Chamber-specific differences in human cardiac fibroblast proliferation and responsiveness toward simvastatin

Abstract

Fibroblasts, the most abundant cells in the heart, contribute to cardiac fibrosis, the substrate for the development of arrythmogenesis, and therefore are potential targets for preventing arrhythmic cardiac remodeling. A chamber-specific difference in the responsiveness of fibroblasts from the atria and ventricles toward cytokine and growth factors has been described in animal models, but it is unclear whether similar differences exist in human cardiac fibroblasts (HCFs) and whether drugs affect their proliferation differentially. Using cardiac fibroblasts from humans, differences between atrial and ventricular fibroblasts in serum-induced proliferation, DNA synthesis, cell cycle progression, cyclin gene expression, and their inhibition by simvastatin were determined. The serum-induced proliferation rate of human atrial fibroblasts was more than threefold greater than ventricular fibroblasts with faster DNA synthesis and higher mRNA levels of cyclin genes. Simvastatin predominantly decreased the rate of proliferation of atrial fibroblasts, with inhibition of cell cycle progression and an increase in the G0/G1 phase in atrial fibroblasts with a higher sensitivity toward inhibition compared with ventricular fibroblasts. The DNA synthesis and mRNA levels of cyclin A, D, and E were significantly reduced by simvastatin in atrial but not in ventricular fibroblasts. The inhibitory effect of simvastatin on atrial fibroblasts was abrogated by mevalonic acid (500 μM) that bypasses 3-hydroxy-3-methylglutaryl coenzyme A (HMG-CoA) reductase inhibition. Chamber-specific differences exist in the human heart because atrial fibroblasts have a higher proliferative capacity and are more sensitive to simvastatin-mediated inhibition through HMG-CoA reductase pathway. This mechanism may be useful in selectively preventing excessive atrial fibrosis without inhibiting adaptive ventricular remodeling during cardiac injury.

cardiac fibroblasts are connective tissue cells of mesenchymal origin that play a variety of roles in the heart, including cardiac development, defining structure and function, wound healing, and extracellular matrix (ECM) remodeling after an injury that affects cardiac mechanical and electrical properties (1, 8, 10, 38). Following neonatal development, fibroblasts maintain low levels of cell division; however, upon physiological or pathological stimulation during normal development, aging, or disease processes, and under the influence of cytokines and growth factors, fibroblasts proliferate and are activated to myofibroblasts that regulate myocardial remodeling, promoting cardiac repair after injury (8, 23, 30, 35, 44). However, excessive fibrosis is detrimental, leading to the development of a substrate that increases predisposition to electrical instability and reentrant arrhythmias, such as atrial fibrillation (AF) (30, 43, 49). Therefore, limiting excessive fibrosis by controlling fibroblast proliferation and activation in the atria may be beneficial in reducing the burden of this common arrhythmia associated with aging and aging-related diseases (45). On the other hand, fibroblasts are essential for wound repair, especially after an injury such as acute myocardial infarction, and inhibition of fibroblast proliferation could be detrimental, limiting the wound repair process in the ventricle (33). Therefore, drugs that can have a chamber-selective effect, preventing fibroblast proliferation in the atria without significantly altering the ventricular fibroblast response, are desirable. Recent studies pointed toward differential responsiveness of fibroblasts from the atria and ventricles to cytokines and growth factors in animal models, but similar information in humans is not available; it is unclear whether the fibroblasts from atria and ventricles respond differently to drugs (6, 49). Statins or 3-hydroxy-3-methylglutaryl coenzyme A (HMG-CoA) reductase inhibitors are cholesterol-lowering drugs widely used for the prevention of ischemic heart disease. They also exert pleiotropic effects against maladaptive tissue remodeling (15, 31, 36, 47) with the potential to be repurposed as a drug to prevent cardiac hypertrophy and fibrosis (13, 28, 30, 31, 51). Animal and human studies have demonstrated statins to be effective in preventing adverse myocardial remodeling (16, 30, 36, 39, 50, 51) by affecting atrial fibroblasts and in reducing the burden of AF in patients with coronary artery disease and heart failure. However, it is unknown whether differences exist in the responsiveness of atrial and ventricular fibroblasts from the human heart to statins (16, 47). Therefore, the present study sought to define the baseline differences in the proliferation of fibroblasts from the atria and ventricles of the human heart, determine their responsiveness to statins, and identify potential targets responsible for these differences.

MATERIALS AND METHODS

All protocols in this study were performed in compliance with institutional guidelines for human research and approved by the local institutional review board.

Materials.

All chemicals were purchased from Sigma-Aldrich Chemicals (St. Louis, MO) unless otherwise indicated. Human atrial and ventricular fibroblasts were acquired from ScienCell Research Laboratories (Carlsbad, CA), Lonza (Walkersville, MD), or Cell Applications (San Diego, CA) from trauma victims who were free of cardiac pathology. Cell culture medium, fetal bovine serum (FBS), and freezing medium (cat. no. 040-50) were from Cell Applications. 5-Ethynyl-2′-deoxyuridine (EdU), TRIzol, Click-iT EdU Alexa Fluor 647 Imaging Kit, LIVE/DEAD Viability, Cytotoxicity Kit, NuPAGE Novex mini-gels, iBlot dry blotting system, and SYBR Green PCR Master Mix all came from Life Technologies (Grand Island, NY). Cyclin D1 (cat. no. sc-450) and anti-mouse immunoglobulin G (IgG) antibodies were from Santa Cruz Biotechnology (Dallas, TX). Anti-α/β-tubulin (cat. no. 2148S) antibody was acquired from Cell Signaling Technology (Danvers, MA), and goat anti-rabbit IgG (cat. no. ab6721) was from Abcam (Cambridge, MA). Clarity Western enhanced chemiluminescence (ECL) substrate came from Bio-Rad Laboratories (Hercules, CA). DNase I, miScript RT II, and RNeasy Mini Kit were manufactured by Qiagen NV (Venlo, The Netherlands). Primers were synthesized by OriGene Technologies (Rockville, MD). Muse cell cycle assay kit was acquired from EMD Millipore (Darmstadt, Germany).

Propagation and storage of human atrial and ventricular fibroblasts.

Experiments were performed on commercially available, disease-free human atrial (4 individuals' samples, in triplicate) and ventricular fibroblasts (3 individuals' samples, in triplicate). Cells were cultured to 50–60% confluence at 37°C in 95% air-5% CO₂ and stored in Cell Applications freezing media, composed of 10% FBS and 10% dimethyl sulfoxide in liquid nitrogen. All atrial and ventricular fibroblasts were used for experiments at passage 3. Cells were plated at the initial density of 2,000 (6-well plates) to 20,000 cells/cm² [96-well and MatTek (MatTek, Ashland, MA) dishes] depending on well size, assay, and experimental conditions. Cells were allowed to attach overnight before treatment. To prevent the fibroblasts from spontaneously differentiating to myofibroblasts, studies were carried out in low-serum (1.25%, 2.50%, or 3.75% FBS) medium after initial plating in HCF media containing 5% FBS + penicillin-streptomycin or Primocin (InvivoGen, San Diego, CA) for 24 h.

Cell proliferation assay.

Human atrial or ventricular fibroblasts (10⁵) were plated in triplicate in a six-well culture plate. Twenty-four hours after seeding, the cells were trypsinized, counted, and grown in parallel sets of six-well plates. Cells were then exposed to low-serum (1.25%, 2.50%, and 3.75% FBS) growth medium with or without simvastatin (100 nM to 10 μM) or atorvastatin (300 nM) for 24, 48, and 72 h. Cells from triplicate wells were counted using a hemocytometer, population doubling was calculated (18), and the number of cells at each time point was expressed as mean ± SE from three or more experiments.

DNA synthesis by EdU incorporation, staining, and imaging.

Atrial and ventricular fibroblast proliferation was also evaluated by EdU incorporation, a nucleoside analog of thymidine incorporated into DNA during active DNA synthesis for measuring cell proliferation (34). Cells were plated at 2,000 cells/cm² in glass-bottom MatTek dishes coated with rat tail type I collagen as previously described (27). After adhesion and quiescence, cells were treated for 48 h with or without simvastatin (1 μM) in the absence or presence of mevalonic acid (MVA, 500 μM). The cells were then fixed (3.7% paraformaldehyde) and permeabilized (0.5% Triton-X 100), and EdU incorporation was determined using the Click-iT EdU Alexa Fluor 647 Imaging Kit according to the manufacturer's protocol. Fluoroshield with 4′,6-diamidino-2-phenylindole (DAPI)-mounted fluorescent images were acquired using an Olympus FV1200 MPE laser confocal microscope (DAPI, excitation 405 nm/emission 470 nm; AlexaFluor EdU, excitation 635 nm/emission 647 nm). Three to five randomly selected fields were analyzed using National Institutes of Health ImageJ software for determining the number of EdU- and/or DAPI-positive cell nuclei. The ratio of EdU- to DAPI-positive nuclei was used as quantification of cell proliferation.

Lactate dehydrogenase assay.

Dose-dependent simvastatin-induced cytotoxicity was measured by cytoplasmic lactate dehydrogenase (LDH) release. Atrial and ventricular fibroblasts plated at 3 × 10⁴ cells/cm² into 96-well culture plates and incubated overnight were exposed to HCF media with 2.5% FBS and different doses of simvastatin (0–20 μM) in the presence or absence of MVA (500 μM) for 48 h. An equal amount of supernatant from each well was collected for measuring LDH release. Fibroblasts adhering to the plate were lysed to release the entire pool of intracellular LDH, and the cell lysates were collected for measuring intracellular LDH levels. LDH activity was measured as an increase in NADH fluorescence (excitation 360 nm/emission 460 nm) in solution containing 0.5 M glycine, 0.4 M hydrazine, 200.0 μM lactate, and 1.0 mM NAD⁺. LDH activity was expressed as relative fluorescence units, and cell cytotoxicity was assessed by calculating the percentage of LDH levels in media to the intracellular LDH levels in lysate.

Cell viability assay.

To monitor simvastatin-induced cytotoxicity, human atrial and ventricular fibroblasts were incubated for 48 h in the presence or absence of 1 μM simvastatin. Cells were treated for 1 h in the dark with calcein-AM (4 mM) and ethidium homodimer-1 (Eth-D1, 2 mM) prepared in 1× HBSS with Ca²⁺ and Mg²⁺. Live/dead cells were measured using the LIVE/DEAD Viability and Cytotoxicity Kit according to the manufacturer's instructions.

Western blot analysis.

Briefly, 30 μg of protein of cell lysates/lane were separated by NuPAGE Novex 4–12% mini-gels and transferred onto polyvinylidene fluoride membranes using the iBlot dry blotting system. The membranes were then blocked for 1 h in Tris-buffered saline with Tween 20 (TBS-T) containing 3% nonfat dry milk at room temperature. Furthermore, membranes were incubated with the mixture of primary antibodies against cyclin D1 (1:200 dilution) and α/β-tubulin (1:1,000 dilution). Following incubation with primary antibodies, the membranes were washed with TBS-T before incubation for 1 h at room temperature with anti-mouse IgG and/or goat anti-rabbit IgG secondary antibodies. The membranes were then washed, and the signals were visualized using clarity Western blot ECL substrate and an Amersham Imager 600RGB (GE Healthcare Life Sciences, Marlborough, MA). The density of the bands was analyzed using ImageQuant TL (GE Healthcare Life Sciences) and/or ImageJ software, and bands of protein of interest were normalized to the density of the respective α/β-tubulin bands.

qRT-PCR assay.

Total RNA was prepared from cells using TRIzol followed by DNA digestion with DNase I and further purified by RNeasy Mini Kit. RNA concentrations were determined using the Infinite 200 NanoQuant (Tecan Group, Männedorf, Switzerland). Equal amounts of RNA from each sample were then reverse transcribed to cDNA by using an miScript RT II kit. qPCR was performed in an ABI 7300 RT-PCR system (Thermo Fisher Scientific, Waltham, MA) using SYBR Green PCR Master Mix. The reactions were incubated at 50°C for 2 min, 95°C for 10 min, 40 cycles at 95°C for 15 s, annealing for 1 min at 60°C, then extension at 72°C for 40 s. Gene-specific primer sequences used were as follows: for cyclin D1: sense, 5′-TCTACACCGACAACTCCATCCG-3′, antisense, 5′-TCTGGCATTTTGGAGAGGAAGTG-3′; for cyclin D2: sense, 5′-GAGAAGCTGTCTCTGATCCGCA-3′, antisense, 5′-CTTCCAGTTGCGAT CATCGACG-3′; for cyclin D3: sense, 5′-AGATCA AGCCGCACATGCGGAA-3′, antisense, 5′-ACGC AAGACAGGTAGCGATCCA-3′; for cyclin A: sense, 5′-CTCTACACAGTCACGGGACA AAG-3′, antisense, 5′-CTGTGGTGCTTTGAGGTAGGTC; for cyclin E: sense, 5′-TGTGTCCT GGATGTTGACTGCC-3′, antisense, 5′-CTCTATGTC GCACCACTGATACC-3′; and for 18S: sense, 5′-ACCCGTTGAACCCCATTCGTGA-3′, antisense, 5′-GCCTCACTAAACCATCCAATC GG-3′. Melt curve analysis was performed by an additional dissociation step of 1 cycle at 95°C for 15 s and ramping data collection at 60°C for 1 min and 95°C for 15 s. Cycle threshold (Ct) values were calculated using application binary interface software. Data were normalized against the 18S signal (internal control). Relative expression values were obtained by normalizing Ct values of the tested genes with Ct values of the 18S genes, using the ΔΔCt method.

Cell cycle progression and distribution assay.

Cell cycle progression and distribution were evaluated by using Muse cell cycle assay kit and Muse cell analyzer per the kit guidelines. Briefly, human atrial or ventricular fibroblasts at a density of 5,000 cells/cm² were seeded in six-well plates and grown for 24 h in HCF media with 5% FBS before being quiescent in serum-deprived medium for another 24 h. Cells were restimulated by HCF media containing 2.5% FBS with/without 1.0 μM simvastatin and/or 500.00 μM MVA for 48 h. Cells were washed in Dulbecco's phosphate-buffered saline (DPBS), detached, counted, and pelleted. Equal numbers of cells were resuspended in DPBS, centrifuged at 300 g for 5 min, and permeabilized/fixed in ice-cold 70% ethanol. Samples were incubated/stored at −20°C for at least 3 h. Cells were pelleted, resuspended, and incubated (room temperature/dark) in 200 μl of Muse cell cycle kit staining reagent containing propidium iodide (PI) at ≈1 × 10⁶ cells/ml. PI fluorescence was acquired with Muse cell analyzer. DNA histogram results were obtained as the percentage of the cell population in G0/G1, S, and G2/M phases of cell cycle for each treatment.

Statistical analysis.

Data are expressed as means ± SE, and n represents the number of repeats. Comparisons between groups were made using the Student's t-test or ANOVA where deemed necessary, and P < 0.05 was considered statistically significant.

RESULTS

Basal proliferation of atrial and ventricular fibroblasts and response to simvastatin.

The proliferation rate of atrial fibroblasts was consistently higher than that of ventricular fibroblasts cultured for the same time period; by 72 h, atrial fibroblasts outnumbered ventricular fibroblasts by more than twofold (P < 0.05) in 2.5% FBS (Fig. 1A). The population doubling measured as the rate of proliferation of atrial fibroblasts was 26.1 ± 0.8 h compared with 47.8 ± 0.14 h for ventricular fibroblasts, a 1.8-fold difference (Fig. 1A, P < 0.001). Treatment with simvastatin (1 μM) inhibited proliferation of both atrial and ventricular fibroblasts, but the sensitivity to growth inhibition was higher in atrial fibroblasts than ventricular fibroblasts (Fig. 1A). The magnitude of slope (rate) for proliferation from 24 to 48 h was reduced by 40.0 ± 7.9% in atrial fibroblasts vs. a reduction of only 9.7 ± 5.6% in ventricular fibroblasts. The proliferative response in ventricular fibroblasts in 3.75% serum (4.83 ± 0.73-fold) was almost similar to what was observed in atrial fibroblasts in 1.25% serum (4.41 ± 0.21-fold), indicating a greater sensitivity of atrial fibroblasts to serum-induced proliferation. The differential responsiveness of atrial and ventricular fibroblast proliferation and inhibition by simvastatin at 72 h was confirmed at varying levels of serum (Fig. 1B), demonstrating a higher sensitivity of atrial fibroblasts toward simvastatin compared with ventricular fibroblasts (41.60 vs. 11.30% inhibition in 1.25% FBS, P < 0.01; 54.40 vs. 18.20% inhibition in 2.50% FBS, P < 0.05, and 46.48 vs. 6.00% inhibition in 3.75% FBS; P < 0.01 Fig. 1B). To quantify differences in the sensitivity of atrial and ventricular fibroblasts to the inhibitory effect of simvastatin, fibroblasts were exposed to an increasing concentration of simvastatin (100 nM to 10 μM). We found a significant decrease in proliferation of ventricular fibroblasts with increasing concentration of simvastatin. The inhibition in the proliferation of ventricular fibroblasts compared with atrial fibroblasts at 300 nM simvastatin (0 vs. 54%) was increased (83 vs. 95%, P = 0.0046) at 10 μM simvastatin concentration (Fig. 1C). The difference in inhibition of atrial and ventricular fibroblast proliferation also was confirmed by atorvastatin, a second HMG-CoA reductase inhibitor (Fig. 1D).

Fig. 1.

Phase-contrast images of atrial and ventricular fibroblasts (FB) cultured for 72 h in the absence or presence of 1 μM simvastatin (A, left). The indicated cells were counted for proliferation at 0, 24, 48, and 72 h using a hematocytometer. Cell proliferation was inhibited more in atrial than ventricular fibroblasts after 72 h of exposure to simvastatin (A, right). Results are displayed as means ± SE from 3 or more independent experiments conducted in triplicate (*P < 0.05). B: inhibition of cell proliferation was quantified by determining the differences between atrial and ventricular fibroblasts for the control group and simvastatin group in different percentages of FBS concentrations (1.25–3.75%) at 72 h. **P < 0.01. C: dose-dependent response of atrial (dashed line) and ventricular fibroblast (solid line) proliferation to simvastatin. Cell count in 100 nM simvastatin concentration was considered 100% proliferation. Cells were grown in cardiac fibroblast growth media with 2.5% FBS for 72 h. ANOVA involving blocking factors using the Friedman test was used (P = 0.0046). The data represent the means ± SE (n = 3). D: inhibition in the proliferation of atrial and ventricular fibroblasts with treatment of 300 nM atorvastatin for 72 h. Ator, atorvastatin; Sim, simvastatin.

Effects of simvastatin on cardiac fibroblast cytotoxicity.

The decrease in the number of fibroblasts with simvastatin treatment could be due to either an inhibition of cellular proliferation or enhanced cell death. To assess whether simvastatin was cytotoxic, atrial and ventricular fibroblasts were exposed to increasing doses of simvastatin (0–20 μM), and cell viability was determined by Calcein-AM/EthD-I assay and release of LDH. The proportion of EthD-1-positive cells indicating cell death in atrial (Fig. 2, A and B) and ventricular (Fig. 2, E and F) fibroblasts not treated with simvastatin was not significantly different than in cells treated with simvastatin for 48 h. Overall, 7.70 and 8.96% (P = 0.81) of atrial fibroblasts were positive for EthD-1 staining in simvastatin-untreated and simvastatin-treated conditions, respectively (Fig. 2I), whereas only 3.25 and 7.56% of ventricular fibroblasts were EthD-1 positive. Similarly, LDH release after 24 h was not statistically different between atrial and ventricular fibroblasts with or without simvastatin treatment (Fig. 2, J and K). Thus the reduction in the number of fibroblasts with simvastatin treatment was not due to cytotoxicity but to inhibition of cardiac fibroblast proliferation.

Fig. 2.

Cardiac fibroblasts were grown to confluence in media containing 2.5% FBS and Primocin. The cells were treated with or without simvastatin (1 μM) for 48 h and assessed for cytotoxicity using cell viability assay by staining the cells with calcein AM and ethidium homodimer (EthD-1) for 30 min at 37°C (A–H). Cytotoxicity was determined by fluorescent microscopy, and cell counts of live and dead cells were expressed as a percentage of the total cells counted. Calcein AM-positive or EthD-1-positive cells in atrial fibroblasts without (A and B) or with 1 μM simvastatin (C and D), and in ventricular fibroblasts without simvastatin (E and F) and with 1 μM simvastatin (G and H) are shown, respectively. I: percentage of calcein AM-positive and EthD-1-positive cells (n = 3, means ± SE; P = NS, not significant). Assessment of cytotoxicity also was determined by lactate dehydrogenase (LDH) release from atrial (J) and ventricular (K) fibroblasts after 24-h exposure to different doses of simvastatin (0–20 μM). Vent, ventricular.

Simvastatin differentially inhibits cardiac fibroblast proliferation and DNA synthesis in atrial vs. ventricular fibroblasts.

Cardiac fibroblasts grown in FBS-containing HCF media demonstrated increased progression through the cell cycle (G0/G1, S, and G2/M phases) as shown by flow cytometry analysis of PI-stained atrial fibroblasts (Fig. 3A). However, this was not seen in ventricular fibroblasts, as the majority of ventricular fibroblasts were restricted to G0/G1 phase with a relatively lower proportion in S and G2/M phases of the cell cycle (Fig. 3B), which correlates with their slower proliferation rates (Fig. 1, I and J). In atrial fibroblasts, treatment with 1 μM simvastatin increased the proportion of cells in G0/G1 phase from an average of 54 to 76% (P < 0.01), decreasing cells in S phase from an average of 18 to 8% (P = 0.014) and cells in G2M phase from 25 to 16% (P < 0.05) (Fig. 3, A and C). In ventricular fibroblasts, simvastatin had no significant effect on the proportion of cells in G0/G1 phase (70 to 80%, P = 0.61), S phase (8.0 to 5.5%; P = 0.07), and G2M phase (16 to 11%; P = 0.11) (Fig. 3, B and D). The overall effects are summarized in Fig. 3, E and F, which demonstrates a statistically significant inhibitory effect of simvastatin on cell cycle progression and DNA synthesis in atrial fibroblasts compared with ventricular fibroblasts. This was further confirmed by the EdU incorporation assay demonstrating increased EdU incorporation into the newly synthesized DNA in the nuclei, as shown in Fig. 4, where active DNA synthesis is represented by red nuclei on confocal microscopy after 48 h of EdU incorporation. The number of EdU-positive nuclei at 48 h was significantly higher in atrial fibroblasts than ventricular fibroblasts (Fig. 4, A and C; P < 0.05). Treatment with simvastatin reduced nuclear EdU incorporation in atrial fibroblasts by 82% (from 85 to 13%; P < 0.01), indicating reduced DNA synthesis compared with 27% (from 65 to 47%, P = 0.04) in ventricular fibroblasts (Fig. 4E). Taken together, these results indicate that atrial fibroblasts have greater sensitivity to simvastatin-mediated inhibition of proliferation and DNA synthesis compared with ventricular fibroblasts.

Fig. 3.

1 × 10⁵ human cardiac fibroblasts were plated in 6-well plates with media containing 5% FBS and were serum starved for 24 h. A 2.5% FBS was reintroduced to the media with or without 1 μM of simvastatin. Cell cycle analysis was performed 48 h afterward. Following propidium iodide DNA staining, fibroblasts were analyzed using MUSE cell analyzer. The histograms show the cell cycle analysis of serum-induced atrial and ventricular fibroblast populations in the absence (A and B) and the presence of simvastatin (C and D). Circles represent the modulation of the cell cycle by simvastatin as determined by MUSE as a percentage of the total number of cells present in G0/G1, S, and G2/M phases. E and F: summary of data from A–D, showing that simvastatin induced cell cycle arrest (G0/G1) (E) and reduction in cell population distribution of S + G2M phase (F) in atrial vs. ventricular fibroblasts (n = 3, *P < 0.05 control vs. simvastatin in atrial fibroblast; **P < 0.01 control atrial vs. ventricular fibroblasts).

Fig. 4.

1 × 10⁴ cells were plated in collagen-coated MatTek dishes and grown in culture for 24 h in human cardiac fibroblast media containing 5% FBS. Cells were serum starved for another 24 h before being cultured in media containing 2.5% FBS + 5-ethynyl-2′-deoxyuridine (EdU) in the absence and the presence of 1 μM simvastatin for another 48 h. Cells were fixed, permeabilized, and stained with DAPI. Incorporation of EdU into nuclear DNA was assessed using confocal microscopy. Images of nuclei from atrial fibroblasts (A, control; B, simvastatin) and ventricular fibroblasts (C, control; D, simvastatin) showed reduced EdU incorporation on simvastatin treatment. E: summary of the percentage of EdU-positive cells (pink) to total nuclei (blue) per field in atrial vs. ventricular fibroblasts in the presence or absence of simvastatin, as determined by ImageJ particle analysis (n = 3, means ± SE; **P < 0.01).

Simvastatin inhibits expression of cell cycle-regulating genes.

To explore the molecular mechanisms underlying the differential responses of atrial and ventricular fibroblasts to serum and simvastatin, we examined the mRNA levels of cyclin A, D1, D2, D3, and E2 that regulate the cell cycle. First, serum deprivation for 24 h was used to synchronize the cell cycle followed by reintroduction of serum for 8 h to stimulate cell proliferation. Serum incubation increased the mRNA levels of all cyclins (cyclin A, D1, D2, D3, and E2) in the atrial fibroblasts by twofold or more compared with the levels in serum-free media. In ventricular fibroblasts, induction of cyclin mRNA by serum occurred at a much lower magnitude (Fig. 5, A–E). The difference in serum-mediated cyclin expression was greatest for cyclin D1 in atrial fibroblasts, with no change in its expression in ventricular fibroblasts. Simvastatin (1 μM) differentially inhibited serum-induced mRNA expression of the cyclin genes by a >50% reduction in atrial fibroblasts with no significant effect on cyclin expression in ventricular fibroblasts (Fig. 5. A–E). Because serum had the most prominent effect on cyclin D1 mRNA expression, the effect of simvastatin on cyclin D1 protein expression also was examined. Simvastatin (1 μM) inhibited cyclin D1 protein expression by 67% (P < 0.01) in atrial fibroblasts with no significant inhibition in ventricular fibroblasts (P = 0.106) (Fig. 5, B and C, P < 0.05). The changes in cyclin D1 protein levels in atrial fibroblasts with serum or simvastatin cotreatment (Fig. 5C) were similar to the changes in mRNA levels under similar experimental conditions (Fig. 5A), indicating transcriptional regulation of cyclin proteins. Thus we demonstrate a differential response by atrial and ventricular fibroblasts to serum-induced cyclin mRNA and protein expression and its inhibition by simvastatin.

Fig. 5.

A: analysis of cell cycle mRNA levels induced by 2.5% FBS for 8 h compared with no FBS shows that cyclin A, cyclin D1, cyclin D2, cyclin D3, and cyclin E were altered more in atrial fibroblasts than in ventricular fibroblasts on simvastatin treatment. Data are expressed as means ± SE, n = 3; *P < 0.05. B: immunoblot results from atrial and ventricular fibroblast cell lysates for cyclin D and α/β-tubulin. C: vertical bars showing the densitometric ratio of cyclin D1 and α/β-tubulin bands in atrial fibroblasts with or without simvastatin. *P < 0.05.

MVA abrogated simvastatin-induced changes in atrial fibroblasts.

MVA is the key metabolite in the mevalonate pathway for cholesterol synthesis, derived from catalysis of HMG-CoA by HMG-CoA reductase, the molecular target of statins (HMG-CoA reductase inhibitors). To test whether the inhibitory effect of simvastatin on serum-induced cardiac fibroblast proliferation is mediated by HMG-CoA reductase, atrial and ventricular fibroblasts were incubated with MVA and then treated with simvastatin. Pretreatment with MVA (500 μM) prevented the inhibitory effect of simvastatin (1 μM) on cell cycle progression by maintaining the G0/G1, S, and G2/M phases (Fig. 6A) similar to serum-treated values in the absence of simvastatin (Fig. 3). This was further confirmed by EdU assay for DNA synthesis (Fig. 6B) in atrial fibroblasts pretreated with MVA before simvastatin treatment. The inhibitory effect of simvastatin (1 μM) on DNA synthesis was prevented by MVA, maintaining the EdU incorporation to the levels similar to serum-treated cells in the absence of simvastatin (Fig. 2). In ventricular fibroblasts that exhibited minimum inhibition of cell progression into S phase or DNA synthesis (EdU incorporation) by simvastatin, MVA pretreatment minimally affected cell cycle progression or EdU incorporation (P = 0.6; Fig. 6B). Thus MVA abrogated the inhibitory effects of simvastatin on serum-treated cell cycle progression and DNA synthesis preferentially in atrial fibroblasts compared with ventricular fibroblasts.

Fig. 6.

Atrial and ventricular fibroblasts plated and grown for 24 h in human cardiac fibroblast media containing 5% FBS were serum starved for 24 h before 2.5% FBS was reintroduced in the absence or presence of 1 μM simvastatin and/or 500 μM mevalonic acid (MVA). Cell cycle was analyzed using MUSE cell analyzer, and EdU incorporation was assessed using confocal microscopy. A: simvastatin-induced cell cycle arrest was prevented by 500 μM MVA in atrial and ventricular fibroblasts. Data are presented (bar graph) as the average percentage of cells in S to the total cell population treated with simvastatin with or without MVA in the presence of 2.5% FBS. B: EdU incorporation to the nuclear DNA indicative of DNA synthesis in atrial and ventricular fibroblasts treated with simvastatin ± MVA. Data are summarized as the differences in ratio of EdU/DAPI-labeled nuclei (Δ change %) in atrial or ventricular fibroblasts treated with simvastatin with or without MVA in 2.5% FBS media for 48 h (n = 3, means ± SE; *P < 0.05, **P < 0.01).

DISCUSSION

In this study, we have demonstrated chamber-specific differences in HCFs obtained from atria and ventricles in terms of their rate of proliferation and pharmacological responsiveness toward simvastatin attributable to a differential effect on the expression of genes controlling cell cycle-regulating cyclin proteins. This inhibitory effect involved the HMG-CoA reductase pathway and could be prevented by MVA.

Progressive fibrosis associated with aging or aging-associated diseases contributes to mechanical and electrical cardiac dysfunction that predisposes to heart failure or arrhythmias such as AF in the elderly (5, 24, 30, 45). Different classes of medications, including antiarrhythmic agents, angiotensin-converting enzyme inhibitors, angiotensin (II) receptor blockers, and aldosterone receptor antagonists, have been used for management of patients with AF with variable effects on the signaling pathways involved in fibrosis and with limited success in preventing atrial fibrillation recurrence or progression (7, 12, 45). Because ECM deposition by fibroblasts and myofibroblasts plays an essential role in atrial disease progress (38), inhibition of fibroblast progression could be effective in limiting the substrate for AF (19). A drug that is selective for atrial fibroblasts is desirable, as it would specifically inhibit atrial fibrosis without altering the response of ventricular fibroblasts, which might be required for effective wound healing after a heart attack, another problem commonly encountered in the elderly (21, 41). Recent studies indicate that fibroblasts, not only differ between organs, but also display heterogeneity within single organs (10). Fibroblasts from the renal medulla differ from those of the renal cortex in the number of lipid droplets and magnitude of myofibroblast transformation in rats (17). Similarly, differences in fibroblast structure, function, and responsiveness to cytokines from skeletal muscle, skin, and lung also have been demonstrated (10, 20). The existence of chamber- and region-specific differences in the heart, with variation in cellular origin, morphology, size, secretion, proliferation, and expression of α-smooth muscle actin in response to growth factors and cytokines, previously has been reported for atrial and ventricular fibroblasts in animal models (2, 6, 25, 49). Our study extends these findings to the human heart, demonstrating a more rapid proliferation of atrial fibroblasts with a faster DNA synthesis rate compared with ventricular fibroblasts and that these differences were related to changes in the expression of genes encoding for cyclin proteins that regulate cell cycle progression (22). This, to our knowledge, is the first comparison of fibroblast function from atrial and ventricular walls of the human heart, and the additional demonstration of differences in the pharmacological responsiveness of atrial and ventricular fibroblasts toward statin-mediated inhibition of their proliferation also is innovative.

Chamber-specific differences in fibroblast proliferation and responsiveness to myocardial injury or profibrotic stimuli can result in differences in fibrotic response (46) and, therefore, susceptibility to excessive fibrosis (49). In this regard, atrial fibroblasts from rat hearts were shown to be more responsive to TGF-β1 compared with ventricular fibroblasts, resulting in greater oxidative stress and fibrosis through the SMAD-dependent NOX4 pathway (49). Similar differences in fibroblasts have been reported in canine hearts (6). The differences in the responsiveness of atrial and ventricular fibroblasts to serum-induced proliferation in our study extend these findings to the human heart. The precise mechanisms underlying these differences are not fully defined but could be related to a differential response to growth factors or cytokines present during various pathological states affecting fibroblast proliferation, mobility, activation, and ECM deposition through differential receptor-dependent pathways (6, 43). In addition, chamber-specific differences in hemodynamics, morphology, wall stress, energy, and metabolic demands could condition a differential gene expression response or intracellular pathway activation via responsiveness to various growth factors or cytokines produced during myocardial injury or hemodynamic stress (3, 4, 6, 9, 11, 14, 26, 37, 40).

Fibroblast proliferation is an essential component of cardiac fibrosis and remodeling as part of the normal repair mechanism after injury, but it is also a contributor to pathological structural alterations that predispose to impaired cardiac filling, contractility, and electrical conduction (38, 42). Therefore, modulating fibroblast proliferation is important to preventing excessive fibrosis and the development or progression of heart failure or cardiac arrhythmias associated with aging or aging-associated diseases. Previously, statins were shown to have an inhibitory effect on proliferation, activation, and ECM synthesis by cardiac fibroblasts (29); however, differential effects of statins on atrial and ventricular fibroblasts from the human heart were not defined. Our study fills this gap in knowledge, demonstrating that atrial and ventricular fibroblasts have a differential responsiveness to statin-mediated inhibition of proliferation that is not due to cytotoxicity but to a cell cycle-regulating mechanism involving transcriptional regulation of cyclins through HMG-CoA reductase-mediated pathway. This inhibition could be prevented by pretreatment with MVA, which bypasses the inhibitory effect of simvastatin on HMG-CoA reductase. The downstream mechanism of the effects of simvastatin in fibroblasts and other cells has been described as involving the geranylgeranyl pyrophosphate pathway by targeting small guanosine-5′-triphosphate (GTP)-binding proteins, such as RhoA to regulate cyclins and/or cyclin-dependent kinase inhibitors independent of the squalene pathway for cholesterol synthesis (31, 32, 48). Simvastatin inhibited proliferation of both atrial and ventricular fibroblasts but with different sensitivity. Cell viability assay and LDH release showed that fibroblast proliferation did not decrease because of cytotoxicity caused by the simvastatin concentration. They demonstrated no incremental effect on cell death beyond the normal cell turnover observed in the absence of the studied drug, which is in agreement with earlier observations (31) but differs from a later study (16) that reported apoptotic cell death in human atrial fibroblasts. The difference in findings could be due to a fourfold lower concentration of simvastatin used in our study and the Porter et al. study (31) compared with the concentration used by Ghavami et al. (16) demonstrating toxic effects. The clinical relevance of the higher concentration of statins used by Ghavami et al. is not clear, and, given the potential toxicity, the U.S. Food and Drug Administration has recommended against the use of higher concentrations for clinical use (http://www.fda.gov/ForConsumers/ConsumerUpdates/ucm293330.htm).

The inhibitory action of both simvastatin and atorvastatin on fibroblast proliferation confirms the pleiotropic property that inhibiting fibroblast proliferation is a class effect of statins rather than simvastatin-specific effect. The differential effect of statins on fibroblast proliferation was determined in in vitro culture conditions, which could be considered a limitation, as the fibroblasts were grown independent of the modulating effects of stretch, hemodynamics, cytokines, and growth factors that may influence the fibroblasts to respond differently in the atria and the ventricles. However, both atrial and ventricular fibroblast responses to serum in the absence or presence of simvastatin were studied in identical conditions and indicated a distinct response to proliferation and inhibition by statins, which cannot be studied in vivo in human hearts. The consistency of the findings of this study in human-derived fibroblasts with other studies in animal models (6, 49) adds to the growing literature highlighting chamber-specific differences in fibroblast response that could be utilized for clinical management.

The study provides direct evidence for chamber-specific differences in proliferation and pharmacological responsiveness of cardiac fibroblasts from human atria and ventricles. These differences were due to a differential stimulatory effect of serum-derived growth factors on the expression of genes controlling cell cycle-regulating cyclin proteins and DNA synthesis that can be reduced by simvastatin through an HMG-CoA reductase-sensitive pathway. This information about HCFs adds to the data from animal models demonstrating a chamber-specific response to stimulatory cytokines and is potentially useful in understanding the functional complexity of atrial and ventricular tissue to injury in health and disease and also provides a tool to selectively modulate atrial fibroblast response without significantly affecting ventricular fibroblasts. Further investigation into the chamber-specific differences in pharmacological modulation of fibroblasts that could be exploited to prevent selectively excessive fibrosis in the atria without inhibiting the adaptive remodeling in the ventricles required to repair injury is needed.

GRANTS

This study was supported by funding from the National Institutes of Health, National Heart, Lung and Blood Institute R01 grant (HL101240 to A. Jahangir) and intramural awards from Aurora Health Care (Cardiovascular Surgery Research Award to F. Rizvi).

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

F.R. and A.J. conception and design of research; F.R., A.D., R.S., U.N., L.E., A.H., and G.R. performed experiments; F.R. and E.L.H. analyzed data; F.R., G.R., and A.J. interpreted results of experiments; F.R. drafted manuscript; F.R., A.D., R.S., U.N., L.E., A.H., G.R., Y.S., E.L.H., D.C.K., A.J.T., and A.J. approved final version of manuscript; Y.S., E.L.H., D.C.K., A.J.T., and A.J. edited and revised manuscript.