, USA).
Groups were compared using Student's t test and Pearson's correlation coefficient was used for assessment of correlation between continuous variables. Results are expressed as means±SD. P-value less than 0.05 was considered significant.   Western blots were prepared and developed as described in the method section.