| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Original Studies |
INSERM U-449, Faculté de Médecine René T.H. Laennec (A.M., C.T., J.O., A.M.M.), 69372 Lyon; and Faculté de Pharmacie, Unité Mixte de Recherche 9921, (F.R.), 34060 Montpellier, France
Address all correspondence and requests for reprints to: Dr. A Mayer, INSERM U-449, Faculté de Médecine René T.H. Laennec, rue Guillaume Paradin, 69372 Lyon Cedex 08. France. E-mail: madec{at}laennec.univ-lyon1.fr
 |
Abstract |
|---|
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
3) above the
mean + 3 SD of 39 healthy controls and of 7 of 15 (47%)
diabetic patients of long duration (P = 0.001).
Responses to insulin were not dictated by specific major
histocompatibility complex class II association and were not observed
in normal subjects with diabetes-associated human leukocyte
antigen-DR/DQ alleles. Whereas no relation of PBMC reactivity with
insulin autoantibodies was found, there was a positive correlation with
the presence of at least one of the four autoantibodies tested and with
IA-2 antibody. An interesting finding was that the proportion of
patients with subsequent low insulin requirement, up to 24 months, was
significantly higher in patients who showed PBMC reactivity to insulin
(8 of 8) than in those who did not (10 of 24, 42%;
P = 0.004). The former had a higher mean
stimulation index than the latter (3.3 ± 2.6 vs.
1.5 ± 0.6; P = 0.006). Furthermore,
interleukin-4 (IL-4) production was lower in type 1 diabetic patients
who proliferated to insulin than in those who did not (23 ± 15
vs. 64 ± 47 pg/mL; P = 0.04),
but interferon-
, IL-2, and IL-10 productions were similar. In
conclusion, these results suggest that proliferation to insulin may
reflect the presence of an higher residual ß-cell mass. |
Introduction |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
(IFN) and interleukin-2
(IL-2), and protective Th2-type T cells, producing IL-4 and IL-10 (7, 8). Because of the obvious lack of availability of intrapancreatic cell
infiltrates in humans, only circulating T cells as peripheral blood
cells can be studied. Peripheral blood mononuclear cells (PBMC) from
individuals at risk of developing diabetes and from recent-onset or
long duration type 1 diabetic patients have been shown to react to
various ß-cell antigens, such as insulin, GAD, IA-2, ß-cell
secretory granules, membrane preparations of RIN cells, islets, and
multiple islet cell proteins (9, 10, 11, 12, 13, 14, 15, 16, 17). However, the results of several
of these studies are conflicting whatever the autoantigen used. Also,
the results of collaborative studies, using the same antigen
preparations distributed to several groups participating in the First
Workshop on Cellular Immunity in Type 1 Diabetes (Canberra,
Australlia, 1996) were very heterogeneous. Nonspecific T cell responses
to contaminants of antigen preparations such as glutanic acid
decarboxylase (GAD) and IA2 have been observed. In the extensive list
of potential pathogenic ß-cell autoantigens in type 1 diabetes,
insulin is of particular interest. Indeed, insulin is the only one
specifically expressed by ß-cell (18, 19). In addition, studies on
the humoral immune markers in the natural history of type 1 diabetes
have suggested that autoantibodies to insulin are among the earliest to
appear, although at lower frequency than that to the other ß-cell
antigens (20). Finally, insulin appears to also be the target of
cellular immunity (9, 13, 21). However, this point is controversial
(11, 13, 17). In the present study, we have combined the analysis of peripheral blood T cell reactivity to human insulin with clinical follow-up data, HLA haplotypes, autoantibodies, and T cell phenotype evaluated by cytokine production in new-onset type 1 diabetic patients in comparison to those in diabetic subjects of long duration and healthy control individuals with diabetes-associated HLA-DR/DQ alleles.
|
Subjects and Methods |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Peripheral venous blood samples were obtained after obtaining informed consent from 44 recent-onset type 1 diabetic patients (24 females and 20 males) at the time of diagnosis. These patients were hospitalized in the Rhône-Alpes area, France, between January 1997 and January 1998. Criteria to be included were: age 30 yr or less (mean, 14 ± 7.1 yr; range, 1–30), clear manifestations of insulinopenia of 6-week duration or less before diagnosis (fasting hyperglycemia, >200 mg/dL, ketosis, polyuria, polydipsia, and weight loss), and interval between admission to the hospital and study 7 days or less (mean, 3.7 ± 1.9 days; range, 1–7). Metabolic stabilization by insulin treatment (<7 days) was obtained before further investigations. A subgroup of 32 of 44 patients was followed up for 6 months. At 6 months, patients were considered to be in remission according to their daily insulin requirements in the face of adequate blood glucose control or hemoglobin A1c, as suggested by Shah et al. (22). In 18 patients (hemoglobin A1c, 5.7 ± 1.2% compared to 7.5 ± 1.3% in the 44 others; P < 0.001), the total daily insulin requirement was less than 0.5 U/kg; these were considered to be in remission and were denoted patients with low insulin requirement. Among them, 4 patients had insulin needs of 0.2 U/kg·day or less. Two other groups of subjects were investigated. Fifteen type 1 diabetic patients of long duration (5 females and 10 males) were studied; disease duration was 1 yr or more (mean, 8.9 ± 6.7; range, 1–23), and their ages ranged from 6–25 yr (mean, 15 ± 5 yr). A total of 39 nondiabetic healthy individuals (20 females and 19 males), aged 17.3 ± 7.9 yr, served as controls. These subjects were matched for age, sex, and ethnic origin with the recent-onset diabetic patients, and they presented no familial history of diabetes or autoimmune diseases. They were also selected to increase the prevalence in the control group of susceptibility to HLA class II haplotypes (see below, HLA determination paragraph).
PBMC cultures
Peripheral venous blood was drawn between 0800–0900 h into sterile heparinized tubes and was processed within 2 h. Mononuclear cells were isolated under sterile conditions by centrifugation in Ficoll Histopaque 1077 (Sigma Chemical Co., St. Quentin Fallavier, France). Cell viability was always greater than 90% using trypan blue stain (BioWhittaker, Inc., Verviers, Belgium) exclusion. After washing, cells were resuspended in RPMI 1640 medium (BioWhittaker, Inc.) supplemented with 50 IU/mL penicillin-streptomicin (Sigma Chemical Co.), 2 mmol/L L-glutamine (Sigma Chemical Co.), 10% pooled human AB+ serum (Sigma Chemical Co.). Pooled human AB+ serum was chosen because of a high level of nonspecific proliferation in control subjects in the presence of autologous serum, as assessed in previous experiments. PBMC (1.5 x 105) were distributed in 180-µL aliquots in round bottomed 96-well plates, and 20 µL antigen solution or medium alone were added to quadruplicate wells. PBMC were incubated in the presence of no antigen, with control antigens BSA at 10 µg/mL or purified tetanus toxoid (TT) at 10 µg/mL (Pasteur, Paris, France), or with polyclonal activator interleukin-2 (IL-2) at 75 U/mL (Hoffmann-La Roche, Inc., Basel, Switzerland). In preliminary experiments, we tested insulin from various sources and selected recombinant human insulin produced in yeast from Sigma Chemical Co.. Three doses of insulin (1, 10, and 20 µg/mL) were used. All of the antigens and media used were tested and certified endotoxin free by the pharmacy laboratory, Herriot Hospital (Lyon, France). PBMC were cultured in a total volume of 200 µL for 7 days at 37 C in 5% CO2. After 6 days, 10 µL RPMI containing 1 µCi [3H]thymidine (NEN, Boston, MA) were added to each well, and the incubation was continued for 16–18 h. Cultures were then harvested, and [3H]thymidine incorporation into the cells was semiautomatically measured by liquid scintillation counting. The mean value of each quadruplicate was calculated. Proliferation was expressed as a stimulation index (SI), which corresponded to the ratio of the counts per min of PBMC in the presence of antigen divided by the counts per min of PBMC without antigen. A SI of 3.0 or more was considered positive, i.e. above the mean + 3 SD of controls. Coefficients of variation (SD/mean) ranged from 4–49%.
Cytokine measurements
Cytokine measurements were performed on supernatants obtained from whole blood or PBMC cultures.
Whole blood assay
Peripheral venous blood from 24 of the 44 new-onset type 1
diabetic patients and from 21 of the 39 healthy control subjects was
placed into sterile heparinized tubes as described. One
hundred-microliter aliquots of whole blood were diluted to a total
volume of 1 mL with RPMI 1640 medium containing 10% FCS
(Boehringer Ingelheim GmbH, Gagny, France) in 24-well
tissue culture plates. Blood cultures were incubated for 48 h (22, 23) at 37 C in 5% CO2 with no antigen, recombinant human
insulin (10 µg/mL), or phytohemagglutinin (PHA; 10 µg/mL;
Sigma Chemical Co.). Supernatants were harvested and
stored at -80 C until assay of IL-2, IFN, IL-4, and IL-10 contents
by enzyme-linked immunosorbent assay (Medgenix EASIA kit, BioSource,
Rungis, France).
PBMC assay
As described, PBMC were isolated and incubated in quadruplicate
wells with no antigen, recombinant human insulin (10 µg/mL), or PHA
(10 µg/mL). After 48 h, supernatants were harvested and stored
at -80 C until assay for IL-2, IFN, IL-4, and IL-10 contents by
enzyme-linked immunosorbent assay (Medgenix EASIA kit).
Autoantibody determinations
Blood samples were also taken for autoantibodies to islet cells (ICA), insulin (IAA), glutamate decarboxylase (GADA), and tyrosine phosphatase (IA2 Ab) in all diabetic and control subjects.
ICA were determined using an indirect immunofluorescence assay as previously described (25). The end-point dilution causing detectable fluorescence was converted into Juvenile Diabetes Foundation Units using a standard curve obtained with an international reference serum. The detection limit was 3 Juvenile Diabetes Foundation Units.
GADA were tested by immunoprecipitation assay using human GAD65 complementary DNA obtained from T. Dyrberg, Bagsvaerd, Denmark. In vitro transcription and translation of 35S-labeled GAD65 used the TNT T7/SP6-coupled reticulocyte lysate system from Promega Corp. (Madison, WI). GADA were expressed as arbitrary units compared with a reference antibody-positive serum that was arbitrary assigned an activity of 1.27 U. GADA activity above 0.11 U was considered positive, which corresponded to the mean + 3 SD value of 95 controls.
IA2 Ab were determined by immunoprecipitation assay using human IA2 complementary DNA donated from M. R. Christie (London, UK). In vitro transcription and translation of 35S-labeled IA2 were performed using the TNT T7/SP6-coupled reticulocyte lysate system from Promega Corp.. IA2 Ab results were expressed as arbitrary units in comparison with a reference antibody-positive sera that was arbitrarily assigned an activity of 100 U. The mean value + 3 SD of 95 normal sera was 1 U.
IAA were assayed by RIA using 125I-labeled human insulin (Sanofi Diagnostic Pasteur, Marnes la Coquette, France) according to a modification of the method described by Soeldner et al. (26). IAA activity of 0.46% was considered positive.
HLA determination
Genetic studies were performed at the Blood Transfusion Centers of Lyon and St. Etienne, France. Distribution of HLA DR and DQ haplotypes among the 34 of 44 (77%) new-onset type 1 diabetic patients was: DR3-DQB1*0201/DR4-DQB1*0302 (n = 5; 15%), at least 1 DR4-DQB1*0302 and no DR3-DQB1*0201 (n = 10; 29%), at least 1 DR3-DQB1*0201 and no DR4-DQB1*0302 (n = 14; 41%), and DRx/x (n = 1). Distribution of haplotypes among the 12 of 15 (80%) long duration type 1 diabetic patients was DR3-DQB1*0201/DR4-DQB1*0302 (n = 3; 25%), at least 1 DR4-DQB1*0302 and no DR3-DQB1*0201 (n = 3; 25%), at least 1 DR3-DQB1*0201 and no DR4-DQB1*0302 (n = 5; 41%), and DRx/x (n = 1). Among the 39 control subjects, 32 (82%) were HLA typed, 6 were DR3-DQB1*0201/DR4-DQB1*0302 (19%), 6 had at least 1 DR4-DQB1*0302 and were not DR3-DQB1*0201 (19%), 5 had at least 1 DR3-DQB1*0201 and were not DR4-DQB1*0302 (16%), and 15 were DRx/x. This control group was a purposely selected normal population with a greater proportion (DR3/DR4, 19%; DR3 and/or DR4, 17 of 32, 53%) than usual (DR3/DR4, <1%; DR3 and/or DR4, 45%) of DR3 and/or DR4 haplotypes encountered in type 1 diabetes (21). HLA class II alleles were determined serologically in all subjects by the standard microlymphocytotoxic technique. HLA-DR and HLA-DQ were typed using PCR amplification with sequence-specific primers (DynAl SSP sets, Compiegne, France). All phenotypically different HLA-DRB1*, -DQA1*, and -DQB1* alleles recognized by the HLA Nomenclature Committee in 1994 were uniquely identified (27).
Statistical analysis
Results were expressed as the mean ± SD. PBMC proliferations of the different groups were compared using Mann-Whitney U test. Correlations of autoantibody titers and PBMC proliferation to insulin were determined according to Fisher’s exact test. Correlations among clinical remission, autoantibodies positivity, PBMC proliferation, high risk HLA-DQ haplotypes, and Th1 and Th2 cytokines were analyzed using 2 x 2 contingency tables and Fisher’s exact test. P < 0.05 was considered significant.
|
Results |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
As shown in Fig. 1
, 14 of the 44
(32%) new-onset type 1 diabetic subjects (group A) had SI of 3.0 or
more compared to 0 of 39 (0%) healthy controls (group B;
P = 0.001). Recent-onset type 1 diabetic subjects had a
significantly higher SI than controls (2.9 ± 2.3; range,
0.7–10.2 vs. 1.3 ± 0.6; range, 0.7–1.9;
P = 0.0001). Five of 44 responded to 1 µg/mL, 8 of 44
responded to 10 µg/mL, and 7 responded to 20 µg/mL. Ten of 14 were
positive for only 1 insulin dose, 2 of 14 were positive for 2 doses (10
and 20 µg/mL), and 2 were positive for all 3 doses. For comparison, 7
of the 15 (47%) type 1 diabetic patients of long duration (group C)
had SI of 3.0 or more with, for the whole group, a mean SI of 4.5
± 5.0 (range, 1.2–21.1; group C vs. group B,
P = 0.01; group A vs. group C,
P = 0.12).
|
HLA and PBMC response to insulin
Diabetes-associated haplotypes, i.e.
DR3/4-DQB1*0201/0302, DR3-DQB1*0201/x, and DR4-DQB1*0302/x, were
equally distributed in the two responder and nonresponder subgroups of
new-onset type 1 diabetic subjects (Table 1). Only one subject had the protective
allele DQB1*0602; he had a positive PBMC response to insulin (SI, 3.4)
and was in clinical remission at 6 months, with insulin requirements of
0.2 IU/kg·day or less.
|
).
In control subjects with high risk HLA alleles, no PBMC proliferation
was observed in the presence of insulin (Table 1).
Autoantibodies and PBMC response to insulin
The prevalences of ICA (67% vs. 53%), GADA (56% vs. 40%), and IAA (39% vs. 28%) were similar in the two subgroups of responder and nonresponder newly diagnosed patients. No correlation was found between IAA titers and proliferative response to insulin (r = -0.04; P = NS). IA2 Ab-positive subjects (n = 34) had significantly higher PBMC SI indexes in the presence of insulin than IA2 Ab-negative subjects (n = 10; 3.2 ± 2.5 vs. 1.9 ± 1.3; P = 0.02). In addition, patients negative for ICA, IAA, and GADA elicited no proliferative response to insulin.
In type 1 diabetic subjects of long duration, there was no relationship between PBMC reactivity to insulin and the presence of any autoantibody. At the time of study, all of these subjects were IAA positive, only 1 of 15 was ICA positive (patient with the shortest diabetes duration), 6 of 15 were GADA positive, and 8 of 15 were IA2 Ab positive.
Subsequent remission in recent-onset type 1 diabetic subjects
Among the 32 recent-onset type 1 diabetic patients followed during
6 months, 18 (56%) were treated with daily insulin doses of 0.50
U/kg·day or less; 4 of them (13%) were treated with insulin doses of
0.20 U/kg·day or less. All patients (n = 8) with PBMC response
to insulin belonged to the group of patients with low insulin
requirements at the 6-month follow-up compared to 10 of 24 (42%)
subjects without PBMC response to insulin (P = 0.004).
Patients with low insulin requirement at 6 months continued to do
better over time. At 9 and 12 months, 12 of 28 (43%) were treated with
daily insulin doses of 0.50 U/kg·day or less; 6 of 8 (75%) patients
with PBMC response to insulin belonged to the group of patients with
low insulin requirement compared to 6 of 20 (30%) subjects without
PBMC response to insulin (P = 0.03). At 18 months, 7 of
21 (33%) were treated with daily insulin doses of 0.50 U/kg·day or
less; 5 of 8 (63%) patients with PBMC response to insulin belonged to
the group of patients with low insulin requirement compared to 2 of 13
(15%) subjects without PBMC response to insulin (P =
0.03). At 24 months, 6 of 20 (30%) were treated with daily insulin
doses of 0.50 U/kg·day or less; 4 of 7 (57%) patients with PBMC
response to insulin belonged to the group of patients with low insulin
requirement compared to 2 of 13 (15%) subjects without PBMC response
to insulin (P = 0.05). At disease onset, subjects with
subsequent low insulin requirement had a higher mean SI than the others
(3.3 ± 2.6 vs. 1.5 ± 0.6; P =
0.006; Fig. 2). It is noteworthy that the
4 patients with insulin requirements of 0.2 IU/kg·day or less showed
a proliferative response to insulin, with SI values of 3.4, 4.1, 5.7,
and 5.8, respectively.
|
In an attempt to characterize T cells that have proliferated
against insulin, cytokine production by PBMC in response to PHA was
determined. As described in Table 2A, IFN and IL-10
productions by PBMC were significantly lower in new-onset type 1
diabetic subjects than in healthy control subjects [IFN, 114
± 68 vs. 323 ± 285 U/mL (P = 0.03);
IL-10, 335 ± 157 vs. 717 ± 502 pg/mL
(P = 0.008)]. However, the subgroup of new-onset type
1 diabetic patients with PBMC proliferation to insulin had a
significantly lower IL-4 response to PHA than those without PBMC
proliferation to insulin (23 ± 15 vs. 64 ± 47
pg/mL; P = 0.04). There was no difference between the
two subgroups of patients in the production of IFN, IL-2, and IL-10.
Interestingly, production of the four cytokines tested was lower in the
non-DR3 and/or -DR4 control subjects than in the DR3 and/or DR4 control
subjects, as shown in Table 2B, but the difference was only significant
for IFN (P = 0.0008) and IL-4 (P =
0.03).
|
|
|
Discussion |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
We observed that specific T cell proliferation was decreased in the presence of insulin concentrations above 20 µg/mL, a result in keeping with the data reported by Harrison et al. (28). Indeed, in a significant number of recent-onset patients, maximal cell proliferation was obtained with insulin concentrations as low as 1 µg/mL. The absence of a correlation between insulin concentration in the culture medium and the level of cell proliferation is against a cellular growth factor effect of insulin and, instead, favors an immune-mediated effect. Moreover, it was our observation that most responders did exhibit a proliferative response to insulin at only one of the three antigen doses; only a few did so at two. Indeed, the antigen dose is the only parameter controlled, whereas the number of autoreactive lymphocytes is unknown as are the number and activity of antigen-presenting cells and the extent of immune response amplification. Although dose-response curves were expected, because of the complexity and the multistep nature of the system, our observations do not exclude such a relationship. It might be that the optimal insulin concentration varies greatly from one patient to another.
Recent-onset diabetic subjects without any autoantibody have shown no proliferation to insulin, underlying the fact that autoimmune responses, like normal immune responses, have coordinated humoral and cellular actors. Moreover, we observed a significant association between IA2 Ab and PBMC reactivity to insulin. IA2 Ab have been shown to be perhaps the most specific marker of ß-cell destruction (29). Although IA2 Ab are associated with the HLA DR4 allele, the response to insulin was not dictated by HLA DR3/DR4 haplotypes, because there was no proliferation to insulin in HLA-DR3/DR4 control subjects. In contrast with some studies (10, 11, 30), but similar to the results reported by Ellis et al. (31), we did not found any correlation between cellular and humoral reactivity. Noteworthy in most studies, culture medium was enriched with autologous serum. The presence of autoantibodies might interfere with the specific T cell proliferative response through idiotypic antiidiotypic interactions, although B and T cell epitopes in the insulin molecule are known to be different (32). In addition, high autoantibody concentrations might reduce the amount of ß-cell autoantigen available for T cells in the supernatants.
Our main finding is that a low insulin requirement was more prevalent in the diabetic patients with reactivity to insulin at disease onset. This observation was confirmed during follow-up. This unexpected result is not explained by differences in the mean glycemic level at the time of study, as metabolic control has been restored in all patients, nor is it related to age or ethnic or HLA characteristics. PBMC stimulability by the control antigen, TT, or by IL-2 was similar in the insulin-responsive and nonresponsive groups. Finally, it could be that cellular reactivity to insulin is related to the persistence of significant residual ß-cell mass, i.e. persistence of the autoantigen in the circulation.
Whole blood or PBMC production of IL-2, IFN, IL-4, and IL-10 was
detectable only after PHA administration, a nonspecific stimulation. We
confirmed that the detection of cytokine production was only feasible
after nonspecific T cell stimulation in human peripheral blood, as
recently described (24, 33, 34). Productions of IFN and IL-10 were
significantly decreased in recent-onset type 1 diabetic patients
compared to those in control subjects, an unexpected and unexplained
finding. Moreover, IL-4 production was lower in those patients with
insulin-reactive PBMC and a subsequently low insulin requirement.
Whether these findings could be related to any Th1/Th2 imbalance in
human type 1 diabetes is presently too speculative to be discussed
further.
In conclusion, we have observed a proliferative response to insulin in the PBMC of a significant proportion of recent-onset type 1 diabetic patients, but in none of the controls. In addition, there was a significant inverse correlation between cellular reactivity to insulin and subsequent low insulin requirement. These data should encourage the search for more sensitive approaches to study circulating T cell autoreactivity to specific antigens to better understand the pathogenesis of type 1 diabetes.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Received July 13, 1998.
Revised November 2, 1998.
Accepted March 23, 1999.
|
References |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
ldner JS, Tutleman M, Srikanta S, Ganda OP,
Eisenbarth GS. 1985 Type 1 diabetic and autoimmunity: islet cell
antibodies, insulin antibodies and ß-cell failure. N Engl J
Med. 313:893–894.[Medline]
24JQ T cells in type 1 diabetes. Nature. 391:177–181.[CrossRef][Medline]
This article has been cited by other articles:
 |
|
 |
|
  S. E. Fineberg, T. T. Kawabata, D. Finco-Kent, R. J. Fountaine, G. L. Finch, and A. S. Krasner Immunological Responses to Exogenous Insulin Endocr. Rev., October 1, 2007; 28(6): 625 - 652. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A Mayer, N Fabien, M C Gutowski, V Dubois, L Gebuhrer, J Bienvenu, J Orgiazzi, and A M Madec Contrasting cellular and humoral autoimmunity associated with latent autoimmune diabetes in adults Eur. J. Endocrinol., July 1, 2007; 157(1): 53 - 61. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Skarsvik, J. Puranen, J. Honkanen, M. Roivainen, J. Ilonen, H. Holmberg, J. Ludvigsson, and O. Vaarala Decreased in vitro type 1 immune response against coxsackie virus b4 in children with type 1 diabetes. Diabetes, April 1, 2006; 55(4): 996 - 1003. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. Lohmann, S. Laue, U. Nietzschmann, T.M. Kapellen, I. Lehmann, S. Schroeder, R. Paschke, and W. Kiess Reduced Expression of Th1-Associated Chemokine Receptors on Peripheral Blood Lymphocytes at Diagnosis of Type 1 Diabetes Diabetes, August 1, 2002; 51(8): 2474 - 2480. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H. Borg, A. Gottsater, P. Fernlund, and G. Sundkvist A 12-Year Prospective Study of the Relationship Between Islet Antibodies and {beta}-Cell Function At and After the Diagnosis in Patients With Adult-Onset Diabetes Diabetes, June 1, 2002; 51(6): 1754 - 1762. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
