Anti-Inflammatory and Antiarrhythmic Effects of Beta Blocker in a Rat Model of Rheumatoid Arthritis.

Background Patients with rheumatoid arthritis are at twice the risk of ventricular arrhythmia and sudden cardiac death as the general population. We hypothesize that β-blocker treatment of rheumatoid arthritis is antiarrhythmic by producing synergistic anticatecholaminergic and anti-inflammatory effects. Methods and Results Collagen-induced arthritis (CIA) was induced in Lewis rats by immunization with type II collagen in Freund's incomplete adjuvant. The treatment with propranolol (4 mg/kg) started on the first day of immunization. We evaluated the ventricular vulnerability to ventricular arrhythmia using in vivo programmed stimulation and performed ex vivo optical mapping to measure the electrical remodeling of the heart. The ventricular tissue was further processed for immunohistochemical staining and protein array analysis. The assessment of ventricular vulnerability showed that the number and duration of the induced ventricular arrhythmia episodes were increased in CIA rats, which were improved with propranolol treatment. The sympathovagal index and the plasma level of catecholamines significantly increased in CIA rats, whereas the use of propranolol attenuated sympathetic hyperactivity. In the optical mapping study, electrical remodeling, characterized by prolonged action potential duration, slow conduction velocity, and steepened action-potential duration restitution, were noted in CIA rats and reversed in the propranolol-treatment group. The propranolol treatment was associated with decreases in paw thickness, fewer inflammatory cell infiltrations in the heart, reduced levels of cardiac inflammatory cytokines, and less cardiac fibrosis as compared with the CIA group. Conclusions CIA increased ventricular arrhythmia vulnerability through sympathetic hyperinnervation and proarrhythmic ventricular electrophysiological remodeling. Treatment with propranolol in CIA rats was both anti-inflammatory and antiarrhythmic.

action potential duration

collagen‐induced arthritis

conduction velocity

heart rate

heart rate variability

intrinsic heart rate

interleukin‐1 beta

pacing-cycle length

rheumatoid arthritis

sympathovagal index

tyrosine hydroxylase

tumor necrosis factor-alpha

ventricular arrhythmia

Clinical Perspective

What Is New?

The present study showed that rheumatoid arthritis increased susceptibility to ventricular arrhythmia by ventricular remodeling, including increased action potential duration and alternans, significant fibrosis, and heterogeneous sympathetic hyperinnervation, whereas treatment with β blockers reversed the proarrhythmic ventricular remodeling and reduced vulnerability to ventricular arrhythmia.

What Are the Clinical Implications?

Patients with rheumatoid arthritis, who receive treatment with a β blocker, experience synergistic anticatecholaminergic and anti‐inflammatory effects, which reduce the risk of ventricular arrhythmia and further cardiovascular events.

Patients with rheumatoid arthritis (RA) have a 2‐fold risk of sudden cardiac death or ventricular arrhythmia (VA) compared with the general population. , , Arthritis could boost the immunoinflammatory process of atherosclerosis; as a result, it could contribute to the incidence of myocardial infarction and heart failure. Notably, proinflammatory cytokines could accelerate coronary atherosclerosis and mediate myocardial electrical instability. One of the possible mechanisms is the chronic activation of inflammatory cytokines in RA, which may elicit deleterious increases in the sympathetic neural outflow. In joint tissue, the production of TNF‐α (tumor necrosis factor‐alpha) and interleukin‐1 beta (IL‐1β) from macrophages could be activated by β adrenoceptors. Cardiac macrophages mediating the production of IL‐1β could cause prolongation of the myocardial action‐potential duration (APD), a decrease in potassium current, and an increase in calcium sparks in cardiomyocytes—all of which contribute to the electrical remodeling underlying the VA propensity in mice with diabetes mellitus. Treatment with the β blocker, propranolol could reduce the gene expression of TNF‐α in mice with viral myocarditis. In this study, we hypothesize that the anticatecholaminergic and anti‐inflammatory effects of propranolol are associated with the attenuation of autonomic imbalance and increased ventricular arrhythmogenicity in a rat model of RA.

Methods

Our data, analytic methods, and study materials are available for the reproduction of our findings from the corresponding authors on reasonable request. All supporting data are provided within the article and its online supplementary files.

Animal Subjects

The present study was performed in accordance with the Guide for the Care and Use of Laboratory Animals and was approved by the Institutional Animal Care and Use Committee at National Chiao Tung University (Hsinchu, Taiwan; NCTU‐IACUC‐106060). At 8 weeks of age, female Lewis rats (weight 210±10 g) were randomly divided into 3 groups: normal control (CON) rats, rats with collagen‐induced arthritis (CIA), and rats with collagen‐induced arthritis treated with propranolol (CIA‐PRO).

Collagen‐induced Arthritis

Type II collagen was dissolved in 0.05 mol/L acetic acid and emulsified with an equal volume of incomplete Freund's adjuvant. , , Rats were injected with 2 mg/kg collagen‐incomplete Freund's adjuvant at the base of the tail (day 0). To ensure a high incidence and severity of arthritis, a boost injection (1 mg/kg) was given on day 7 after the initial immunization in the same manner (Figure 1A). The severity of arthritis can be evaluated with a clinical score from 0 to 4 depending on paw inflammation. The normal condition is 0, apparent redness and swelling of the ankle is 1, moderate redness and swelling of the ankle or the wrist is 2, severe redness and swelling of the entire paw including digits is 3, and maximally inflamed limb involving multiple joints is 4.

Figure 1

Lewis rats with collagen‐induced arthritis.

A, Timeline of the experiment's design and the treatment progress of the CIA rats. B, Inflammation of hind paw. C, Effects of weight gain and loss (n=6). D, The thickness of CIA hind paw (n=6). E, Clinical score of CIA symptoms (n=6). Data are shown as mean with ±SD. *P<0.05 and ***P<0.001 vs CON group, † P<0.05 and ††† P<0.001 vs CIA group. CIA indicates collagen‐induced arthritis group; CIA‐PRO, CIA rats treated with propranolol; and CON, control group.

Propranolol Administration

Propranolol was administered intragastrically using a rodent feeding tube with a dose of 1 mg/kg at day 1, then 3 times daily until day 30 of the experiment.

Heart Rate Variability Analysis

The single‐lead ambulatory ECG was recorded for 30 minutes every 2 days. The standard deviation (SD) of N‐N intervals and the root mean square of successive differences of N‐N interval were calculated as the time‐domain measures of heart rate variability (HRV). In the HRV spectral analysis, the low‐frequency power (0.2–0.8 Hz) and high‐frequency power (>0.8 Hz) were derived from the sum of the area within a specific frequency range under the power spectrum density curve of the entire 5‐minute segment.

Sympathetic, Vagal Tone, and Intrinsic Heart Rate

The basal heart rate (HR) was recorded for 10 minutes, and methylatropine (3 mg/mL) was injected into the jugular vein by catheter. HR recording was continued for 10 minutes. Propranolol (4 mg/kg) was injected 10 minutes after the methylatropine injection, and the intrinsic heart rate (IHR) was evaluated. After 1 day, the sequence of drug treatment was reversed. In addition, the sympathovagal index (SVI) was defined as basal HR divided by IHR, indicating the autonomic balance of the heart.

In Vivo Ventricular Arrhythmia Induction

Animals were anesthetized by an injection of 50 mg/kg of Zoletil (Virbac, Carros, France) and a tracheotomy was performed to assist breathing after the thoracotomy during programmed electric stimulation. The programmed electric‐stimulation protocol, designed as 10 stimuli of S1, was delivered, followed by an extra stimulus S2, starting at a coupling interval of 70 ms with 5‐ms decrements. VA episodes were calculated by 10 stimuli per rat and VA duration was defined as the duration of VA in each episode.

Optical Mapping and Study Protocol

We performed high‐speed optical mapping on Langendorff‐perfused rat hearts. The rat heart was isolated immediately and perfused with Tyrode's solution in a warm bath at 37°C. The heart was stained with Di‐4‐ANEPPS and blebbistatin to promote observation of the electrical response to electrical stimulation at a constant pacing rate. The heart was illuminated with a green light‐emitting diode (wavelength 505±20 nm). The induced fluorescence was collected through a 600‐nm long‐pass filter by a CMOS (complementary metal oxide semiconductor) camera (SciMedia, Costa Mesa, CA). The digitized fluorescence images were then transferred to a computer for further analysis. ,

Protein Array

The expression of cytokines in heart tissue, including TNF‐α and IL‐1β, was quantified by RayBio Rat Inflammation Array 1 (Raybiotech, Inc, Norcross, GA; catalog no.: QAR‐INF‐1).

The plasma expression of epinephrine and norepinephrine were detected using commercial enzyme‐linked immunosorbent assay kits (Abnova, Taipei, Taiwan; catalog no.: KA1877).

Immunohistochemical Studies

For the analysis of immune cells and nerve formation, myocardial sections (20 μm) were fixed with acetone for 15 minutes at −20°C (tyrosine hydroxylase [TH]) or 5% paraformaldehyde for 10 minutes at room temperature (CD45 cells). Sections were incubated with primary antibodies against TH (Millipore, Burlington, MA; catalog no.: AB152) and CD45 cells (Abcam, Cambridge, UK; catalog no.: AB23910) overnight at 4°C. For the investigation of myocardial fibrosis, hearts were fixed overnight in 4% paraformaldehyde and embedded in paraffin. Tissue sections were stained with Masson trichrome. The density of the stained areas was determined by Image‐Pro Plus software (Media Cybernetics, Rockville, MD). Each slide was examined under a microscope with ×40 objectives to select 3 fields with the highest density in each section.

Statistical Analysis

Comparisons of data between each pair of 3 groups were performed by using the Mann–Whitney U test when the data were lacking normal distribution, or by unpaired Student t test. Overall, the 3 groups were also compared by using the Kruskal–Wallis test when the data were lacking normal distribution, or by 1‐way ANOVA. A P<0.05 was considered statistically significant. All data were reported as the median with 25th to 75th interquartile ranges except HRV, weight, paw thickness, clinical score, APD70 (APD at 70% of repolarization), and conduction velocity (CV), which were presented as mean±SD. Restitution curve and maximum slope were calculated by OriginLab (OriginLab Corp, Northampton, MA) and GraphPad Prism 7 software (GraphPad Software, Inc, La Jolla, CA).

Results

Joint Inflammation Induced by Collagen‐induced Arthritis

Severe joint inflammation was induced by CIA after 4 weeks (Figure 1B). In the CIA group, there was a significant decrease in body weight (P<0.05; Figure 1C). In the 3 groups, there were significant differences at day 15 and day 20 (P<0.05). Paw swelling was profound in CIA rats from day 14 to day 21 (Figure 1D). In addition, there were significant differences among the 3 groups at day 15 and day 20 (P<0.05). The clinical scores of the CIA rats were also higher after immunization than those of the CON rats, indicating that the symptom of arthritis was induced to exhibit paralysis and maximally inflamed limbs with the involvement of multiple joints (Figure 1E). There were significant differences among the 3 groups from day 10 to day 25 (P<0.05). Of note, in the CIA‐PRO group, the inflammation of joints was similar to that of the CIA rats, complicated with paw swelling and increased clinical scores, but mildly relieved in the late course of inflammation (Figure 1D and 1E).

Sympathetic Predominance Determined by HRV and SVI

The effects of double pharmacological vagal and sympathetic blockade on HR in the 3 groups are depicted in Figure 2. The IHR was significantly decreased in the CIA group versus that of the CON rats (CIA: 293.94 [33.96], CON: 348.85 [49.62], P<0.001), but the IHR was significantly increased after administration of propranolol (CIA‐PRO: 334.59 [27.57], P<0.001; Figure 2A). Moreover, the SVI suggested that a sympathetic autonomic component was predominant in the determination of the basal HR in CIA rats, which was relieved by the treatment with propranolol (CON: 0.98 [0.08], CIA: 1.29 [0.22], CIA‐PRO: 1.12 [0.12], P<0.001; Figure 2B). Additionally, in the 4 weeks of HRV analysis, mean HR and low‐frequency and low‐frequency/high‐frequency power were significantly increased in the CIA group (Table 1 and Table S1; P<0.05). In the fourth week, compared with the CON‐group rates, a reduced root mean square of successive differences of N‐N interval and deceleration capacity were noted in the CIA rats; these findings were consistent with findings from our previous study (Table 1; P<0.05). Moreover, the plasma level of epinephrine was significantly increased in the CIA rats versus the CON rats (CIA: 1.25 [0.19], CON: 1.07 [0.12]; P<0.01). Conversely, norepinephrine was significantly reduced in the CIA group versus the CON and CIA‐PRO groups (CIA: 0.30 [0.51], CON: 1.90 [2.06], CIA‐PRO: 2.96 [1.46], P<0.01), as shown in Figure 2C and 2D. The differences in epinephrine levels among the groups were in accordance with the differences of SVI among the groups. Hence, the use of propranolol appears to attenuate sympathetic hyperinnervation in RA.

Figure 2

The measurement of the sympathovagal index imbalance.

A, The intrinsic heart rate demonstrates the net autonomic nerve activities after heart‐rate responses to methylatropine and propranolol (n=6). B, The sympathovagal index expresses the autonomic balance of the heart (n=6). C, Sympathetic overactivity is presented as a high level of epinephrine expression in CIA rats (n=6). D, Treatment with propranolol increases the plasma level of norepinephrine and balances the autonomic dysfunction (n=6). Data are shown as the median with 25th to 75th interquartile ranges. In the 3 groups, there are revealing significant differences in intrinsic heart rate, sympathovagal index, and epinephrine and norepinephrine expression (all P<0.05). **P<0.01 and ***P<0.001 vs CON group, †† P<0.01 and ††† P<0.001 vs CIA group. CIA indicates collagen‐induced arthritis group; CIA‐PRO, CIA rats treated with propranolol; and CON, control group.

Table 1

Effects of Propranolol on HRV Parameters in the Fourth Week

Parameters	CON	CIA	CIA‐PRO	ANOVA P Value
Linear HRV parameters
Mean HR, 1/min	414.47±11.3	489.98±9.3***	423.72±23.7†††	<0.001
SDNN, ms	4.76±1.1	4.31±1.0	4.65±0.9	0.34
RMSSD, ms	3.68±0.7	2.40±0.0*	4.20±1.3††	0.03
LF (n.u)	55.09±4.1	66.64±5.7***	49.37±2.0†††	0.002
HF (n.u)	39.75±3.9	41.77±5.6	50.40±3.5**†	0.02
LF/HF	1.24±0.2	1.80±0.6*	0.98±0.1††	0.03
Nonlinear HRV parameters
DFAα1	0.76±0.2	0.33±0.1***	0.70±0.1††	<0.001
DFAα2	0.99±0.3	0.78±0.1	0.73±0.3	0.13
DC, ms	10.50±1.6	6.75±2.4*	14.49±2.1*†††	<0.001
AC, ms	−8.32±1.7	−4.95±1.6**	−11.38±1.8*†††	<0.001

Data are shown as means±SD. An unpaired Student t test was performed for each pair of 3 groups. AC indicates acceleration capacity; CIA, collagen‐induced arthritis; CIA‐PRO, collagen‐induced arthritis treated with propranolol; CON, normal control; DC, deceleration capacity; DFA, detrended fluctuation analysis; HF, high frequency; HR, heart rate; HRV, heart‐rate variability; LF, low frequency; n.u., normalized units; RMSSD, the root mean square of successive differences of N‐N interval; and SDNN, the standard deviation of N‐N intervals.

* P<0.05, **P<0.01, and ***P<0.001 vs CON group.

† P<0.05, †† P<0.01, and ††† P<0.001 vs CIA group.

Changes in VA Inducibility and Electrophysiological Parameters

VA was induced by the S1‐ to S2‐stimulation protocol described in the Methods section. The time course of VA induction among the 3 groups is shown in Figure 3A. To analyze VA reproducibility, the probability of the arrhythmias was calculated as the number of events of VA after 10 S1‐ to S2‐induction episodes for each rat. The probability of VA was significantly different between the CIA and CON groups (CIA: 2.00 [1.00], CON: 0.00 [0.00]; P<0.01), and was also decreased by propranolol treatment (CIA‐PRO: 0.5 [1.00], P<0.01; Figure 3B). Furthermore, we observed a significantly longer duration of VA in CIA versus CIA‐PRO rats, respectively (P<0.001; Figure 3C). Our finding shows that the vulnerability to VA was increased in CIA rats, and that treatment with propranolol was associated with a reduction in VA inducibility.

Figure 3

Probability and duration of induced VA episodes.

A, Examples of VA induction attempts in the 3 groups. B, The number of VA episodes was increased in CIA rats (n=6). In a comparison of the 3 groups, there are significant differences (P=0.02). C, The overall duration of VA was significantly decreased in CIA‐PRO rats (n=6). Overall comparison also shows a significant difference (P=0.004). Data are shown as the median with 25th to 75th interquartile ranges. **P<0.01 and ***P<0.001 vs CON group, †† P<0.01 and ††† P<0.001 vs CIA group. CIA indicates collagen‐induced arthritis group; CIA‐PRO, CIA rats treated with propranolol; CON, control group; and VA, ventricular arrhythmia.

Figure 4A shows APDs obtained with different pacing‐cycle lengths (PCLs) of the 3 groups. Compared with CON rats, the APD70 of the CIA and CIA‐PRO rats was significantly prolonged at all PCLs (P<0.05). After propranolol administration, the APD70 was shorter in the CIA rats at PCLs of 250, 180, and 140 ms (P<0.05; Figure 4B). Furthermore, the CV was decreased in the CIA rats at all PCLs as opposed to the CON rats (Figure 4C). Although the CV was not significantly different between the CIA and CIA‐PRO rats, the phase maps showed that the electric propagations were gradually delayed in the CIA group (Figure 5A) Also, enhanced heterogeneity of myocardial conduction and severe cardiac instability were found in CIA rats, but administration of propranolol could restore the heterogeneous conduction as shown in the isochrone maps (Figure 5B). Figure 4D shows the APD restitution curves. The APD restitution slope in CIA rats was constantly higher than in CON rats from slow (250 ms) to fast (120 ms) PCLs. The maximum slope of APD restitution was also significantly increased in the CIA rats as compared with CON rats (CIA: 1.46 [1.77], CON: 0.28 [0.12]; P<0.01), but the maximum slope was lower than 1 in CIA‐PRO rats (CIA: 0.50 [0.23], P<0.05), meaning that treatment with propranolol could reduce VA vulnerability (Figures 3 and 4E). The aforementioned parameters of the CIA‐PRO rats were improved with the propranolol treatment, showing less APD prolongation and restoration of CV heterogeneity as compared with CIA rats.

Figure 4

Prolongation of APD and arrhythmogenic ventricular substrate in CIA rats.

A, The left ventricles of the CIA rats have longer APDs than the CON rats. B, APDs measured in 70% repolarization with different PCLs (n=6). C, CV in the CIA rats at different PCLs (n=6). D, APD restitution curve and (E) maximum slope (n=6). Data are shown as mean with ±SD, expect maximum slope, which are shown as the median with 25th to 75th interquartile ranges. The overall 3‐group comparison shows a significant difference in maximum slope (P=0.02). *P<0.05, and **P<0.01 vs CON group, † P<0.05, and †† P<0.01 vs CIA group. APD indicates action potential duration; CIA, collagen‐induced arthritis group; CIA‐PRO, CIA rats treated with propranolol; CON, control group; CV, conduction velocity; and PCL, pacing‐cycle length.

Figure 5

Schematic and effect of propranolol on ventricular CV and electrical propagation in CIA.

A, Consequent phase maps of electrical propagation in the left ventricle. B, Isochronal map was fragmented in CIA rats, but it was recognizable by the administration of propranolol. CIA indicates collagen‐induced arthritis group; CIA‐PRO, CIA rats treated with propranolol; and CON, control group.

Increased Expression of Inflammatory Cytokines in Myocardium

Because of the arthritis‐induced systemic inflammation, the serum levels of TNF‐α and IL‐1β were significantly elevated in CIA and CIA‐PRO rats compared with CON rats, but there was no difference between CIA and CIA‐PRO rats (data not shown). With respect to the expression of inflammatory cytokines in the left ventricle (LV) of rat hearts, the TNF‐α (CIA: 181518.75 [56236.35], CON: 87504.86 [13814.64]; P<0.01) and IL‐1β (CIA: 995.06 [550.03], CON: 674.76 [110.25]; P<0.05) levels in the CIA rats were significantly elevated versus the CON rats (Figure 5). However, the TNF‐α (CIA‐PRO: 112688.11 [30150.53]; P≤0.05) and IL‐1β (CIA‐PRO: 659.11 [76.66], P<0.05) levels in CIA‐PRO rats were significantly reduced when compared with CIA rats (Figure 6).

Figure 6

The measurement of inflammatory cytokines in cardiac tissue.

Relative expression of TNF‐α (A) and IL‐1β (B) in the left ventricle (n=6). There are significant differences among the 3 groups in the expression of TNF‐α (P=0.003) and IL‐1β (P=0.03). Data are shown as the median with 25th to 75th interquartile ranges. *P<0.05, and **P<0.01 vs CON group, † P<0.05 vs the CIA group. CIA indicates collagen‐induced arthritis group; CIA‐PRO, CIA rats treated with propranolol; CON, control group; IL‐1β, interleukin‐1 beta; and TNF‐α, tumor necrosis factor‐alpha.

Infiltration With Inflammatory Cells and Cardiac Fibrosis

Because arthritis triggers a systemic inflammatory response, we examined the number of CD45‐positive cells in the myocardium of the CIA rats and the CON rats. As expected, arthritis induced a prominent and long‐lasting myocardial infiltration of inflammatory cells (CIA: 0.15 [0.08], CON: 0.03 [0.04]; P<0.001) in the CIA rats, but this was significantly reduced in the CIA‐PRO rats (CIA‐PRO: 0.08 [0.04], P<0.001; Figure 7A). Immunohistochemical staining showed that sympathetic nerves, immunopositive to TH (the sympathetic nerve marker), were more abundant in rats with CIA than in the controls (CIA: 0.06 [0.07], CON: 0.01 [0.02]; P<0.001). After the propranolol treatment, the severity of sympathetic hyperinnervation was alleviated in CIA‐PRO rats (CIA‐PRO: 0.08 [0.04], P<0.01; Figure 7B). To confirm the specific expression of the sympathetic nerve, cardiac tissues were double‐stained with TH and PGP9.5 (protein gene product 9.5), which is a known neuronal marker. We found that the sympathetic neurons were overactive in CIA rats and that the administration of propranolol was able to reduce the hyperinnervation (Figure S1). We further examined the level of ventricular fibrosis among the 3 groups, and found increased fibrosis in the CIA group (CIA: 4.61 [1.95], CON: 0.54 [0.19]; P<0.001), but the fibrosis dramatically decreased after administration of propranolol (CIA‐PRO: 0.76 [0.12], P<0.001; Figure 7C).

Figure 7

Effect of collagen‐induced arthritis on ventricular fibrosis and sympathetic nerve density and infiltration of lymphocytes.

A, Examples of left ventricle tissue from the 3 groups of rats stained with anti‐CD45 antibody. There is increased infiltration of CD45‐positive lymphocytes in CIA rats compared with CON and CIA‐PRO rats (n=6). B, Example of ventricular immunostaining of tyrosine hydroxylase‐positive nerves. Compared with CON rats, there is increased sympathetic nerve density in CIA and CIA‐PRO rats. Of note, there was a reduced proportion of tyrosine hydroxylase‐positive nerves in the CIA‐PRO rats vs the CIA rats (n=6). C, Examples of fibrosis percentage in the 3 groups using Masson's trichrome staining. Increased fibrosis of CIA rats was noted (n=6). Overall, the 3 groups have significant differences (P<0.001). Data are shown as the median with 25th to 75th interquartile ranges. ***P<0.001 vs CON group, †† P<0.01 and ††† P<0.001 vs CIA group. CIA indicates collagen‐induced arthritis group; CIA‐PRO, CIA rats treated with propranolol; and CON, control group.

Discussion

Our study is the first to demonstrate the interaction between autonomic function and arrhythmogenic cardiac electrophysiology during highly inflammatory stages in RA. CIA rats exhibited severe arthritis, sympathetic hyperinnervation, and increased ventricular arrhythmogenicity, which was characterized by prolonged APD, slow CV, and steepened APD restitution. As a result, the vulnerability to VA increased in CIA rats. The possible mechanism underlying the proarrhythmic electrical remodeling may be explained by the sympathetic hyperinnervation, increased infiltration of inflammatory cells, elevated inflammatory cytokines, and a higher proportion of fibrosis in the LV of rat hearts with CIA. On the other hand, we observed that the levels of TNF‐α and IL‐1β in the LV were reduced, and the SVI imbalance was attenuated in CIA‐PRO rats. These findings suggest that there might be synergic anticatecholaminergic and anti‐inflammatory effects of β blockers, which protect the heart from VA in rats with CIA. To our knowledge, this study is the first to demonstrate the impact of a neuro‐immune interaction in cardiac tissue on ventricular arrhythmogenesis and the protective effect of β blockers in RA.

Inflammation and Ventricular Arrhythmia in Rheumatoid Arthritis

Although the underlying mechanisms accounting for the proarrhythmogenic substrate in RA are intricate, a leading role seems to be played by chronic systemic inflammatory activation. Many basic studies have demonstrated significant direct effects of inflammatory cytokines on cardiac electrophysiology. Previous studies have also shown that increased TNF‐α could decrease the transient outward current and delayed rectifier potassium channel, resulting in prolonged APD. , Perfused hearts from transgenic mice overexpressing TNF‐α exhibited a prolonged APD and re‐entrant VA. In addition, human recombinant IL‐1 prolonged APD by enhancing an L‐type calcium current in guinea pig ventricular cells. Inflammasome activation in cardiac macrophages mediates the production of IL‐1β in mice with diabetes mellitus. IL‐1β causes the prolongation of APD, and induces a decrease in potassium current and an increase in calcium sparks in cardiomyocytes, which are changes that underlie diabetes mellitus‐induced VA. The clinical evidence that inflammatory cytokine levels correlate with the QTc interval in patients with RA strongly indicates that in vivo these mechanisms are of crucial importance. Systemic inflammation could also indirectly produce proarrhythmogenic changes in RA by inducing autonomic dysfunction. Many basic and clinical studies have demonstrated that inflammatory cytokines increase the sympathetic outflow by targeting the autonomic centers of the brain. , Local inflammation is also detected by vagal and sensory nerve fibers with express receptors for inflammatory mediators like IL‐1β. An afferent signal is generated and transmitted to the brain, which in turn leads to activation of the sympathetic nervous system. On the other hand, local inflammation in the joint or heart could occur also as a consequence of activation of the central sympathetic nervous system through a proinflammatory influence on adaptive immune cells, for example, the recruitment of leukocytes and the induction of proinflammatory cytokines. Overactivation of the β‐adrenergic system results in the induction of an inflammatory cascade and eventually activates production of IL‐1β and IL‐18.

Cardiac fibrosis is an important structural remodeler in RA. It contributes to arrhythmogenesis by altering source‐sink relationships to create a vulnerable substrate, while simultaneously facilitating the emergence of triggers such as depolarization‐induced premature ventricular complexes, both factors combining synergistically to promote initiation of reentrant ventricular tachycardia. A network of proinflammatory cytokines (such as TNF‐α and IL‐1β) and cardiac fibroblasts is responsible for initiating and maintaining the fibrotic process. Meanwhile, β‐adrenergic‐receptor overexpression and an increased level of β‐adrenergic‐receptor agonists, including epinephrine and norepinephrine, could produce cardiac hypertrophy and fibrosis in vivo. Cardiac fibroblasts also have adrenergic receptors, and stimulation of the β2‐adrenergic receptor leads to increased proliferation of human and rodent cardiac fibroblasts.

Our study provides evidence linking myocardial inflammation and VA in RA. We observed significant sympathetic hyperactivity, prolonged APD, decreased CV, and altered APD restitution with a steepened maximal slope in CIA rats as compared with CON rats. A steepened APD restitution resulted in increased APD alternans, promoting breakup of the electrical wave into fibrillation‐like states and arrhythmogenesis. As a result, the vulnerability to VA increased in CIA rats. This proarrhythmic remodeling was associated with sympathetic hyperinnervation, increased infiltration of inflammatory cells, elevated inflammatory cytokines, and a higher proportion of fibrosis in the LV of the hearts of rats with CIA.

Anti‐Inflammation and Antiarrhythmic Effects of β Blockers

A previous clinical study has shown that treatment with nonselective β blockers is associated with reduced severity of systemic inflammation and improved survival of patients with acute‐on‐chronic liver failure. In a canine model of ischemic cardiomyopathy, the anti‐inflammation effect of a β blocker was shown by decreased levels of myocardial proinflammatory cytokine IL‐1β, increased myocardial levels of the anti‐inflammatory cytokine IL‐10, and less myocardial oxidative stress, leukocytosis, and fibrosis. Local sympathetic denervation by bilateral cervical sympathetic ganglionectomy attenuated myocardial inflammation in mice with myocardial infarction. Reduced inflammatory cell infiltration of the LV and reduced levels of proinflammatory cytokine TNF‐α coinciding with increased levels of anti‐inflammatory molecule IL‐10 upon sympathetic denervation were found. Our results are consistent with prior studies showing that the levels of TNF‐α and IL‐1β in the LV were reduced, and myocardial infiltration of inflammatory cell and cardiac fibrosis were decreased after propranolol treatment in rats of the CIA‐PRO group. There is a reciprocal relationship between inflammation and increased sympathetic activity, with inflammation predisposing to increases in sympathetic activity and sympathetic overactivity worsening inflammation. Treatment with β‐blocker propranolol could block this vicious cycle at both the central sympathetic nerve system and the local cardiac sympathetic innervation level. We observed CIA rats treated with propranolol had attenuated sympathetic overactivity and similar IHR as the controls. The reduced level of TH in the cardiac tissue of CIA‐PRO suggested less local distribution of the sympathetic nerve compared with the CIA rats. Thus, propranolol decreased sympathetic tone and attenuated local inflammation in CIA rats. In addition to the attenuation of joint swelling of CIA rats after treatment with propranolol, the proarrhythmic electrical remodeling in CIA rats was prevented or reversed by the anti‐inflammatory effects of the β blocker, resulting in less vulnerability to VA in CIA‐PRO rats.

The proportion of cardiac fibrosis was also improved by the administration of propranolol in CIA rats. Beta blockers have been demonstrated to prevent cardiac fibrosis and improve LV function in a hypertensive rat model and canine model of ischemic cardiomyopathy. , The plausible mechanisms underlying the reduced cardiac fibrosis after treated by propranolol in CIA rats is that a β blocker could prevent cardiac fibrosis by directly inhibiting the β‐adrenergic receptor in cardiac fibroblasts and indirectly attenuating the effects of inflammatory cytokines induced by sympathetic overactivity in fibrosis.

Conclusions

We found that RA increased vulnerability to VA by a broad range of ventricular remodeling, including sympathetic hyperinnervation, prolonged APD, slow CV, steepened slope of APD restitution, and increased cardiac fibrosis; thus providing further insight into the mechanisms underlying clinical RA‐induced VA and sudden death. Furthermore, our results show that the synergistic anticatecholaminergic and anti‐inflammatory effects of β blockers could reverse proarrhythmic ventricular remodeling and reduce vulnerability to VA in a rat model of RA.

Sources of Funding

This work was supported by the Science and Technology Unit, Ministry of Health and Welfare, Executive Yuan, Taiwan (grant nos. 106‐2314‐B‐002‐046‐MY2 and 105‐2314‐B‐002‐072); the National Taiwan University Hospital Hsinchu Branch (grant nos. 107‐HCH010, 105‐HCH065, and HCH104‐073); and the National Chiao Tung University (grant nos. 104W970 and 105W970).

Disclosures

None.

Table S1

Figure S1

Click here for additional data file.