Skip Navigation

Increased Level of Interferon-α in Blood of Patients with Insulin-Dependent Diabetes Mellitus: Relationship with Coxsackievirus B Infection

  1. Wassim Chehadeh1,a,
  2. Jacques Weill2,
  3. Marie-Christine Vantyghem3,
  4. Gunar Alm4,
  5. Jean Lefèbvre3,
  6. Pierre Wattré1,a and
  7. Didier Hober1,a
  1. 1Laboratoire de Virologie, Institut Gernez-Rieux, Lille, France
  2. 2Service d'Endocrinologie Pédiatrique, Hôpital Jeanne de Flandre, Lille, France
  3. 3Service d'Endocrinologie et Maladies Métaboliques, Clinique Marc-Linquette, Centre Hospitalier Régional et Universitaire, Lille, France
  4. 4Department of Veterinary Immunology, Biomedical Center, Uppsala, Sweden
  1. Reprints or correspondence: Dr. Didier Hober, Laboratoire de Virologie, CHRU, Institut Gernez-Rieux, 59037 Lille Cedex, France (dhober{at}chru-lille.fr).
 
Next Section

Abstract

The activation of the interferon (IFN)-α system and its relationship with coxsackievirus B (CVB) infection has been analyzed in 56 patients with insulin-dependent diabetes mellitus (IDDM; 25 children and 31 adults). Elevated levels of IFN-α were found in plasma of 70% of patients (39/56), and a positive detection of IFN-α mRNA in blood cells by reverse transcriptase-polymerase chain reaction (RT-PCR) was observed in 75% of patients (42/56). Enterovirus (EV) RNA assayed by seminested RT-PCR was detected in the blood of 50% of IFN-α-positive patients but not in any IFN-α-negative patients. The results of genotype analysis of amplified EV RNA sequences (5 CVB2, 8 CVB3, and 8 CVB4) were concordant with the results of CVB-neutralization tests. The comparison between IFN-α, EV RNA, and serology suggested that the proportion of CVB infection associated with IFN-α positivity might be higher than is predicted from the investigation of EV RNA. Together, the results suggest that, in a majority of cases, a CVB infection is associated with clinical IDDM.

Insulin-dependent diabetes mellitus (IDDM) is an autoimmune disease characterized by the selective destruction of pancreatic β cells by autoreactive T lymphocytes. Although genetic markers for susceptibility to IDDM have been clearly established, studies on discordance for IDDM between monozygotic twins have suggested the role of environmental factors, such as viruses, in the expression of the disease [1]. Epidemiological data showed an increased incidence of IDDM after epidemics caused by enteroviruses (EVs) [2]. Virus-specific IgM responses, thought to indicate recent or continuing infection by coxsackievirus B (CVB), have been described in serum samples from children and adults, at onset of IDDM [3, 4]. In addition, a CVB was cultured from the autopsied pancreas of a child who died of diabetes [5]. A temporal association of CVB infections and increases in islet cell antibody has also been reported [6]. Experimental studies have demonstrated that the CVB4 strain could cause IDDM in animals through acute cytolytic destruction of β cells [7]. Furthermore, it has been shown that the CVB4 strain E2, isolated from a patient with IDDM, induces β cell autoimmunity and hyperglycemia in some strains of mice [8]. A strong correlation has been reported between the persistence of viral RNA in the pancreases of CD1 mice infected with CVB4 strain E2 and the development of diabetes at 6 months after infection [9].

Evidence for EV involvement in human IDDM was based largely on the detection of specific anti-CVB IgM as well as on the detection of neutralizing antibodies (NAs) in serum samples from children at the onset of disease. Immunological methods based on antigen or antibody detection, however, have been limited by the absence of a single common antigen among the EVs. These problems have stimulated the development of nucleic acid-based methods [10]. Thus, using the reverse transcriptase-polymerase chain reaction (RT-PCR), we and other authors have reported the presence of the EV genome in 27%–64% of diabetic patients at the clinical manifestation of the disease [11–13]. RT-PCR and antibody assays may not detect all EV infections, however, particularly if there is a local infection and the level of virus replication is low. Interferon (IFN)-α is known to play an important role in the primary defense against viral infections. The production of IFN-α is a specific and reliable marker of ongoing virus infection; indeed, elevated levels of IFN-α have been detected in serum samples from subjects with viral infections [14, 15]. In addition, IFN-α has been detected in patients with autoimmune disorders of suspected viral etiology [16]. Thus, IFN-α can be used as an additional indicator of the presence of a virus in patients. The aim of this study was to determine whether subjects with IDDM had signs of increased IFN-α production as a marker of EV infection.

Previous SectionNext Section

Materials and Methods

Patient Group

Blood samples were obtained from 56 patients with IDDM. They were collected from June 1997 to September 1998. Patients admitted to the hospital and outpatients were grouped according to their ages, in children and adult groups. The children (n = 25, 16 boys and 9 girls; median age, 13 years; range, 3–17 years) were enlisted from the Department of Pediatric Endocrinology, Jeanne de Flandre Hospital (Lille, France). They were classified according to the course of the disease: 12 children newly diagnosed with IDDM (6 who were at the onset of the disease and had metabolic decompensation [diabetic ketoacidosis or diabetic ketosis] and 6 who had no metabolic decompensation) and 13 children previously diagnosed with IDDM (1 who had metabolic decompensation at the time of diagnosis and 12 who had no metabolic decompensation). The adults (n = 31, 21 men and 10 women; median age, 37 years; range, 20–69 years) were enlisted from the Department of Endocrinology and Metabolic Diseases, Marc-Linquette Hospital (Lille, France); they included 20 patients newly diagnosed with IDDM (16 who were at the onset of the disease and suffered from metabolic decompensation and 4 who had no metabolic decompensation) and 11 patients previously diagnosed with IDDM (7 who had metabolic decompensation at the time of diagnosis and 4 who had no metabolic decompensation). All patients with IDDM included in this study were treated with insulin. IDDM was diagnosed according to the National Diabetes Data Group [17]. Informed consent was obtained from the adults and the parents of the children, and the committee on research ethics at each participating hospital approved the study protocol.

Control Groups

One control group included 14 children (7 boys and 7 girls; median age, 13 years; range, 7–17 years) and 10 adults (6 men and 4 women; median age, 28 years; range, 23–45 years) without any suspected immunological, infectious, or metabolic disease, who either were hospitalized or were outpatients at the hospital. None of these controls had a family history of IDDM. Another control group included 13 patients with non-insulin-dependent diabetes mellitus (NIDDM) who either were hospitalized or were outpatients at the hospital: 3 children (1 boy and 2 girls; median age, 15 years; range, 15–17 years) and 10 adults (5 men and 5 women; median age, 50.5 years; range, 26–66 years). Blood samples from controls were collected from June 1997 to September 1998.

Immunoassay for IFN-α

Venous blood was collected from patients and controls in sterile 7.5-mL tubes (Becton Dickinson, Meylan, France) containing 20 IU heparin/mL (Heparin Choay, Sanofi, France). Plasma was immediately separated by centrifugation and stored at −80°C until assayed. The concentration of IFN-α was determined by a specific and sensitive dissociation-enhanced lanthanide fluoroimmunoassay (DELFIA) based on the direct sandwich technique using a mixture of 2 murine monoclonal antibodies to human IFN-α LT27:273 and LT27:293, which bind >90% of natural IFN-α subtypes coated onto microtiter plate wells (LKB WALLAC, Turku, Finland), and an europium-labeled murine anti-human IFN-α, as described elsewhere [18]. The fluorescence in the microtitration wells was measured in a time-resolved fluorometer (1230 Arcus Fluorometer, LKB WALLAC). The leukocyte reference IFN G-23-902-530 of the National Institutes of Health (Bethesda, MD) was used as a standard. The assay detection limit was 0.5 IU IFN-α/mL.

Bioassay for IFN-α

Plasma samples from patients and controls were assayed for antiviral activity by protection of Madin-Darby bovine kidney (MDBK) cells (Biowhittaker, Verviers, Belgium) against vesicular stomatitis virus (VSV)-induced cytopathic effects, as described elsewhere [19]. IFN-α concentrations (in international units per milliliter) were inferred from a natural human IFN-α standard (Alferon N Injection; Interferon Sciences, New Brunswick, NJ), kindly provided by Dr. M. S. Liao (Interferon Sciences). The assay detection limit was 2 IU IFN-α/mL.

Characterization of the IFN-α

Typing of IFN by NAs

The antiviral activity of samples was neutralized with a rabbit antiserum to IFN-α (ICN Immunobiological, Orsay, France) at a final concentration of 2 × 103 neutralizing units (NU)/mL (1 NU/mL of antibodies neutralizing 1 IU/mL of IFN-α activity). Samples were preincubated for 30 min at 37°C with the relevant antibody before assayed for antiviral activity by protection of MDBK cells against VSV.

Acid lability testing of IFN

One milliliter of plasma from children and adults with IDDM was adjusted to pH 2 by addition of 1 mL of 0.1 M HC1 and incubated for 24 h at 4°C. The mixture was then adjusted to pH 7 with addition of 0.04 mL of 2 M NaOH and was retitered for IFN-α activity in MDBK cells. As a control test, plasma samples from healthy controls, which were negative for IFN-α, were mixed with natural human IFN-α at 16 IU/mL before exposure to pH 2 and were similarly tested.

RNA Extraction

Venous whole blood samples in sterile EDTA tubes were taken for each control subject and each patient. Blood samples of 250 μL were aliquoted and mixed with 50 IU RNase inhibitor (Boehringer, Mannheim, Germany) and were stored at −80°C until use. Native RNA was extracted from 200 μL of whole blood specimens by the acid guanidinium-thiocyanatephenol-chloroformextraction procedure using a commercial system (RNAgents Total RNA Isolation System; Promega, Madison, WI). Extracted RNA was then dissolved in 50 μL of diethylpyrocarbonate (DEPC)-treated water (Sigma-Aldrich, Saint Quentin Fallavier, France) and used in the RT-PCR assays. Two positive extraction controls were used: the first was used for IFN-α RNA detection and consisted of RNA extracted from whole blood cultivated as described elsewhere [20] and activated by Sendai virus (provided by Dr. D. Garcin, Department of Genetics and Microbiology, University of Geneva) for 5 h to induce IFN-α mRNA synthesis. The second positive control consisted of RNA extracted from Hep-2 cells (Biowhittaker) infected with CVB3 (ATCC) for 4 h and was used for EV RNA detection.

Oligonucleotide Primers

Human IFN-α1 primers were purchased from Clontech (Palo Alto, CA). The upstream primer 5′-TGATGGCAACCAGTTCCAGAAGGCTCAAG-3′ (positions 244–272) and the downstream primer 5′-ACAACCTCCCAGGCACAAGGGCTGTATTT-3′ (positions 546–518) generate a 303-bp PCR product.

The sequences of the EV-specific primers were selected in the regions of the 5′ untranslated part of the viral genome, which are highly conserved among all EV strains. The external primers (EV1, 5′-CAAGCACTTCTGTTTCCCCGG-3′, and EV2, 5′-ATTGTCACCATAAGCAGCCA-3′) generate a 435-bp fragment, whereas the use of primer EV1 and internal primer EV3 (5′-CTTGCGCGTTACGAC-3′) generate a 362-bp PCR product. These primers were synthesized by Eurogentec (Seraing, Belgium) and were described elsewhere by Leparc et al. [21].

Glyceraldehyde phosphate dehydrogenase (GAPDH) primers were synthesized by Eurogentec (upstream primer: 5′-GTCTTCACCACCATGGAGA3′, positions 361–379; downstream primer: 5′-CCAAAGTTGTCATGGATGACC-3′, positions 566–546). The amplified PCR fragment size was 206 bp.

One-Step RT-PCR Reaction

The cDNA synthesis, as well as the cDNA amplification, was performed in a single tube by using the Titan 1-Tube RT-PCR system (Boehringer). The 1-step reaction system uses an enzyme mix consisting of an avian myeloblastosis virus reverse transcriptase for first-strand synthesis and Taq DNA polymerase and Pwo DNA polymerase for the PCR assay. The system includes dithiothreitol (DTT) solution and a single 5× RT-PCR buffer with 7.5 mM MgCl2 and dimethyl sulfoxide. The 1-step RT-PCR reaction mixture consists of 0.2 mM each of dATP, dCTP, dGTP, and dTTP (Boehringer), 20 pmol of each primer, 5 mM DTT solution, 20 IU RNase inhibitor, 1 μL of enzyme mix, 10 μL of 5× RT-PCR reaction buffer with 1.5 mM MgCl2, and 1 μg of extracted RNA. The mixture was then adjusted to a volume of 50 μL with DEPC-treated water, and the RT-PCR reaction was done using an ABI GeneAmp 9700 thermocycler (Applied Biosystems, Foster City, CA).

For each RNA sample, GAPDH mRNA was transcribed in cDNA and then amplified in an RT-PCR reaction. A negative control (no RNA) was also included in each PCR. The absence of introns within IFN-α genes precludes the distinction between amplification products resulting from reverse-transcribed IFN-α1 mRNA and those arising from residual genomic DNA. Therefore, a control RT-PCR reaction with GAPDH primers that anneal to sequences in exons on both sides of 2 introns has been used for each RNA sample. With this approach, PCR products generated from genomic DNA were 388 bp, which is much larger than products derived from intronless mRNA (206 bp). Only samples without contaminating genomic DNA have been considered.

Seminested Amplification

A second-run PCR for EV detection was done as described by Leparc et al. [21]. In brief, 1 α1 of the amplified products was added to the PCR mixture consisting of 5 μL of 10α PCR reaction buffer (Boehringer), 0.2 mM of the previously described dNTPs, 1.25 IU Taq DNA polymerase (Boehringer), and 20 pmol each of primers EV1 and EV3. The mixture was adjusted to a volume of 50 μL with DEPC-treated water, and DNA amplification was undertaken as described above.

Detection of PCR Products

The amplified RT-PCR products were analyzed on 2% agarose gel containing 0.5 μg/mL of ethidium bromide (Sigma-Aldrich) and were visualized using the Gel Doc 2000 system (Bio-Rad, Ivry-surseine, France). Image processing and analysis operations of DNA bands were performed using Quantity One software (Bio-Rad). A 100-bp DNA ladder (Gibco BRL, Paris) was used as a molecularweight marker.

Sequence Analysis of Amplified Products

The nucleotide sequences of fragments amplified in our experiments were different from one serotype to another, which enables their identification with a rough estimate >90%. The nucleotide sequences were determined by double-strand DNA cycle sequencing using an AmpliTaq FS Big Dye Terminator dichlororhodamine cycle sequencing kit (Applied Biosystems), according to the manufacturer's instructions. Electrophoresis and analysis of DNA sequence reactions were done using an automated DNA sequencer (model 377 XL; Applied Biosystems). The sequencing primers for IFN-α were the same as those used in the RT-PCR reaction. The RT-PCR products were sequenced, and analysis data demonstrated 99% homology with the IFN-α1 sequence. The sequencing primers for EV were the primers used in the first run of PCR amplification. Prior to sequencing, the PCR products were purified using the Wizard PCR Preps DNA Purification system (Promega). Both strands of the DNA fragments were sequenced. The derived sequences were then analyzed and compared to each other and to sequences available from the GenBank database, by use of the Usedit and Sequence Navigator programs (Applied Biosystems).

Assays of CVB Antibodies

Plaque-neutralization assay

Two-fold serial dilutions of serum samples, obtained from control subjects and subjects with IDDM by veniculture into sterile tubes, in Eagle's minimum essential medium 1640 (Gibco BRL, Eragny, France) supplemented with 2% fetal calf serum and 1% L-glutamine were incubated with 100 TCID50 of CVB serotypes 1–6 (American Type Culture Collection, Manassas, VA, and Centre National de Réference des Entérovius, Lyon, France) in 96-well microtiter plates for 2 h at 37°C. Hep-2 cells (Biowhittaker) in suspension were then added at ∼ 1 × 104 cells per well, and the plates were reincubated for 3 days at 37°C in a humidified incubator with 5% CO2. Results were expressed as the inverse final dilution of serum that totally inhibited the viral cytopathic effect.

ELISA test

Two SERION ELISA classic kits (Institut Virion-Serion, Wyrzberg, Germany) were used for quantitative detection of specific human anti-CVB IgG and IgM antibodies. Measurable antibody activities were expressed in units per milliliter. Serum IgG levels ⩾90 U/mL and IgM levels ⩾50 U/mL were considered to be positive results, according to the manufacturer's guidelines.

Statistical Analysis

Data are summarized as mean ± SD. The significance of the differences of levels of IFN-α was determined by the Mann-Whitney U test. The odds ratio test (P value for α error of .05) and the χ2 test or χ2 test with Yates correction (χ2c) were used when appropriate, and correlations were evaluated with Spearman's test.

Previous SectionNext Section

Results

IFN-α levels in plasma of patients with IDDM

The IFN-α levels in plasma were determined by the DELFIA method. The mean levels of IFN-α in children with IDDM were significantly higher than those in control children. The mean levels of IFN-α in adults with IDDM were also significantly higher than those in controls. The criterion for a positive detection of IFN-α in plasma of patients was a value greater than the mean +3 SD of IFN-α values for healthy controls. Thus, 16 (64%) of 25 children with IDDM and 23 (74%) of 31 adults with IDDM were IFN-α-protein positive (table 1).

The 303-bp amplification product of human IFN-α1 mRNA was detected in extracts of whole blood from all patients with IDDM who had high circulating IFN-α, whereas no IFN-α1 transcripts were detected in normal individuals or in individuals with NIDDM (figure 1). Samples of RNA whole blood extracts obtained from 3 patients (1 child and 2 adults) contained IFN-α1 transcripts, whereas IFN-α protein was not detected in their plasma (table 1). Thus, by use of the RT-PCR method, 68% of children with IDDM and 81% of adults with IDDM were IFN-α mRNA positive. The difference between patients newly diagnosed and those previously diagnosed with IDDM, with or without metabolic decompensation, was not significant (χ2c test, P < .1; table 2).

Figure 1
View larger version:
Figure 1

Reverse transcriptase-polymerase chain reaction (RT-PCR) products with interferon (IFN)-α1-specific primers. RT-PCR was performed as described in Materials and Methods, and products were analyzed by 2% agarose gel electrophoresis and stained with ethidium bromide. Template RNA used was from whole blood of healthy children and adults (Janes 2 and 3, respectively), children and adults with non-insulin-dependent diabetes mellitus (lanes 4 and 5, respectively), children previously diagnosed with insulin-dependent diabetes mellitus (IDDM; lanes 6–8), children newly diagnosed with IDDM (lanes 9–11), adults previously diagnosed with IDDM (lanes 12 and 13), and adults newly diagnosed with IDDM (lanes 14 and 15). Lane 1, Molecular-weight markers (100-bp DNA ladder); lane 16, positive control RNA extracted from Sendai-activated whole blood cells; lane 17, negative control (no RNA). Glyceraldehyde phosphate dehydrogenase (GAPDH) RNA served as an external control for each sample.

Figure 2
View larger version:
Figure 2

Relationship between the values of interferon (IFN)-α assayed by the dissociation-enhanced lanthanide fluoroimmunoassay (DELFIA) method and antiviral titers of plasma samples in patients with insulin-dependent diabetes mellitus. IFN-α values are expressed in international units per milliliter, and the titers were determined by the inverse of the dilution of plasma samples that caused 50% protection of Madin-Darby bovine kidney cells challenged with vesicular stomatitis virus. The box indicates contents of IFN-α and antiviral titers after absorption with anti-IFN-α polyclonal antibodies.

Figure 3
View larger version:
Figure 3

Individual representation of anti-coxsackievirus B (CVB) antibody titers in patients with insulin-dependent diabetes mellitus, grouped according to the presence of enterovirus (EV) RNA and interferon (IFN)-α in their peripheral blood. Antibodies were detected by neutralization assay, and the titer corresponds to the inverse of the last dilution that inhibits replication of CVB1–6. The maximum titer of serotype-neutralizing antibody detected against ⩾1 CVB was plotted. EV RNA was detected by seminested reverse transcriptasepolymerase chain reaction (RT-PCR) in blood. Patients were segregated as IFN-α positive or IFN-α negative according to the presence or the absence of IFN-α mRNA in their blood, detected by RT-PCR.

Table 1
View larger version:
Table 1

Interferon (IFN)-α in circulating blood of patients with insulin-dependent diabetes mellitus (IDDM).

Table 2
View larger version:
Table 2

Interferon (IFN)-α in circulating blood of patients with insulin-dependent diabetes mellitus (IDDM) during the course of the disease.

To further examine the biological activity of IFN-α detected by DELFIA in patients, we determined its antiviral activity by protection of MDBK cells against VSV-induced cytopathic effects. Results were expressed as IFN-α concentrations (in international units per milliliter) and were inferred from the natural human IFN-α standard. A significant correlation was obtained, for patients with IDDM, between the IFN-α levels detected with the DELFIA method and those detected with the antiviral activity assay (P < .0001, n = 56; figure 2). Plasma samples from patients with IDDM were incubated with anti-IFN-α NAs (vol/vol). This treatment resulted in a complete removal of antiviral effect as assayed in MDBK cells (figure 2). In addition, the biological activity of IFN-α was completely stable after pH 2 treatment of plasma samples of patients (data not shown).

EV infection in patients with IDDM

Forty-two patients with IDDM (17 children and 25 adults) were IFN-α positive, whereas 14 patients (8 children and 6 adults) were IFN-α negative. To explore the relationship between the activation of the IFN-α system and EV infection in patients with IDDM, we looked for the presence of EV RNA and anti-CVB antibodies in their blood.

In the group of 40 controls, the detection of EV RNA was negative. In IFN-α-negative patients with IDDM, the detection of EV RNA was negative (0%). In IFN-α-positive patients with IDDM, 21 (50%) of 42 subjects were positive for EV RNA. The levels of anti-CVB IgM antibodies were not different in controls and in most patients. In only 2 adults with IDDM were high levels of anti-CVB IgM antibodies detected (table 3). One of these patients was EV RNA positive. There was a significant relationship between the detection of IFN-α and the presence of EV RNA in the blood of patients with IDDM (χ2 test, P = .0008; n = 56). EV RNA was found in patients previously diagnosed with IDDM and in patients newly diagnosed with IDDM (7 children and 14 adults). There was no relationship between EV RNA detection and either the course of the disease or the presence of metabolic decompensation (χ2c test, P < .1; table 3). EV RNA was detected in blood samples taken from 6 of 12 patients previously diagnosed with IDDM without metabolic decompensation who were seen at a routine visit; they were clinically stable without concurrent illness at the time of blood collection. EV sequences were found in samples collected from June 1997 to September 1998, without any relationship with epidemic outbreaks.

Table 3
View larger version:
Table 3

Enterovirus (EV) infection in interferon (IFN)-α-positive patients with insulin-dependent diabetes mellitus (IDDM), grouped according to the detection of IFN-α and EV infection during the course of the disease.

The IFN-α-positive detection was compared with the detection of EV RNA from the blood and with the levels of anti-CVB NAs in patients. In serum samples from controls, the titers of serotype NA were ⩽1 : 32 directed against virus types 1–6. For each patient, the maximum titer of antibodies directed against ⩾1 serotypes of the CVB viruses was taken, and the value was plotted with regard to the detection of IFN-α and EV RNA. From figure 3, one can see that higher levels of antibodies were found in the IFN-α-positive group than in the IFN-α-negative group. The difference in antibody levels was statistically significant between IFN-α-positive patients and controls (P < .0001) but not between IFN-α-negative patients and controls (P > .05). There was a significant difference in the geometric mean of the CVB NA titer between the IFN-α-positive/EV RNA-positive (398 ± 5, n = 21) or the IFN-α-positive/EV RNA-negative patients (74 ± 6, n = 21) and the negative group (4.8 ±6, n = 14; P < .0001 and <.003, respectively). When the 6 values of antibody titers ⩾256 in the IFN-α-positive/EV RNA-negative group were eliminated, the geometric mean of antibody titers (33.1 ± 3.9, n = 15) was still significantly higher than the one for the negative group (P = .02). Of the 42 IFN-α-positive patients, 23 had CVB NA titers >128. Of the 14 IFN-α-negative patients, no one had CVB NA titers >128. The relationship between positive detection of IFN-α and high levels of antibodies (titers >128) in the blood of patients with IDDM was statistically significant (χ2 test, P < .001; n = 56). The low antibody titers detected in certain patients with IDDM by use of the CVB-microneutralization test prompted us to determine the levels of CVB antibodies with another method. In serum samples from controls (healthy and with NIDDM), the levels of anti-CVB IgG antibodies detected by ELISA were <80 U/mL. Of the 23 samples of patients with NA titers ⩾256, 100% were ⩾90 U/mL—that is, positive for anti-CVB IgG antibodies by ELISA. Of the 33 remaining samples of patients with antibody titers <256, 100% were <90 U/mL—that is, negative for anti-CVB IgG antibodies. The levels of antibody were significantly higher in IFN-α-positive patients, compared with IFN-α-negative patients and controls, but the difference between IFN-α-negative patients and controls was not significant. The ELISA results were concordant with the CVB-neutralization test results (data not shown).

Of the 21 EV RNA-positive patients, 17 had CVB NA titers ⩾128. Of the 35 EV RNA-negative patients, 6 had CVB NA titers >128. Statistical analysis showed a significant difference at the 95% level between the EV RNA-positive and -negative patients with IDDM, with respect to positive detection of CVB NA titers >128 (odds ratio, 20.54; 95% confidence interval, 4.30–11.15). The relationship between positive detection of EV RNA and high levels of antibodies (titers >128) in the blood of patients with IDDM was statistically significant (χ2 test, P < .001; n = 56).

The detection of EV RNA in the circulating blood of IFN-α-positive patients with IDDM prompted us to identify the genotype of amplified EV RNA sequences. The sequence analysis of PCR amplicons derived from amplifying the 5′ untranslated regions (UTRs) of circulating EV demonstrated that patient sequences were distributed throughout the CVB groups: 5 CVB2 (92%–98% of homology), 3 children and 2 adults; 8 CVB3 (90%–99% of homology), 3 children and 5 adults; 8 CVB4 (91%–94% of homology with nondiabetogenic JBV strain and 97%–99% of homology with diabetogenic E2 strain), 1 child and 7 adults (table 4). There was a relationship between the genotype of CVB, which was determined by sequence analysis of PCR amplicons, and high titers of serotype-specific NA (χ2 test, P < .0001; n = 56). For example, in patient e, the maximum titer of serotype NA was 512 directed against CVB4, and the EV RNA sequence had 97% homology with CVB4. In 4 patients (B, F, H, and N), however, the serology was not discriminating, since we obtained maximum titers of antibodies against 2 serotypes.

Table 4
View larger version:
Table 4

Enterovirus (EV) RNA genotype and coxsackievirus B (CVB) antibody titers in patients with insulin-dependent diabetes mellitus (IDDM).

Previous SectionNext Section

Discussion

The role of viruses in the pathogenesis of IDDM has been suggested, and previous studies, including ours, reported the detection of EV RNA in the blood of patients with IDDM [11–13]. The present study, however, is different in many respects from those of other investigators: first, the plasma levels of IFN-α were studied in each patient; second, the plasma levels of IFN-α were investigated in the same patients whose blood was used for detecting EV RNA; third, the relationship between IFN-α levels, EV RNA, and serology was studied.

We report that a factor was present in the plasma of patients with IDDM that was able to protect cells from in vitro viral infection. This antiviral factor has been identified as IFN-α, since absorption with anti-IFN-α antibodies could eliminate its antiviral activity. In light of the fact that the results were obtained by DELFIA and antiviral assay, we conclude that an elevated plasma level of IFN-α can be found in patients with IDDM. The elevated levels of immunoreactive and biologically active IFN-α in patients with IDDM strongly indicate activation of the IFN-α system during IDDM.

This is the first report of detectable IFN-α in the blood of patients with type 1 diabetes. The IFN-α detected had an antiviral activity, and, like all known IFN-α subtypes, its antiviral activity was not inactivated by exposure to acidic pH. The results of our experiments demonstrate that the patients who displayed elevated levels of endogenous IFN-α in their serum samples also had detectable IFN-α mRNA in their cells. The results confirm the activation of the IFN-α-producing cells in the blood of patients with IDDM. A rapid degradation of IFN-α in circulating blood may occur; in addition, this cytokine may be released and consumed locally at the site where the immune reaction occurs. Therefore, it can be produced spontaneously by cells but be undetectable in plasma, which may explain the results obtained for 3 patients in this study (IFN-α mRNA positive, IFN-α-protein negative). The results of our experiments provide direct evidence for the activation of the IFN-α system and can explain that IFN-dependent 2′,5′ oligoadenylate synthetase activity was high in the blood lymphocytes of patients with IDDM in studies by Bonnevie-Nielsen et al. [22, 23].

A method that requires whole blood has been used for studying the presence of EV RNA in patients with IDDM. RNA was extracted from whole blood samples instead of from plasma because it has been reported that peripheral blood cells can harbor EVs during and after a viremic phase [24]. This method, previously successfully used by us for detecting EV RNA in adult patients with IDDM [11], is well adapted to blood collection from infants, since small-volume blood samples (100 μL) are required.

Our results show that, in every patient with detectable EV RNA, the level of IFN-α was high. The data support the hypothesis of the role of circulating virus in the increased secretion of IFN-α observed in patients. The cellular and molecular mechanisms of the raised levels of IFN-α in these patients are unknown, although the IFN-α present in plasma may be produced by blood cells, because IFN-α mRNA was detectable in these cells, and previous studies demonstrated that EVs, in particular CVB, were capable of inducing the release of IFN-α from human monocytes in vitro [25, 26]. The mechanisms of increased expression of IFN-α in patients with IDDM and the role of CVB are under investigation in our laboratory.

Investigation of coxsackie viral infection in EV RNA-negative patients has been based on the combination of the positive detection of IFN-α and the use of the neutralization test. Anti-CVB antibody titers ⩾256, considered as evidence of CVB infection in the recent past [27], were found in 6 of 21 patients who were IFN-α positive but EV RNA negative. Moreover, for IFN-α-positive/EV RNA-negative patients, the mean of the CVB NA titers was higher than that for IFN-α-negative/EV RNA-negative patients. In 2 patients who were IFN-α positive, the CVB NA titers were low (32 and 0) and the ELISA for EV-specific IgM was negative, whereas EV PCR provides evidence for circulating viral sequences. Two different procedures were used for detecting CVB antibodies, and good correlation was found for antibody levels obtained with the neutralization test and ELISA. Therefore, the low levels of CVB NAs in certain patients with IDDM were not due to technical problems of the neutralization test or reduced sensitivity of the assay. Our results agree with those of other authors who showed that patients with confirmed CVB infection did not seroconvert for any CVB and that the humoral immune response to CVB varies markedly between individuals [28, 29]. Together, these results suggest that low CVB antibody titers in patients with IDDM do not exclude a CVB infection and that the detection of IFN-α may reflect a CVB infection even when viral RNA and increased levels of CVB antibodies are not found.

The levels of CVB IgM are thought to reflect recent or persistent infection [30]; however, the IgM antibody assay cannot distinguish CVB-infected individuals from other patients with IDDM, since, in most patients in this study, the detection of IgM was negative. Our results agree with those of Palmer et al. [31], who reported that the CVB infection elicited a weak IgM response in patients newly diagnosed with IDDM. Therefore, the results of CVB IgM assays in studies of the role of CVB infection in IDDM must be interpreted with caution.

One explanation for IFN-α positivity but EV RNA negativity and the low CVB antibody levels, within the limits of the assays, could be a low EV load and/or an infection of secondary target organs, such as the pancreas, without viremia and, consequently, with reduced levels of NA. Furthermore, we cannot exclude the role of other viruses in the activation of the IFN-α system in certain patients.

Computer analysis of sequencing data showed that amplified EV RNA in diabetic patients was related closely to CVB (5 CVB2, 8 CVB3, and 8 CVB4 sequences). It is not possible to determine, on the basis of the 5′ UTR sequence, the exact types of EVs involved just from their belonging to the CVB group. This region does not always correlate with the serotype, the sequence determinants of which are located downstream of the 5′ UTR, in the region coding for the capsid proteins. We have observed, however, a strong correlation between specific anti-CVB antibodies, determined by plaque-neutralization assay and sequence analysis of RT-PCR-amplified products. There is an antibody cross-reactivity between various CVB serotypes, and heterotypic responses can be observed. This study, however, demonstrates that high anti-CVB antibody levels in patients with IDDM (titers >128 in a neutralization test) are found in subjects with EV RNA in the blood. Thus, in diabetic patients, serological tests are of particular clinical importance, since increased levels of anti-CVB antibodies may be indicative of ongoing CVB infection.

In previous studies, we reported that amplified EV RNA in diabetic adult patients was related closely to CVB3, CVB4, or CVB4E2. Our current results, obtained from adults and children, are consistent with earlier results. In some cases, the EV sequences were related to CVB2. In other studies using serological and/or RT-PCR methods, other EV serotypes in addition to CVB, such as CVA and echoviruses, have been suggested to play a role in the pathogenesis of IDDM [13, 32–35]. These discrepancies may be due to the epidemiology of EV, since the results were obtained from different European countries.

The high frequency of positive IFN-α detection and/or CVB infection among outpatients with previously diagnosed IDDM and without clinical manifestation or metabolic decompensation indicates asymptomatic acute viral infections or, alternatively, that a persistent infection is present in these patients. In order to elucidate whether an acute, rather than a persistent, CVB infection is associated with IDDM, studies of the variations of IFN-α, virus detection, and specific virus antibodies in blood samples, obtained consecutively over several weeks and months, from patients with IDDM should be performed.

IFN-α has not been detected in 25% of the patients with IDDM in this study. It is not always possible to detect IFN-α in patients with viral infections, because the half-life of IFN-α is short [36]. By measuring IFN-α-induced genes or proteins, the activation of the IFN-α system may be detected much more consistently [37, 38]. The negative detection of IFN-α and EV RNA associated with low levels of antibodies to CVB may be due to a low EV load or may have been due to the presence of viruses other than CVB that would not have induced the synthesis of IFN-α. Further investigations are needed to explore these hypotheses. Viral infections are listed first among the possible environmental factors playing a role in the pathogenesis of type 1 diabetes; however, dietary factors such as cow's milk proteins and toxins such as N-nitroso compounds are thought to be involved in some cases [39], which may explain the negative detection of direct or indirect markers of viral infection in certain patients in this study.

The results of this study suggest that ongoing CVB infections are associated either with triggering or with the precipitation of diabetes. The mechanism by which CVB may induce diabetes in humans is unknown. A sequence homology between the islet antigen, glutamate decarboxylase, and CVB P2-C, the most strongly conserved EV protein [40], could contribute to the diabetogenicity of the viral infection, according to the hypothesis of molecular mimicry [41]. Horwitz et al. [42], however, showed that molecular mimicry alone is not sufficient and that diabetes-induced CVB infection in mice is a consequence of the infection of the pancreas, leading to the release of islet antigen, which results in the restimulation of resting autoreactive T cells. Thus, the pancreatropic nature of the virus plays a role in the pathogenesis of IDDM. CVBs, especially CVB3 and CVB4 [43, 44], can infect β cells in vitro, and CVB antigens have been found in β cells from patients with fatal CVB infections [45]. Therefore, CVB infections, observed in patients in the current study, can result in β cell damage that may play a role in the expression of the disease.

The presence of a cytokine molecule in a biological fluid is an important part of proving its role in a disease. The increased level of IFN-α in most of the patients with IDDM (75%; 25 adults and 17 children) in our study raised the question of the role of the molecule in disease progression. It is currently thought that development and maintenance of some immune disorders are correlated to abnormal activation of the IFN system [16]. It is interesting that IFN-α has been suggested as an important cofactor in the development of Th1-type immune reaction, which can contribute to the development of autoimmune diseases [46]. Moreover, induction of autoimmunity as a consequence of the therapeutic use of recombinant and natural IFN-α is now documented [47], and case reports of the appearance of type 1 diabetes subsequent to the use of IFN-α in therapy have been published [48, 49]. IFN-α has been associated with the pathogenesis of type 1 diabetes in animal models of autoimmune diabetes [50]. The mechanisms underlying the sensitization of a host to IFN-α are not fully understood, and such influences, together with virulence of CVB strains, could explain the outcome of CVB infection in each patient. To elucidate whether the antiviral effect or the pathophysiological role of IFN-α predominates in CVB infection in IDDM, further studies are required.

In conclusion, the expression of IFN-α is associated in 50% of cases with the presence of EV sequences in circulating blood. Higher levels of CVB NAs were detected in patients who were IFN-α positive and EV RNA negative, compared with IFN-α-negative patients; however, low levels of CVB antibodies (IgG and/or IgM) may be associated with CVB infection in IFN-α-positive patients with IDDM. Together, the results suggest that the proportion of CVB infection associated with IFN-α positivity may be higher than predicted from the investigation of EV RNA and serology. The results of this study show that an ongoing CVB infection, evidenced by increased plasma levels of IFN-α, is associated with IDDM in adults and children in the majority of cases. The measurement of endogenous IFN-α combined with EV RNA and serology may overcome sensitivity limitations in the evaluation of the role of CVB in triggering β cell autoimmunity and clinical IDDM.

Previous SectionNext Section

Acknowledgment

We thank V Chieux for helpful discussions.

Previous SectionNext Section

Footnotes

  • ↵a The Laboratoire de Virologie, Centre Hospitalier Régional et Universitaire Lille, is a part of the “Equipe d'Accueil 1048: Nouvelles Thérapeutiques du Diabète de Type 1 et Pathogenèse Virale de la Maladie” of the Ministère de l'Education Nationale de la Recherche et de la Technologie, France.

  • Financial support: the Fondation de France, the Conseil Régional Nord Pas-De-Calais, Centre Hospitalier Régional et Universitaire Lille, and the Ministère de l'Education Nationale de la Recherche et de la Technologie (Unité Propre de Recherche de l'Enseignement Supérieur: Equipe d'Accueil 1048).

  • Received November 29, 1999.
  • Revision received February 16, 2000.
Previous Section
 

References

  1. 1.↵
    1. Yoon JW
    . Role of viruses in the pathogenesis of IDDM. Ann Med 1991;23:437-45.
    MedlineWeb of Science
  2. 2.↵
    1. Fohlman J,
    2. Friman G
    . Is juvenile diabetes a viral disease. Ann Med 1993;25:569-74.
    MedlineWeb of Science
  3. 3.↵
    1. Banatvala JE,
    2. Bryant J,
    3. Schernthaner G,
    4. et al
    . Coxsackie B, mumps, rubella, and cytomegalovirus specific IgM responses in patients with juvenile-onset insulin-dependent diabetes mellitus in Britain, Austria, and Australia. Lancet 1985;1:1409-12.
    CrossRefMedlineWeb of Science
  4. 4.↵
    1. Frisk G,
    2. Friman G,
    3. Tuvemo T,
    4. Fohlman J,
    5. Diderholm H
    . Coxsackie B virus IgM in children at onset of type 1 (insulin-dependent) diabetes mellitus: evidence for IgM induction by a recent or current infection. Diabetologia 1992;35:249-53.
    CrossRefMedlineWeb of Science
  5. 5.↵
    1. Yoon JW,
    2. Austin M,
    3. Onodera T,
    4. Notkins AL
    . Isolation of a virus from the pancreas of a child with diabetic ketoacidosis. N Engl J Med 1979;300:1173-9.
    MedlineWeb of Science
  6. 6.↵
    1. Hiltunen M,
    2. Hyoty H,
    3. Knip M,
    4. et al
    . Islet cell antibody seroconversion in children is temporally associated with enterovirus infections: Childhood Diabetes in Finland (DiMe) Study Group. J Infect Dis 1997;175:554-60.
    Abstract/FREE Full Text
  7. 7.↵
    1. Szopa TM,
    2. Titchener PA,
    3. Portwood ND,
    4. Taylor KW
    . Diabetes mellitus due to viruses: some recent developments. Diabetologia 1993;36:687-95.
    CrossRefMedlineWeb of Science
  8. 8.↵
    1. Gerling I,
    2. Chatterjee NK,
    3. Nejman C
    . Coxsackievirus B4-induced development of antibodies to 64,000-Mr islet autoantigen and hyperglycemia in mice. Autoimmunity 1991;10:49-56.
    CrossRefMedlineWeb of Science
  9. 9.↵
    1. See DM,
    2. Tilles JG
    . Pathogenesis of virus-induced diabetes in mice. J Infect Dis 1995;171:1131-8.
    Abstract/FREE Full Text
  10. 10.↵
    1. Graves PM,
    2. Norris JM,
    3. Pallansch MA,
    4. Gerling IC,
    5. Rewers M
    . Perspectives in diabetes: the role of enteroviral infections in the development of IDDM: limitations of current approaches. Diabetes 1997;46:161-8.
    Abstract
  11. 11.↵
    1. Andreoletti L,
    2. Hober D,
    3. Hober-Vandenberghe C,
    4. et al
    . Detection of coxsackie B virus RNA sequences in whole blood samples from adult patients at the onset of type I diabetes mellitus. J Med Virol 1997;52:121-7.
    CrossRefMedlineWeb of Science
  12. 12.
    1. Clements GB,
    2. Galbraith DN,
    3. Taylor KW
    . Coxsackie B virus infection and onset of childhood diabetes. Lancet 1995;346:221-3.
    CrossRefMedlineWeb of Science
  13. 13.↵
    1. Nairn C,
    2. Galbraith DN,
    3. Taylor KW,
    4. Clements GB
    . Enterovirus variants in the serum of children at the onset of type 1 diabetes mellitus. Diabet Med 1999;16:509-13.
    CrossRefMedlineWeb of Science
  14. 14.↵
    1. Green JA,
    2. Charette RP,
    3. Yeh TJ,
    4. Smith CB
    . Presence of interferon in acuteand convalescent-phase sera of humans with influenza or influenza-like illness of undetermined etiology. J Infect Dis 1982;145:837-41.
    Abstract/FREE Full Text
  15. 15.↵
    1. Flowers D,
    2. Scott GM
    . How useful are serum and CSF interferon levels as a rapid diagnostic aid in virus infections? J Med Virol 1985;15:35-47.
    CrossRefMedlineWeb of Science
  16. 16.↵
    1. Schattner A
    . Interferons and autoimmunity. Am J Med Sci 1988;295:532-44.
    CrossRefMedlineWeb of Science
  17. 17.↵
    National Diabetes Data Group. Classification and diagnosis of diabetes mellitus and other categories of glucose intolerance. Diabetes 1979;28:1039-57.
    MedlineWeb of Science
  18. 18.↵
    1. Cederblad B,
    2. Blomberg S,
    3. Vallin H,
    4. Perers A,
    5. Alm GV,
    6. Ronnblom L
    . Patients with systemic lupus erythematosus have reduced numbers of circulating natural interferon-alpha-producing cells. J Autoimmun 1998;11:465-70.
    CrossRefMedlineWeb of Science
  19. 19.↵
    1. Lebon P,
    2. Commoy-Chevalier MJ,
    3. Robertgalliot B,
    4. Morin A,
    5. Chany C
    . Production d'interferon humain de type 1 par des lymphocytes au contact de cellules infectées par le virus herpes et fixées par le glutaraldehhyde. C R Seances Acad Sci D 1980;290:37-40.
    Medline
  20. 20.↵
    1. Chehadeh W,
    2. Hober D,
    3. Chieux V,
    4. et al
    . Biological properties of IFNα produced ex-vivo by whole blood of HIV-1 infected patients. Scand J Immunol 1999;49:660-6.
    CrossRefMedlineWeb of Science
  21. 21.↵
    1. Leparc I,
    2. Aymard M,
    3. Fuchs F
    . Acute, chronic, and persistent enterovirus and poliovirus infections: detection of viral genome by seminested PCR amplification in culture-negative samples. Mol Cell Probes 1994;8:487-95.
    CrossRefMedlineWeb of Science
  22. 22.↵
    1. Bonnevie-Nielsen V,
    2. Larsen ML,
    3. Frifelt JJ,
    4. Michelsen B,
    5. Lernmark A
    . Association of IDDM and attenuated response of 2′,5′-oligoadenylate synthetase to yellow fever vaccine. Diabetes 1989;38:1636-42.
    Abstract
  23. 23.↵
    1. Petrovsky N,
    2. Harrison LC,
    3. Kyvik KO,
    4. Beck-Nielsen H,
    5. Green A,
    6. Bonnevie-Nielsen V
    . Evidence for the viral aetiology of IDDM. Autoimmunity 1997;25:251-2.
    CrossRefMedlineWeb of Science
  24. 24.↵
    1. Vuorinen T,
    2. Vainionpää,
    3. Kettinen H,
    4. Hyypiä T
    . Coxsackievirus B3 infection in human leucocytes and lymphoid cell lines. Blood 1994;84:823-9.
    Abstract/FREE Full Text
  25. 25.↵
    1. Henke A,
    2. Mohr C,
    3. Sprenger H,
    4. et al
    . Coxsackievirus B3-induced production of tumor necrosis factor-alpha, IL-1 beta, and IL-6 in human monocytes. J Immunol 1992;148:2270-7.
    Abstract
  26. 26.↵
    1. Pitkäranta A,
    2. Linnavuor K,
    3. Hovi T
    . Virus-induced interferon production in human leukocytes: a low responder to one virus can be a high responder to another one. J Interferon Res 1991;11:17-23.
    MedlineWeb of Science
  27. 27.↵
    1. Nairn C,
    2. Galbraith DN,
    3. Clements GB
    . Comparison of coxsackie B neutralization and enteroviral PCR in chronic fatigue patients. J Med Virol 1995;46:310-3.
    MedlineWeb of Science
  28. 28.↵
    1. Lennette EH,
    2. Shinomoto TT,
    3. Schmidt NJ,
    4. Magoffin RL
    . Observation on the neutralizing antibody response to group B coxsackie viruses in patients with central nervous system disease. J Immunol 1960;86:257-66.
    Web of Science
  29. 29.↵
    1. Danes L,
    2. Havelkova M,
    3. Prochazkova I
    . Prevalence of infection with nonpolio enteroviruses among healthy children from a children's home and in a summer camp. J Hyg Epidemiol Microbiol Immunol 1983;24:191-202.
  30. 30.↵
    1. Tilzey AJ,
    2. Signy JM,
    3. Banatvala JE
    . Persistent coxsackie B virus specific IgM response in patients with recurrent pericarditis. Lancet 1986;1:1491-2.
    MedlineWeb of Science
  31. 31.↵
    1. Palmer JP,
    2. Cooney MK,
    3. Ward RH,
    4. Hansen JP,
    5. et al
    . Reduced coxsackie antibody titers in type I (insulin-dependent) diabetic patients presenting during an outbreak of coxsackie B3 and B4 infection. Diabetologia 1982;22:426-9.
    CrossRefMedlineWeb of Science
  32. 32.↵
    1. Frisk G,
    2. Nilsson E,
    3. Tuvemo T,
    4. Friman G,
    5. Diderholm H
    . The possible role of coxsackie A and echo viruses in the pathogenesis of type I diabetes mellitus studied by IgM analysis. J Infect 1992;24:13-22.
    CrossRefMedlineWeb of Science
  33. 33.
    1. Roivainen M,
    2. Knip M,
    3. Hyoty H,
    4. et al
    . Several different enterovirus serotypes can be associated with prediabetic autoimmune episodes and onset of overt IDDM: Childhood Diabetes in Finland (DiMe) Study Group. J Med Virol 1998;56:74-8.
    CrossRefMedlineWeb of Science
  34. 34.
    1. Lönnrot M,
    2. Knip M,
    3. Roivainen M,
    4. Koskela P,
    5. Akerblom HK,
    6. Hyoty H
    . Onset of type 1 diabetes mellitus in infancy after enterovirus infections. Diabet Med 1998;15:431-4.
    CrossRefMedlineWeb of Science
  35. 35.↵
    1. Hyöty H,
    2. Hiltunen M,
    3. Knip M,
    4. et al
    . A prospective study of the role of coxsackie B and other enterovirus infections in the pathogenesis of IDDM: Childhood Diabetes in Finland (DiMe) Study Group. Diabetes 1995;44:652-7.
    Abstract
  36. 36.↵
    1. Taylor JL,
    2. Grossberg SE
    . Recent progress in interferon research: molecular mechanisms of regulation, action, and virus circumvention. Virus Res 1990;15:1-26.
    CrossRefMedlineWeb of Science
  37. 37.↵
    1. Roers A,
    2. Hochkeppel HK,
    3. Horisberger MA,
    4. Hovanessian A,
    5. Haller O
    . MxA gene expression after live virus vaccination: a sensitive marker for endogenous type 1 interferon. J Infect Dis 1994;169:807-13.
    Abstract/FREE Full Text
  38. 38.↵
    1. Chieux V,
    2. Hober D,
    3. Chehadeh W,
    4. et al
    . MxA protein in capillary blood of children with viral infections. J Med Virol 1999;59:547-51.
    CrossRefMedlineWeb of Science
  39. 39.↵
    1. Akerblom HK,
    2. Knip M
    . Putative environmental factors in type 1 diabetes. Diabetes Metab Rev 1998;14:31-67.
    MedlineWeb of Science
  40. 40.↵
    1. Argos P,
    2. Kamer G,
    3. Nickiin MJH,
    4. Wimmer E
    . Similarity in gene organization and homology between proteins of animal picornaviruses and plant comovirus suggest common ancestry of these virus families. Nucleic Acids Res 1984;12:7251-67.
    Abstract/FREE Full Text
  41. 41.↵
    1. Atkinson MA,
    2. Maclaren NK
    . The pathogenesis of insulin-dependent diabetes mellitus. N Engl J Med 1994;331:1428-36.
    CrossRefMedlineWeb of Science
  42. 42.↵
    1. Horwitz MS,
    2. Bradeley LM,
    3. Harbertson J,
    4. Krahl T,
    5. Lee J,
    6. Sarvetnick N
    . Diabetes induced by coxsackie virus: initiation by bystander damage and not molecular mimicry. Nat Med 1998;4:781-5.
    CrossRefMedlineWeb of Science
  43. 43.↵
    1. Yoon J-W,
    2. Onodera T,
    3. Notkins AL
    . Virus induced diabetes mellitus: beta cell damage and insulin dependent hyperglycaemia in mice infected with coxsackievirus B4. J Exp Med 1978;148:1068-80.
    Abstract/FREE Full Text
  44. 44.↵
    1. Szopa TM,
    2. Ward T,
    3. Dronfield ND,
    4. Portwood ND,
    5. Taylor KW
    . Coxsackie B4 viruses with the potential to damage β-cells of the islets are present in clinical isolates. Diabetologia 1990;33:325-8.
    CrossRefMedlineWeb of Science
  45. 45.↵
    1. Jenson AB,
    2. Rosenberg HS,
    3. Notkins AL
    . Pancreatic islet cell damage in children with fatal viral infections. Lancet 1980;1:354-8.
    CrossRefMedlineWeb of Science
  46. 46.↵
    1. Rogge L,
    2. Barberis-Maino L,
    3. Biffi M,
    4. et al
    . Selective expression of an interleukin-12 receptor component by human T helper 1 cells. J Exp Med 1997;185:825-31.
    Abstract/FREE Full Text
  47. 47.↵
    1. Conlon KC,
    2. Urba WJ,
    3. Smith JW II.,
    4. Steis RG,
    5. Longo DL,
    6. Clark JW
    . Exacerbation of symptoms of autoimmune disease in patients receiving alpha-interferon therapy. Cancer 1990;65:2237-42.
    CrossRefMedlineWeb of Science
  48. 48.↵
    1. Fabris P,
    2. Betterle C,
    3. Floreani A,
    4. et al
    . Development of type 1 diabetes mellitus during interferon alpha therapy for chronic HCV hepatitis. Lancet 1992;340:548.
    MedlineWeb of Science
  49. 49.↵
    1. Murakami M,
    2. Iriuchijima T,
    3. Mori M
    . Diabetes mellitus and interferon-alpha therapy. Ann Intern Med 1995;123:318.
    FREE Full Text
  50. 50.↵
    1. Rabinovitch A,
    2. Suarez-Pinzon W,
    3. El-Sheikh A,
    4. Sorensen O,
    5. Power RE
    . Cytokine gene expression in pancreatic islet-infiltrating leukocytes of BB rats: expression of Th1 cytokines correlates with beta-cell destructive insulitis and IDDM. Diabetes 1996;45:749-54.
    Abstract
| Table of Contents

This Article

  1. J Infect Dis. (2000) 181 (6): 1929-1939. doi: 10.1086/315516
  1. AbstractFree
  2. » Full Text (HTML)Free
  3. Full Text (PDF)Free

Classifications

Share

  1. Email this article

Navigate This Article

Current Issue

  1. March 1, 2012 205 (5)
  1. Journal of Infectious Diseases
  1. Alert me to new issues

Most