Journal
of General Virology (2002), 83, 2169–2176. Printed in Great Britain
...................................................................................................................................................................................................................................................................................
Circulating and cell-bound antibodies increase coxsackievirus
B4-induced production of IFN-α by peripheral blood
mononuclear cells from patients with type 1 diabetes
Didier Hober,1 Wassim Chehadeh,1, 2 Jacques Weill,3 Christine Hober,4 Marie-Christine Vantyghem,5
Pascale Gronnier6 and Pierre Wattre! 1
1
Laboratory of Virology, UPRES EA ‘ Viral Pathogenesis of Diabetes Type 1’, University Hospital, and University Lille 2, 59037 Lille,
France
2
SEDAC Therapeutics, IBL-CNRS, 59021 Lille, France
3
Pediatric Endocrine Unit, University Hospital, Lille, France
4
Department of Medicine, Endocrinology and Metabolism, Charlon Hospital, Henin-Beaumont, France
5
Department of Endocrinology and Metabolism, University Hospital, Lille, France
6
Department of Pediatrics, Saint-Antoine Hospital, Catholic University, Lille, France
Increased levels of IFN-α have been found in patients with type 1 diabetes who have detectable
levels of coxsackievirus B4 (CVB4) RNA in their blood. The IFN-α-inducing activity of CVB4 in vitro
is weak but can be enhanced by human IgGs. Therefore, it was investigated in vitro whether a
preferential IFN-α response of peripheral blood mononuclear cells (PBMCs) to CVB4 exists in
patients with type 1 diabetes (n U 56) compared with healthy subjects (n U 20) and whether
antibodies play a role. In patients, the levels of IFN-α obtained after stimulation by PBMCs with
CVB4 were higher (P U 0n008), an individual IFN-α response by PBMCs to CVB4 was more frequent
(P U 0n0004) and increased levels of IFN-α were observed in CVB4-infected whole blood cultures.
The IFN-α-inducing activity of patients plasma and IgGs mixed with CVB4 and then added to PBMCs
was high in comparison with healthy subjects (P 0n001) and was inhibited by preincubating the
cells with anti-FcγRII, anti-FcγRIII and anti-CAR (coxsackievirus and adenovirus receptor) antibodies. The strong IFN-α responsiveness of PBMCs to CVB4 suggested that IgGs bound to the cell
surface might play a role. A short 56 SC incubation of PBMCs from patients responsive to CVB4
generated supernatants, which, when added to cells, exhibited IFN-α-enhancing activity in
combination with CVB4, whereas those of controls did not. Specific antibodies for FcγRI, FcγRII and
CAR inhibited this activity. These studies demonstrate that CVB4, through interactions with
circulating and/or cell-bound IgGs, can strongly induce the production of IFN-α by PBMCs from
patients with type 1 diabetes.
Introduction
IFN-α is a marker of ongoing virus infection (Green et al.,
1982 ; Flowers & Scott, 1985). Recently we reported that
increased levels of IFN-α in plasma were associated in 50 % of
cases with the presence of enterovirus sequences, particularly
Author for correspondence : Didier Hober. Correspondence address :
Laboratoire de Virologie, Ba# t. Paul Boulanger, Ho# pital Albert Calmette,
Centre Hospitalier Re! gional et Universitaire (CHRU), 59037 Lille
Cedex, France.
Fax j33 3 20 44 52 81. e-mail dhober!chru-lille.fr
0001-8450 # 2002 SGM
coxsackievirus B4 (CVB4), in the circulating blood of adults
and children with type 1 diabetes (Chehadeh et al., 2000a).
Moreover, IFN-α mRNA was detected in blood cells from
patients with IFN-α in their plasma. These results suggest that
IFN-α is produced during the course of CVB4 infection. In so
far as CVB4 components were found in the circulating blood
from patients, the synthesis of IFN-α in patients with type 1
diabetes could result from the interaction of CVB4 with
peripheral blood mononuclear cells (PBMCs).
Enteroviruses like polioviruses and CVB are referred to as
weak IFN-α inducers, as compared to strong IFN-α inducers
like Sendai virus (SV) and herpes simplex virus type 1 (HSV-1)
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sun, 18 Jun 2017 05:31:46
CBGJ
D. Hober and others
(Pitkaranta & Hovi, 1993 ; Feldman et al., 1994). However, it
has been shown that human polyvalent IgGs enhance the IFNα-inducing capacity of poliovirus in vitro (Palmer et al., 2000).
Furthermore, we have reported recently that CVB4-induced
synthesis of IFN-α by PBMCs in vitro can be enhanced through
interactions between CVB4, specific antibodies isolated from
the plasma of healthy subjects, FcγR and a receptor for
CVB called CAR (coxsackievirus and adenovirus receptor)
(Chehadeh et al., 2001 ; Bergelson et al., 1997). This suggests
that antibodies can play a role in the IFN-α response to CVB4.
Infection by a CVB serotype induces production of
serotype- and CVB group-specific antibodies (Frisk et al.,
1989). Epidemiological studies have indicated that CVB
infection is frequent in patients with type 1 diabetes (Yoon,
1990 ; Hyoty et al., 1995) ; therefore, these individuals may
have a higher prevalence of CVB antibodies compared with
appropriate controls (Szopa et al., 1993 ; Hyoty et al.,
1995 ; Hiltunen et al., 1997). Thus, CVB-specific antibodies can
be found in the circulating blood of diabetes patients and these
antibodies may play a role in the interaction between CVB4
and PBMCs.
The mechanism of the increased expression of IFN-α in
patients with type 1 diabetes is unknown. In the current study,
we investigated in vitro whether a preferential IFN-α response
by PBMCs to CVB4 exists in patients with type 1 diabetes
compared with healthy subjects and whether antibodies can be
involved.
Methods
Patients and healthy subjects. Blood samples were obtained
from 56 patients with type 1 diabetes : 12 newly diagnosed and 13
previously diagnosed children (median age 13 years, range 3–17 years)
and 20 newly diagnosed and 11 previously diagnosed adults (median age
37 years, range 20–69 years). Of the 56 patients, 30 had metabolic
decompensation (diabetic ketoacidosis or diabetic ketosis) and the
remaining 26 had no metabolic decompensation. All patients included in
this study were treated with insulin. Informed consent was obtained from
the adults and the parents of the children ; the committee on research
ethics approved the study protocol.
Blood samples were obtained from 20 healthy subjects : 10 adults (6
males and 4 females ; median age 28 years, range 23–45) and 10 children
(5 males and 5 females ; median age 12 years, range 7–16), who had not
any suspected immunological, infectious or metabolic disease. These
subjects were hospitalized or were outpatients at the hospital.
Viruses. The JBV strain of CVB4, kindly provided by J. W. Almond
(Aventis Pasteur, Marcy-L’Etoile, France), was grown in Hep-2 cells
(BioWhittaker) in Eagle’s minimum essential medium (MEM) (Gibco BRL)
supplemented with 10 % heat-inactivated foetal calf serum (FCS) (Eurobio)
and 1 % -glutamine (Eurobio). Supernatants were collected 3 days postinfection (p.i.) and then clarified at 1000 r.p.m. for 10 min. Virus titres
were determined by plaque-formation assay on Hep-2 cells and expressed as p.f.u.\ml. Aliquots of virus preparations were stored frozen
at k80 mC.
SV was kindly provided by D. Garcin (Department of Genetics and
Microbiology, University of Geneva, Switzerland). HSV-1 (a laboratory
CBHA
strain) was cultivated in Vero cells (ATCC) in MEM supplemented with
5 % FCS and 1 % -glutamine.
IFN-α induction
Whole blood cultures. Venous blood was collected in sterile 7n5 ml
tubes (Beckton Dickinson) containing 20 IU\ml heparin (Heparin
Choay). Samples (25 µl) of heparinized whole blood were dispensed
into each of triplicate wells of a 96-well tissue culture plate containing
225 µl RPMI 1640 (Gibco BRL) or 225 µl RPMI 1640 containing
10( p.f.u.\ml CVB4. The blood cultures were incubated in a humidified
incubator at 37 mC with 5 % CO .
#
PBMCs. PBMCs from heparinized blood were separated over a Ficoll–
paque solution (Diatrizoate Ficoll, Eurobio). Mononuclear cells were collected, washed three times with RPMI 1640, adjusted to a concentration of 2i10' cells\ml in RPMI supplemented with 10 % FCS and 1 %
-glutamine and then distributed as 0n1 ml aliquots into 96-well tissue
culture plates. Thereafter, 0n1 ml of medium containing virus at an
m.o.i. of 1 or medium containing virus at an m.o.i. of 1 and plasma at
different dilutions were added to wells. Cultures were then incubated
for 48 h at 37 mC in a 5 % CO atmosphere.
#
After this incubation period, whole blood and PBMC cultures were
irradiated with UV light for 1 h to inactivate residual viruses. Supernatants of culture were then harvested, clarified by centrifugation for
10 min at 1000 r.p.m. and stored at k20 mC until assayed for the
presence of IFN-α. Cultures not treated with viruses served as controls.
Immunoassay for IFN-α. The concentration of IFN-α was
determined by a specific and sensitive dissociation-enhanced lanthanide
fluoroimmunoassay (DELFIA), as described previously (Chehadeh et al.,
2000b). Antibodies were kindly provided by G. Alm, Biomedical Center,
Uppsala, Sweden. The assay detection limit was 0n5 IU IFN-α\ml.
Bioassay for IFN-α. Samples of whole blood and PBMC cultures
were assayed for antiviral activity by protection of MDBK cells
(BioWhittaker) against vesicular stomatitis virus-induced cytopathic
effects, as described previously (Chehadeh et al., 2000b). The assay
detection limit was 2 IU IFN-α\ml. The antiviral activity of supernatant
samples was neutralized with a rabbit antiserum to IFN-α (Biosource).
Protein A-affinity chromatography. IgGs were obtained from
the plasma of healthy subjects or patients using a protein A–Sepharose
CL-4B column, as described previously (Chehadeh et al., 2001).
Antibody-dependent assays. PBMCs were distributed into 96well tissue culture plates as described above. Virus–antibody complexes
were made by incubating plasma or IgGs at various dilutions with CVB4
for 1 h at 37 mC. The optimal dilution of plasma differed between subjects
and ranged from 1 : 100 to 1 : 1000, as described previously (Chehadeh
et al., 2001). CVB4 or CVB4–antibody complexes were added to cells at
an m.o.i. of 1 before incubation for 2 h at 37 mC. In certain experiments,
PBMCs were incubated with IgGs for 1 h and, after washing, were
incubated with CVB4 at an m.o.i. of 1 for 90 min. The cells were then
washed three times and cultured in RPMI 1640 containing 10 % FCS for
48 h at 37 mC in a 5 % CO atmosphere. After the incubation period,
#
supernatants were harvested, clarified and stored until assayed for the
presence of IFN-α.
Isolation of cytophilic antibodies. PBMCs (5i10(\ml) were
incubated for 30 min in a 56 mC water bath. Cells were removed by
centrifugation at 300 g for 10 min. The cell-free supernatants generated
(eluted fraction by PBMCs) were used in assays of CVB4-induced
production of IFN-α by PBMCs to detect the presence of cytophilic
antibody dissociated from the cell surface during the 56 mC incubation.
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sun, 18 Jun 2017 05:31:46
IFN-α production in type 1 diabetes patients
(a)
IFNa (IU/ml)
Antibodies. Monoclonal IgG1 neutralizing anti-human FcγRI
(CD64), clone 10.1 (Biotest), monoclonal IgG2a neutralizing anti-human
FcγRII (CD32), clone 2E1, and monoclonal IgG1 neutralizing anti-human
FcγRIII (CD16), clone 3G8, antibodies were purchased from Coulter
Immunotech. Monoclonal IgG1 anti-human CAR antibody (RmcB) was
kindly provided by J. Bergelson (Division of Infectious Disease, Children’s
Hospital of Philadelphia, Philadelphia, Pennsylvania, USA). Anti-D
human monoclonal IgG1 and IgG3 immune sera were kindly provided by
Etablissement de Transfusion Sanguine (Lille, France) and used as
irrelevant human IgG antibodies. Murine IgG1 and IgG2a (Coulter
Immunotech) were used as control antibodies.
PBMC
Whole Blood
1250
1000
750
500
250
150
100
50
Assays of CVB antibodies
0
Healthy subjects
n = 20
Plaque neutralization assay. The presence of anti-CVB4 neutralizing
ELISA. The Serion ELISA Classic kit (Institut Virion-Serion) was used
for quantitative detection of specific human anti-CVB IgG antibodies,
according to the manufacturer’s instructions. Results were expressed in
IU\ml.
Statistical analysis. Data are summarized as meanpSD. The
significance of the differences in IFN-α levels was determined by the
Mann–Whitney U-test. The χ#-test was used when appropriate.
Results
(b)
900
Diabetic patients
n = 56
plasma
CVB4
plasma + CVB4
750
IFNa (IU/ml)
antibodies (NAs) in plasma were routinely assessed according to the
method described elsewhere (Chehadeh et al., 2000a). Results were expressed as the inverse of the final dilution (titre) of the sample that
totally inhibited the cytopathic effect of CVB4 in Hep-2 cells.
600
350
300
150
0
CVB4-induced production of IFN-α in vitro
In healthy subjects, the mean level of CVB4-induced IFN-α
in the supernatant of PBMC cultures was significantly lower
than the one in the supernatant of whole blood cultures
(2n6p2n76 versus 46p15 IU\ml ; P 0n001, n l 20) (Fig.
1a). In patients with type 1 diabetes, the range of IFN-α levels
was large ; high individual IFN-α levels were observed. The
mean level of IFN-α in the supernatant of PBMC cultures was
significantly higher than the one in healthy subjects (85p
134 IU\ml ; P l 0n008, n l 56). An individual IFN-α response
by PBMCs of 32 IU\ml (meanj3 SD of healthy subjects l
11 IU\ml) was obtained in 24 of 56 patients (43 %) and 0 of
20 healthy subjects (0 %) (P l 0n0004, χ#-testing). The mean
level of IFN-α in the supernatant of whole blood cultures from
patients was higher than the one in healthy subjects but the
difference was not significant (109p158 IU\ml ; P l 0n07,
n l 56 versus healthy subjects). An individual IFN-α level in
whole blood cultures of 128 IU\ml (meanj3 SD of healthy
subjects l 91 IU\ml) was obtained in 17 of 56 patients (30 %)
and 0 of 20 healthy subjects (0 %) (P 0n02, χ#-testing). In
patients, high levels of IFN-α in CVB4-infected whole blood
cultures (up to 1024 IU\ml) were associated with low levels of
IFN-α in CVB4-infected PBMC cultures (0 IU\ml). In contrast
with CVB4, when PBMCs and whole blood were infected with
SV or HSV-1, the mean levels of IFN-α in patients were not
significantly higher than those in healthy subjects (SV-infected
PBMCs : 903p1069 versus 716p328 IU\ml, respectively,
P l 0n45 ; HSV-1-infected PBMCs : 830p525 versus 691p
300 IU\ml, respectively, P l 0n26) (data not shown).
Healthy subjects
n = 20
Diabetic patients
n = 56
Fig. 1. CVB4-induced production of IFN-α by PBMCs from patients with
type 1 diabetes. (a) Individual representation of IFN-α levels in CVB4infected whole blood and PBMC cultures. (b) Production of IFN-α by
patients’ PBMCs infected with CVB4 preincubated in the absence or
presence of their own plasma diluted at 1 : 100. Blood samples were
obtained from 20 healthy subjects and 56 patients with type 1 diabetes.
Culture supernatant samples were harvested 48 h p.i. IFN-α levels in
culture supernatants were determined by biological assay and the results
were expressed as IFN-α concentrations in IU/ml. Horizontal bars indicate
the mean in (a), whereas the meanjSD are shown in (b).
High IFN-α-inducing activity of plasma from patients
with type 1 diabetes combined with CVB4
When plasma samples from patients with type 1 diabetes
were preincubated with CVB4 and then added to the patient’s
own PBMCs, the levels of IFN-α were higher than those
obtained in the absence of plasma (P 0n001) (Fig. 1b).
Moreover, the mean level was higher than the one obtained
with plasma and PBMCs from healthy subjects (472p348 IU\
ml, n l 56, versus 134p72 IU\ml, n l 20 ; P 0n001). Fig.
2(a) shows that the levels of IFN-α were higher when plasma
from patients, instead of plasma from healthy subjects, were
added to PBMCs from healthy subjects (P 0n001), whereas
the levels of IFN-α were lower when plasma from healthy
subjects, instead of plasma from diabetes patients, were added
to PBMCs from diabetes patients (P 0n001).
To assess whether IgGs were involved in the enhancement
of CVB4-induced IFN-α production in these experiments,
plasma from controls or patients were passed over a protein
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sun, 18 Jun 2017 05:31:46
CBHB
D. Hober and others
(a)
Healthy subjects plasma + CVB4
IFNa (IU/ml)
1500
Patients plasma +CVB4
1000
500
0
Healthy subjects PBMC
n =3
(b)
Patients PBMC
n=5
1500
IFNa (IU/ml)
Healthy subjects n = 3
Patients n =5
1000
500
0
Flow-through fraction
of plasma + CVB4
Eluted fraction
of plasma + CVB4
Fig. 2. Role of plasma in the CVB4-induced production of IFN-α by PBMCs
from patients with type 1 diabetes. (a) High IFN-α-inducing activity of
CVB4 combined with plasma from patients with type 1 diabetes. Plasma
samples diluted at 1 : 100 were preincubated with CVB4 before being
added to PBMCs. IFN-α levels in culture supernatant samples at 48 h p.i.
were determined by biological assay. Plasma samples of five patients with
type 1 diabetes were tested with their PBMCs (n l 17) and with PBMCs
from three healthy subjects (n l 15). Plasma samples of three healthy
subjects were tested with their PBMCs (n l 9) and plasma samples of
two of them were tested with PBMCs from five patients with type 1
diabetes (n l 10). MeanjSD are presented. (b) IFN-α levels in cultures
by PBMCs infected with CVB4 preincubated with eluted or flow-through
fractions obtained from plasma passed over a protein A-affinity
chromatography column. Plasma samples were obtained from three
healthy subjects and from five patients with type 1 diabetes. The
experiments were repeated with PBMCs from three different donors.
MeanjSD of seven independent experiments with plasma fractions from
healthy subjects and 10 independent experiments with plasma fractions
from patients with type 1 diabetes are presented.
A-affinity chromatography column. The eluted fractions containing IgG antibodies and the flow-through fraction (free from
IgG1, -2 and -4) were preincubated at 20 µg\ml with CVB4 for
1 h at 37 mC before being added to PBMC cultures. As seen in
Fig. 2(b), eluted fractions enhanced CVB4-induced IFN-α
production, whereas flow-through fractions did not. The
enhancing activity obtained with IgGs from patients was
significantly higher than the one obtained with IgGs from
controls (P 0n001). In contrast, irrelevant human IgGs had
no IFN-α-enhancing activity.
The enhancing effect of IgGs from controls and patients on
CVB4-induced IFN-α production in PBMC cultures was
suppressed significantly when PBMCs were preincubated with
2 µg\ml of either anti-human FcγRII IgG2a antibodies or antihuman FcγRIII IgG1 antibodies but not with anti-human FcγRI
IgG1 antibodies (Table 1). Irrelevant control antibodies did not
inhibit IgG-mediated enhancement of CVB4-induced IFN-α
production. To examine the role of CAR on the efficiency of
IgG-mediated IFN-α production, PBMCs were preincubated
for 1 h with anti-human CAR antibodies before challenging
with CVB4 preincubated with IgGs from healthy subjects or
patients. IFN-α production was inhibited in the presence of
anti-CAR antibodies [80p12 (n l 3) and 93p5 % (n l 3)
inhibition in experiments with IgGs from donors and patients,
respectively], whereas control isotypes did not inhibit IgGmediated enhancement of CVB4-induced IFN-α production
(data not shown). When PBMCs were preincubated with antiFcγRII, anti-FcγRIII or anti-CAR antibodies before adding SV
or HSV-1, the levels of IFN-α were not reduced (data not
shown), which demonstrated that these antibodies did not give
a negative signal to IFN-α-producing cells that blocked IFN-α
synthesis.
Role of cytophilic IgGs in CVB4-induced production of
IFN-α by PBMCs from patients with type 1 diabetes
We observed that PBMCs from patients with type 1
diabetes produced up to 512 IU\ml IFN-α in response to
CVB4, whereas those from healthy subjects were low re-
Table 1. Antibodies to FcγR inhibit the production of IFN-α by PBMCs infected with CVB4 preincubated with IgGs
PBMCs isolated from the blood of three healthy subjects were preincubated with anti-FcγR or control antibodies for 1 h before adding CVB4
mixed with IgGs obtained from the plasma of healthy subjects (n l 3) or diabetic patients (n l 3). IFN-α levels in culture supernatants at 48 h p.i.
were measured by DELFIA. Results are expressed as meanpSD.
Preincubation of PBMCs with
IFN-α (IU/ml) produced by
PBMCs infected with CVB4jIgG from
Medium
IgG1
IgG2a
Anti-FcγRI
antibody
Anti-FcγRII
antibody
Anti-FcγRIII
antibody
Healthy subject
Patient
219p99
980p135
220p97
1071p121
233p112
1018p82
213p93
1009p42
37p18
40p20
51p27
51p23
CBHC
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sun, 18 Jun 2017 05:31:46
IFN-α production in type 1 diabetes patients
40
(a)
20
Medium
CVB4
Medium
CVB4
30
IFNa (IU/ml)
20
10
0
PBMC
(n =10)
(b)
0
PBMC + IgG
(n = 10)
PBMC +
PBMC +
PBMC +
Eluted fraction Eluted fraction Eluted fraction
of PBMC
of PBMC
of PBMC
from
from patients
from patients
controls
responsive
non responsive
to CVB4
to CVB4
(n =3)
(n =3)
(n =3)
Medium
CVB4
20
(b)
15
sponders ( 8 IU\ml) (see Fig. 1). In so far as the IFN-αinducing activity of CVB4 can be enhanced by IgGs, as shown
in our experiments, we investigated whether antibodies bound
to the surface of cells played a role in the IFN-α response by
PBMCs to CVB4.
Preincubation of PBMCs from healthy subjects with IgGs,
after washing and before adding CVB4, resulted in the
production of IFN-α (Fig. 3a). These results suggested that
CVB4-specific antibodies bound to the surface of cells. To
dissociate and recover antibodies from PBMCs, the PBMCs
were heated at 56 mC for 30 min. Supernatants were generated
by subsequent removal of cells and cell debris by centrifugation. CVB4-induced IFN-α production by PBMCs was used as
a test to assess IFN-α-enhancing activity in these supernatants.
As seen in Fig. 3(b), a significant IFN-α-enhancing activity was
detected in supernatants of IgG-treated PBMCs heated at
56 mC. The preincubation of PBMCs used for testing super-
Anti-CAR
antibodies
Anti-FcçRIII
antibodies
0
Anti-FcçRII
antibodies
Fig. 3. CVB4-induced production of IFN-α in cultures by PBMCs. (a) IFN-α
levels in cultures of PBMCs preincubated with IgGs before adding CVB4.
The cells were preincubated in the absence or presence of 20 µg/ml IgG
for 1 h, then they were washed and mock-infected or infected with CVB4.
IgGs were obtained from the plasma of healthy subjects by using protein
A-affinity chromatography. The experiments were repeated with IgGs from
10 different donors. MeanjSD are presented. (b) IFN-α levels in cultures
of PBMCs preincubated with supernatants generated from heated PBMCs.
PBMCs from healthy subjects were preincubated for 1 h in the absence or
presence of cell-free supernatant of PBMCs from healthy subjects either
treated or not with IgGs, heated for 30 min in a 56 mC water bath. Then,
the cells were washed three times and mock-infected or infected with
CVB4 at an m.o.i. of 1 for 90 min. MeanjSD of four independent
experiments are presented. (a, b) The IFN-α levels in culture supernatants
at 48 h p.i. were measured by DELFIA.
5
Anti-FcçRI
antibodies
PBMC +
PBMC +
supernatant of
supernatant of
PBMCs heated at 56 °C IgG-treated PBMCs
(n=4)
heated at 56 °C
(n=4)
IgG2a
0
10
IgG1
10
IFNa (IU/ml)
IFNa (IU/ml)
30
10
Medium
IFNa (IU/ml)
(a)
Fig. 4. CVB4-induced production of IFN-α enhanced with supernatant
generated from heated PBMCs from patients with type 1 diabetes. (a)
PBMCs were isolated from blood samples of : healthy subjects (n l 3) ;
patients with type 1 diabetes whose PBMCs produced IFN-α in response
to CVB4 (responsive, n l 3) ; patients with type 1 diabetes whose PBMCs
did not produce IFN-α in response to CVB4 (nonresponsive, n l 3).
Eluted fractions of PBMCs were cell-free supernatants of PBMCs heated
for 30 min in a 56 mC water bath. These eluted fractions were incubated
for 1 h with PBMCs from healthy subjects, then the cells were washed
three times and mock-infected or infected with CVB4 at an m.o.i. of 1 for
90 min. IFN-α levels in culture supernatants at 48 h p.i. were measured by
DELFIA. MeanjSD of three independent experiments are presented. (b)
Role of FcγR and CAR in the CVB4-induced production of IFN-α enhanced
with supernatant generated from heated PBMCs from patients with type 1
diabetes. Eluted fractions generated by PBMCs from patients responsive to
CVB4 (n l 3) were added to PBMCs from healthy subjects preincubated
for 1 h with NAs directed against FcγRI, FcγRII, FcγRIII and CAR or with
control antibodies. MeanjSD of three independent experiments are
presented.
natants with anti-FcγRI, anti-FcγRII and anti-CAR but not antiFcγRIII decreased by approximately 100 % the level of IFN-α
(data not shown).
These results suggested that cell-bound antibodies can play
a role in the production of IFN-α by PBMCs infected with
CVB4. Therefore, attempts were made to recover an IFN-αenhancing activity from PBMCs from controls and patients
with type 1 diabetes. As seen in Fig. 4(a), no significant activity
was detected in the supernatant (eluted fraction) of heated
PBMCs from controls and patients producing no IFN-α in the
presence of CVB4, whereas significant activity was detected in
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sun, 18 Jun 2017 05:31:46
CBHD
D. Hober and others
(a)
Discussion
IFN (IU/ml)
103
102
101
100
10–1
10–2
10–1
100
101
102
103
104
Anti-CVB neutralizing antibodies (titer)
(b)
IFN (IU/ml)
104
103
102
101
10–2
10–1
100
101
102
103
104
Anti-CVB neutralizing antibodies (titer)
Fig. 5. Relationship between the values of CVB4-induced production of
IFN-α by PBMCs and anti-CVB antibodies titres of plasma samples,
determined by plaque-neutralization assay in healthy subjects ($) and
patients with type 1 diabetes (#). (a) PBMCs infected with CVB4. (b)
PBMCs infected with CVB4 preincubated with plasma diluted at 1 : 100.
supernatants of heated PBMCs from patients producing IFN-α
in the presence of CVB4 (256, 64 and 64 IU\ml). Anti-FcγRI,
anti-FcγRII and anti-CAR but not anti-FcγRIII inhibited the
IFN-α-inducing activity of these supernatants (Fig. 4b).
Relationship between CVB4-induced production of
IFN-α by PBMCs and the level of anti-CVB antibody
The production of IFN-α by PBMCs in response to CVB4
was compared with the levels of anti-CVB4 NAs in healthy
subjects and patients with type 1 diabetes (Fig. 5). Higher
levels of NAs were found in patients compared with controls.
For each individual, the titre of antibody was plotted with
regard to the level of IFN-α. There was no correlation between
IFN-α induced by CVB4, or CVB4 preincubated with plasma,
and the titres of NAs. In certain patients, the detection of NAs
was negative or the titres were very low, whereas the level of
IFN-α was high ; therefore, the levels of CVB antibodies were
determined using another method. However, using ELISA,
the levels of anti-CVB IgG antibodies were concordant with
the CVB-neutralization test results and there was still no
correlation between the level of antibodies detected by ELISA
and the level of IFN-α (data not shown).
CBHE
The current study demonstrates that CVB4 can strongly
induce the production of IFN-α by PBMCs from patients with
type 1 diabetes in the absence or presence of plasma, whereas
the levels of CVB4-induced synthesis of IFN-α by PBMCs from
healthy controls are low or moderate in the same conditions.
There are several noteworthy considerations for the
systems used in the study. The measurement of IFN-αproducing capacity by the whole blood method has been
reported previously by us and other teams (Hober et al.,
1998 ; Uno et al., 1996). The production of IFN-α per
mononuclear cell was higher in whole blood than in PBMC
cultures from healthy controls and patients with type 1
diabetes, although the cell number in whole blood cultures was
about three times as low as that in PBMC cultures (data not
shown). The discordant IFN-α levels in whole blood and
PBMC cultures infected with CVB4, in particular the high
values in whole blood cultures from patients, prompted us
to study the role of plasma in CVB4-induced production of
IFN-α.
The production of IFN-α by PBMCs from healthy controls
and patients with type 1 diabetes, obtained with a mixture of
CVB4 and plasma, depended on the antibodies contained in
the plasma, as evidenced by : (1) the eluted fractions containing
IgGs obtained from plasma passed over a protein A–Sepharose
column ; and (2) the suppression of the enhancing effect of
virus\eluted fraction by anti-FcγRII and anti-FcγRIII antibodies.
The activity of plasma and IgGs obtained in our experiments
suggested that the level of IFN-α-enhancing antibodies was
higher in patients with type 1 diabetes than in healthy subjects.
It may be assumed that IgGs from diabetics formed immune
complexes with CVB4, since the adsorption of CVB4 complexed to patients IgGs on C1q-coated microwell plates,
according to the procedure described previously (Chehadeh et
al., 2001), inhibited the IFN-α-enhancing activity of IgGs (data
not shown). Furthermore, our experiments with neutralizing
anti-CAR antibodies suggested that CAR, a specific receptor
for CVB4, in addition to FcγRII and FcγRIII, played a role in the
induction of IFN-α by CVB4–antibody complexes in PBMCs
from patients and healthy controls. This result showed that
IFN-α production was not just a consequence of cell activation
by binding of CVB4–IgG complexes but also depended on
binding of CVB4 to a specific cell surface receptor. Our results
agree with those of Palmer et al. (2000), suggesting that FcγRII
plays a role in the poliovirus–antibody complex-induced
production of IFN-α by PBMCs. However, in the present
study, FcγRIII played a role, whereas in their experiments, it did
not ; the difference could be related to the virus used and\or to
the nature of IFN-α-producing cells. Indeed, previous works
from our laboratory showed that the major IFN-α-producing
cells in response to CVB4–IgG complexes were CD14+
monocytes (Chehadeh et al., 2001), whereas Palmer and
colleagues suggested that poliovirus–antibody complexes
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sun, 18 Jun 2017 05:31:46
IFN-α production in type 1 diabetes patients
activated the synthesis of IFN-α by natural IFN-α-producing
cells characterized as type 2 dendritic cell precursor (Siegal
et al., 1999 ; Palmer et al., 2000). In an extension of the present
work we will attempt to determine whether monocytes are
involved in the production of IFN-α observed in vivo and in
vitro in response to CVB4 infection in individuals with type 1
diabetes.
We demonstrated that PBMCs armed in vitro with
cytophilic antibodies produced IFN-α in the presence of CVB4.
Moreover, we provide evidence of the presence of antibodies
on the surface of freshly isolated PBMCs from patients
producing IFN-α in response to CVB4. The role of antibodies
was evidenced by the following experiments : (1) a 30 min
incubation at 56 mC was used to recover cell surface-associated
antibodies, as described previously (Tyler et al., 1989) ; (2) the
supernatants generated in this manner contained significant
levels of IFN-α-enhancing activity in combination with CVB4,
as measured by IFN-α assays. A role for IgGs in CVB4-induced
production of IFN-α in these experiments was indicated by the
ability of monoclonal antibodies against FcγRI and FcγRII to
significantly inhibit the production of IFN-α. The anti-FcγRII
monoclonal antibody (clone 2E1) used in this study binds well
to FcγRIIa-expressing cells but is marginally reactive with cells
expressing FcγRIIb (Van de Winkel & Anderson, 1995). Thus,
the effect of the anti-FcγRII monoclonal antibody in our
experiments do not result from a possible activation of FcγRIIb,
which negatively regulates intracellular signalling passing
through immunoreceptor tyrosine-based activation motifs and
their cytoplasmic ligands like FcγRI-mediated activation of
IFN-α (Daeron, 1997). IFN-α production was inhibited by antiFcγRII but not by anti-FcγRIII antibodies, suggesting that
FcγRIII, which, like FcγRII, binds immune complexes, is not
involved in IFN-α production nor is it expressed on the IFN-αproducing cells in experiments with PBMCs armed with
cytophilic antibodies. The presence of both FcγR(I and II) and
CAR molecules on cells was required for CVB4-induced
IFN-α synthesis by PBMCs armed with cytophilic antibodies ;
otherwise, blocking of FcγR or CAR independently could not
inhibit IFN-α synthesis. The results of our experiments suggest
that IgGs bound FcγRI and FcγRII, which is in agreement with
the fact that these receptors bind monomeric IgGs. In contrast
with FcγRI, FcγRII binds monomeric IgGs weakly but reacts
mostly with IgG complexes (Saute' s, 1997 ; Carayannopoulos
& Capra, 1993). However, when isolated from blood, FcγRbearing cells are saturated by serum IgGs. In our system, one
possible explanation for the effect of antibodies preincubated
with freshly isolated PBMCs already saturated by IgGs is the
following : IgGs added to cells bind to cell surface molecules (glycosyltransferases and\or lectin-like molecules)
(Carayannopoulos & Capra, 1993), then, when CVB4 is added,
immune complexes formed compete with monomeric IgGs
for receptors and displace bound monomeric IgGs. Indeed,
conformational changes in the Fc region by virus–antibody
complex formation facilitates interactions with the Fc receptor
(Frey & Einsfelder, 1984) ; therefore, immune complexes are
several times more effective than monomeric IgGs at competing for receptors.
The IFN-α response mediated by the supernatants of heated
PBMCs from patients producing IFN-α in the presence of
CVB4 was reduced compared with the responses of these
PBMCs to CVB4 (13p4 versus 128p64 IU\ml). This may be
due to the relatively low concentrations of enhancing antibodies in the preparations. Moreover, the effects of supernatants of heated PBMCs from patients can be underestimated.
Indeed, they were tested with cells from healthy controls ;
whether these cells can bind IgGs as well as cells from patients
is not known.
When PBMCs from patients with type 1 diabetes were
exposed to CVB4 in the absence or presence of their own IgGs,
the levels of IFN-α were higher than those obtained with cells
and IgGs from healthy subjects. The high IFN-α response to
CVB4 in diabetic patients was not related to age of onset of
diabetes, duration of diabetes or metabolic decompensation
(data not shown), which indicates intrinsic hyperresponsiveness to CVB4 in type 1 diabetes. It has been reported that there
is an hyper-IFN-α responsiveness by PBMCs to IFN-α inducers
like polyinosinic: polycytidilic acid [poly(I: C)] in type 1
diabetes (Toms et al., 1991) ; however, in the current work we
found no difference between patients and healthy subjects in
response to IFN-α inducers like HSV-1 and SV. The difference
observed in our study of CVB4 response in patients and
healthy subjects can be explained by a history of CVB
infections responsible for the presence of higher levels of IFNα-enhancing anti-CVB antibodies in patients than in healthy
subjects. Nevertheless, there was no correlation between the
titres of anti-CVB antibodies detected in patients by neutralization assay or by ELISA and the production of IFN-α by their
PBMCs. These results are in agreement with our previous
studies showing that IFN-α-enhancing antibodies directed
towards CVB4 detected in healthy controls were different
from NAs against that virus (Chehadeh et al., 2001) and a priori
suggest that IFN-α-enhancing antibodies do not bind H
antigen, which is incorporated into the immunoenzymatic
assay used in our experiments for detecting anti-CVB antibodies. Together these data are consistent with the existence
of anti-CVB antibodies in plasma from patients with type 1
diabetes not detected by methods usually available.
The preferential IFN-α response by PBMCs from patients
with type 1 diabetes to CVB4 supports the hypothesis that
CVB4 can play a role in the development of type 1 diabetes,
although the mechanism has not been elucidated. Further
studies are needed to know whether the observed IFN-α
hyperactivity towards CVB4 is an event prior to (predisposing)
or post clinical onset of type 1 diabetes.
In conclusion, the current studies demonstrate that CVB4
can strongly induce the in vitro production of IFN-α by PBMCs
from patients with type 1 diabetes in comparison with healthy
subjects. The IFN-α-inducing activity of CVB4 is strongly
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sun, 18 Jun 2017 05:31:46
CBHF
D. Hober and others
enhanced by IgGs from patients with type 1 diabetes in
comparison with IgGs from healthy subjects. Moreover, it has
been shown that the CVB4-induced production of IFN-α by
PBMCs from patients can result from the presence of IFN-αenhancing-specific antibodies on the cell surface. Studies are
needed to characterize the antibodies involved in the CVB4induced production of IFN-α and to define their role in the
pathogenesis of CVB4 infection. Future works will be directed
along this line in our laboratory.
The authors thank Delphine Caloone for technical assistance. This
work was supported by CHRU Lille (AO98\1911) and Ministe' re de
l’Education Nationale de la Recherche et de la Technologie, Universite! de
Lille II, France.
Hiltunen, M., Hyoty, H., Knip, M., Ilonen, J., Reijonen, H., Vahasalo, P.,
Roivainen, M., Lonnrot, M., Leinikki, P., Hovi, T. & Akerblom, H. K.
(1997). Islet cell antibody seroconversion in children is temporally
associated with enterovirus infections. Childhood Diabetes in Finland
(DiMe) Study Group. Journal of Infectious Diseases 175, 554–560.
Hober, D., Benyoucef, S., Chehadeh, W., Chieux, V., De La Tribonnie' re,
X., Mouton, Y., Bocket, L. & Wattre! , P. (1998). Production of
interleukin-4, interferon(IFN)-γ and IFN-α in human immunodeficiency
virus-1 infection : an imbalance of type 1 and type 2 cytokines may reduce
the synthesis of IFN-α. Scandinavian Journal of Immunology 48, 436–442.
Hyoty, H., Hiltunen, M., Knip, M., Laakkonen, M., Vahasalo, P.,
Karjalainen, J., Koskela, P., Roivainen, M., Leinikki, P., Hovi, T. and
others (1995). 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 44, 652–657.
Palmer, P., Charley, B., Rombaut, B., Daeron, M. & Lebon, P. (2000).
References
Bergelson, J. M., Cunningham, J. A., Droguett, G., Kurt-Jones, E. A.,
Krithivas, A., Hong, J. S., Horwitz, M. S., Crowell, R. L. & Finberg, R.
W. (1997). Isolation of a common receptor for coxsackie B viruses and
adenoviruses 2 and 5. Science 275, 1320–1323.
Carayannopoulos, L. & Capra, J. D. (1993). Immunoglobulins : structure
and function. In Fundamental Immunology, 3rd edn, pp. 283–314. Edited
by W. E. Paul. New York : Lippincott Williams & Wilkins.
Chehadeh, W., Weill, J., Vantyghem, M. C., Alm, G., Lefe' bvre, J.,
Wattre! , P. & Hober, D. (2000a). Increased level of interferon-α in blood
of patients with insulin-dependent diabetes mellitus : relationship with
coxsackievirus B infection. Journal of Infectious Diseases 181, 1929–1939.
Chehadeh, W., Kerr-Conte, J., Pattou, F., Alm, G., Lefe' bvre, J., Wattre! ,
P. & Hober, D. (2000b). Persistent infection of human pancreatic islets
by coxsackievirus B is associated with alpha interferon synthesis in β
cells. Journal of Virology 74, 10153–10164.
Chehadeh, W., Bouzidi, A., Alm, G., Wattre! , P. & Hober, D. (2001).
Human antibodies isolated from plasma by affinity chromatography
increase the coxsackievirus B4-induced synthesis of interferon-α by
human peripheral blood mononuclear cells in vitro. Journal of General
Virology 82, 1899–1907.
Daeron, M. (1997). Fc receptor biology. Annual Review of Immunology
15, 203–234.
Feldman, S. B., Ferraro, M., Zheng, H. M., Patel, N., Gould-Fogerite,
S. & Fitzgerald-Bocarsly, P. (1994). Viral induction of low frequency
interferon-α producing cells. Virology 204, 1–7.
Flowers, D. & Scott, G. M. (1985). How useful are serum and CSF
interferon levels as a rapid diagnostic aid in virus infections? Journal of
Medical Virology 15, 35–47.
Frey, J. & Einsfelder, B. (1984). Induction of surface IgG receptors in
cytomegalovirus-infected human fibroblasts. European Journal of Biochemistry 138, 213–216.
Frisk, G., Nilsson, E., Ehrnst, A. & Diderholm, H. (1989). Enterovirus
IgM detection : specificity of mu-antibody-capture radioimmunoassays
using virions and procapsids of coxsackie B virus. Journal of Virological
Methods 24, 191–202.
Green, J. A., Charette, R. P., Yeh, T. J. & Smith, C. B. (1982). Presence
of interferon in acute- and convalescent-phase sera of humans with
CBHG
influenza or influenza-like illness of undetermined etiology. Journal of
Infectious Diseases 145, 837–841.
Antibody-dependent induction of type I interferons by poliovirus in
human mononuclear blood cells requires the type II Fcγ receptor (CD32).
Virology 278, 86–94.
Pitkaranta, A. & Hovi, T. (1993). Induction of interferon in human
leukocyte cultures by natural pathogenic respiratory viruses. Journal of
Interferon Research 13, 423–426.
Saute' s, C. (1997). Structure and expression of Fc receptors (FcR). In Cellmediated Effects of Immunoglobulins, pp. 29–66. Edited by W. H. Fridman
& C. Saute' s. Heidelberg : Springer-Verlag.
Siegal, F. P., Kadowaki, N., Shodell, M., Fitzgerald-Bocarsly, P. A.,
Shah, K., Ho, S., Antonenko, S. & Liu, Y. J. (1999). The nature of the
principal type 1 interferon-producing cells in human blood. Science 284,
1835–1837.
Szopa, T. M., Titchener, P. A., Portwood, N. D. & Taylor, K. W. (1993).
Diabetes mellitus due to viruses : some recent developments. Diabetologia
36, 687–695.
Toms, G. C., Baker, P. & Boucher, B. J. (1991). The production of
immunoreactive α- and γ-interferon by circulating mononuclear cells in
type 1 diabetes. Diabetic Medicine 8, 547–550.
Tyler, D. S., Nastala, C. L., Stanley, S. D., Matthews, T. J., Lyerly, H. K.,
Bolognesi, D. P. & Weinhold, K. J. (1989). GP120 specific cellular
cytotoxicity in HIV-1 seropositive individuals. Evidence for circulating
CD16+ effector cells armed in vivo with cytophilic antibody. Journal of
Immunology 142, 1177–1782.
Uno, K., Nakano, K., Maruo, N., Onodera, H., Mata, H., Kurosu, I.,
Akatani, K., Ikegami, N., Kishi, A., Yasuda, Y., Tanaka, K., Setoguchi, J.,
Kondo, M., Muramatsu, S. & Kishida, T. (1996). Determination of
interferon-α-producing capacity in whole blood cultures from patients
with various diseases and from healthy persons. Journal of Interferon and
Cytokine Research 16, 911–918.
Van de Winkel, J. G. J. & Anderson, C. L. (1995). CD32 cluster
workshop report. In Leucocyte Typing V, pp. 823–826. Edited by S. F.
Schlossman. New York : Oxford University Press.
Yoon, J. W. (1990). The role of viruses and environmental factors in the
induction of diabetes. Current Topics in Microbiology and Immunology 164,
95–123.
Received 11 March 2002 ; Accepted 7 May 2002
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sun, 18 Jun 2017 05:31:46