| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
The Barbara Davis Center (R.G., A.P., T.S., L.Y., D.M., R.G.G., J.B., P.S., A.V., G.S.E., J.H., P.G., A.W.) and Department of Pathology (R.G.), University of Colorado at Denver and Health Science Center, Aurora, Colorado 80010; Department of Pathology (R.G.), The Children’s Hospital, Denver, Colorado 80218; and Donor Alliance (A.I., S.D.), Denver, Colorado 80246
Address all correspondence and requests for reprints to: Roberto Gianani, The Barbara Davis Center, 1775 North Ursula Street, Aurora, Colorado 80010. E-mail: roberto.gianani{at}uchsc.edu.
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Objective: In this manuscript, we describe a novel strategy in obtaining pancreata and pancreatic lymph nodes from islet autoantibody-positive organ donors that involves careful coordination among the laboratory and the organ donor provider organization.
Design: We developed a rapid screening protocol for islet autoantibodies measurement of organ donors to allow identification of positive subjects before organ harvesting. In this way we were able to obtain pancreata and pancreatic lymph nodes from subjects with and without islet autoimmunity.
Setting: The organ donors used in this study were obtained from the general community.
Subjects: The population studied consisted of 112 organ donors (age range 1 month to 86 yr, mean age 39 yr).
Main Outcome Measure: The main outcome measure of this study consisted of evaluating the pancreatic histology and identify T cells autoreactive for islet antigens in the pancreatic lymph nodes.
Results: To date we have identified three positive subjects and obtained the pancreas for histological evaluation from one of the autoantibody-positive donors who expressed ICA512 autoantibodies. Although this subject did not exhibit insulitis, lymphocytes derived from pancreatic lymph nodes reacted to the islet antigen phogrin.
Conclusion: In summary, these results indicate that it is possible to screen organ donors in real time for antiislet antibodies, characterize pancreatic histology, and obtain viable T cells for immunological studies.
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Whereas autoantibodies are a clear marker of autoimmunity, autoreactive T cells to islet autoantigens are most likely responsible for islet destruction in type 1A diabetes. However, the identification and quantification of T cell antigen-specific responses in the peripheral blood of diabetic or prediabetic subjects has been difficult. Previous studies have demonstrated a T cell response to the islet autoantigen IA-2ß (phogrin) in prediabetic subjects (3, 4). In particular, the authors showed T cell response to two distinct epitopes of IA-2ß designated peptide 2 and peptide 7, which are also targeted by diabetogenic CD4 T cell clones in the NOD mouse. IA-2ß is structurally related to the tyrosine phosphatase IA-2 molecule, and the two molecules are both targets of diabetes-associated autoantibodies with overlapping specificities (5).
Several reports in animal models of islet destruction and/or inflammation suggest that loss of insulin-producing cells may be accompanied by formation of new ß-cells (regeneration) through either replication of preexisting ß-cells (proliferation) or differentiation of nonendocrine precursors (neogenesis) (6, 7, 8). A recent report by Dor et al. (9) with fate marking indicated proliferation but not neogenesis of ß-cells. Because autoantibodies to islet antigens can be present several years before the development of diabetes, we assumed that during this phase, one might identify pancreatic histological changes. Cadaveric organ donors represent a potential source of pancreatic specimens from islet autoantibody-positive subjects. In a preliminary retrospective study of stored sera of 777 organ donors, we identified autoantibodies reacting with either insulin or GAD65 in 23 of 777 donors (10, 11), and two additional donors had multiple islet autoantibodies (to GAD65 and IA-2). Islet antibody screening was performed with a rapid assay that we hypothesized would allow prospective rapid research analysis.
In the current study, we screened organ donors who became available for pancreas donation with biochemical islet autoantibody radioassays. To date, we have prospectively screened 112 organ donors for antibodies to GAD65, IA-2, and insulin and identified three subjects to be autoantibody positive. We obtained the pancreas of one IA-2 autoantibody-positive individual and 14 antibody-negative subjects.
For a subset of subjects (one antibody positive and three antibody negative subjects), we analyzed pancreatic lymph node-derived lymphocytes by Enzyme-Linked Immunospot (ELISPOT) assay to determine the presence of IA-2ß reactive T cells and assess whether viable T lymphocytes could be recovered from pancreatic lymph nodes of cadaveric organ donors.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
To date, we have obtained serum specimens from 112 nondiabetic cadaveric organ donors for islet antibody screening.
The cadaveric donors were obtained through a collaborative arrangement with our local donor provider organization (Donor Alliance). Although all the sera were screened for autoantibodies, procurement of the organ was dependent on final clinical donation of organs as well as availability from the islet isolation laboratory. Consent for research use was a prerequisite for both serum and organ analysis.
Screening of organ donors with biochemical autoantibody assays
After institutional review board protocol review, serum specimens for autoantibody measurements were obtained after organ donor certification and evaluation. Autoantibody assays for GAD65, IA-2, and insulin were performed on receipt of the sera. The GAD65, IA-2, and insulin autoantibody assays are based on immunoprecipitation of radioactively labeled molecules, and they have been previously described (12). For these three assays the levels were expressed as an index with positivity set above the usual cutoff [>0.2 for GAD65 autoantibodies, > 0.1 for IA-2 autoantibodies, and >0.02 for insulin antibodies of the 99th percentile to enhance specificity (99th percentile index of 0.032 for GAD65 autoantibodies, 0.049 for IA-2 autoantibodies, and 0.01 for insulin antibodies)]. One hundred twelve organ donors were screened for autoantibodies to IA-2, GAD65, and insulin.
Preparation of tissue for histological studies
Blocks of pancreatic tissue from one autoantibody-positive and 14 control antibody-negative organ donor subjects were obtained immediately after surgical removal. Tissue blocks (n = 22) were prepared from the pancreas of the IA-2 autoantibody-positive organ donor and paraffin embedded after fixation in 10% buffered formalin. Seven-micrometer tissue sections were obtained for histological and immunohistochemical studies. For each tissue block at least one hematoxylin and eosin-stained section was obtained. Control pancreata were obtained from 14 antibody-negative (six male and eight females) organ donors ranging in age between 18 and 52 yr (mean age 39.6 yr). Three donors were matched with the antibody-positive donors for whom the pancreas was available for both sex and age (mean age 54.28 yr). For 102 of 112 subjects, the human leukocyte antigen (HLA) DR type (provided by Donor Alliance) was available.
Table 1
summarizes the age, sex, body mass index, and HLA type for the subjects whose pancreata were available for histological studies.
|
Immunohistochemical staining for insulin and glucagon
In each case, sections were stained with antibodies directed to insulin and glucagon. The pancreatic sections were simultaneously stained with an antiglucagon mouse monoclonal antibody (Jackson Immunoresearch, West Grove, Pa) and an anti-insulin guinea pig polyclonal antibody (Jackson Immunoresearch) followed by detection with Texas Red-conjugated donkey antimouse IgGs (Jackson Immunoresearch) and aminomethylcoumarin acetate-conjugated donkey antiguinea pig IgGs (Jackson Immunoresearch).
Immunostaining for leukocyte common antigen (LCA), and insulin
Each donor section was double stained with a mouse monoclonal antibody to the LCA (CD45) and insulin (as described above). The CD45 staining was detected with an antimouse IgG alkaline phosphatase conjugated and developed with Fast Red (Sigma, St. Louis, MO). The sections were analyzed by immunofluorescence microscopy, allowing simultaneous visualization of both the insulin-stained tissue and the inflammatory cells (LCA positive cells).
Morphometric analysis
For each case we also determined the ratio between insulin and total area and glucagon and total area (expressed as percentage) as well the ratio between insulin and glucagon-positive areas in five randomly selected microscopic fields. In each section, the stained area for insulin and glucagon was quantified by computer-assisted image analysis (Image Pro Plus, Silver Spring, MD). The number of CD45-positive cells in the pancreas was determined for each donor by calculating the average cell density (cells per square millimeter) in 10 randomly selected fields. To evaluate for insulitis, we determined for each of the subjects the average number of leukocytes that was present within the 10 islets within the largest mononuclear infiltrate (if less than 10 islets contained at least one leukocyte, the remaining islets were considered to contain 0 leukocytes). We compared both the density of pancreatic leukocytes and the number of leukocytes within the islets between the antibody-positive subject and each of the negative controls using the Whitney Mann rank test.
The difference between the values obtained for the IA-2 antibody-positive and each of the control donors for insulin to glucagon ratio, insulin to total pancreatic area ratio, and glucagon plus insulin to total pancreatic area was also analyzed for statistical significance using the Whitney Mann rank test.
Immunohistochemical staining for cytokeratin 19 and chromogranin
Formalin-fixed paraffin-embedded sections were stained simultaneously with a polyclonal rabbit antibody to chromogranin A and B (Abcam, Cambridge, UK) followed by detection with Texas Red-conjugated antimouse donkey (Jackson Immunoresearch) and a mouse monoclonal to cytokeratin 19 (Abcam) followed by detection with the TSA plus system (PerkinElmer, Wellesley, MA).
Isolation of lymphocytes from pancreatic lymph nodes
Pancreatic lymph nodes could be obtained from one antibody-positive (donor 7) and three antibody-negative control subjects.
Lymph nodes were removed from pancreatic fat and processed using a tissue homogenizer. After centrifugation of the material, supernatants were collected and strained to remove any residual fatty deposits. Cells were harvested, counted, and frozen in a solution containing 30% RPMI 1640 (CellGro, Herndon, VA), 60% fetal calf serum (Hyclone, Logan, UT), and 10% dimethylsulfoxide. Long-term storage in liquid nitrogen preserved cells for use in the ELISPOT experiments.
ELISPOT assay
Ninety-six-well nitrocellulose-backed plates were coated with antihuman interferon (IFN)-
mAb (Endogen, Inc., Cambridge, MA) or antihuman IL-5 (Endogen) overnight at 4 C. The plates were washed, and blocking buffer (PBS/BSA) was added to the wells for 1 h. Plates were then washed again, and 300,000-thawed pancreatic lymph node cells/well were plated in the presence of antigen. Candidate antigens included were B9–23 peptide of insulin, recombinant human insulin, GAD peptides 271–285 and 556–581, peptide 7 of the human IA2ß molecule (the structure of this peptide, highly conserved between human and mouse, is KNRSLAVLTYDHSRI), and the mitogen phytohemoagglutinin as a positive control.
The lymphocytes from each subject were run in triplicate. After 16–60 h incubation at 37 C, 5% CO2 cells were washed and stained with antihuman IFN (Endogen) or antihuman IL-5 (BD PharMingen, San Diego CA), a secondary biotinylated antibody (Pierce, Rockford, IL), and peroxidase-conjugated avidin (Dako, Carpinteria, CA). The reaction was then developed with N, N-dimethylformamide (Pierce) and 3-amino-9-ethyl carbazole (Pierce). Spots derived from cytokine-producing cells were quantified using the Series-1 immunospot and satellite analyzers (CTL Inc., Cleveland, OH). For each well the results were expressed as the difference between the spots counted and the background for that particular subject (i.e. the number of spots counted in the absence of the stimulated peptide). The values obtained from the antibody-positive subject and the three negative controls were compared using the Whitney Mann rank test.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
None of these donors had a clinical diagnosis of pancreatic disease (i.e. chronic or acute pancreatitis or pancreatic cancer). At the time of the hospitalization, the mean glucose level in this group was 191 mg/ml (with a SD of 80) and the median level was 169 mg/ml. As shown in Fig. 1, one donor had autoantibodies reacting with IA-2 (donor 7) (autoantibodies reacted with multiple IA-2 constructs including IA-2 iC, ICA512 BDC, and full length IA-2); one had autoantibodies reacting to GAD65 (donor 66); and a third donor had autoantibodies to both GAD65 and IA-2 (donor 37). The pancreas of donor 7 was obtained, whereas the pancreata of donors 37 and 66 could not be obtained, in one case because of metabolic decompensation preceding donation of the organs and in the other because of surgical unavailability. Among the 102 organ donor antibody positive for whom HLA DR typing was available, only five subjects were positive for the type 1A diabetes high-risk genotype DR3/DR4, whereas 43% were DR4 or DR 3 positive. The DR HLA genotype of the IA-2 antibody-positive subject (donor 7) was DR2/DR3 (having one high risk allele and the protective DR2 allele), whereas the DRs of the other two antibody-positive subjects, donors 37 (positive for GAD65 and IA-2 antibodies, age 63 yr) and 66 (positive for GAD65 antibodies, age 23 yr), were DR8/DR14 and DR4/DR3, respectively. Thus, one of the donors positive for islet antibodies was one of five with the high-risk DR4/3 genotype.
|
The pancreas of donor 7 contained numerous small and medium-size islets that were often associated with ducts (islet ductal complexes) (Fig. 2A) as well as clusters of glucagon or insulin-positive cells immediately adjacent to the epithelium lining the ducts (Fig. 2B). Furthermore, in several ducts the epithelial cells lining the ducts exhibited strong positivity for both cytokeratin 19 (a marker of ductal cells) and chromogranin (a marker of endocrine cells) [Fig. 3, A (cytokeratin 19) and B (chromogranin)]. Cytokeratin 19-chromogranin double-positive cells were not seen in any of the control adult pancreata (n = 14) from antibody-negative subjects but were seen in fetal pancreas [Fig. 3, C (cytokeratin 19), and D (chromogranin)].
|
|
Table 2 summarizes the weight of each pancreas, the ratio between insulin- and glucagon-stained tissue, and the ratio between endocrine tissue (expressed as the sum of insulin and glucagon stained areas) and total pancreatic area in each case.
|
After examining multiple sections of donor 7 and all of the control pancreata, we could not detect more than five CD45-positive cells (up to four leukocytes found within the islets of antibody negative subjects) within any of the islets examined. Furthermore, there was no statistical significance difference between the antibody-positive and control subjects in the number of CD45-positive cells within the 10 islets with the largest mononuclear infiltrate. Table 3 shows the average number of leukocytes in the pancreatic tissue and the islets for each case examined. Figure 4 shows double immunostaining of this pancreas and a control pancreas for LCA and insulin.
|
|
and IL-5 productions in ELISPOT assay.
The cells from the antibody-positive donor responded with IFN production only to the IA-2 ß epitope tested (IFN spots = 19 ± 3. 2/300,000 cells) but not to any of the other antigens. None of the three controls responded to this peptide (all < 5 ± 1 spots/300,000 cells) (P < 0.05) or any of the other antigens. We did not detect an IL-5 response to the islet antigens tested in donor 7 or in the antibody-negative controls. Figure 5 shows the ELISPOT measuring IFN secretion from the IA-2 autoantibody-positive organ donor lymphocytes in response to stimulation with peptide 7 of IA-2ß.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Pancreas was obtained from one of three autoantibody-positive subjects. The pancreatic histology in this IA-2 autoantibody-positive case for which the pancreas could be obtained revealed alteration in the islet architecture.
The pancreas of this subject showed features that are found in normal fetal but not adult pancreas. These included numerous islet-like clusters associated with ducts and the positivity of some ductal cells for both chromogranin and cytokeratin 19.
Morphometric analysis revealed that in this pancreas there was no loss of ß-cells. This may signify either that ß-cell loss did not occur in this subject or that if it did occur, it was compensated by regeneration or new ß-cell formation. Interestingly, morphometry also revealed that individuals whose pancreata contained areas of fibrosis had an increased ratio between area of endocrine tissue and total pancreatic area. Because the pancreata with fibrosis had normal weight (see Table 2
), it is unlikely that this was due to a loss of exocrine tissue (as seen in patients with chronic pancreatitis).
To our knowledge, this is the first report that describes the histology of the pancreas in an islet autoantibody-positive nondiabetic cadaveric donor. Given the pancreatic histology in this subject’s pancreas, there was unusual endocrine tissue potentially resulting from replication of duct-associated endocrine cells or neogenesis.
The results of the ELISPOT experiments indicate that the pancreatic lymph node of the IA-2 autoantibody subject contains IA-2ß-reactive T cells producing IFN. In this subject we did not identify T cell reactivity (by ELISPOT) for insulin and GAD65. A recent report (13) shows that insulin autoreactive T cells could be recovered from pancreatic lymph nodes of diabetic subjects. It is possible that the lack of T cell reactivity for insulin and GAD in this subject is related to his lack of humoral immunity for these autoantigens. In addition, we did not detect insulitis, raising the possibility that the presence of the single islet autoantibody was not specifically related to pancreatic disease. This subject had a DR genotype DR2/DR3 that has been associated with presence of autoantibodies but strong protection from progression to type 1 diabetes (14). Bottazzo and co-workers (15) similarly did not find insulitis in two pancreata from autopsy specimens of nondiabetic GAD65 autoantibody-positive subjects. We have evaluated 22 pancreatic blocks and more than 44 sections from this pancreas but still could have potentially missed focal insulitis.
It is also likely that individuals expressing more than one autoantibody (or diabetic subjects as in the study cited above) will have a higher probability of insulitis with T cell reactivity to multiple islet antigens.
Given the expected frequency of positive donors, it will be important to expand similar screening to multiple sites and screen a larger number of autoantibody-positive and negative subjects with the major aim of characterizing both normal pancreas and the pancreatic response to injury and obtain islet infiltrating T lymphocytes.
|
Footnotes |
|---|
The authors have no conflict of interest.
First Published Online February 14, 2006
Abbreviations: ELISPOT, Enzyme-linked immunospot; GAD, glutamic acid decarboxylase; HLA, human leukocyte antigen; IA-2, islet autoantigen 2; ICA512, islet cell antigen 512; IFN, interferon; LCA, leukocyte common antigen.
Received May 25, 2005.
Accepted February 8, 2006.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
-Cell neogenesis in an animal model of IDDM. Diabetes 46:599–606[Abstract]This article has been cited by other articles:
 |
|
 |
|
  S. Diamantopoulos, G. Allende, J. M. Ferreira, G. Ciancio, G. W. Burke, and A. Pugliese Retrospective Assessment of Islet Cell Autoantibodies in Pancreas Organ Donors Diabetes Care, September 1, 2008; 31(9): 1741 - 1742. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Kobayashi, J. Jasinski, E. Liu, M. Li, D. Miao, L. Zhang, L. Yu, M. Nakayama, and G. S. Eisenbarth Conserved T cell receptor {alpha}-chain induces insulin autoantibodies PNAS, July 22, 2008; 105(29): 10090 - 10094. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. K. Steck and G. S. Eisenbarth Genetic Similarities Between Latent Autoimmune Diabetes and Type 1 and Type 2 Diabetes Diabetes, May 1, 2008; 57(5): 1160 - 1162. [Full Text] [PDF]
|
|
|
|
|
|
P. In't Veld, D. Lievens, J. De Grijse, Z. Ling, B. Van der Auwera, M. Pipeleers-Marichal, F. Gorus, and D. Pipeleers Screening for Insulitis in Adult Autoantibody-Positive Organ Donors Diabetes, September 1, 2007; 56(9): 2400 - 2404. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. S. Eisenbarth Update in Type 1 Diabetes J. Clin. Endocrinol. Metab., July 1, 2007; 92(7): 2403 - 2407. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Pietropaolo, E. Barinas-Mitchell, and L. H. Kuller The Heterogeneity of Diabetes: Unraveling a Dispute: Is Systemic Inflammation Related to Islet Autoimmunity? Diabetes, May 1, 2007; 56(5): 1189 - 1197. [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 |
