Silencing of CCR2 in myocarditis

Myocarditis is a common cause for sudden cardiac death in young adults. The infiltration of monocytes into the myocardium represents a hallmark of the disease. Here we investigate the importance of chemokine (C-C motif) receptor 2 (CCR2) in mice with myocarditis and employ in vivo siRNA silencing in combination with non-invasive molecular imaging. We identify an alteration on leukocyte progenitor trafficking by siRNA silencing and present human data indicating the clinical importance of the chemokine receptor CCR2.

Abstract

Background

Myocarditis is characterized by inflammatory cell infiltration of the heart and subsequent deterioration of cardiac function. Monocytes are the most prominent population of accumulating leucocytes. We investigated whether in vivo administration of nanoparticle-encapsulated siRNA targeting chemokine (C-C motif) receptor 2 (CCR2)—a chemokine receptor crucial for leucocyte migration in humans and mice—reduces inflammation in autoimmune myocarditis.

Methods and results

In myocardium of patients with myocarditis, CCL2 mRNA levels and CCR2⁺ cells increased (P < 0.05), motivating us to pursue CCR2 silencing. Flow cytometric analysis showed that siRNA silencing of CCR2 (siCCR2) reduced the number of Ly6C^high monocytes in hearts of mice with acute autoimmune myocarditis by 69% (P < 0.05), corroborated by histological assessment. The nanoparticle-delivered siRNA was not only active in monocytes but also in bone marrow haematopoietic progenitor cells. Treatment with siCCR2 reduced the migration of bone marrow granulocyte macrophage progenitors into the blood. Cellular magnetic resonance imaging (MRI) after injection of macrophage-avid magnetic nanoparticles detected myocarditis and therapeutic effects of RNAi non-invasively. Mice with acute myocarditis showed enhanced macrophage MRI contrast, which was prevented by siCCR2 (P < 0.05). Follow-up MRI volumetry revealed that siCCR2 treatment improved ejection fraction (P < 0.05 vs. control siRNA-treated mice).

Conclusion

This study highlights the importance of CCR2 in the pathogenesis of myocarditis. In addition, we show that siCCR2 affects leucocyte progenitor trafficking. The data also point to a novel therapeutic strategy for the treatment of myocarditis.

See page 1434 for the editorial comment on this article (doi:10.1093/eurheartj/ehu302)

Translational Perspective.

Myocarditis is a common cause for sudden cardiac death in young adults. The infiltration of monocytes into the myocardium represents a hallmark of the disease. Here we investigate the importance of chemokine (C-C motif) receptor 2 (CCR2) in mice with myocarditis and employ in vivo siRNA silencing in combination with non-invasive molecular imaging. We identify an alteration on leukocyte progenitor trafficking by siRNA silencing and present human data indicating the clinical importance of the chemokine receptor CCR2.

Introduction

Myocarditis is a major cause of sudden death in young adults.¹ Even if the initial presentation suggests a mild course of disease, progression to heart failure frequently occurs.² Progress in our understanding of myocarditis pathophysiology has yet to translate into improved clinical treatment options.³ Conventional immunosuppressive therapy has not proven effective in the treatment of myocarditis.⁴ Experimental autoimmune myocarditis (EAM) in mice mimics certain aspects of inflammatory cardiomyopathy in humans, and has proven useful in studying myocarditis and resulting heart failure.⁵ While T-cell responses are crucial, CD11b⁺ monocytes/macrophages represent the majority of accumulating leucocytes⁶ and subserve many effector functions in tissue damage. Infiltration of CD68⁺ macrophages is a diagnostic hallmark for human disease evaluation.⁷

The chemokine (C-C motif) receptor 2 (CCR2) mediates the egress of inflammatory monocytes from the bone marrow⁸ and is essential for recruitment to the site of inflammation.⁹ A recent study reported that haematopoietic and myeloid progenitor cells also express CCR2, and that it regulates their migration to inflammatory sites in the liver.¹⁰ Mice lacking CCR2 exhibit a reduced severity of myocarditis.¹¹ A previously developed lipid nanoparticle siRNA carrier^12–14 delivers siRNA to myeloid cells after intravenous injection. Encapsulating siRNA that targets CCR2 (siCCR2) into this nanoparticle reduced CCR2 expression in monocytes and decreased their accumulation in acute and chronic inflammation.^12,15

As part of the current experimental murine study, we also evaluated CCR2⁺ levels in human patients with myocarditis. We report that CCR2⁺ cells enrich in hearts of patients with myocarditis. Chemokine (C-C motif) receptor 2 therefore may represent a promising therapeutic target in this disease. In mice with autoimmune myocarditis, we found that silencing CCR2 reduced monocyte numbers in the heart and improved outcome. In addition to dampening monocyte traffic, siCCR2 also reduced granulocyte macrophage progenitor (GMP) efflux from the bone marrow into the blood. Macrophage magnetic resonance imaging (MRI) non-invasively detected myocarditis in mice and followed the effects of RNAi.

Methods

Human studies

Biopsy samples were collected from seven patients with clinically suspected myocarditis (mean age, 40 ± 6 years; three men). Specimens were obtained from the apical part of the free left ventricle in patients undergoing cardiac catheterization, using a standardized protocol. The study was conducted in accordance with the Declaration of Helsinki, and the institutional medical ethics committee approved the study protocol. Biopsies were washed with NaCl (0.9%) and immediately transferred and stored in liquid nitrogen until RNA was extracted. RNA was extracted using the RNeasy kit, according to the manufacturer's protocol (Qiagen, Germany). RNA purity and concentration were determined using the Bioanalyzer 2100 (Agilent Technologies, Berkshire, UK) with a Eukaryote Total RNA Pico assay chip. RNA integrity number (RIN) >3 was defined as the minimum requirement for further analyses. Commercially available samples from healthy donors (BioServe Biotechnologies, USA) were used as controls for quantitative real-time polymerase chain reaction (PCR) (q-PCR), performed to measure CCR2 and MCP-1 expression. Primers were designed using NCBI Primer-Blast and synthesized by Eurofins MWG Operon (Ebersberg, Germany). Sixty nanograms of RNA extracted from biopsies of patients (n = 7; 3 excluded due to RIN <3) and controls (n = 7) were reverse transcribed using SuperScript III first strand cDNA synthesis kit (Invitrogen). Quantitative real-time PCR was carried out according to standard protocols with the SYBR-Green method (Thermo Scientific) using an ABI 7000 system (ABI). The specificity of each primer-pair was monitored by dissociation curve analysis. Threshold cycle (CT) values were assessed in the exponential phase of amplification and the data were analysed using the delta-CT method. The mean value of the reference genes RPL13, and β-actin was used as a reference.

Immunoenzyme staining of CCR2 was performed on 2-µm paraffin sections of formalin-fixed tissues using standard avidin–biotin anti-alkaline phosphatase techniques (Vectastain; Vector Laboratories). Antigen retrieval was achieved by steam-cooking the slides in 10 mM citrate buffer (pH 6.1; Dako) for 30 min. A solution of 10% Earle's balanced salt solution (Sigma–Aldrich) supplemented with 1% HEPES, 0.2% BSA, and 0.1% saponin (all from Sigma–Aldrich), pH 7.4, was used as a washing and permeabilization buffer. Primary Ab dilutions also were prepared in this buffer with 4% γ-venin (Behring) added and incubated overnight at 4°C.

Mouse anti-CCR2 (clone 48607, R&D Systems) was used as the primary antibody. Biotinylated sheep anti-mouse IgG was applied as a secondary reagent for 30 min at room temperature. Naphthol AS-biphosphate (Sigma–Aldrich) with New Fuchsin (Merck) was used as the substrate for alkaline phosphatase. Control tissue was provided by the Tissue Bank for Inflammatory Diseases (GEZEH). Patients who had no history of heart disease and died due to non-cardiac causes served as controls (mean age, 70 ± 3 years; three men).

Induction of autoimmune myocarditis in mice

A/J mice (5–8 weeks of age) were purchased from The Jackson Laboratory. The Subcommittee on Animal Research Care at Massachusetts General Hospital approved all procedures. Myocarditis was induced by subcutaneous injection of 120 μg of the Troponin I peptide VDKVDEERYDVEAKVTKN, which was synthesized by the MGH Peptide Core with a purity of >90%. The peptide was injected in an emulsion with complete Freund's adjuvant (Sigma–Aldrich) containing 5 mg/mL Mycobacterium tuberculosis H37Ra (Sigma–Aldrich) on Days 0, 7, and 14, as described previously.¹⁶ For adjuvant controls an emulsion of phosphate buffered saline (PBS) and complete Freund's adjuvant (Sigma–Aldrich) containing 5 mg/mL M. tuberculosis H37Ra (Sigma–Aldrich) was injected on Days 0, 7, and 14 (parallel to myocarditis induction groups).

In vivo RNAi

siRNA targeting CCR2 was synthesized and encapsulated into lipidoid nanoparticles (LNPs), as described previously.¹² Briefly, nanoparticle siRNA formulations were prepared using the lipid C12-200.¹³ A spontaneous vesicle formation formulation procedure was performed using disteroylphosphatidyl choline, cholesterol, PEG-DMG, and siRNA.^14,17 The final lipid:siRNA weight ratio was ∼12:1. The mean particle diameter was ∼60–80 nm, and the siRNA entrapment efficiency was ∼95%. Mice enrolled in longitudinal imaging trial were treated with 0.5 mg/kg/day siCCR2 intravenously twice per week, starting on the day of the first sensitization. For the evaluation of treatment induction in ongoing cardiac inflammation, mice were treated with the same dose from Day 14 to 28 after the initial troponin peptide induction. Lipid nanoparticle-encapsulated siRNA targeting luciferase—an enzyme not present in the mice used in this study—was employed as a control siRNA (siCON). For bio-distribution studies, a single dose of 1 mg/kg of LNP-encapsulated Alexa Fluor 647-labelled siRNA was injected intravenously.

Histological analysis of mouse tissue

Directly after euthanasia on Days 21 or 60, respectively, hearts were excised and rinsed in PBS and embedded in OCT (Sakura Finetek, Torrance, CA, USA). Serial 6 µm thick sections were cut and stained with haematoxylin and eosin. Severity of myocarditis was evaluated according to a 6-tier scoring system: Grade 0, no inflammation; Grade 1, cardiac infiltration in up to 5% of the cardiac sections; Grade 2, 6–10%; Grade 3, 11–30%; Grade 4, 31–50%; and Grade 5, >50%.¹⁸ Masson's trichrome staining was used for the evaluation of fibrosis. Immunohistochemistry for CD11b was done with the avidin–biotin peroxidase method. The reaction was visualized with a 3-amino-9-athyl-carbazol substrate (AEC, DAKO CA, USA). For immunofluorescence staining, sections were incubated with anti-CD11b or CD4 followed by biotinylated secondary antibody, and texas red-conjugated streptavidin (GE Healthcare). DAPI (4′,6-diamidino-2-phenylindole, Vector Laboratories) identified cell nuclei. Microscopy was performed on a Nikon 80i upright fluorescence scope. Percentage positive area or cell numbers were quantified with IPLab (version 3.9.3; Scanalytics, Inc., Fairfax, VA, USA) analysing five high-power fields per section and per animal at ×200 or ×400 magnification.

Quantitative polymerase chain reaction

Real-time PCR from cardiac tissue at Day 60 was performed using the Universal Probe Library from Roche Diagnostics. Primers and probes were HPRT: 5′-gtcaagggggacataaaag-3′, 5′-tgcattgttttaccagtgtcaa-3′ and COL3A1: 5′-tgagtcgaattggggaga at-3′, 5′-tcccctggaatctgtgaatc-3′.

Cell isolation

Heart tissue from mice with myocarditis was harvested at the expected peak of inflammation (Day 21 after immunization). After weighing, tissue was minced with fine scissors, placed into a cocktail of 450 U/ml collagenase I, 125 U/ml collagenase XI, 60 U/ml DNase I, and 60 U/ml hyaluronidase (Sigma–Aldrich) and shaken at 37°C for 1 h. Cells were then triturated through nylon mesh, centrifuged and resuspended in HBSS supplemented with 0.2% (w/v) BSA and 1% (w/v) FCS. Spleens were triturated in HBSS (Mediatech, Inc.) at 4°C and filtered through nylon mesh (BD Biosciences) with the end of a 3-mL syringe plunger. The cell suspension was centrifuged at 300 × g for 10 min at 4°C. Red blood cells were lysed with ACK lysis buffer, and splenocytes were washed with HBSS and resuspended in HBSS supplemented with 0.2% (w/v) BSA and 1% (w/v) FCS. Femurs were flushed with HBSS to isolate bone marrow cells. For analysis of bio-distribution of the probe, organ harvest was performed 2 h after injection of fluorochrome-labelled siRNA. Spleens were harvested after 7 days of siRNA treatment and processed as described above for analysis of CCR2 expression after cell sorting. Total leucocyte numbers were determined using Trypan blue (Mediatech, Inc.).

Flow cytometry

Cell suspensions were incubated with a cocktail of mAbs against T cells (CD90-PE, 53–2.1), B cells (B220-PE, RA3-6B2), NK cells (CD49b-PE, DX5 and NK1.1-PE, PK136), granulocytes (Ly-6G-PE, 1A8), myeloid cells (CD11b-APC, M1/70), antigen-presenting cells (I-Ab (AF6-120.1)-biotin-Strep-PerCP), dendritic cells (CD11c-Alexa 700), macrophages (F4/80-PE-Cy7), and monocyte subsets (Ly-6C-FITC, AL-21) (all antibodies from BD Biosciences or Biolegend). Monocytes were identified as CD11b^high (CD90/B220/CD49b/NK1.1/Ly-6G)^low (F4/80/I-Ab/CD11c)^low Ly-6C^high/low. Macrophages/dendritic cells were identified as CD11c^high (CD90/B220/CD49b/NK1.1/ Ly-6G)^low (F4/80/I-A^b/CD11c)^high Ly-6C^low. Neutrophils were identified as CD11b^high (CD90/B220/ CD49b/NK1.1/Ly-6G)^high (F4/80/I-A^b/CD11c)^low Ly-6C^int. GMP were identified as lineage (CD90/B220/CD49b/NK1.1/Ly-6G/Ter119/CD11b/CD11c/IL7Rα)^low c-kit^high Sca-1^neg CD16/32^high CD34^high. Alexa Fluor 647-labelled siRNA encapsulated in LNPs was detected in the APC channel. T cells were identified as CD3^high after incubation with a mAb against CD3 (Biolegend). We used a primary antibody from Abcam (ab21667) and a secondary anti-rabbit IgG antibody (FITC) from Abbiotec for CCR2 staining. Reported cell numbers were calculated as the product of total live cells and percent cells within the respective FACS gate. Data were acquired on an LSRII (BD Biosciences) and analysed with FlowJo v.8.8.6 (Tree Star, Inc.).

In vitro migration assay

Migration experiments using MCP-1 as a chemo-attractant were performed in BD BioCoat invasion chambers (BD Bioscience). Granulocyte macrophage progenitors were sorted on an FACS Aria from the bone marrow of mice that had received 0.5 mg/kg/day siCCR2 or siCON for 3 days prior to sacrifice. Cells were suspended in RPMI 1640 media (Cellgro, Mediatech, Inc, VA, USA) supplemented with 0.2% FCS (Valley Biomedical, Inc.); 1×10⁵ cells were placed on the matrigel-coated 8 µm pore size PET membrane and incubated in a humidified incubator at 37°C, 5% CO₂ for 1 h, allowing the cells to attach to the matrigel. Migration was induced by addition of 25 nM of MCP-1 (BD Bioscience) to the lower compartment. After 2 h, the number of migrated cells in the lower compartment was determined by cell counts using a haemocytometer.

Intra-vital microscopy of haematopoietic progenitor cells in the bone marrow

Granulocyte macrophage progenitors were isolated by FACS sorting from the bone marrow of C57BL/6-Tg(UBC-GFP)30Scha/J mice. After sorting of Lin⁻ c-kit⁺ Sca1⁻ CD16/CD32⁺ CD34⁺ GMP, cells were injected into non-irradiated recipient C57BL/6 mice. Each imaged mouse received 25 000–40 000 GMP. Mice were anaesthetized and prepared for in vivo imaging. Prior to the imaging session, 2 nmol of AngioSense^® 750 EX (PerkinElmer) was injected intravenously to visualize blood vessels. Cells within the skull bone marrow cavity were imaged as described previously.^19,20 Data were acquired as Z-stacks at 5 µm steps containing three separate channels: transplanted GFP-positive GMP (assigned to the green channel), nanoparticle-encapsulated siRNA (assigned to the red channel), and blood pool (assigned to the blue channel). Image processing was performed using ImageJ software. After in vivo imaging, mice were euthanized for subsequent flow cytometric analyses.

Magnetic resonance imaging

We used different MRI approaches to visualize cells and myocardial function: (i) measurement of phagocytic cells via a magnetic nanoparticle and (ii) measurement of cardiac volumes and ejection fraction (EF). Twenty-four hours prior to MRI, mice were injected intravenously with 30 mg/kg of the magnetofluorescent nanoparticle CLIO-VT680 (cross-linked iron-oxide nanoparticles derivatized with the near infra-red fluorochrome VT680). The chemistry core at the Center for Systems Biology prepared the nanoparticles as described previously.²¹ CLIO enhances image contrast on T2* weighted MRI.²² On Day 21 after induction of myocarditis, bright-blood cine images were obtained with ECG and respiratory gating (SA Instruments, Stony Brook, NY, USA) using a gradient echo FLASH-sequence on a 7 T horizontal bore scanner (Pharmascan, Bruker, Billerica, MA, USA). Imaging parameters were as follows: echo time (TE) 5 ms; 16 frames per RR interval (TR 7.0–10.0 ms); in-plane resolution 200/200 µm; slice thickness 1 mm. The myocardial contrast-to-noise ratio (CNR) was calculated as CNR = (myocardial signal-background signal)/(standard deviation of the noise). On Day 60 after induction of myocarditis, mice were imaged again and cardiac volumes were quantitated from 6 to 8 short-axis imaging slices covering the left ventricle as described previously.²³ After completion of MRI, hearts were excised, cut in short-axis rings and imaged using a planar fluorescent reflectance imaging system (OV-110, Olympus) at excitation/emission wavelengths of 610–645/675–715 nm, respectively and exposure times of 60–75 ms.

Transthoracic echocardiography

Echocardiography was performed using a Vevo 2100 with a mouse transducer (15 MHz). The echocardiographer was blinded with respect to the treatment group. Mice were shaved and imaged without anaesthesia. Left ventricular parasternal long- and short-axis views were obtained in 2D imaging. The EF was calculated using the Simpson's rule.

Statistics

Results are expressed as mean ± SEM unless stated otherwise. Statistical comparisons between two groups were evaluated by Student's t-test and one-way ANOVA for more than two groups. A value of P < 0.05 indicated statistical significance.

Results

Cardiac accumulation of CCR2⁺ cells in patients with acute myocarditis

The leucocyte infiltrate in human myocarditis contains CD68⁺ cells, a surface-marker expressed on monocytes and macrophages.⁷ A subset of monocytes expresses CCR2, a receptor crucial for the accumulation of the pro-inflammatory monocyte subset.⁹ To evaluate the presence of CCR2⁺ cells in human myocarditis, we collected heart tissue biopsies from patients with myocarditis and compared them with control specimens obtained from individuals who died due to trauma with no history of cardiac disease. Histological analysis revealed that while in control cardiac tissue 1.0 ± 0.4 CCR2⁺ cell was found per mm² heart tissue, 11.6 ± 0.9 CCR2⁺ cells were present in samples from patients with myocarditis (P < 0.0001, Figure 1A and B). These findings were corroborated by quantitative PCR analysis that showed a 5.4- and a 4.6-fold increase of CCR2 and its major cognate chemokine ligand MCP-1 (also known as CCL2) mRNA in heart tissue from patients with myocarditis compared with controls (Figure 1C).

Figure 1.

CCR2⁺ cells accumulate in the heart in patients with acute myocarditis. (A) Histology of heart samples from patients with acute myocarditis and control patients stained for chemokine (C-C motif) receptor 2. (B) Bar graph shows enumeration of CCR2⁺ cells per mm² heart tissue. (C) Quantitative polymerase chain reaction analysis of heart biopsies for chemokine (C-C motif) receptor 2 and MCP-1 comparing patients with myocarditis to patients without a history of heart disease. Data are mean ± SEM (n = 4–7 per group), *P < 0.05.

siRNA silencing of CCR2 prevents cardiac accumulation of monocytes in mice with myocarditis

Motivated by CCR2⁺ cells present in clinical myocarditis, we tested the effects of in vivo RNAi silencing of this protein¹² in mice with autoimmune myocarditis. Twenty-one days after induction of myocarditis with the pathogenic peptide, hearts from mice with acute disease contained 1.9×10⁵ Ly6C^high monocytes, roughly an order of magnitude more when compared with adjuvant-injected control mice (0.2 × 10⁵ cells per heart, P = 0.04, Figure 2A). siRNA silencing of CCR2 significantly reduced accumulation of Ly6C^high monocytes (0.58 × 10⁵, P = 0.01, Figure 2A), cells that rely on CCR2 for their migration. In mice injected with siCCR2, analysis of H&E stained hearts showed a reduction in the disease score, while immunohistochemistry revealed a strong decrease in the myeloid cell marker CD11b⁺ (Figure 2B and C).

Figure 2.

In vivo silencing of chemokine (C-C motif) receptor 2 reduces the number of monocytes in mice with myocarditis. (A) Representative dot plots from FACS analysis of hearts. Gates on top row depict monocyte/macrophage population; bottom row shows gating for Ly6C^high monocytes. Bar graph on the right enumerates Ly6C^high monocytes in the heart. (B) Haematoxylin and eosin staining of the heart, with disease severity grading plotted on the right. (C) CD11b immunohistochemistry quantification. Data are mean ± SEM (n = 3–10 per group), *P < 0.05.

MNP MRI detects RNAi effects in mice with myocarditis

To investigate the impact of RNAi non-invasively, we used magnetic nanoparticle enhanced MRI to quantitate phagocytes residing in the inflamed myocardium. In validation experiments with fluorescently labelled, macrophage-avid nanoparticles, T2*-weighted MRI contrast increased in hearts with acute myocarditis on Day 21 (Figure 3A). Ex vivo fluorescence reflectance imaging confirmed uptake of magnetofluorescent nanoparticles in the myocardium of mice with myocarditis (Figure 3B). Microscopic fluorescence nanoparticle signal co-localized with immunostained CD11b⁺ cells (Figure 3C). We next designed a longitudinal study combining macrophage MRI during the peak inflammatory activity on Day 21, with MRI volumetry follow-up on Day 60 (Figure 4A illustrates the study design). Three groups of mice were investigated: (i) control mice without myocarditis, in which we administered only the adjuvant; (ii) mice with myocarditis injected with siCON; and (iii) mice with myocarditis that were injected with siCCR2. The CNR on nanoparticle MRI changed from 14 ± 4 in control mice to −0.1 ± 2 in mice with myocarditis (P = 0.008, Figure 4B and C). siRNA silencing of CCR2 mostly prevented CNR changes (10.6 ± 3; P = 0.01 vs. siCON). These imaging findings during the acute phase of myocardial inflammation correlated with the outcome, as cardiac function significantly declined in mice with autoimmune myocarditis, while injection of siCCR2 preserved left ventricular EF (64 ± 2% vs. 72 ± 2%, P = 0.02, Figure 4D).

Figure 3.

Nanoparticle imaging validation in mice with myocarditis. (A) Magnetic resonance imaging shows uptake of cross-linked iron-oxide at the peak of cardiac inflammation on Day 21. Arrows point towards hypointense areas, indicating iron-oxide nanoparticle accumulation. Fluorescent labelling of cross-linked iron-oxide allows ex vivo probe detection by fluorescence reflectance imaging as shown in (B). (C) Immunofluorescence histology of heart sections from a mouse with myocarditis 24 h after injection with cross-linked iron-oxide-VT680. Staining with CD11b co-localizes with nanoparticle signal.

Figure 4.

Macrophage magnetic resonance imaging detects myocarditis and reports on treatment efficacy. (A) Experimental design. (B) Representative short-axis magnetic resonance images on Day 21 after disease induction. (C) Bar graph for contrast-to-noise ratio detected after nanoparticle administration on Day 21 after disease induction. (D) Quantification of left ventricular ejection fraction on Day 60. Data are mean ± SEM (n = 9–10 per group), *P < 0.05.

Initiation of RNAi treatment during ongoing cardiac inflammation

At Day 14 after induction of EAM, many leucocytes are present in the heart.⁶ To evaluate whether RNAi can still alter the outcome when siCCR2 treatment is initiated after cardiac inflammation established, we set up an additional trial starting treatment at this later time point (Figure 5A–F). We found that delayed siCCR2 treatment still improved the outcome on Day 60, indicated by better cardiac function (left ventricular EF measured by echocardiography, 66 ± 4% vs. 79 ± 2%, P = 0.03, Figure 5B). We detected reduced inflammation at this late disease stage (P = 0.01, Figure 5E). We also observed a reduction of fibrosis by histological analysis in mice treated with siCCR2 compared with control-treated mice (19 ± 1% vs. 12 ± 2%, P = 0.02, Figure 5D). These findings were corroborated by quantitative PCR analysis that showed a 10-fold reduction of collagen mRNA in heart tissue from siCCR2-treated mice (P = 0.01, Figure 5C). Previous studies addressed the importance of CD4⁺ T cells in EAM.²⁴ Probing for these cells on Day 60 showed reduced myocardial numbers after siCCR2 treatment (Figure 5F). To test whether siCCR2 directly targets T cells, we injected fluorescently labelled siRNA encapsulated in nanoparticles intravenously and harvested spleens from these mice. By using CD3 as a general T-cell marker, we detected no uptake of siRNA into T cells (Supplementary material, Figure S1). Because no siRNA was delivered to T cells, we can exclude that the observed therapeutic effects were mediated through direct silencing in T cells.

Figure 5.

Initiation of RNAi during later disease stages. (A) Experimental design. (B) Left ventricular ejection fraction on Day 60 after late initiation of siRNA silencing of chemokine (C-C motif) receptor 2. (C) Quantitative polymerase chain reaction analysis of collagen 3 in heart tissue on Day 60. (D) Quantification of left ventricular fibrosis (Masson's trichrome staining) and representative images. (E) Haematoxylin and eosin staining of the heart, with disease severity grading plotted on the left. (F) CD4 immunofluorescence quantification and representative images. Data are mean ± SEM (n = 7–8 per group), *P < 0.05.

siRNA silencing of CCR2 targets myeloid progenitor migration

Splenic myelopoiesis relies on transfer of bone marrow haematopoietic progenitors to the organ. This traffic occurs through blood and may be an interesting target for anti-inflammatory therapy.²⁵ A recent study showed that haematopoietic progenitors express CCR2, and that the cells rely on it for migration.¹⁰ We therefore tested with flow cytometry whether bone marrow GMP take up nanoparticle-encapsulated siRNA after intravenous injection, and found this to be the case (Figure 6A). We further employed intravital microscopy to visualize siRNA uptake into GMP residing in the haematopoietic bone marrow niche (Figure 6B). Flow cytometry validated the in vivo microscopy data, demonstrating siRNA uptake into GMP in the bone marrow and also into splenic GMP after systemic delivery (Figure 6C). In vivo treatment with nanoparticle-encapsulated siCCR2 resulted in a significant reduction of CCR2 expression in bone marrow GMP at the mRNA and protein levels (Figure 6D and E). We used an in vitro migration assay to evaluate if the knock down affects GMP migration. Granulocyte macrophage progenitors isolated from mice treated with siCCR2 showed significantly reduced migration towards the chemo-attractant MCP-1 (control treatment: 780 ± 139 vs. siCCR2: 260 ± 24, P = 0.006, Figure 6F).

Figure 6.

RNAi in haematopoietic progenitor cells. (A) FACS analysis shows uptake of fluorescently labelled siRNA encapsulated in lipidoid nanoparticles. Representative gating strategy for the identification of GMP. Histograms illustrate the uptake of siRNA (red) into GMP. Black indicates fluorescence in cells harvested from non-injected control mice. Quantification of mean fluorescent intensity is shown in the bar graph on the right. (B) Intravital microscopy of siRNA uptake into GMP residing in bone marrow. One day after GFP⁺ GMP adoptive transfer, nanoparticle-encapsulated, fluorescently labelled siRNA was injected intravenously, and siRNA uptake into GMP was imaged 2 h later in mouse calvarium bone marrow. Green, GFP-positive GMP; red, AF647 tagged siRNA; blue, blood pool imaging agent. Scale bar: 50 µm. (C) FACS analysis confirmed siRNA uptake into adoptively transferred granulocyte macrophage progenitors in the bone marrow (upper panels) and in the spleen (lower panels). (D) Quantitative polymerase chain reaction analysis of chemokine (C-C motif) receptor 2 expression in isolated granulocyte macrophage progenitors from the bone marrow shows knock down. (E) Flow cytometric evaluation of chemokine (C-C motif) receptor 2 protein expression on granulocyte macrophage progenitors. Cells from control-treated animals are displayed in blue and siRNA silencing of chemokine (C-C motif) receptor 2-treated animals in green. Mean fluorescence intensity is shown in the bar graph on the right. (F) Quantification of in vitro migration assay towards MCP-1 of granulocyte macrophage progenitors isolated from mice treated with siRNA silencing of chemokine (C-C motif) receptor 2 in comparison with control treatment. Data are mean ± SEM (n = 4–5 per group), * P < 0.05.

siRNA silencing of CCR2 prevents monocyte release and myeloid progenitor traffic

Following inflammatory stimuli, haematopoietic progenitor cells migrate from the bone marrow to the spleen, leading to extramedullary haematopoiesis.^20,26 To evaluate this phenomenon in the context of autoimmune myocarditis, we first explored whether the spleen increased myelopoiesis in mice with myocarditis. A colony forming unit assay of splenocytes revealed more capacity to form colonies when splenocytes were harvested from mice with autoimmune myocarditis (Figure 7A). We further observed a significant increase of splenic GMP (to 10 × 10⁴ ± 2 × 10⁴, P = 0.005 vs. control, Figure 7B). siRNA silencing of CCR2 drastically reduced spleen size (Figure 7C) and decreased splenic monocytes from 4.9 × 10⁶ ± 2 × 10⁵ to 3.4 × 10⁶ ± 1 × 10⁵ (P = 0.001, Figure 7D) and splenic GMP (15.4 × 10⁴ ± 2 × 10⁴ to 7.1 × 10⁴ ± 1 × 10⁴ (P = 0.01, Figure 7E). In contrast, the number of bone marrow GMP increased after siCCR2 treatment (siCON, 2.9 × 10⁵ ± 9 × 10⁴; siCCR2, 5.9 × 10⁵ ± 7 × 10⁴; P < 0.05, Figure 7F and G), likely reflecting impaired release from this compartment. The number of monocytes in the bone marrow also increased (siCON, 1.3 × 10⁶ ± 3 × 10⁵; siCCR2: 2.5 × 10⁶ ± 2 × 10⁵, P < 0.05, Figure 7H). Circulating Lin⁻ c-kit⁺ progenitor cells were reduced after siCCR2 (siCON, 1.8 × 10³ ± 0.4 × 10³; siCCR2, 0.8 × 10³ ± 0.2 × 10³, P < 0.05, Figure 7I). Taken together, these data suggest that silencing CCR2 attenuated not only monocyte but also progenitor cell migration from the bone marrow into the blood.

Figure 7.

Myocarditis induces splenic myelopoiesis. (A) Colony forming unit assay of splenocytes from control mice and mice with myocarditis. (B) Enumeration of granulocyte macrophage progenitors in the spleen from naive, adjuvant controls, or mice with myocarditis. (C) Size comparison of spleens from mice with myocarditis after treatment with control siRNA (left) or siRNA silencing of chemokine (C-C motif) receptor 2 (right). (D) Bar graphs show quantification of monocytes/macrophages (D) and granulocyte macrophage progenitors (E) in the spleen in myocarditis (Day 21). (F) Dot plots from bone marrow samples of mice with myocarditis (Day 21) treated with control siRNA or siRNA silencing of chemokine (C-C motif) receptor 2. Blue gate illustrates granulocyte macrophage progenitors gate after prior gating for c-kit⁺, lineage⁻, CD115⁻, and Sca-1⁻ cells. (G) Enumeration of granulocyte macrophage progenitors in the bone marrow. (H) Quantification of monocytes/ macrophages (CD11b⁺, lineage⁻) in bone marrow of mice with autoimmune myocarditis (day 21). (I) The bar graph shows quantification of c-kit⁺/lin⁻/Sca-1⁻ haematopoietic progenitors in the blood of mice with autoimmune myocarditis. Mean ± SEM (n = 5–8 per group), *P < 0.05.

Discussion

The high variability of the clinical manifestations and aetiology of myocarditis²⁷ may explain why several clinical trials of immunosuppression for acute myocarditis have shown negative results.^4,28 An alternative explanation could be that the investigated drugs resulted in non-selective immunosuppression, also limiting the function of leucocytes that may mediate inflammation resolution. Our finding of increased cardiac CCR2⁺ cells in patients with myocarditis suggests that the MCP-1/CCR2 axis could be a more selective therapeutic target in this disease. Previous studies support this notion—for example, by reporting increased MCP-1 serum levels in patients with myocarditis.²⁹

Based on these clinical observations, we silenced CCR2 expression in mice with autoimmune myocarditis. Administration of nanoparticle-encapsulated siRNA reduced monocyte accumulation in the heart and consequently prevented the chronic decline of left ventricular function. These data extend a previous report in mice with genetic lack of CCR2, which also exhibited reduced inflammation in autoimmune myocarditis.¹¹ The use of RNAi potentially allows to target the MCP-1/CCR2 axis therapeutically in patients with myocarditis. Of note, monocytes may give rise to antigen-presenting cells,³⁰ therefore, the findings in our study may, at least partially, also be explained with a reduced sensitization. At late disease stages, we found reduced myocardial numbers of CD4⁺ cells after siCCR2 treatment. Altered antigen presentation in local lymph nodes and subsequently dampened T-cell responses may ameliorate disease progression. On the other hand, we detected no uptake of nanoparticle-encapsulated siRNA into T cells. This indicates that, while siRNA treatment has no direct effect on T cells, diminishing the myeloid response during inflammation also altered T-cell presence.

To correlate the reduction of cardiac inflammation by in vivo RNAi with the outcome, we explored non-invasive MRI as a clinically viable companion imaging tool. Currently, cardiac biopsy provides the only definite modality to diagnose myocarditis.³¹ This invasive procedure involves patient discomfort, bears the risk of complications, and is prone to sampling error. A post-mortem study calculated that an average of 17.2 samples are needed to diagnose myocarditis in more than 80% of cases.³² Van Heesvijk and colleagues reported that fluorine-19 (19F) cardiac magnetic resonance imaging detects inflammation in a mouse model of EAM.³³ The avidity of magnetic nanoparticles for monocytes and macrophages can serve to image their accumulation in myocardial infarction using MRI.^21,22 This non-invasive technique allows an evaluation of the distribution of inflammation of the whole heart in combination with analysis of cardiac morphology and function. Iron-oxide nanoparticles detect myocarditis in rats,³⁴ and human studies followed cardiac inflammation after myocardial infarction in patients, suggesting translatability.^35,36 In the current study, iron-oxide nanoparticle MRI evaluated therapeutic efficacy, a strategy that could augment evaluation of new therapeutics in future clinical trials.

Treatment with siCCR2 is unsuitable for patients with active viral myocarditis as circulating monocytes may have important roles in defence against infection. In line with current clinical guidelines for the treatment of myocarditis,³⁷ a cardiac biopsy would be necessary to rule out active infection prior to initiation of any anti-inflammatory therapy.

We here describe in vivo RNAi in haematopoietic progenitors for the first time. Granulocyte macrophage progenitors incorporated the delivery nanoparticles, resulting in a knockdown of the target gene. Granulocyte macrophage progenitors—which give rise to monocytes, macrophages and neutrophils—³⁸express the receptor CCR2 and rely on it for migration.¹⁰ The development of autoimmune myocarditis and consequent heart failure takes several weeks. Haematopoietic progenitor cells may impact the inflammatory activity in the heart over such a period by increasing the supply of inflammatory leucocytes. Kania et al. recently identified bone marrow derived, inflammatory progenitor cells that infiltrate the heart during acute myocarditis and contribute to fibrosis. In vitro, these CD45⁺ progenitors differentiated into cells with a macrophage-like phenotype.^39,40 Therapy with siCCR2 may attenuate migration of these progenitors.

In mice with myocarditis, we observed extramedullary leucocyte production in the spleen. Future studies will have to verify translatability of this finding and evaluate the importance of progenitor cell migration for disease progression. Our observations indicate that targeting haematopoietic cells with RNAi modulates leucocyte supply to the inflamed myocardium via the bone marrow-blood-heart axis. The improved long-term outcome after siCCR2 treatment suggests that this strategy may provide an avenue for translating the findings in CCR2-deficient mice to human patients.

Supplementary material

Supplementary material is available at European Heart Journal online.

Funding

This work was funded in part by grants from the Deutsche Herzstiftung, DFG, and from the National Heart, Lung, and Blood Institute (R01-HL114477, R01-HL095629, R01-HL117829, and HHSN268201000044C).

Conflict of interest: T.I.N., A.B., and K.F. are Alnylam Pharmaceuticals employees; D.G.A. receives research funding from Alnylam.