Effects of Intranasal Administration of Recombinant Murine Interferon-γ on Murine Acute Myocarditis Caused by Encephalomyocarditis Virus
Background—Viral myocarditis has been strongly implicated in the pathogenesis of dilated cardiomyopathy as well as acute myocarditis. Among the antiviral therapies, interferons (IFNs) have been widely studied and become very important in clinical practice.
Methods and Results—To investigate the possibilities of IFN therapy in viral myocarditis, we analyzed the effects of recombinant murine interferon (mIFN)-γ and natural mIFN-α/β by the intranasal and intramuscular routes on the development of acute murine myocarditis caused by encephalomyocarditis virus. Both mIFN-γ and mIFN-α/β treatment by either route significantly increased the survival rate; none of the mIFN-γ–treated mice died. The effect of mIFN-γ was significantly greater than that of mIFN-α/β. Furthermore, intranasal administration of mIFN-γ significantly suppressed virus replication and inflammation in the heart.
Conclusions—Our data demonstrate that IFN therapy, especially intranasal administration of IFN-γ, dramatically improved the prognosis of acute murine viral myocarditis by suppressing virus replication and raises the possibility of antiviral therapy with IFN-γ in patients with acute myocarditis.
The EMC virus, which is a member of Cardiovirus of the family of Picornaviridae (which includes Coxsackievirus), causes acute myocarditis occasionally with fatal outcome in humans and various animal species. DBA/2 mice infected with EMC virus show marked dilatation and hypertrophy of the heart in the chronic stage of myocarditis, which is characterized by lesions similar to those seen in dilated cardiomyopathy in humans.1 2
Among the antiviral therapies currently available in clinical practice, IFNs have become one of the most important and are under investigation to explore other potential uses. It has been demonstrated that treatment of EMC virus–infected mice with anti-IFN globulin resulted in the death of all animals with marked virus replication in the visceral organs.3 This indicates that IFNs are one of the essential components of the host immune system against viral infection. Regarding in vivo anti-EMC virus activity, human IFN-α A/D administered intraperitoneally has been mainly evaluated.2 4 5 6 In addition, there is a report in which recombinant human leukocyte IFN-α or mIFN-γ given intraperitoneally prevented the death of mice that were inoculated with a lethal dose of EMC virus.7 However, studies of IFN by the intranasal route are scarce. To our knowledge, there is the only study in which mIFN-γ was protective against Sendai virus replication in mice when given intranasally but not intravenously.8
Recent advances in medical science have revealed that the lungs are capable of absorbing various substances of high or low molecular weight that are inhaled or administered for local or systemic distribution.9 10 In fact, the bioavailability of IFN-α, which has a high molecular weight, was as high as 56% of the dose given to anesthetized rats by the intratracheal route.11
This study was conducted to compare the protective effect of mIFN-γ with that of natural mIFN-α/β on EMC virus–induced myocarditis by the intranasal and intramuscular routes on the basis of survival rate and duration, virus titers in the heart, and pathology of the heart determined in DBA/2 mice.
Recombinant mIFN-γ (lot 004003; titer, 6.52×105 IU/mL; specific activity, 6.27×105 IU/mg of protein) and natural mIFN-α/β (lot 005; titer, 6.95×107 IU/mL; specific activity, 3.56×107 IU/mg of protein) were generously given by Hayashibara Biochemical Laboratories, Inc (Okayama, Japan) and stored at −80°C until use.
Virus and Inoculation
The M variant of EMC virus ATCC VR-1314 was obtained from the American Type Culture Collection and was used after three passages in DBA/2 slc mice (SLC Japan). Hearts of the infected mice were excised and emulsified in 10% heat-inactivated FBS in 10% Eagle’s MEM, pH 7.6, through the use of a model K Polytron (Kinematica AG) and was clarified by centrifugation at 4000 rpm for 10 minutes at 4°C. The supernatant was divided into aliquots and frozen at −80°C. The mean titer of the virus preparation was 1.5×105 pfu/mL when determined on L929 monolayered cells. The diluent, which contained 500 pfu/mL, was administered to 3-week-old mice in a volume of 0.1 mL IP.
For histopathological examination of the infected mice, the virus was diluted to 300 pfu/mL and administered to 6-week-old DBA/2 slc mice in a volume of 0.1 mL IP.
Administration of IFNs
Mice were anesthetized with ketamine HCl (Ketaral-50; Sankyo Zoki Co); then, mIFN-γ or mIFN-α/β solution was administered into the right nasal cavity once daily in a volume of 20 μL with the use of a micropipette (Pipetman p-20; Gilson) or injected intramuscularly into the thigh in a volume of 0.1 mL. Treatment with mIFNs lasted for 7 days, starting 2 days before virus inoculation (days −2 to 4) to determine the life-prolonging effect and for histopathological study.
In the experiment to determine the effect of timing of mIFN administration, treatment started 2 days before virus inoculation (days −2 to 4), 1 day after virus inoculation (days 1 to 7), or 3 days after inoculation (days 3 to 9).
Survival time was monitored until 14 days after virus inoculation.
Determination of EMC Virus Titer in Viscera
To determine the effect of mIFNs in reducing virus yield in the viscera, IFN treatment lasted for 6 days, starting from 2 days before virus inoculation (days −2 to 3).
The heart, brain, and pancreas were removed from individual animals 4 days after inoculation, rinsed with physiological saline, and preserved at −80°C until plaque assay. Tissues of the infected viscera were homogenized with 9 vol of 10% Eagle’s MEM using a Polytron and centrifuged at 4000 rpm for 10 minutes at 4°C. The supernatant was serially diluted in 10 vol of Eagle’s MEM, and 0.1-mL aliquots of the dilutions were overlaid on the monolayered L929 cells. The confluent L929 cell monolayers were incubated with 0.1 mL of serial 10-fold dilutions of virus-containing supernatant in 10% Eagle’s MEM. The viruses were adsorbed onto the L929 cell monolayers for 60 minutes at 37°C in 5% CO2 with tilting at 10-minute intervals. After adsorption, the cells were washed with phosphate-buffered saline without Ca2+ and Mg2+. Then, 2 mL of the 10% Eagle’s MEM containing 1% methylcellulose was overlaid and incubated for 4 days at 37°C in 5% CO2. After the overlay medium was discarded, the cells were exposed to 0.05% crystal violet in 10% buffered formalin, pH 6.3, overnight at ambient temperature for fixing and staining. After the dye solution was discarded and the cell plates were rinsed with chilled physiological saline, the plates were dried, plaques formed on the monolayer were counted, and the virus titer was calculated (detection limit, <102 pfu/g of tissue). The virus yield in the IFN-treated group was compared with that in the infected control group.
Histopathology of the Heart
Treated, nontreated, and mock-infected mice were killed, and their hearts were removed 10 days after virus inoculation. The hearts were rinsed immediately with physiological saline and sectioned transversely into three parts. The portion with the largest ventricular circumference was fixed in 10% buffered formalin solution. The tissues were embedded in paraffin and stained with hematoxylin and eosin.
On the basis of microscopic examination of the ventricular tissue appearance, myocardial damage was evaluated as follows: the absence of damage was scored as 0; lesions in <25% of the myocardium, +1; lesions in 25% to 50% of the tissues, +2; lesions in 50% to 75% of the tissues, +3; and lesions in >75% of the tissues, +4. Pathological scores were assessed independently by three experts in pathology.
Determination of ED50 In Vitro by Plaque Assay
For the study of the antiviral efficacy of mIFN-α/β and mIFN-γ, EMC virus was prepared by three passages in DBA/2 mice and two passages in FL cells. The supernatant obtained was stored at −80°C. The virus was thawed and diluted with 10% Eagle’s MEM to a concentration of 250 pfu/mL immediately before use. The host cells used for this assay were monolayered L929 cells, which were prepared by incubation of 1.25×105 cells on 12-well plates for 2 days at 37°C in 5% CO2/95% air. mIFN-α/β and mIFN-γ were prepared by twofold dilution from 50 to 0.4 IU/mL with 10% Eagle’s MEM.
Next, 2 mL of the mIFN suspension was added to three wells at each concentration. Two milliliters of 10% Eagle’s MEM was added to the control wells. After the cells were exposed to mIFN for 24 hours, the medium was discarded and the cells were rinsed with 10% Eagle’s MEM. EMC virus was placed in a volume of 0.2 mL each and allowed to adsorb for 60 minutes at 37°C with tilting at 10-minute intervals. The cells were rinsed with 10% Eagle’s MEM, followed by incubation with 2 mL of 10% Eagle’s MEM containing 1% methylcellulose at 37°C for 4 days in 5% CO2/95% air. After the overlay medium was discarded, the cells were fixed and stained with 0.05% crystal violet in 10% buffered formalin solution, pH 6.3, overnight at an ambient temperature. After rinsing with chilled physiological saline, the numbers of plaques on the monolayer cells were counted, and the 50% plaque reduction rate was calculated as the ED50 value.
Survival curves was determined according to the Kaplan-Meier method, and then probability values were calculated according to the generalized Wilcoxon test to evaluate the life-prolonging effects of IFNs. One-way ANOVA followed by two-way Dunnett’s test was used to evaluate the differences of virus yields in the viscera between the groups. The Mann-Whitney U test was used to evaluate the histopathological gradings in the heart. Bonferroni’s correction was used for the multiple comparisons. The level of significance was set at P<.05.
Effect of mIFN Treatment on the Survival of Mice
In the intranasal administration experiment, all infected control mice died 5 to 7 days after inoculation. In sharp contrast, none of the mice that received mIFN-γ at 10 000 IU/d from days −2 to 4 died by the end of the 14th day after inoculation. The life-prolonging effect of mIFN-γ was significant at P=.0002. The survival rate for the mice treated with mIFN-α/β at 100 000 IU/d for the same treatment periods was only 10%, but the life-prolonging effect of mIFN-α/β was also significant at P=.0352 (Fig 1A⇓). The effect of mIFN-γ at 10 000 IU/d was significantly (P=.0015) greater than that of mIFN-α/β at 100 000 IU/d.
In the intramuscular administration experiment, the survival rate of the infected control mice through the 14-day observation period was 6.7% (Fig 1B⇑). Again, none of the mice that received mIFN-γ at 1000 IU/d from days −2 to 4 had died by the end of the 14-day postinoculation observation period, demonstrating a significant (P=.0002) life-prolonging effect. The survival rate for the animals treated with mIFN-α/β administered at 10 000 IU/d for the same treatment periods was 53.3%, and the life-prolonging effect of mIFN-α/β was significant at P=.0004 (Fig 1B⇑). The effect of mIFN-γ at 1000 IU/d was significantly (P=.0219) greater than that of mIFN-α/β at 10 000 IU/d.
Effect of mIFN Treatment on Virus Yields in Visceral Organs
The antiviral efficacies of mIFN-γ and mIFN-α/β administered intranasally or intramuscularly in the target organ, the heart of infected mice, were analyzed through the use of the plaque assay. The virus yield in the hearts of mice treated with mIFN-γ 10 000 IU/d intranasally was 3.7±1.0 log10 pfu/g of tissue, which is significantly (P<.01) lower than that of infected control mice (6.3±0.2 log10 pfu/g of tissue) (Fig 2A⇓). Treatment with mIFN-α/β at 100 000 IU/d intranasally also significantly (P<.05) reduced the virus yield to 4.0±2.1 log10 pfu/g of tissue, but the effect was less than that of mIFN-γ (Fig 2A⇓).
In the intramuscular administration experiment, there was no significant effect of virus yield reduction in the mIFN-γ or mIFN-α/β treatment (Fig 2B⇑).
Other Visceral Organs
We also analyzed the antiviral efficacies of mIFN-γ and mIFN-α/β administered intranasally or intramuscularly in the brain and pancreas through the use of the plaque assay. mIFN-γ at 10 000 IU/d intranasally also reduced the virus yield to 4.4±1.5 and 5.1±1.5 log10 pfu/g of tissue, respectively, which again was significantly (P<.01 and <.05, respectively) lower than that of infected control mice (7.0±0.3 and 7.2±0.3 log10 pfu/g of tissue, respectively). mIFN-α/β at 100 000 IU/d intranasally also significantly reduced the virus yield to 3.5±1.3 and 4.5±2.3 log10 pfu/g of tissue, respectively. Both again were significantly(P<.01 and <.05, respectively) lower than that of infected control mice. In the intramuscular administration experiment, there again was no significant effect of virus yield reduction in the mIFN-γ or mIFN-α/β treatment.
These results indicate that protection against EMC virus infection was achieved through an antiviral effect of mIFN-γ or mIFN-α/β administered intranasally in the major viscera.
Effect of mIFN Treatment on Histology of the Heart
There were extensive necrotic foci of myocytes with calcification and cellular infiltration in the hearts of infected control mice (Fig 3B⇓). The cellular infiltrates mainly consisted of mononuclear cells. Polymorphonuclear leukocytes were also observed near the large necrotic lesions, whereas only minimal cardiac lesions were observed in the mice treated with mIFN-γ at 10 000 IU/d intranasally (Fig 3A⇓).
Necrosis of Cardiac Myocytes
The administration of mIFN-γ at 10 000 IU/d intranasally significantly (P=.0462) suppressed necrosis of myocytes in the ventricular myocardium, as demonstrated by the pathology scores of +0.1±0.4 in the mIFN-γ–treated group and +1.0±0 in the control group (Table⇓). The scores for mIFN-γ and mIFN-α/β administered intramuscularly were comparable to the scores for infected controls (Table⇓).
Similar to the examination of necrotic myocytes, the administration of mIFN-γ to the animals at 10 000 IU/d intranasally significantly suppressed (P=.00462) cellular infiltration (inflammation) into the ventricular myocardium, as demonstrated by the pathology scores of +0.1±0.4 in the mIFN-γ–treated group and +1.0±0 in the infected control group (Table⇑). The scores for mIFN-γ and mIFN-α/β administered intramuscularly were comparable to the scores of infected control animals (Table⇑).
Calcification in Ventricular Myocardium
The administration of mIFN-γ and mIFN-α/β to the animals intranasally produced no significant differences in the pathology score for calcification in the myocardium compared with the infected control group (Table⇑). The scores for mIFN by the intramuscular route also were comparable to those of the infected control animals (Table⇑).
These results indicate that protection against viral myocarditis was achieved through the effect of mIFN-γ administered intranasally.
Effect of Administration Route of IFN-γ on Survival of Mice
In an experiment in which we used four doses by the intranasal route and three doses by the intramuscular route, the intranasal route was slightly more effective than the intramuscular route in prolonging the life of infected animals in all doses, even at the lowest dose of 500 IU/d for 7 days starting from 2 days before virus inoculation. The effect was much more apparent at the higher doses of 1000, 5000, and 10 000 IU/d intranasally at significance levels of P=.0028, .0004, and .0004, respectively (Fig 4A⇓). Because all animals treated at the higher two doses survived throughout the observation period, a dose-effect relationship was not established. The administration of mIFN-γ by the intramuscular route also was clearly effective at all tested doses of 100, 500, and 1000 IU/d for 7 days at significance levels of P=.0063, .0018, and .0018, respectively (Fig 4B⇓). The survival rates obtained by the intramuscular route appeared to be dependent on the dose, but dose-dependency was not demonstrated statistically.
Effect of Timing of IFN-γ Administration on Survival of Mice
mIFN-γ was administered to the infected mice at doses of 100 000 and 1000 IU/d by the intranasal or intramuscular routes for 7 days starting 2 days before, 1 day after, or 3 days after virus inoculation (Fig 5A⇓ and 5B⇓, respectively). The life-prolonging effect was apparent by both routes at a significance level of P=.0003 when the treatment was started at 2 days before or 1 day after virus inoculation. The protective effect was also seen when the treatment was started at 3 days after virus inoculation, but the difference between the treated and the infected control group was not statistically significant.
Although acute myocarditis usually is idiopathic, it has been generally accepted that most cases of the disease are caused by picornaviruses such as EMC virus. Because patients with acute myocarditis present with acute heart failure, and some of them develop dilated cardiomyopathy later in the course of the disease, much effort has been made to develop new therapeutic interventions. Various recombinant cytokines, including mIFN-γ, have been recognized to significantly augment host resistance against virus infection in mice when administered by the intranasal route compared with the intravenous route.8 The lung, which is known to be highly permeable to high-molecular-weight molecules,9 10 seems to be an efficient administration route for bioactive peptides. A study of intratracheal administration of peptides and proteins in rats demonstrated that human IFN-α had an absolute bioavailability of >56%.11 Therefore, we tried to treat murine acute myocarditis induced by EMC virus with mIFNs by the intranasal route.
In this study, we found that both mIFN-γ and mIFN-α/β by either intranasal or intramuscular route were effective in prolonging survival time of the mice with viral myocarditis, when the treatment was started from 1 day after virus inoculation. However, the intranasal inhalation was found to be more effective than the pain-inflicting intramuscular injection in suppressing viral growth in the target organs, including heart, brain, and pancreas. In addition, intranasal administration of mIFN-γ was more effective in suppressing the disease than that of mIFN-α/β. In a separate in-house experiment, the 50% plaque reduction assay of the antiviral activity using L929 cells and EMC virus in this study provided ED50 values for mIFN-γ and mIFN-α/β of 1.0 and 1.3 IU/mL (95% confidence limits, 0.75020 and 1.38097 and 0.80069 and 1.83707), respectively (Fig 6⇓). The two types of mIFN exhibited similar activity against the myocarditis-inducing virus. Similar results have been reported regarding the antiviral efficacies of human type I (IFN-α/β) and type II (IFN-γ) IFNs using human fibroblast FS-4 cultures.12 Regarding IFN-γ-therapy in other animal models, Iida et al8 demonstrated that intranasal but not intravenous administration of IFN-γ protected mice against Sendai virus infection. This suggests that factors other than bioavailability may be involved in the differences in effectiveness among the administration routes. The authors suggested that the activation of alveolar macrophages and neutrophils by the intranasal route was one of the mechanisms involved. However, the precise mechanism by which that intranasal route was more effective than the intravenous or intramuscular route is still unknown and remains to be clarified.
The immunomodulating activities of mIFN-γ, including macrophage activation in lung alveoli, are unique and not provided with mIFN-α or -β by the pulmonary route. Recent advances in cell biology revealed that IFN-γ is the most potent macrophage activator; it produces macrophages to kill virus-infected cells and proteins to destroy pathogens.13 14 The antiviral activity of macrophages appears to be mediated by nitric oxide and possibly other reactive nitrogen intermediates synthesized by nitric oxide synthase, which is induced by macrophage-activating agents such as IFN-γ. After ingestion by macrophages, microbes such as viruses are digested and degraded into peptide fragments in the cytoplasm and then presented on the cell surface by major histocompatibility complex molecules. T cells specifically recognize processed antigens through their T-cell receptors and then proliferate and release chemicals that help other immune cells to combat or eradicate infections. Because IFN-γ is a much more potent inducer of major histocompatibility complex class II antigen expression on macrophages than IFN-α/β,15 the decreased virus population in the target organs revealed in the present study seems to be well understood.
Regarding murine viral myocarditis caused by coxsackievirus B3, IFN-γ is mainly synthesized by the infiltrating natural killer (NK) cells in the early stage of acute myocarditis.16 17 It was shown that depletion of NK cells by treatment with anti-asialo GM1 antibody plus complement aggravates coxsackievirus B3–induced murine myocarditis by enhancing virus replication in the heart.18 This strongly suggests that NK cells play a critical role in limiting virus replication by killing virus-infected cells as well as synthesizing IFN-γ. These data support the protective effect of IFN-γ in viral myocarditis because IFN-γ is known to activate NK cells as well as directly suppress virus replication. In the present study, we also analyzed the relative distribution of phenotypic markers among the infiltrating cells, especially NK cells, cytotoxic T lymphocytes (CTLs), and T-helper cells in the heart with viral myocarditis with or without IFN-γ treatment. We found that there was a tendency toward decrease in the percentage of NK cells and CTLs among the infiltrating cells as well as a marked decrease in the total number of the infiltrating cells in the IFN-γ–treated group compared with the control group (data not shown). This supports the virus-suppressing effect of IFN-γ because NK cells and CTLs are thought to kill virus-infected myocardial cells.
Thus, the striking effects of mIFN-γ achieved by the intranasal route in suppressing virus replication in the heart and other organs, improving myocardial inflammation, and prolonging life of the infected mice have demonstrated that this cytokine is very promising as a biotherapeutic agent against viral myocarditis. However, in contrast to the experimental viral myocarditis, viruses frequently cannot be detected in patients with acute myocarditis because such patients usually come to the hospital after they developed clinical manifestations that are mainly the result of inflammatory responses, by which time the viruses have disappeared. In the present study, the effectiveness of IFN therapy was significantly reduced when given after virus inoculation. This may limit the clinical application of the IFN therapy to some extent.
Regarding the effect of IFN therapy on human myocarditis, until now there have been only a few reports. Siegener et al19 reported that enteroviral RNA was detected by in situ hybridization in 20 of 77 patients with dilated cardiomyopathy who underwent endomyocardial biopsy. Four enterovirus-positive patients with hemodynamic deterioration were given IFN-α subcutaneously for 6 months. Then, all 4 patients had improved hemodynamic and clinical parameters, and in 2 of them enteroviral RNA was not detectable in a subsequent biopsy. Mirić et al20 also reported that treatment with IFN-α subcutaneously was effective in improving cardiac function in 11 of 14 patients with dilated cardiomyopathy. Considering that these results were obtained through subcutaneous administration of IFN-α, intranasal administration of IFN-γ will have a remarkable value in the treatment of patients with acute viral myocarditis and dilated cardiomyopathy.
It is well known that in acute viral myocarditis, cell-mediated autoimmune mechanisms induced by viral infection as well as the direct cytopathic effect of viruses are important in the pathogenesis of the myocardial damage. Evidence has accumulated that the persistent myocardial cell damage involved in viral myocarditis may cause continuous destruction of contractile proteins and facilitate fibrosis, which may lead to dilated cardiomyopathy. Therefore, immunomodulating therapy against the cell-mediated autoimmunity induced by virus infection along with IFN therapy would be the best way to treat patients with acute myocarditis and chronic ongoing myocarditis.
Selected Abbreviations and Acronyms
|ED50||=||50% effective dose|
|MEM||=||minimal essential medium|
The authors thank Masahisa Kyogoku, MD, for his support of this investigation. We also thank Masatoshi Sakashita, Tomoko Doi, Atsumi Kawai, and Naoko Ise for their excellent technical assistance.
- Received June 19, 1997.
- Revision received October 21, 1997.
- Accepted October 23, 1997.
- Copyright © 1998 by American Heart Association
Matsumori A, Kawai C. An animal model of congestive (dilated) cardiomyopathy: dilatation and hypertrophy of the heart in the chronic stage in DBA/2 mice with myocarditis caused by encephalomyocarditis virus. Circulation. 1982;66:355–360.
Kanda T, Nagaoka H, Kaneko K, Wilson JE, McManus BM, Imai S, Suzuki T, Murata K, Kobayashi I. Synergistic effects of tacrolimus and human interferon-α A/D in murine viral myocarditis. J Pharmacol Exp Ther. 1995;274:487–493.
Gresser I, Tovey MG, Bandu M, Maury C, Brouty-Boye D. Role of interferon in the pathogenesis of virus diseases in mice as demonstrated by the use of anti-interferon serum, I: rapid evolution of encephalomyocarditis virus infection. J Exp Med. 1976;144:1305–1315.
Weck PK, Rinderknecht E, Estell DA, Stebbing N. Antiviral activity of bacteria-derived human alpha interferons against encephalomyocarditis virus infection of mice. Infect Immun. 1982;35:660–665.
Connel EV, Cerruti RL, Sim IS. Interactions between recombinant interferons alpha and gamma in the treatment of experimental virus infections in mice. In: Stewart WE, Schellekens H, eds. The Biology of the Interferon System 1985. Amsterdam: Elsevier Science Publishers BV; 1986;419–422.
Schanker LS, Mitchell EW, Brown RA Jr. Species comparison of drug absorption from the lung after aerosol inhalation or intratracheal injection. Drug Metab Dispos. 1986;14:79–88.
Patton JS, Trinchero P, Platz RM. Bioavailability of pulmonary delivered peptides and proteins: α-interferon, calcitonins and parathyroid hormones. J Control Releas. 1994;28:79–85.
Rubin BY, Gupta SL. Differential efficacies of human type I and II interferons as antiviral and antiproliferative agents. Proc Natl Acad Sci U S A. 1980;77:5928–5932.
Johnson HM, Bazer FW, Szente BE, Jarpe MA. How interferons fight disease. Sci Am. 1994;270:40–47.
Huang S, Hendriks W, Althage A, Hemmi S, Bluethmann H, Kamijo R, Vilček J, Zinkernagel RM, Ague M. Immune response in mice that lack the interferon-γ receptor. Science. 1993;259:1742–1745.
Kamijo R, Shapiro D, Le J, Huang S, Aguet M, Vilček J. Generation of nitric oxide and induction of major histocompatibility complex class II. Proc Natl Acad Sci U S A. 1993;90:6626–6630.
Seko Y, Matsuda H, Kato K, Hashimoto Y, Yagita H, Okumura K, Yazaki Y. Expression of intercellular adhesion molecule-1 in murine hearts with acute myocarditis caused by coxsackievirus B3. J Clin Invest. 1993;91:1327–1336.
Godeny EK, Gauntt CJ. Murine natural killer cells limit coxsackievirus B3 replication. J Immunol. 1987;139:913–918.
Siegener MS, Heim A, Figulla HR. Subclassification of dilated cardiomyopathy and interferon treatment. Eur Heart J. 1995;16(suppl O):147–149.
Mirić M, Mišković A, Vasiljević JD, Keserović N, Pešić M. Interferon and thymic hormones in the therapy of human myocarditis and idiopathic dilated cardiomyopathy. Eur Heart J. 1995;16(suppl O):150–152.