HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Citation Map Services Related articles in The JI Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Hayward, S. L. Articles by Elliott, J. F. Search for Related Content PubMed PubMed Citation Articles by Hayward, S. L. Articles by Elliott, J. F.

CD4 T Cells Play Major Effector Role and CD8 T Cells Initiating Role in Spontaneous Autoimmune Myocarditis of HLA-DQ8 Transgenic IAb Knockout Nonobese Diabetic Mice¹

Sarah L. Hayward^*,, Norma Bautista-Lopez^2,*, Kunimasa Suzuki^*,, Alexey Atrazhev^3,*, Peter Dickie^* and John F. Elliott^4,*,,

^* Department of Medical Microbiology and Immunology, Alberta Diabetes Institute, and Division of Dermatology and Cutaneous Sciences, Department of Medicine, University of Alberta, Edmonton, Alberta, Canada

	Abstract

Top Abstract Introduction Materials and Methods Results Discussion Disclosures References

 
In humans, spontaneous autoimmune attack against cardiomyocytes often leads to idiopathic dilated cardiomyopathy (IDCM) and life-threatening heart failure. HLA-DQ8 transgenic IAb knockout NOD mice (NOD.DQ8/Ab⁰; DQA1*0301, DQB1*0302) develop spontaneous anticardiomyocyte autoimmunity with pathology very similar to human IDCM, but why the heart is targeted is unknown. In the present study, we first investigated whether NOD/Ab⁰ mice transgenic for a different DQ allele, DQ6, (DQA1*0102, DQB1*0602) would also develop myocarditis. NOD.DQ6/Ab⁰ animals showed no cardiac pathology, implying that DQ8 is specifically required for the myocarditis phenotype. To further characterize the cellular immune mechanisms, we established crosses of our NOD.DQ8/Ab⁰ animals with Rag1 knockout (Rag1⁰), Ig H chain knockout (IgH⁰), and ₂-microglobulin knockout (₂m⁰) lines. Adoptive transfer of purified CD4 T cells from NOD.DQ8/Ab⁰ mice with complete heart block (an indication of advanced myocarditis) into younger NOD.DQ8/Ab⁰ Rag1⁰ animals induced cardiac pathology in all recipients, whereas adoptive transfer of purified CD8 T cells or B lymphocytes had no effect. Despite the absence of B lymphocytes, NOD.DQ8/Ab⁰IgH⁰ animals still developed complete heart block, whereas NOD.DQ8/Ab⁰₂m⁰ mice (which lack CD8 T cells) failed to develop any cardiac pathology. CD8 T cells (and possibly NK cells) seem to be necessary to initiate disease, whereas once initiated, CD4 T cells alone can orchestrate the cardiac pathology, likely through their capacity to recruit and activate macrophages. Understanding the cellular immune mechanisms causing spontaneous myocarditis/IDCM in this relevant animal model will facilitate the development and testing of new therapies for this devastating disease.

	Introduction

Top Abstract Introduction Materials and Methods Results Discussion Disclosures References

 
Idiopathic dilated cardiomyopathy (IDCM)⁵ is a condition where previously healthy individuals develop life-threatening heart failure associated with cardiac enlargement, but with no apparent underlying cause (1, 2). In humans, the cardiac dilation seen in IDCM typically occurs secondary to an inflammatory process within the heart muscle (i.e., myocarditis); it is thought to be triggered either by a prior subclinical viral infection or by a primary autoimmune attack against cardiomyocytes. A better understanding of the immune mechanisms acting to cause myocarditis and IDCM could potentially lead to advances in specific treatments for this disorder, which at the present time are largely supportive and nonspecific. Historically, two murine models have been used to investigate immune mechanisms of myocarditis: susceptible mouse strains develop postviral myocarditis following infection with Coxsackievirus B3 (3, 4, 5) and (in the same strains) myocarditis can be induced by immunization with cardiac myosin in adjuvant (6, 7, 8). Although both models show histological evidence of myocarditis and develop antiheart autoantibodies, the myocarditis does not arise spontaneously and is typically mild and, in addition, the models are limited because prior virus infection or immunization potentially biases the immune phenotype.

Research into IDCM caused by primary anticardiac autoimmune processes has been advanced in the past few years by the development of several different lines of transgenic and/or gene knockout mice which develop spontaneous myocarditis and dilated cardiomyopathy. These include: 1) several lines which overexpress TNF- specifically within cardiomyocytes (9, 10, 11); 2) a line of PD-1 receptor knockout mice crossed onto the BALB/c background (12, 13); and 3) several lines of NOD mice which express the human MHC class II molecule HLA-DQ8 (DQA1*0301, DQB1*0302) in the absence of endogenous mouse class II MHC (i.e., NOD.DQ8/Ab⁰) (14, 15). Of the three available spontaneous models, the latter displays pathology most similar to what has been observed in humans with IDCM.

Our group was the first to describe a single line of NOD.DQ8/Ab⁰ mice where 100% of the animals developed spontaneous anticardiac autoimmunity and IDCM. Features of the model include rising titers of anticardiac autoantibodies, focal mononuclear cell infiltrates associated with cardiomyocyte destruction in all layers of the myocardium, escalating heart block on electrocardiogram (ECG), and gradual cardiac dilation with progression to end-stage heart failure (15). Subsequently, Lipes and colleagues (14) described three additional independent NOD.DQ8/Ab⁰ lines with a similar phenotype. Although Lipes’ NOD.DQ8/Ab⁰ lines showed somewhat lower disease penetrance, their results indicated that the myocarditis/IDCM phenotype observed in our NOD.DQ8/Ab⁰ line was not due to a founder effect. Given the generality of the observation that NOD.DQ8/Ab⁰ animals provide an accurate model of spontaneous autoimmune myocarditis/IDCM in humans, we proceeded to characterize the immune mechanisms acting to cause disease in our transgenic line. In undertaking these investigations, we assumed it would be of interest to compare the results obtained with what is known about immune mechanisms acting to cause diabetes in NOD mice, because the same NOD background genes (i.e., genes other than MHC) are likely important for autoimmunity in both models. For these experiments we crossed our animals onto the Rag1 knockout (Rag1⁰), Ig H chain knockout (IgH⁰), and ₂-microglobulin knockout (₂m⁰) backgrounds, all of which were readily available in the NOD genetic background. In the present study, we also used a NOD.DQ6/Ab⁰ line (originally constructed by our group for studies into diabetes resistance) which is closely matched to the NOD.DQ8/Ab⁰ line in every aspect except for the DQ haplotype (DQA1*0102, DQB1*0602 vs DQA1*0301, DQB1*0302). These latter investigations show that DQ6 mice do not develop any cardiac pathology; thus, implicating the DQ8 haplotype as an important contributing factor in the autoimmune myocarditis/IDCM phenotype seen in the DQ8 animals.

Four different cell types (CD4 T cells, CD8 T cells, B lymphocytes, and macrophages; Ref. 15 and our unpublished result) are all found within the mononuclear cell infiltrates observed in older NOD.DQ8/Ab⁰ animals with heart block and it is unclear which of these cell types are primarily responsible for the cardiomyocyte destruction. In our first study, we showed that adoptive transfer of splenocytes (but not antisera) from NOD.DQ8/Ab⁰ animals with complete heart block into younger irradiated animals could cause accelerated onset of autoimmune myocarditis and heart block (15); however, the irradiation model is cumbersome and suffers from certain limitations. In the present study, we establish the NOD.DQ8/Ab⁰Rag1⁰ adoptive transfer model and reproduce results obtained with the irradiation model; we then go on to show that adoptive transfer of CD4 T cells alone can induce disease, whereas adoptive transfer of purified CD8 T cells or B lymphocytes has no effect. Phenotyping of NOD.DQ8/Ab⁰ IgH⁰ animals indicates that they still develop complete heart block and myocarditis, but with slightly delayed kinetics, whereas NOD.DQ8/Ab⁰₂m⁰ mice fail to develop any cardiac pathology. Taken together, these results suggest that CD8 T cells are important for initiating myocarditis in NOD.DQ8/Ab⁰ mice (although a role for NK cells cannot be excluded), whereas once autoimmune CD4 T cell responses have been induced, these responses alone are capable of inducing the full disease phenotype, through the recruitment and activation of macrophages. B lymphocytes are not needed for disease to occur (although they may play an ancillary Ag-presenting role, resulting in slightly earlier disease onset when they are present), and Abs play little or no role in the disease process. These results have significant parallels with immune mechanisms causing diabetes in NOD mice and also with the post-Coxsackievirus myocarditis model in mice. The central role of CD4 T cells and macrophages in the final disease process suggests that therapies targeted at these cell subsets may be of most benefit for patients with advanced disease.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion Disclosures References

 
Mice

NOD/LtJ mice were purchased from The Jackson Laboratory (JAX). The IAb knockout NOD line (designated NOD/Ab⁰) was provided by A. M. Jevnikar (University of Western Ontario, London, Ontario, Canada), and the DQ8 transgenic NOD line by L. Wicker (University of Cambridge, Cambridge, U.K.). Details we published previously regarding the construction of this particular DQ8 transgenic NOD line (16) need to be corrected in one aspect (L. Burkly, Biogen Idec, Cambridge, MA, unpublished observation). The DQA1*0301 and DQB1*0302 genes were not microinjected as two separate genomic fragments (i.e., from cosmids H11A and X10A; Ref. 17), but rather the entire DQ8 gene (containing both and segments plus upstream regulatory elements) was injected as a single large DNA fragment, having been isolated from the cosmid pDC2 (17). This DQ8 transgenic NOD line has been assigned the formal designation NOD.Tg(CD2-CD4,HLA-DQA1*0301,HLA-DQB1*0302)1Ell. For the sake of brevity, we will henceforth refer to this line as NOD.DQ8.

Intercrossing of the NOD.DQ8 and IAb knockout NOD lines to create the NOD.Cg-H2-Ab1^tm1GruTg(CD2-CD4,HLA-DQA1*0301,HLA-DQB1*0302)1Ell (accession ID MGI:3590082) line has been described previously (16). In our previous two studies (15, 16), we used the designations "human CD4, DQA1*0301, DQB1*0302 transgenic IA null NOD" and subsequently "HLA-DQ8 transgenic IAb knockout NOD"; henceforth, we will designate this line simply as NOD.DQ8/Ab⁰.

The Rag1 knockout NOD (NOD.129S7(B6)-Rag1^tm1Mom/J), Ig H chain knockout NOD (NOD.129S2(B6)-Igh-6^tm1Cgn), and ₂m knockout NOD (NOD.129P2(B6)-2m^tm1Unc/J) lines were all provided by D. Serreze (The Jackson Laboratory, Bar Harbor, ME). The NOD.DQ8/Ab⁰Rag1⁰, NOD.DQ8/Ab⁰IgH⁰, and NOD.DQ8/Ab⁰₂m⁰ lines were created by crossing the NOD.DQ8/Ab⁰ animals with the relevant knockout lines, intercrossing the F₁ generation and selecting for desired F₂ offspring using flow cytometric analysis and PCR (see below). All breeding colonies were housed in conventional specific pathogen-free or (in the case of Rag1 knockout) virus Ab-free facilities. Care and handling was done in accordance with the guidelines of the Canadian Council on Animal Care, and all experimental protocols were approved by our institutional Health Sciences Animal Policy and Welfare Committee.

Establishment of DQ6 transgenic NOD and DQ6 transgenic IAb knockout NOD lines

NOD mice transgenic for both HLA-DQ6 (DQA1*0102, DQB1*0602) and human CD4 were created by microinjecting NOD/LtJ single-cell embryos with a mixture of two different DNA molecules: 1) a human genomic fragment containing the entire DQ6 gene, and 2) a human CD4 minigene construct with mouse CD4 enhancer, previously demonstrated to yield transgenic mice with tissue-appropriate expression of human CD4 (18). To obtain the DQ6 gene, we constructed a bacterial artificial chromosome (BAC) genomic library using DNA from the homozygous B-lymphoblastoid cell line IWB (National Institutes of General Medical Sciences Human Genetic Mutant Cell Repository). Initially, 192 separate BAC pools (500 independent BAC clones per pool) were screened by PCR using sequence-tagged site primer pairs for loci 9–10 and 2D7 (19); the two resulting positive pools were then randomly plated and screened by filter hybridization to obtain two positive clones. End sequencing revealed that the most favorable BAC clone (74-kb insert) extended from 30 kb upstream of the DQA1 gene to 12-kb upstream of the DQB1 gene (both promoters drive inwards). BAC DNA from this clone was purified using a Qiagen 500 column, digested with NotI, and the insert was separated from the vector by sedimentation over sucrose gradients (5–30% sucrose in a SW 55Ti rotor at 50,000 rpm x 80 min, 18°C). The buffer was changed to 10 mM Tris-HCl (pH 8.0), 0.1 mM EDTA by passing the insert DNA over a Sephadex G25 column, the resulting material was mixed with an equimolar quantity of the 40-kb human CD4 minigene insert (NotI digested and prepared by the same method as above), and the final concentration of DNA was adjusted to 1 µg/ml in 10 mM Tris-HCl (pH 8.0), 0.1 mM EDTA. Embryo harvesting, microinjection, and transfer were all performed using standard methods; we used exclusively NOD/LtJ embryos derived from animals supplied by JAX, because, in our experience, embryos from other lines of NOD mice from other suppliers cannot be used to create transgenic animals. PCR analysis of tail DNA was used to identify transgenic offspring which were then further characterized for DQ expression levels by staining of PBMC and FACScan analysis using monoclonal SPV-L3 (anti-DQ monomorphic). Of 21 live births arising from 103 transplanted embryos, 8 animals tested positive for the DQ6 transgene by PCR. However, only one animal (female N8) expressed DQ and human CD4 at appropriate levels and this was used to establish the final transgenic line homozygous for both human CD4 and DQ6. This DQ6 transgenic NOD line has been assigned the formal designation NOD.Tg(CD4,HLA-DQA1*0102,HLA-DQB1*0602)1Ell. Again for brevity, we will henceforth use the designation NOD.DQ6.

To generate IAb knockout NOD.DQ6 animals, IAb knockout NOD animals (see above) were crossed with the NOD.DQ6 line, F₁ animals were intercrossed, and F₂ animals were screened by flow cytometry for the absence of IA and the presence of two copies of DQ6 and CD4 (animals homozygous for the DQ transgene consistently stained 2-fold brighter than hemizygotes). A single breeding pair from the F₂ generation was used to establish the NOD.Cg-H2-Ab1^tm1GruTg(CD4,HLA-DQA1*0102,HLA-DQB1*0602)1Ell (accession ID MGI:3590083) line, for brevity denoted NOD.DQ6/Ab⁰ (note that both this line and the "NOD.DQ8/Ab⁰" line are transgenic for human CD4; the human CD4 transgene is cointegrated with DQ8 or DQ6 and is observed to always transmits alongside of DQ.)

Genotyping

Presence of the HLA-DQ transgene was detected by PCR using the primers TGAGCACAGTGGATTGAG and TGGTAGTTGTCAGGAAGG which target the locus/STS 9–10 (19). Genomic DNA was prepared from tail biopsies using the DNeasy Tissue kit (Qiagen Sciences), and PCRs were done under standard conditions (annealing at 48°C x 1 min). Additional genotyping experiments were also performed on representative transgenic animals to confirm that the HLA-DQ8 (DQA1*0301, DQB1*0302) and HLA-DQ6 (DQA1*0102, DQB1*0602) genes were present as expected. Sequence-specific primers as described by Bunce et al. (20) were used and details are available upon request.

Abs and FACS analysis

Expression of HLA-DQ, IA, TCR/CD3, CD4, CD8, B220, H-2K^b, and/or H-2K^d Ags on various cell populations was analyzed by flow cytometry using a FACScan instrument (BD Biosciences) and CellQuest software, with gates adjusted for the lymphocyte population. Fluorescently tagged Abs used for staining were either prepared in-house or (where indicated) purchased commercially. They included: SPV-L3 (mouse anti-human DQ monomorphic); 10-3-6 (mouse anti-mouse IAg7); H57-597 (hamster anti-mouse TCR); GK1.5 (rat anti-mouse CD4); OKT4 (rat anti-human CD4); CL169F (rat anti-mouse CD8; Cedarlane Laboratories); RA3-6B2 (rat anti-mouse B220); CL9013 (mouse anti-mouse H-2K^b; Cedarlane Laboratories); and SF1-1.1 (mouse anti-mouse H-2K^d). In generating the NOD.DQ8/Ab⁰Rag1⁰ line, as well as screening for the absence of IA, we also followed the animals for the presence of H-2K^b and the absence of H-2K^d. Because the IAb knockout was generated in the 129 strain, the interrupted IAb gene has H-2K^b nearby, whereas any animals retaining an intact copy of IAb would have derived it from NOD, which has H-2K^d nearby.

Bone marrow-derived dendritic cells (DCs)

Bone marrow cells were washed out from dissected, open-ended femurs of 6- to 8-wk-old mice, suspended in endotoxin free RPMI 1640 medium (Invitrogen Life Technologies) and collected by centrifugation (250 x g, 5 min, 20°C). The cells were resuspended in DC culture medium (RPMI 1640 supplemented with 10% FCS (very low endotoxin; Invitrogen Life Technologies), 10 ng/ml mouse IL-4, and 10 ng/ml mouse GM-CSF (both cytokines from PeproTech), diluted to 1 x 10⁶ cells/ml and plated in 6-well tissue culture-treated plates at 4 x 10⁶ cells/well. The cells were incubated at 5% CO₂ and 37°C and 24 h later, all the media in each well were gently aspirated and replaced with fresh medium. On days 2, 4, and 6, half of the media in each well was gently aspirated and replaced. On day 8, the culture was stimulated with mouse TNF- (PeproTech) and LPS (Sigma-Aldrich) (both at 10 ng/ml). On day 9, the remaining cells were harvested and triply stained for the DC marker CD11c (monoclonal HL3; BD Pharmingen), HLA-DQ (monoclonal SPV-L3), and mouse IA (monoclonal 10-3-6) and then analyzed by FACScan.

Monitoring mice for myocarditis by ECG

These methods have been described previously (see supporting text for Ref. 15). In brief, nonsedated mice were placed in a syringe barrel which had been modified to allow access to the feet. Electrodes, with limb leads, attached were clipped to each of the four foot pads and ECG leads I, II, III, augmented voltage right, and augmented voltage left were simultaneously recorded. Parameters measured included sinus rate (P-P interval), PR interval (onset of P wave to onset of QRS complex) and ventricular rate (QRS-QRS interval). First degree heart block was defined as a PR interval >36 msec, a value that is >3 SD above the mean for a large group of normal animals. Complete heart block was diagnosed whenever the P waves (i.e., atrial depolarization events) were completely dissociated from the QRS complexes (i.e., ventricular depolarization events).

ELISA

Falcon 353911 microtiter plates (BD Labware) were coated overnight at 4°C with 50 µl/well of 5 µg/ml porcine cardiac myosin (Sigma-Aldrich) in PBS. The plates were washed with PBS + 0.1% Tween 20 (PBS-T), blocked with 1% w/v BSA in PBS-T and incubated with mouse sera (various dilutions in PBS-T). The plates were then incubated with HRP-conjugated goat anti-mouse IgG (H+L; 1:5000; Jackson ImmunoResearch Laboratories) and developed by using ABTS (Sigma-Aldrich) as a substrate. OD 405 nm – OD 490 nm was measured on a plate reader and expressed as the mean of triplicate wells.

Adoptive transfers

Donor lymphocytes were harvested from 16- to 24-wk-old NOD.DQ8/Ab⁰ mice demonstrating at least first-degree heart block, with the exception of the heart-infiltrating lymphocytes which were harvested from 6-wk-old NOD.DQ8/Ab⁰ animals before the onset of heart block. Splenic lymphocytes were obtained by passing total splenocytes over Lympholyte-M gradients (Cedarlane Laboratories); these cells or enriched lymphocyte populations (see below) were washed in RPMI 1640, pelleted, and finally resuspended in PBS and adoptive transfers were accomplished by tail vein injection into 8-wk-old NOD.DQ8/Ab⁰Rag1⁰ mice. Onset of heart block in recipients was monitored via biweekly ECG.

Echocardiography

M-mode echocardiograms were performed on nonsedated animals using a 15-mHz probe and SONOS 5500 instrument (Agilent Technologies). Moistened fur on the left thorax was shaved using a scalpel blade and LiquaSonic ultrasound transmission gel (Chester Laboratories) was applied to the bare skin before probing. The left ventricle was visualized in diastole and the two-dimensional parasternal short axis was measured.

MACS enrichment for CD4 T cells

A Mouse CD4⁺ T Cell Isolation kit (Miltenyi Biotec) was initially used to obtain a population of enriched CD4 T cells from the spleens of NOD.DQ8/Ab⁰ mice with heart block. This method accomplishes enrichment by magnetic depletion of non-CD4 T cells, including CTLs, B cells, NK cells, DCs, macrophages, granulocytes, and erythroid cells, using a mixture of five different biotin-conjugated Abs against relevant cell surface markers in combination with ferromagnetic anti-biotin microbeads. Lymphocytes were purified from spleens using Lympholyte-M gradients (Cedarlane Laboratories). In accordance with the kit instructions, the cells were incubated with the mixture of biotinylated Abs, washed, incubated with the antibiotin microbeads, and finally passed over a magnetic separation column mounted within the magnetic field of a MACS separator (Miltenyi Biotec). Purity of the enriched CD4 cells which passed through the column was evaluated by flow cytometry (FACScan; BD Biosciences) using Abs against mouse CD4, CD8, TCR, and B220. Cell populations enriched by this method consisted of 91% CD4 T cells and 9% B220 cells (the latter being negative for the three other surface markers).

Cell enrichment by fluorescent cell sorting

Splenic lymphocytes from NOD.DQ8/Ab⁰ mice with heart block were obtained using Lympholyte-M gradients (Cedarlane Laboratories). Lymphocytes from 5 to 10 animals were pooled and aliquots of cells incubated with fluorescently tagged anti-mCD4, anti-mCD8, or anti-B220 and then positively sorted on a FACS Aria instrument (BD Biosciences). Purity of cell populations selected by this method was 99% as assessed by subsequent FACScan analysis.

Isolation of lymphocytes from mouse hearts

Cardiac-infiltrating lymphocytes were isolated from the hearts of 6-wk-old NOD.DQ8/Ab⁰ mice using an adaptation of the method of Afanasyeva et al. (21). Briefly, mice were injected via tail vein with 100 U heparin 10 min before sacrifice, euthanized by CO₂ inhalation, and their hearts were rapidly removed and placed in petri plates containing 5 ml of calcium-free bicarbonate buffer (CFBB) at 37°C. CFBB consisted of 120 mM NaCl, 5.4 mM KCl, 1.2 mM MgSO₄, 1.2 mM NaH₂PO₄, 5.6 mM glucose, 20 mM NaHCO₃, 10 mM 2,2-butanedione monoxime, and 5 mM taurine (pH 7.2); all solutions were equilibrated with 95% O₂-5% CO₂ before use and, unless specified, all reagents were purchased from Sigma-Aldrich. The aorta was cannulated with a 22-gauge needle attached to a syringe containing 3 ml of CFBB at 37°C and the heart was retrogradely perfused slowly over a 3-min period. Thereafter, both cardiac chambers were rapidly flushed with CFBB to ensure that all blood had been removed (heparin prevented any clot formation) and the heart was transferred to a second petri plate containing 5 ml of CFBB-3DE (CFBB supplemented with three different digestive enzymes: 0.5 mg/ml collagenase type D; 0.5 mg/ml collagenase type B (both from Roche); and 0.02 mg/ml protease type XIV). The heart was again retrogradely perfused through the aorta over a 3-min period, this time using 3 ml of CFBB-3DE at 37°C. Perfused organs were incubated in a standard tissue-culture incubator (5% CO₂, 37°C) for 7 min to allow for digestion and then manually dispersed into single cells using sterile razor blades. The cell suspension was filtered through a 40-µm cell strainer (BD Falcon) and the lymphocytes were isolated by centrifugation over Lympholyte-M (Cedarlane Laboratories). Purified lymphocytes were stained for CD4 and FACS sorted, as above, before adoptive transfer.

Histology/immunohistochemistry

For H&E staining, hearts were fixed in 10% buffered formalin (Fisher), embedded in paraffin, sectioned, and stained with H&E using standard methods. For CD4 and CD11b staining, frozen sections were made using OCT Tissue Tek compound (Fisher) and slides were fixed in ice-cold acetone, blocked in 2%FCS/PBS and treated with both avidin and biotin solutions (Vector Laboratories). Rat anti-mouse CD4 (H129.19; BD Pharmingen) and CD11b (M1/70; BD Pharmingen) were used at 1:100 and 1:500, respectively, and detected using a biotin-conjugated goat anti-rat IgG (H+L) at 1:200. Staining was visualized using ABC kit (Vector Laboratories) followed by diaminobenzidine (Sigma-Aldrich) and the slide was counterstained with hematoxylin, dehydrated, and mounted.

	Results

Top Abstract Introduction Materials and Methods Results Discussion Disclosures References

 
To provide a DQ control strain which could be directly compared with our NOD.DQ8 animals, we made use of a second transgenic line that expressed DQ6 (DQA1*0102, DQB1*0602) rather than DQ8, but which we anticipated would be closely matched in all other aspects. Splenocytes and bone marrow-derived DCs from the two different transgenic lines showed appropriate expression of their respective DQ transgenes (i.e., mouse class II MHC (IA) and human class II MHC (DQ) were expressed on the same cells; Fig. 1, A and B, E and F), and levels of DQ were comparable between the two lines. Crossing of the NOD.DQ8 and NOD.DQ6 lines with NOD/Ab⁰ mice (i.e., IAb knockout) resulted in lines expressing the expected human MHC, but without IA (Fig. 1, C and D, G and H). The NOD.DQ8/Ab⁰ and NOD.DQ6/Ab⁰ lines had comparable levels of DQ expression in both splenocytes and DCs and they had comparable numbers of CD3, CD4, and CD8 T cells, as well as B cells, with these numbers being close to those found in wild-type NOD mice (Table I). The NOD.DQ8/Ab⁰ and NOD.DQ6/Ab⁰ lines were also comparable in terms of proliferative responses to staphylococcal enterotoxin B and to recall Ags (both mouse lines showed robust recall responses), and these responses could be blocked with anti-DQ mAbs (data not shown).

View larger version (42K): [in this window] [in a new window]   FIGURE 1. The DQ8 and DQ6 transgenic NOD lines show comparable levels of tissue appropriate DQ transgene expression. Two color FACScan analysis of class II MHC-DQ and -IA expression was performed on splenic lymphocytes and bone marrow-derived DCs from the NOD.DQ8 and NOD.DQ6 lines, as well as the derived IAb knockout (Ab0) lines. Splenocytes and DCs from NOD.DQ8 animals (A and B), NOD.DQ8/Ab0 animals (C and D), NOD.DQ6 animals (E and F), and NOD.DQ6/Ab0 animals (G and H) are shown.

View larger version (19K): [in this window] [in a new window]   FIGURE 2. Unlike the NOD.DQ8/Ab0 line, NOD.DQ6/Ab0 mice do not develop anticardiac autoimmunity. A, Incidence of heart block vs age in the NOD.DQ6/Ab0 line compared with the NOD.DQ8/Ab0 line (n = 16 animals in each group, analyzed concurrently). B, Titers of anticardiac myosin autoantibodies in the NOD.DQ8/Ab0 line (at 12 wk of age) compared with wild-type NOD (at 12 wk), and to the NOD.DQ6/Ab0 line at two different ages (12 wk = •, 40 wk = ). Each serum sample tested was a pool from four individual animals at the specified age.

View larger version (21K): [in this window] [in a new window]   FIGURE 3. Anticardiac autoimmunity occurs in NOD.DQ8/Ab0Rag10 mice following adoptive transfer of splenic lymphocytes from NOD.DQ8/Ab0 mice. Heart block, left ventricular enlargement, and anticardiac myosin autoantibodies are all evident by 12–14 wk posttransfer. A, Incidence of heart block following adoptive transfer of 1 x 106 lymphocytes (n = 8 animals). B, Echocardiographic data for the same mice as in A, measured 14 wk posttransfer (n = 4 animals of each sex), with concurrent controls being age- and sex-matched NOD.DQ8/Ab0Rag10 animals which did not receive adoptive transfer. C, Individual titration curves measuring anticardiac myosin autoantibodies present in each of the adoptive transfer recipients in A, measured 14 wk posttransfer. These are compared with titration curves for NOD.DQ8/Ab0 mice (; serum pooled from three 12-wk-old animals) and to curves for NOD.DQ8/Ab0 Rag10 mice which did not receive adoptive transfers (•; serum pooled from four 22-wk-old animals).

View larger version (49K): [in this window] [in a new window]   FIGURE 5. Gross and/or microscopic pathology of the hearts from the various experiments described in this study. A, Representative hearts from 22-wk-old NOD.DQ8/Ab0Rag10 mice which either did not receive adoptive transfer of lymphocytes (i), received 1 x 106 NOD.DQ8/Ab0 splenocytes 14 wk prior (ii), or received 1 x 106 CD4-enriched splenocytes 14 wk prior (iii). The scale indicates 0.5 cm. B, H&E-stained cardiac sections (x200 magnification) from 22-wk-old NOD.DQ8/Ab0 (i), NOD.DQ8/Ab0Rag10 (ii), NOD.DQ8/Ab0Rag10 14 wk after receiving 1 x 106 NOD.DQ8/Ab0 splenocytes (iii), and NOD.DQ8/Ab0Rag10 14 wk after receiving 1 x 106 CD4-enriched NOD.DQ8/Ab0 splenocytes (iv). C, A high percentage of macrophages can be detected in the infiltrate invading the hearts of NOD.DQ8/Ab0Rag10 mice that received CD4 T cells alone. i–iii, Infiltrate at x100 magnification stained by H&E (i), anti-CD4 (ii), or anti-CD11b (iii). iv–ix, Infiltrate at x200 magnification stained by H&E (iv and vii), anti-CD4 (v and viii), or anti-CD11b (vi and ix). Also shown (D) are heart sections (x200 magnification) from 22-wk-old Ig H chain knockout (i) and 2m knockout (ii) animals (respectively, NOD.DQ8/Ab0IgH0 and NOD.DQ8/Ab02m0).

In the first series of experiments, CD4 T cell enrichment was achieved using a negative selection strategy, which used a mixture of mAbs and Miltenyi Biotec ferromagnetic microbeads. One million CD4 splenic lymphocytes enriched by this method were adoptively transferred to 8-wk-old NOD.DQ8/Ab⁰Rag1⁰ mice via tail vein injection and results for a representative experiment are shown in Fig. 4A. The recipients began to develop complete heart block as early as 4 wk posttransfer and by 12 wk the entire cohort had developed disease (Fig. 4A). Although disease onset appeared to be somewhat accelerated in the animals that received the CD4-enriched cells compared with animals who received the same number of total lymphocytes (total lymphocytes and CD4 cells were taken from the same pool of donor splenocytes), at only one time point (8 wk) did the difference between the two groups reach statistical significance (p < 0.02). All recipient mice were euthanized at 14 wk posttransfer and the hearts were examined for gross and microscopic pathology. Interestingly, hearts from animals that had received the CD4-enriched cell populations were consistently larger than those found in animals that had received the same number of whole splenocytes (Fig. 5Aiii). In fact, hearts from the Rag1⁰ animals that received the CD4-enriched cell transfers were often larger than those typically seen in NOD.DQ8/Ab⁰ mice undergoing spontaneous disease (15). Histopathological examination of the myocardium from animals that had received the CD4-enriched cell transfers showed that mononuclear cells were more numerous and lymphocyte infiltrates more widespread/generalized than those seen in animals receiving transfer of the same number of whole splenocytes (Fig. 5B, iv vs iii).

View larger version (20K): [in this window] [in a new window]   FIGURE 4. Adoptive transfer of CD4 T cells alone (from the spleen or heart) induces myocarditis and heart block in NOD.DQ8/Ab0Rag10 mice, whereas splenic CD8 T cells or B cells do not. A, Incidence of complete heart block following concurrent transfer of 1 x 106 total splenocytes or 1 x 106 enriched CD4 splenocytes (magnetically enriched by negative selection, resulting in 91% CD4, 9% B220 cells) (n = 8 animals/group). B, Incidence of complete heart block following concurrent transfer of 1 x 106 total splenocytes (n = 10) or 1 x 106 FACS-sorted splenic CD4 T cells, CD8 T cells, or B cells (purity 99%, n = 5 animals/group). C, Incidence of complete heart block following transfer of 2.5 x 105 FACS-sorted CD4 lymphocytes isolated from the hearts of 6-wk-old NOD.DQ8/Ab0 mice (n = 3).

Because 100% of our NOD.DQ8/Ab⁰ animals develop myocarditis, and because the autoimmune process begins early in life (mononuclear cell infiltrates against pulmonary vein ensheathing cardiomyocytes are first observed at 3 wk of age, and are well-established in all animals by 5 wk of age (15)), it is reasonable to assume that in all animals, autoreactive lymphocytes are also entering the heart well before any ECG abnormalities are detected. To explore the hypothesis that the early heart-infiltrating lymphocytes from young NOD.DQ8/Ab⁰ animals would be especially virulent, we isolated lymphocytes from the hearts of 6-wk-old animals, FACS sorted these into CD4, CD8, and B220 lymphocyte populations and adoptively transferred the purified cells to young NOD.DQ8/Ab⁰Rag1⁰ mice. Only the CD4 cells were capable of inducing heart block (Fig. 4C); although somewhat lower numbers of cells were transferred (2.5 x 10⁵ CD4 cells/animal, roughly the same number as would be present in 1 x 10⁶ total splenocytes), the heart-infiltrating cells did not appear to induce cardiac pathology any faster than the CD4 cells obtained from spleens.

The mononuclear cell infiltrates present in the hearts of NOD.DQ8/Ab⁰Rag1⁰ recipients of splenic CD4 T cells were further analyzed by immunohistochemistry (Fig. 5C). Staining revealed the presence of scattered CD4 T cells (Fig. 5C, ii, v, and viii), whereas the predominant cell types stained positive for CD11b (Fig. 5C, iii, vi, and ix). Examination of these infiltrates at high magnification revealed predominance of macrophages and, to a lesser extent, lymphocytes, with only the occasional neutrophils observed.

Although purified CD8 T cells or B lymphocytes were not capable of inducing myocarditis in our Rag1 knockout adoptive transfer model, it is still possible that these immune cells play a critical role in initiating the anticardiomyocyte autoimmune response in immune competent NOD.DQ8/Ab⁰ mice. To test this possibility, we crossed our NOD.DQ8/Ab⁰ animals with two different knockout lines: 1) a NOD Ig H chain knockout (IgH⁰) line, and 2) a NOD ₂m knockout (₂m⁰) line. As expected, the resulting NOD.DQ8/Ab⁰IgH⁰ mice were deficient in B lymphocytes (Fig. 6A). Cohorts of NOD.DQ8/Ab⁰IgH⁰ animals at four different ages (8, 12, 18, and 24 wk) were first assessed for the presence of heart block by ECG (Fig. 6C) and then sacrificed and their hearts examined for gross and microscopic pathology (Fig. 5Di). The appearance of heart block was slightly delayed in the IgH⁰ animals compared with a concurrent group of NOD.DQ8/Ab⁰ animals (Fig. 6B); however, this difference did not reach statistical significance (p > 0.1). Histological examination of NOD.DQ8/Ab⁰ IgH⁰ hearts from 22-wk-old animals showed mononuclear cell infiltrates which were very similar to those seen in NOD.DQ8/Ab⁰ animals (Fig. 5, Di vs Bi) and the magnitude of cardiac enlargement was also comparable between the two groups.

View larger version (19K): [in this window] [in a new window]   FIGURE 6. Ig H chain knockout (IgH0) or 2m knockout (2m0) mutations of NOD.DQ8/Ab0 mice suggests differing etiological roles for B vs CD8 T lymphocytes. A, FACScan of B cell populations in NOD.DQ8/Ab0 (clear peaks) vs NOD.DQ8/Ab0IgH0 (filled peak). B, FACScan of CD8 T cell populations in NOD.DQ8/Ab0 (clear peaks) vs NOD.DQ8/Ab02m0 (filled peak). Analysis of peripheral blood lymphocytes is shown; splenic lymphocytes gave the same result. C, Incidence of heart block in the three strains at various ages. D, Comparison of anticardiac myosin autoantibodies present in the serum of the strains indicated. The wild-type NOD (wt NOD) and NOD.DQ8/Ab0 serum samples were pooled from four 12-wk-old animals, whereas the NOD.DQ8/Abb02m0 samples were pooled from 12 animals at each of the two ages. As expected, NOD.DQ8/Ab0IgH0 animals had no measurable serum Abs.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion Disclosures References

 
We serendipitously discovered that NOD.DQ8/Ab⁰ animals develop spontaneous anticardiomyocyte autoimmunity and fatal dilated cardiomyopathy. With only one transgenic line, our initial hypothesis was that this was due to a founder effect, arising possibly because the transgene integration site caused inappropriate overexpression of the DQ within cardiomyocytes. Although RNase protection experiments quickly ruled out elevated levels of DQ mRNA within the heart (data not shown), definitive evidence that the myocarditis was not due to a founder effect came from the work of Lipes and colleagues (14). All three of their independently derived DQ8 transgenic lines developed spontaneous myocarditis and fatal dilated cardiomyopathy. When the three lines described by the Lipes group (14) are considered together with our line, mortality rates appear to be roughly correlated with levels of DQ8 transgene expression. In one published NOD.DQ8/Ab⁰ line, myocarditis was not reported (22) and we speculate that transgene expression levels in these mice are insufficient to induce an autoimmune response against cardiomyocytes, which may reflect unbalanced expression of the two DQ chains (a mixture of separate DQ8 and DQ8 genomic fragments was microinjected in this case; in contrast to our model in which a large human genomic fragment, containing, on a single DNA insert, all DQ and DQ coding exons as well as corresponding human promoters and upstream regulatory elements, was used).

In NOD.DQ8/Ab⁰ mice, there is no apparent reason why the cardiomyocyte is targeted and with our present understanding we could not have predicted this outcome. Therefore, we hypothesized that, in the NOD/Ab⁰ background, perhaps expression of any one of a number of different DQ molecules (i.e., encoded by different haplotypes) would lead to autoimmune myocarditis. To address this possibility, we examined another DQ transgenic line (NOD.DQ6/Ab⁰) for signs of anticardiac autoimmunity (note that the DQ6 animals provide a control for the DQ8 with respect to mouse class I and class III MHC genes). Despite being matched closely to the NOD.DQ8/Ab⁰ line in every aspect except DQ itself, NOD.DQ6/Ab⁰ animals failed to show any signs of myocarditis providing additional evidence that it is the DQ8 molecule which dictates the "choice" of cellular target for the autoimmune response. Although absence of myocarditis in our NOD.DQ6/Ab⁰ mice could again be due to a founder effect, the idea that non-DQ8 haplotypes do not develop myocarditis in this animal model is also supported by preliminary results from our most recent DQ transgenic NOD/Ab⁰ animals, where two independent founder lines (both NOD.DQ8aDQ2b/Ab⁰, i.e., expressing the transheterodimer DQA1*0301, DQB1*0201) have shown no cardiac pathology (K. Suzuki and J. F. Elliott, unpublished observations).

Direct experimental evidence regarding the role of lymphocytes and various lymphocyte subsets in mediating myocarditis in our NOD.DQ8/Ab⁰ animals was obtained by establishing a Rag1 knockout adoptive transfer model. In their initial publication, Lipes and colleagues (14) also determined that splenocytes, but not serum, could transfer disease, with disease appearance assessed by gross and microscopic pathology and (with cell transfers) by the occurrence of anticardiac myosin autoantibodies. Although these authors injected a different number of splenocytes than we used in our experiments (1.5- to 2-fold more, or possibly 15- to 20-fold more as cited later in the manuscript), the fact that our results are very similar to theirs supports the notion that the basic immunological mechanism of myocarditis in the NOD.DQ8/Ab⁰ model will be the same for DQ8 transgenic animals from diverse origins.

CD4 T cells very likely play a role in the autoimmune myocarditis of NOD.DQ8/Ab⁰ mice based on the simple fact that this particular autoimmune phenotype requires DQ8, a molecule whose function it is to present processed Ag to CD4 T cells. Our adoptive transfer model allowed us to demonstrate that indeed purified CD4 T cells alone could induce disease, whereas purified CD8 T cells or B lymphocytes did not. These findings are virtually identical to those described previously for the adoptive transfer of type 1 diabetes in NOD mice (23) and they suggest that in both disease models, once the autoimmune response has become established, CD4 T cells alone are sufficient to orchestrate the destruction of the target cells or tissues (24). In NOD.DQ8/Ab⁰ animals, the destruction is likely accomplished through a proinflammatory delayed-type hypersensitivity-like mechanism, whereby pathogenic CD4 T cells entering the myocardium secrete a number of cytokines including IFN- and IL-17 and these, in turn, cause the recruitment and activation of macrophages (25, 26). Inflammation and tissue destruction then ensue via the production of additional proinflammatory cytokines by the macrophages (e.g., TNF-) and through the release of proteolytic enzymes and reactive oxygen species from macrophages and occasional granulocytes. Immunohistochemical analysis of hearts from NOD.DQ8/Ab⁰Rag1⁰ animals that had received purified CD4 T cells yielded images entirely consistent with the scenario outlined above. The predominance of macrophages bears some resemblance to the myocarditis seen in transgenic mice with cardiomyocyte-specific overexpression of TNF- (9, 11) and suggests that the TNF- and DQ8 transgenic models may share some downstream cellular effector mechanisms. In addition to macrophage-mediated damage, it has been shown in experimental autoimmune myocarditis that CD4 T cells within the myocardial infiltrates express high levels of FasL (27) and this may be another complementary mechanism acting in NOD.DQ8/Ab⁰ animals to cause direct cardiomyocyte destruction and inflammatory cell recruitment.

In comparison to adoptive transfer of total splenocytes, adoptive transfer of purified CD4 T cells into NOD.DQ8/Ab⁰ Rag1⁰ mice consistently induced a greater degree of cardiac enlargement and more extensive mononuclear cell infiltration. There are several possible explanations for this phenomenon. The simplest is that in purified form, 4-fold more of the pathogenic CD4 T cells were actually transferred (i.e., 1 x 10⁶ cells were transferred in each experiment, but spleens contain only 25% CD4 T cells) and the larger number of cells caused more pronounced cardiac destruction. However, other possible explanations can be envisaged, such as a suppressive effect mediated by CD8 T cells which are cotransferred in the total splenocyte populations. This effect might be analogous to what has been observed in rheumatoid arthritis where CD8 T cells are known to have a regulatory or protective effect (28). These ideas could potentially be tested in the future by adoptive transfer of specific mixtures of purified cell populations.

For the CD4 adoptive transfer experiments, we used both a negative selection strategy involving magnetic microbeads and FACS sorting to obtain enriched CD4 T cells. Interestingly, the magnetically enriched CD4 T cells (with contaminating B cells) appeared to induce disease slightly faster than total splenocytes, whereas with the more highly purified CD4 T cells disease occurred with slightly delayed kinetics. Although these differences were not statistically significant (except at one data point), they suggest that B cells may play a minor, ancillary role in disease pathogenesis, likely by presenting cardiac Ags to CD4 T cells.

Results obtained with NOD.DQ8/Ab⁰ mice crossed onto the Ig H chain knockout showed that B lymphocytes are not needed to generate myocarditis, although disease onset was slightly delayed in the NOD.DQ8/Ab⁰IgH⁰ animals, again suggesting that B lymphocytes may play an ancillary role in disease pathogenesis. Our results are similar to those described for experimental autoimmune myocarditis (29), as well as for several other murine autoimmune models, where the full disease phenotype can occur in the absence of B lymphocytes. This finding appears to be in contrast to what is generally reported for NOD mice, where Ig H chain knockout animals fail to develop autoimmune diabetes (30, 31). We suggest that this difference in the apparent requirement for B cells is due to differing threshold requirements for B cell Ag presentation to reach the respective end points. In NOD.DQ8/Ab⁰ animals, the autoimmune response against cardiomyocytes appears to be very robust, with 100% of animals developing fatal disease. In these animals, DCs alone (or in concert with macrophages) can apparently present sufficient cardiac Ags to initiate and amplify a lethal CD4 autoimmune response. The APC function of B lymphocytes is not required; although, if present, it can contribute to earlier onset of heart block and cardiac failure. In contrast, a certain proportion of wild-type NOD mice (15–50% depending on the colony) never progress to diabetes. Thus, the CD4 autoimmune response in NOD is not as robust and we infer that the APC activity provided by B lymphocytes is needed in nearly all animals to stimulate CD4 autoimmune responses beyond the critical threshold needed to trigger diabetes. This idea is consistent with reports that in NOD mice it is the APC activity of B lymphocytes (rather than Ab production) that is required to achieve the full diabetes phenotype (32, 33, 34). Thus, we hypothesize that the differential requirement for B lymphocytes between NOD.DQ8/Ab⁰IgH⁰ and NOD/IgH⁰ animals relates to the fact that the NOD requires the highly efficient APC function of B lymphocytes for diabetes whereas, for cardiomyopathy, this same function is detectable, but not critical, in the NOD.DQ8/Ab⁰ animals.

The results obtained with NOD.DQ8/Ab⁰₂m⁰ mice suggest that CD8 T cells, NKT cells, or NK cells are critical for the spontaneous anticardiomyocyte autoimmune response to arise. Absence of NKT cells (which recognize Ag on CD1d) seems least likely to be the cause of the myocarditis-free phenotype. Both NOD mice (35, 36) and our NOD.DQ8/Ab⁰ mice (data not shown) are already deficient in NKT cells and it has been shown that a further reduction in NKT cells tends to exacerbate rather than mitigate anti-islet autoimmunity in the NOD (35, 36, 37, 38). Recently Yokoyama and colleagues (39) have shown that NK cells acquire functional competence through a specific interaction with self-MHC molecules (termed "licensing"; Ref. 39). In the absence of class I MHC, functional maturation will not occur, thus absence of myocarditis in our NOD.DQ8/Ab⁰₂m⁰ mice could potentially be due to some defect in NK cell development and this idea remains to be explored. In contrast, a central role for CD8 T cells would be consistent with the role that these cells are thought to play in autoimmune diabetes in NOD mice, where NOD/₂m⁰ animals fail to develop insulitis or diabetes and the cloning of islet-infiltrating CD8 T cells has shown them to play a key role in early disease pathogenesis (40, 41, 42, 43). From our results (and by analogy with what is thought to happen in NOD mice), we hypothesize that it is autoreactive CD8 cells which are responsible for the initial killing of cardiomyocytes within the pulmonary veins and atria of young NOD.DQ8/Ab⁰ animals. This releases significant quantities of cardiac Ags, which are presented to CD4 T cells in the context of DQ8 by APCs, triggering and amplifying the CD4 autoimmune response. With time, the CD4 autoimmune response becomes self-perpetuating and CD8 T cells are no longer required, although they likely continue to have some cytotoxic activity against cardiomyocytes.

In comparing the cellular pathogenic mechanisms and progression of autoimmunity in NOD vs NOD.DQ8/Ab⁰ mice, we favor the simple hypothesis that because the two strains share the NOD background, the same types of autoimmune effector cells, arising in the same basic temporal sequence, will occur in both types of mice. In this context, the key similarities between the NOD and NOD.DQ8/Ab⁰ include: 1) a critical role for CD8 T cells early in the autoimmune process; 2) a central role for CD4 T cells in orchestrating the final tissue destruction; and 3) little or no role for Abs in the disease process. As discussed above, the difference between NOD and NOD.DQ8/Ab⁰ with respect to the requirement for B lymphocytes appears to be quantitative rather than qualitative and it does not reflect a fundamental difference in the cellular pathogenic mechanisms acting to cause autoimmunity in the two different strains. The question as to why cells are targeted in the NOD and cardiomyocytes in NOD.DQ8/Ab⁰ has no answer at the present time. This difference in target organ selection must arise in part from differences in MHC class II molecules (IAg7 vs DQ8), because CD4 T cells are involved and because NOD.DQ6/Ab⁰ animals (matched at class I MHC) do not develop anticardiomyocyte autoimmunity. However, because CD8 cells are also involved in pathogenesis, differences in MHC class I molecules (Kd in NOD vs Kb in NOD.DQ8/Ab⁰; both strains express Db) may well play a role as well. In this context, it is interesting that Kb is considered to be protective against diabetes in wild-type NOD mice (44). If IAg7 knockout/DQ8 knockin transgenic NOD animals (i.e., replace IAg7 with DQ8 directly in the NOD MHC locus) can be generated, they will be useful to resolve the issue.

In humans, IDCM is a diagnosis of exclusion, reserved for patients in whom no apparent cause for the myocarditis and downstream cardiac dilation can be identified. Although there are likely a variety of root causes for IDCM (e.g., prior inapparent viral infection, de novo anticardiac autoimmune response, etc.), by the time the patient presents to the clinic these may be impossible to discern. Therefore, at a practical level the challenge for the field is to develop an approach to subclassify the disease based on the dominant cellular immune mechanism causing cardiac damage at the time of presentation and to test therapies that are most likely to affect the relevant immune pathways. We suggest that the end-stage CD4 T cell-driven, macrophage-mediated damage seen in NOD.DQ8/Ab⁰ mice may be representative of one form of immune pathology seen in a subset of humans (arbitrarily named "immunotype 1 myocarditis" for the sake of discussion). In such patients with active disease, we would predict that therapies which inhibit macrophage function and signaling (e.g., agents that block TNF- and/or IL-1 signaling) would be of greatest immediate benefit, therapies that inhibit CD4 T cell activation and function would be useful in the slightly longer term and therapies that primarily affect Ab production would be of little value. These ideas can now be tested using our NOD.DQ8/Ab⁰ and NOD.DQ8/Ab⁰Rag1⁰ adoptive transfer models at various stages of disease progression. Although it is possible that HLA-DQ8 will be found more often in patients with "immunotype 1 myocarditis," it is difficult to extrapolate from our NOD.DQ8/Ab⁰ animals to the human situation. A number of reports have suggested that the DQ8 haplotype (or linked DR4) is associated with idiopathic myocarditis or dilated cardiomyopathy in humans (45, 46, 47, 48), while other studies have implicated non-DQ8/non-DR4 haplotypes (49, 50, 51). There is no consensus on this issue, perhaps a reflection of the fact that in humans other class II MHC molecules (i.e., a second DQ molecule as well as DP and DR molecules) are always expressed alongside of the DQ8 and these may have a significant influence on the global CD4 T cell repertoire. Because the autoimmunity occurs rapidly and reproducibly in the NOD.DQ8/Ab⁰ and NOD.DQ8/Ab⁰Rag1⁰ adoptive transfer models, and because ECG provides a powerful noninvasive method to monitor disease progression, the animals provide a useful in vivo system to test new immunosuppressive agents, either DQ8-specific agents (e.g., copolymers) or general agents (e.g., small molecules, costimulatory blocking agents, etc.).

	Acknowledgments

 
We thank K. Kavanagh for establishing the mouse ECG method and for allowing us to use her equipment and software, L. Elder for the preparation of the H&E slides, E. Michelakis and S. Archer for allowing us the use of their echocardiography instrument and for allowing S. McMurtry to assist with the echocardiogram analysis. We also thank W. Min for advice on isolating DCs, D. Littman for the human CD4 construct, P. Nation for examination of the cell types within the infiltrate, and C. Anderson and L. Guilbert for critical reading of the manuscript.

	Disclosures

Top Abstract Introduction Materials and Methods Results Discussion Disclosures References

 
The authors have no financial conflict of interest.

	Footnotes

 
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

¹ This work was supported by an operating grant (MOP-53198) from the Canadian Institutes of Health Research. J.F.E. was supported by a Research Scientist Award from the Albert Heritage Foundation for Medical Research.

² Current address: Department of Medicine, University of Alberta, Edmonton, Alberta T6G 2B7, Canada.

³ Current address: Department of Oncology, Cross Cancer Institute, University of Alberta, 11560 University Avenue, Edmonton, Alberta T6G 1Z2, Canada.

⁴ Address correspondence and reprint requests to Dr. John F. Elliott, University of Alberta, 1–21 Medical Sciences Building, Edmonton, Alberta, Canada. E-mail address: john.elliott{at}ualberta.ca

⁵ Abbreviations used in this paper: IDCM, idiopathic dilated cardiomyopathy; ECG, electrocardiogram; ₂m, ₂-microglobulin; BAC, bacterial artificial chromosome; DC, dendritic cell; CFBB, calcium-free bicarbonate buffer.

Received for publication December 2, 2005. Accepted for publication April 4, 2006.

	References

Top Abstract Introduction Materials and Methods Results Discussion Disclosures References

 

1. Dec, G. W., R. W. DeSanctis. 1998. Scientific American Medicine. D. C. Dale, and D. D. Federman, eds. Scientific American, New York.
2. Manolio, T. A., K. L. Baughman, R. Rodeheffer, T. A. Pearson, J. D. Bristow, V. V. Michels, W. H. Abelmann, W. R. Harlan. 1992. Prevalence and etiology of idiopathic dilated cardiomyopathy (summary of a National Heart, Lung, and Blood Institute workshop). Am. J. Cardiol. 69: 1458-1466. [Medline]
3. Herskowitz, A., L. J. Wolfgram, N. R. Rose, K. W. Beisel. 1987. Coxsackievirus B3 murine myocarditis: a pathologic spectrum of myocarditis in genetically defined inbred strains. J. Am. Coll. Cardiol. 9: 1311-1319. [Abstract]
4. Rose, N. R., L. J. Wolfgram, A. Herskowitz, K. W. Beisel. 1986. Postinfectious autoimmunity: two distinct phases of coxsackievirus B3-induced myocarditis. Ann. NY Acad. Sci. 475: 146-156. [Medline]
5. Wolfgram, L. J., K. W. Beisel, N. R. Rose. 1985. Heart-specific autoantibodies following murine coxsackievirus B3 myocarditis. J. Exp. Med. 161: 1112-1121. [Abstract/Free Full Text]
6. Neu, N., N. R. Rose, K. W. Beisel, A. Herskowitz, G. Gurri-Glass, S. W. Craig. 1987. Cardiac myosin induces myocarditis in genetically predisposed mice. J. Immunol. 139: 3630-3636. [Abstract]
7. Neu, N., B. Ploier, C. Ofner. 1990. Cardiac myosin-induced myocarditis: heart autoantibodies are not involved in the induction of the disease. J. Immunol. 145: 4094-4100. [Abstract]
8. Smith, S. C., P. M. Allen. 1991. Myosin-induced acute myocarditis is a T cell-mediated disease. J. Immunol. 147: 2141-2147. [Abstract]
9. Kubota, T., C. F. McTiernan, C. S. Frye, S. E. Slawson, B. H. Lemster, A. P. Koretsky, A. J. Demetris, A. M. Feldman. 1997. Dilated cardiomyopathy in transgenic mice with cardiac-specific overexpression of tumor necrosis factor-. Circ. Res. 81: 627-635. [Abstract/Free Full Text]
10. Kubota, T., C. F. McTiernan, C. S. Frye, A. J. Demetris, A. M. Feldman. 1997. Cardiac-specific overexpression of tumor necrosis factor- causes lethal myocarditis in transgenic mice. J. Card. Fail. 3: 117-124. [Medline]
11. Bryant, D., L. Becker, J. Richardson, J. Shelton, F. Franco, R. Peshock, M. Thompson, B. Giroir. 1998. Cardiac failure in transgenic mice with myocardial expression of tumor necrosis factor-. Circulation 97: 1375-1381. [Abstract/Free Full Text]
12. Nishimura, H., T. Okazaki, Y. Tanaka, K. Nakatani, M. Hara, A. Matsumori, S. Sasayama, A. Mizoguchi, H. Hiai, N. Minato, T. Honjo. 2001. Autoimmune dilated cardiomyopathy in PD-1 receptor-deficient mice. Science 291: 319-322. [Abstract/Free Full Text]
13. Okazaki, T., Y. Tanaka, R. Nishio, T. Mitsuiye, A. Mizoguchi, J. Wang, M. Ishida, H. Hiai, A. Matsumori, N. Minato, T. Honjo. 2003. Autoantibodies against cardiac troponin I are responsible for dilated cardiomyopathy in PD-1-deficient mice. Nat. Med. 9: 1477-1483. [Medline]
14. Taylor, J. A., E. Havari, M. F. McInerney, R. Bronson, K. W. Wucherpfennig, M. A. Lipes. 2004. A spontaneous model for autoimmune myocarditis using the human MHC molecule HLA-DQ8. J. Immunol. 172: 2651-2658. [Abstract/Free Full Text]
15. Elliott, J. F., J. Liu, Z. N. Yuan, N. Bautista-Lopez, S. L. Wallbank, K. Suzuki, D. Rayner, P. Nation, M. A. Robertson, G. Liu, K. M. Kavanagh. 2003. Autoimmune cardiomyopathy and heart block develop spontaneously in HLA-DQ8 transgenic IA knockout NOD mice. Proc. Natl. Acad. Sci. USA 100: 13447-13452. [Abstract/Free Full Text]
16. Liu, J., L. E. Purdy, S. Rabinovitch, A. M. Jevnikar, J. F. Elliott. 1999. Major DQ8-restricted T-cell epitopes for human GAD65 mapped using human CD4, DQA1*0301, DQB1*0302 transgenic IA(null) NOD mice. Diabetes 48: 469-477. [Abstract]
17. Okada, K., J. M. Boss, H. Prentice, T. Spies, R. Mengler, C. Auffray, J. Lillie, D. Grossberger, J. L. Strominger. 1985. Gene organization of DC and DX subregions of the human major histocompatibility complex. Proc. Natl. Acad. Sci. USA 82: 3410-3414. [Abstract/Free Full Text]
18. Killeen, N., S. Sawada, D. R. Littman. 1993. Regulated expression of human CD4 rescues helper T cell development in mice lacking expression of endogenous CD4. EMBO J. 12: 1547-1553. [Medline]
19. Ellis, M. C., A. H. Hetisimer, D. A. Ruddy, S. L. Hansen, G. S. Kronmal, E. McClelland, L. Quintana, D. T. Drayna, M. S. Aldrich, E. Mignot. 1997. HLA class II haplotype and sequence analysis support a role for DQ in narcolepsy. Immunogenetics 46: 410-417. [Medline]
20. Bunce, M., C. J. Taylor, K. I. Welsh. 1993. Rapid HLA-DQB typing by eight polymerase chain reaction amplifications with sequence-specific primers (PCR-SSP). Hum. Immunol. 37: 201-206. [Medline]
21. Afanasyeva, M., D. Georgakopoulos, D. F. Belardi, A. C. Ramsundar, J. G. Barin, D. A. Kass, N. R. Rose. 2004. Quantitative analysis of myocardial inflammation by flow cytometry in murine autoimmune myocarditis: correlation with cardiac function. Am. J. Pathol. 164: 807-815. [Abstract/Free Full Text]
22. Kudva, Y. C., G. Rajagopalan, R. Raju, R. S. Abraham, M. Smart, J. Hanson, C. S. David. 2002. Modulation of insulitis and type 1 diabetes by transgenic HLA-DR3 and DQ8 in NOD mice lacking endogenous MHC class II. Hum. Immunol. 63: 987-999. [Medline]
23. Christianson, S. W., L. D. Shultz, E. H. Leiter. 1993. Adoptive transfer of diabetes into immunodeficient NOD-scid/scid mice: relative contributions of CD4⁺ and CD8⁺ T-cells from diabetic versus prediabetic NOD.NON-Thy-1a donors. Diabetes 42: 44-55. [Abstract]
24. DiLorenzo, T. P., R. T. Graser, T. Ono, G. J. Christianson, H. D. Chapman, D. C. Roopenian, S. G. Nathenson, D. V. Serreze. 1998. Major histocompatibility complex class I-restricted T cells are required for all but the end stages of diabetes development in nonobese diabetic mice and use a prevalent T cell receptor chain gene rearrangement. Proc. Natl. Acad. Sci. USA 95: 12538-12543. [Abstract/Free Full Text]
25. Park, H., Z. Li, X. O. Yang, S. H. Chang, R. Nurieva, Y. H. Wang, Y. Wang, L. Hood, Z. Zhu, Q. Tian, C. Dong. 2005. A distinct lineage of CD4 T cells regulates tissue inflammation by producing interleukin 17. Nat. Immunol. 6: 1133-1141. [Medline]
26. Witowski, J., K. Ksiazek, A. Jorres. 2004. Interleukin-17: a mediator of inflammatory responses. Cell. Mol. Life Sci. 61: 567-579. [Medline]
27. Ishiyama, S., M. Hiroe, T. Nishikawa, T. Shimojo, S. Abe, H. Fujisaki, H. Ito, K. Yamakawa, N. Kobayashi, T. Kasajima, F. Marumo. 1998. The Fas/Fas ligand system is involved in the pathogenesis of autoimmune myocarditis in rats. J. Immunol. 161: 4695-4701. [Abstract/Free Full Text]
28. Taneja, V., N. Taneja, T. Paisansinsup, M. Behrens, M. Griffiths, H. Luthra, C. S. David. 2002. CD4 and CD8 T cells in susceptibility/protecti