Auloiiiiniiaiiir. Vol. 28. pp 91 107 Reprints awilahle directly from thc puhliahcr Photocopying permitted by licen.;e only

I IYYX OPA (Overseas Puhliahers A.;sociation) N.V. Publishcd by Iiccnsc under

the Harwood Academic Puhli\hcrs imprint. part OS The Gordon and Breach Publishing Group.

Printed in Malaysia.

Syngeneic Islet Transplantation in Prediabetic BB-DP Rats - A Synchronized Model for Studying ,@Cell

Destruction during the Development of IDDM ULLA BJERRE CHRISTENSEN", THOMAS SPARRE", ANNE COOKE ',

HENRIK ULLITS ANDERSEN ", THOMAS MANDRUP-POULSEN and J 0 R N NERUP",*

During development of IDDM mononuclear cell infiltration is seen in the islets of Langerhans in both man and rodent models. This process is not synchronized in time and space. To create a synchronized model for investigation of the cellular and molecular events during IDDM development, we isolated and transplanted 200 neonatal BB-DP rat islets under the kidney capsule of 30 day old BB-DP rats. Islet transplantations were also carried out from Wistar Furth (WF) to WF rats, from WF to Wistar Kyoto (WK) rats and from WK to BB-DP rats to compare disease occurrence in an islet syngraft with changes in islet syngrafts or allografts in non-diabetes prone recipients and with changes in islet allografts in diabetes prone recipients, respectively. Pancreata and grafts were harvested a t pre-scheduled time points before onset of diabetes and at onset of diabetes, and stained for insulin, M H C class I, M H C class 11, aLj-TCR, CD4, CD8 or EDl. Diabetes incidence in the syngrafted BB-DP rats was 75% at 78 rt 5 days of age. The incidence and time of onset of IDDM was unaffected by islet syngrafting. Positive correlations were found between the percentage of infiltrated islets in situ and the number of infiltrating cells in the islet syngraft from the same BB-DP rats (p = 0.003-p < 0.0001, r = 0.5 - 0.7). The number of infiltrating cells regard- less of cell type in the graft was inversely correlated to the graft insulin content ( p = 0.0003- p < 0.0000, Y = -0.6 to - 0.8). The graft insulin content was 70% and 90% in BB-DP rats before onset of diabetes and BB-DP rats not developing diabetes respectively, and 30% in the diabetic rats (p < 0.01). Interestingly only 5% of the allografted BB-DP rats developed diabetes. No correlation was found between the number of infiltrating cells in the graft and islets in situ in the BB-DP rats not developing diabetes. Only baseline infiltration was seen in grafts from syngrafted W F rats. In allografted WF islet to WK rats graft rejection was seen 12 days after transplantation. No correlation was found between the number of infil- trating cells in the graft and islets in situ. In conclusion the cellular infiltration in syngeneic but not allogeneic islets grafted to 30 day old BB-rats mirrors that seen in islets in situ. Syngeneic islet grafting in BB-DP rats may be useful for studying the cellular and molecular events during the development of IDDM.

Kc,~wwt/.s: IL-I, Allografts, Prevention of diabetes

* Corresponding author. Tel.: + 45 44 43 93 89. Fax: + 45 44 43 92 32

91

Aut

oim

mun

ity D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y M

onas

h U

nive

rsity

on

10/3

0/14

For

pers

onal

use

onl

y.

U.B. CHRISTENSEN of trl.

~hhrcv iC l l i~JnS: Biobreeding diabetes prone (BB-DP), Wistar Furth (WF), Wistar Kyoto WK- (WK), Insulin dependent diabetes mellitus (IDDM).

INTRODUCTION

The development of insulin dependent diabetes mel- litus (IDDM) in man[' and rodent modelsL4 is characterized by mononuclear cell infiltration in the islet of Langerhans and a selective destruction of the pancreatic @-cells (insulitis). Insulitis is a process desynchronized in time and space. Islets from recent-onset diabetic patients can be divided into three categories: normal insulin-containing islets unaffected by any infiltrative process, insulin containing islets with insulitis and insulin deficient islets devoid of infiltration.[21 Insulin containing islets are encountered in pancreatic sections from 33% of recent-onset diabetics, and 23% of the in- sulin containing islets also show signs of insulitis."]

BB-DP rats normally develop diabetes at 60- 120 days of age. At 70-90 days of age pancreatic islets from BB-DP rats exhibit all stages of insulitis as well as normal islets at the same time.[41 The NOD mouse develops diabetes at 12-30 weeks of age. Mononuclear cell infiltration is observed from at least the 4th week of age,["."'] and the different stages of the insulitis process can be found within the same pancreas at the same time.["]

The fact that the mononuclear cellular infiltra- tion is desynchronized in time and space hampers the study of the cellular and molecular changes in the ij-cells during IDDM development in vivo. To investigate if islet syngrafts in prediabetic BB-DP rats mirror the changes in the islets in .situ at any given time-point and hence represent a model of the mononuclear cell infiltration, we studied the insulin content and cellular infiltrates in islet syngrafts and the cellular infiltrates in islets in .situ before and at onset of IDDM in BB-DP rats. These parameters were also evaluated in islet syngrafts in non- diabetes prone rats and in islet allografts in both diabetes-prone and non diabetes-prone rats for comparison.

We report a strong correlation between cellular infiltration in islet syngrafts, but not allografts, and the islets in situ in BB-DP rats. The pattern and time-course of infiltration in the syngrafts differed markedly from islet allograft rejection. Surpris- ingly, islet allografting protected development in BB-DP rats.

against IDDM

MATERIALS AND METHODS

Reagents

RPMI 1640 (containing 1 I mmol/l glucose and 1.15 mniol/l-arginine) and Hanks' balanced salt solution (HBSS) were purchased from Gibco (Paisley, Scotland). RPMI was supplemented with 20 mmol/l HEPES buffer, 100,000 lU/l penicillin and 100 mg/l streptomycin. Purified mouse anti-rat monoclonal antibodies, CD8 (clone OX-8), CD4 (clone OX-35), aP-TCR (clone R73), R T l B (clone OX-6), and RTlA (clone OX-18) were from Pharmigen (San Diego, CA). Mouse anti-rat mono- cyte/niacrophage antibody (clone ED- 1 ) was from Serotec (Oxford, UK). Biotinylated goat anti- mouse immunoglobulin and FITC-conjugated rabbit anti-guinea pig immunoglobulin were pur- chased from DAKO A/S (Glostrup, Denmark). Monoclonal guinea-pig anti-rat insulin antibody (GP-3 I ) was kindly supplied by Novo Nordisk ( Bagsvmd, Denmark). Vectarstain ABC kit was purchased from Vector Lab. (Peterborough, UK). Diaminobenzidine, aluminum potassium sulphate 12-hydrate, potassium iodate analar, and hema- toxylin were froin Merck, Darmstad, Germany, thyniol and glycerol from BDH (Poole, UK) and streptozotocin from Sigma (St. Louis, MO, USA). Ketamin was purchased from Park-Davis (Barcelona, Spain), xylazin from Bayer (Leverkusen. Germany), Temgesic:icl from Reckitt and Coleinann

Aut

oim

mun

ity D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y M

onas

h U

nive

rsity

on

10/3

0/14

For

pers

onal

use

onl

y.

A N I N V f V O MODEL FOR THE STUDY OF IDDM DEVELOPMENT 93

(Hull, UK) and isopentan from Fluka Chemie (Buchs, Germany).

Animals

Twenty-eight to thirty-two day old BB/Wor/ Mol-BB2 (BB-DP), Wistar Furth (WF), and Wistar Kyoto (WK) rats were purchased from Mdlegirden, LI. Skensved, Denmark, where the BB-DP rats were housed separately in a specific pathogen-free environment. All rats were housed under controlled conditions of light (light on from 6:OO am to 6:OO pm), humidity (60-70%) and temperature (20-22°C) from five days prior to trans- plantation until sacrifice. The rats were offered a standard rat chow (Altromin, Chr. Petersen A/S, Ringsted, Denmark) and free access to tap water. Four to five day-old BB-DP, W F and WK rats were picked up at Mullegirden in the morning on the day of islet isolation, and transported in animal transport boxes.

Islet Isolation and Culture

Pancreatic tissue was removed and placed in tubes containing icecold HBSS. Islets were isolated by hand-picking after collagenase digestion of the pancreata."*] Following isolation, islets were pre- cultured for 4 days in RPMI 1640 with 10% fetal calf serum at 37°C. Islets were then washed twice and precultured for 24h in RPMI 1640 with 0.5% human serum (HS) at 37°C before transplantation.

Transplantation Procedure and Graft Retrieval

The neonatal islets were transplanted under the kidney capsule to 28-32 day-old recipient rats under sterile conditions as previously described." 'I Since diabetes incidence is similar in male and female BB-DP rats"41 both sexes were equally used. To reduce a possible risk of graft rejection due to any gender-related incompatibility, male islets were transplanted to male recipients and female islets to female recipients. Two-hundred (range: 187 -21 5 )

islets were hand picked using a 10ml Transpfer- pettor (Brand, Germany) pipette. The islets were allowed to sediment, and as much supernatant as possible was removed prior to the transplantation of a volume of 0.1-0.2 ml islets. The recipient rats were weighed, and blood glucose (BG) was measu- red. The rats were then anesthesized with ketamin 8.75 mg/100 g and xylazin 0.7 mg/100 g ip. A small incision was made over the left kidney, the kidney was lifted forward, and a small incision using a knife and a pocket using a blunt instrument was gently made in the capsule. The islets were placed under the capsule towards the lower kidney pole. The muscle and skin were sutured, the rats allowed to recover in a cage in the operating facilities and given Temgesic 0.01 mg sc in the neck before returning to the animal facility. The rats were given Temgesic twice daily for 3 days post transplanta- tion. Blood glucose and weight were measured every third day.

The rats were killed by cervical dislocation and immediately afterwards the splenic part of the pan- creas and the grafted kidney were removed. The grafts were carefully dissected from the kidney and capsule under a microscope and snap frozen in isopentan together with the pancreas and stored at -80°C until further use.

Study Design

To ensure that destruction of the islets in situ in the transplanted rats would not be masked by the insulin production in the graft, six W F rats (weight 122- 136 g) were made diabetic by iv injection of streptozotocin (STZ) (65 mg/kg). Rats were consid- ered diabetic upon having BG > 14mmol/l for two consecutive days. Eight days after injection the diabetic rats (BG 25.3 i 3.5 mmol/l) were trans- planted with 200 syngeneic W F islets. Rats were sacrificed six days after transplantation at which timepoint BG (23.5 i 2.4 mmol/l) was not signifi- cantly reduced, indicating that insulin production from 200 grafted islets would not be sufficient to substitute insulin production from destroyed islets it1 s i tu .

Aut

oim

mun

ity D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y M

onas

h U

nive

rsity

on

10/3

0/14

For

pers

onal

use

onl

y.

94 U . B . CHRISTENSEN c t ( I /

To investigate if islet syngrafts in BB-DP rats mirror the insulitis process seen in the islets in situ during IDDM development, 30 day-old (range 28-32) BB-DP rats were grafted with 200 (range 187--215) BB-DP rat islets under the kidney capsule. Groups of six rats were killed on days 7, 12,23,37,48 and 173 (day 173 n = 4) or at onset of diabetes. Onset of diabetes was defined as BG > 14 mmol/l for two consecutive days.

(A) To compare the disease occurence in islet syngrafts in BB-DP rats with an islet allograft response in BB-DP rats, WK islets (RTI ' ) were transplanted to BB-DP rat (RT1"). Groups of three rats were killed on days 7, 12, 23, 37, 48, 90 and 173.

(B) The syngraft response in BB-DP rats was compared with alloresponse in non- diabetes prone rats by grafting W F islets (RTI") to WK rats. Groups of six rats were killed on days 7, 10, 12 and 17 after trans- plantation.

(C) To investigate if BB-DP rat islets are re.jected to the same degree and with the same pattern of infiltrating cells as W F islets transplanted

to WK rats, BB-DP rat islets were trans- planted to WK rats. Groups of three rats were killed at days 7 and 12 after transplan- tation.

Control Experiments

(A) To control for the effect of transplantation procedure per se W F islets were transplanted to WF rats. Groups ofthree were killed on days 7, 10, 12, 17, 37 and 48 after transplantation. As expected the number ofinfiltrating cells were low in W F syngrafts and remained constant through out the observation period (Fig. I ) . Insulin staining was 85 f- 10% of total graft area at all timepoints.

(B) To characterize the number and types of passenger inflammatory cells in neonatal islets at the time of transplantation, 100 BB-DP and W F islets from 6 different isolates were placed in Eppendorph tubes, the supernatant was removed, the islets snap frozen and embedded in Tissue-tec for cryo-section. For each islet pellet two section were stained, the area of the

0

7 10 12 17 37 days aner transplantalion

MHC Class II H MHC Class I 0 CD4 0 ED1 H 4-TCR acd8

FIGURE 1 W F islets (17 = 3 ) .

Number of infiltrating ce l l sh i i ' i n the grafts from WF rats killed on days 7, 10. 12, 37. 48. after syngratling with

Aut

oim

mun

ity D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y M

onas

h U

nive

rsity

on

10/3

0/14

For

pers

onal

use

onl

y.

A N IN VIVO MODEL FOR THE STUDY OF IDDM DEVELOPMENT 95

islet mass was measured and number of infil- trating cells counted. In both rat strains all islets stained positive for insulin. Islets stained posi- tive for MHC class I1 and ED-1, and only in low amounts. No positive staining was seen of MHC class I, a/?-TCR, CD4 or CD8. In islets from four of six BB-DP rats, 4 f 2 MHC class 11 positive cells/mm' and in three of six BB-DP rats 3 f 2 mm2 ED1 positive cell were found. In islets from three of six WF rats 5 * 2 MHC class I1 positive cells/mm' and in three of six WF rats 3 f 2 ED- 1 positive cells/mm2 were found.

(C) To detect the number of infiltrating cells in the islets iri sifu before transplantation, 30 day-old BB-DP and WF rat pancreata were removed, prepared as for grafts for histology and exam- ined by light microscopy (n = 6 in each group). From each pancreata two sections, at least 150 mm apart, comprising at least 10 islets were examined. All islets in situ in the examined sec- tions stained positively for insulin. An average of 2 f 1 MHC class TI positive cells/islet were found in 60-70% of the islets. Hence, only islets in situ of grafted animals containing more than two MHC class I1 positive cells were con- sidered infiltrated. No CD8, MHC I or a$- TCR positive cells were seen. Two BB-DP and one WF islet contained one CD4 positive cell each, and two BB-DP and WF islets contained one ED1 cell each. No significant differences between the examined cell types were observed in the islets in the two strains.

lmmunoperoxidase and lmmunofluorescence Staining

The frozen pancreatic tissue and grafts were cut in 5 p m cryostat sections, allowed to dry in open air for 16 - 20 h, fixed in 100°/~ acetone for 10 min, left to dry for loinin, wrapped in aluminum foil and stored at -80°C until use.

Serial tissue sections from graft or pancreas were double stained with insulin ab and one of the following mab's: Anti-MHC class I IgGI, (OX- 1 8), reacting with MHC class I antigen on a majority of

nucleated cells;"51 anti-MHC class I1 IgG1, (OX-6), reacting with non-polymorphic rat MHC class IT antigens (I-a);[I6] anti cxP-TCR IgGl, (R73), reacting with the constant determinant of the tip- TCR receptor found on most peripheral T lympho- cytes, intestinal intraepithelial lymphocytes and thymocyte~;"'~ anti CD4 IgG2, (OX-35), reacting with CD4 antigen on T helper cells, thymocytes, and macrophages;"81 anti CD8 IgGI, (OX-8), reacting with the chain on the CD8 antigen on T cytotoxic/suppressor cells and NK cells;""211 and anti ED-I IgGl (EDl), recognizing a single chain glycoprotein expressed predominantly on lyso- soma1 membrane of myeloid cells and by the majority of tissue macrophages (Serotec).

Fixed cryostat tissue was incubated with 0.4 M glycine for 30 min and subsequently with 3% H202 for 30 min, and finally with 10% normal goat serum for 30min. The sections were incubated for one hour with the relevant mouse-anti rat monoclonal antibody (mab) followed by goat anti-mouse IgG biotin conjugate and with avidin-biotin horseradish peroxidase (Vectarstain ABC kit, Peterborough, UK). The stain was developed with 3.3'-diamino- benzidine tetrahydrochloride at 0.6 mg/ml in PBS containing 0.01 YO H202. The tissue sections were then incubated for one hour with guinea pig- anti-rat insulin, followed by FITC labelled rabbit anti-guinea pig IgG and lightly counterstained with hematoxylin (potassium alum 62.5 g/l, hema- toxylin 1.25 g/l, glycerol 250 mg/l, potassium iodate 0.25g/l, one drop of thymol in distilled water). All reagents were added sequentially with three washes in 1 % PBS in between and titrated before use.

Microscopy

Five pm stained coded sections from grafts and pancreata were evaluated by two independent observers using an Olympus 5H light microscope. The area of the graft was measured using a computer based morphometric analysis and the number of positive cells per graft was counted visually and expressed as number of cells per mm2.

Aut

oim

mun

ity D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y M

onas

h U

nive

rsity

on

10/3

0/14

For

pers

onal

use

onl

y.

96 U.B. CHRISTENSEN et t i / .

25

20 - 1 E l 5 E

10

5 '

An interobserver variance of more than 20% of the cell count resulted in re-evaluation by both observers and a third observer. All observers were blinded to protocol. Insulin staining was evaluated

RESULTS

Effects of Islet Transplantation on Time of Onset and Incidence of Diabetes

.

'

'

as O%, 0-33%, 33-66'30 or 66-100% staining of the graft area. The pancreatic islets in situ was scored as number of insulin positive islets per section. The number of islets in situ with cellular infiltration was counted and expressed as percen- tage of the total number of islets. A graft was considered infiltrated when the number of infil- trating cells exceeded the mean+3SD of the number of cells seen in syngrafted WF islets.

The diabetes incidence in the syngrafted BB-DP rats and non-transplanted control rats was 75% and 86%, respectively (NS). The mean time of onset was 7 8 f 5 days and 6 9 f 2 5 days of age, respectively (NS). Diabetes developed on day 48 f 5 after transplantation. Interestingly the diabetes incidence was significantly lower (6.6%, p < 0.005) in the allografted BB-DP rats than in the syngrafted BB-DP rats.

Statistical Analysis Syngeneic Met Transplantation to BB-DP Rats

Results are expressed as means f SD unless stated Except for a transient decrease in BG at day 18 otherwise. Spearman rank correlations and the after transplantation (from 6.5 f 1.7 mmol/l to Mann-Witney u-test were used for statistical 4.5 f 0.3 mmol/l) ( p < 0.05) in the rats that devel- analysis. oped diabetes no significant differences were seen

in BG concentrations between the groups, until onset of diabetes (Fig. 2).

3 0 I . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . .

0 0 7 12 23 37 48 173

Days a f t e r transplantat ion

FIGURE 2 islets. Diabetes developed on day 48 ?C 5 days. I Syngrafted BB-DP rats developing diabetes, developing diabetes, A allograftcd BB-DP rats.

Blood glucose concentration of allografted BB-DP rats with WK islets and syngrafted BB-DP rats with BB-DP syngrafted BB-DP rats noi

Aut

oim

mun

ity D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y M

onas

h U

nive

rsity

on

10/3

0/14

For

pers

onal

use

onl

y.

AN I N VlVO MODEL FOR THE STUDY OF IDDM DEVELOPMENT 91

Islets In Situ

The percentage of islets in situ infiltrated with all cell-types examined increased significantly from day 7 to onset of diabetes (Fig. 3(a)). All islets in situ in the prediabetic rats stained positive for insulin, whereas only 312 .5% of the islets in diabetic rats stained positive for insulin.

Cvajts

The number of infiltrating cells found on day 7 in the syngrafts in BB-DP rats was not different from the number and relative proportions of infiltrating cells pooled for all observation periods in syngrafts in WF rats (Figs. 1 and 3(b)). The number of infiltrating cells in the BB-DP syngrafts did not change between day 7 and 48 after transplantation, apart from a significantly reduced number of ED1 positive cells on day 48 (Fig. 3(b)). At onset of diabetes the number of all cell types examined rose significantly compared to day 7 (Fig. 3(b)), and compared to the data pooled from all observation periods in the WF syngrafts.

The degree of infiltration for MHC class 11, MHC class 1, CD4 and ED1 positive cells in the BB-DP graft was strongly and positively correlated to the percentage of islets showing infiltration in situ (Table I). Infiltration with aP-TCR or CD8 positive cells in the islets in situ were apart from 3 rats only seen at onset of diabetes (Fig. 3(a)), hence correlation was calculated only for the diabetic rats.

The insulin staining in the BB-DP syngrafts was identical in the prediabetic rats and decreased to 33 f 22% in the diabetic rats (Fig. 4).

The degree of graft infiltration was inversely correlated to the insulin staining in the same graft (Table I ) .

Allogeneic Transplantation of WK Islets to BB Rats

The blood glucose levels in the non-diabetic allografted animals were not significantly different from syngrafted BB-DP rats not developing

diabetes, at all time-points (Fig. 2). The body weight was comparable in allo-and syngrafted BB-DP rats throughout the observation period (Data not shown).

Islets In Situ

In the pancreatic islets in situ only a minor proportion (two to six percent) of the islets stained for MHC class TI, rup-TCR and ED1 showed signs of infiltration (Fig. 5(a)). All islets in situ examined at days 7, 12, 23, 37, 48, 90 and 173 after grafting, stained positively for insulin.

Grafts

In the graft, infiltration tended to be higher on days 12 and 23 after grafting compared to syngrafted BB-DP rats (Figs. 3(b) and 5(b)). Insulin staining in the WK allografts was not significantly different throughout the observation period (Fig. 6). No correlation was found between MHC class I1 staining cells in the graft and percentage of MHC class I1 staining islets in situ showing infiltrates ( r : 0.4, p = 0.09). Since islets in situ did not contain (uP-TCR or ED1 positive cells until onset of diabetes (Fig. 5(a)), no correlation was calculated for these antibodies.

Thus, islet allografting to BB-DP rats protected against both insulitis and diabetes. Graft infiltra- tion was not significantly different with regard to quantity ( p > 0.23) or quality from that observed in BB-DP islets syngrafted on days 7, 12,23 37,48 or 173, despite the fact that insulin content in the WK islet allografts was preserved.

Allogeneic Transplantation of WF Islets to WK Rats

Islets In Situ

Only few (three to seven percent) of the islets in situ stained for MHc class 11, np-TCR and ED1 showed signs of infiltration (Fig. 7(a)). All islets stained positively for insulin at all timepoints examined (data not shown).

Aut

oim

mun

ity D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y M

onas

h U

nive

rsity

on

10/3

0/14

For

pers

onal

use

onl

y.

98

120

1W

en

60

40

M

0

3ooo

2500

x)o

1500

loo0

500

0

U.B. CHRISTENSEN e/ d.

- .............................................................. ..................................................................................................................... percentage inlinraled iskts

7 12 23 37 48 OM

D a p aHer lranq3anlalion

( a )

173

FIGURE 3 (a) Percentage of infiltrated islets iri ,si/u from RB-I)P rats killed on days 7, 12, 23, 37, 4X, 173 after islet syngralting o r at onset of diabetes ( D M ) ( ! I - 6, day 173 r 1 = 4 ) . * / I < 0.01 comparing onset of DM with day 7. (b) Nuniher of iiifiltratlng cclls~mm' in the grafts from BR-DP rats killed on d a y s 7. 12. 23. 37. 48. 173 aftel- islet syngrafting or a t onset o f diabetes ( D M ) ( I ? : 6. day 173 / I = 4), list of signs iis for Figure 3(a) . * IJ < 0.01 comparing onset of D M with day 7. # / J < 0.0.5 cornparing day 48 t o day 7.

Aut

oim

mun

ity D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y M

onas

h U

nive

rsity

on

10/3

0/14

For

pers

onal

use

onl

y.

AN I N VIVO MODEL FOR THE STUDY OF IDDM DEVELOPMENT 99

TABLE I Correlation coefficients of number of infiltrating cells/mm2 versus percentage of stained islets in .v i f z r from BB-DP rats grafted with BB-DP islets and killed at days 7, 12, 23, 37, 48, 173 and at onset of diabetes. Apart from the diabetic rats, only 3 rats showed infiltration with tu/-l-TCR and CDX positive cells in the islets in s i f u (Fig. 3(a)) hence. correlation was only calculated diabetic rats

Type of infiltrating cell Graft infiltrate versus YO infiltrated islets Bi vim

Graft infiltrate versus YO insulin staining in graft

MHC cLi\\ 11 MHC c l m I c D4 ED I rv/j-TCR CD8

r : 0.6, p = 0.0001 r : 0.5. /I = 0.003 r : 0.7. p < 0.0000I r: 0.5, p=0.001 r : 0.5.p=0.2

r: -0.2, p = 0.7

r : -0.8, p < 0.00001 r : -0.7, p < 0.00001 r: -0.8. p < 0.00001 r : -0.8, p < 0.00001 I': -0.6, p < 0.0003 r : -0.x. p < 0.00001

120 7

T

80

60

40

20

0 ? 12 23 37 40 173

Days after transplantation

FIGURE 4 onset of D M

Graft insulin staining in percentage of total graft area from syngrafted BB-DP rats. * p = 0.05 comparing day 7 with

Thus, islets allografted to non-diabetes prone rats did not sensitize the recipients against the islet iii s i t i 4 .

Gvufis

On day 10 significantly more ED1 positive cells and on day 12 significantly more MHC class 11, CD4, CD8 and tr,!j-TCR positive cells were seen com- pared to day 7 (Fig. 7(b)). On day 17 the number of infiltrating cells had returned to baseline (Fig.

7(b)). At day 17 insulin staining was significantly increased compared to day 12 (Fig. 8). On day 7 the insulin staining in allografted W K islets and syngrafted BB islets were identical, but had significantly decreased in the W F rats allografted with WK islets at day 12 ( p = 0.007). No significant differences were seen in infiltration in the graft between allografted W F islets day 12 and BB-DP syngrafted islets at onset of diabetes, but infiltra- tion in islets it? sitzr from W K rats grafted with W F islets was significantly lower compared to islets

Aut

oim

mun

ity D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y M

onas

h U

nive

rsity

on

10/3

0/14

For

pers

onal

use

onl

y.

100-

80 ~

60.

40.

20

0

percentage of infiltrated islets

-

7 12 23 37 48 90 1 7 3

Days after transplantation

I MHC Class II BlaP-TCR OED1

2500

2000

1500

1000

500

0

No. cells in the transplant

T

7 12 23 37 48 90 173

Days after transplantation

FIGURE 5 (a) Percentage of infiltrated islets i r r , s ir i t l‘i-oiii H H - D I ’ rills killed o n days 7. 12, 2 3 . 37, 48, 00 and 173 alter- allogrnfting with WK islets ( a = 3 ) . (b) Number 01‘ inliltrating cclls,’nim’ in the grafts f‘rom allo-grafted RR-DP rat\ killed on days 7. 12. 23. 37, 4X. 90 and 173 after allografting with WK islets ( i t : 3).

Aut

oim

mun

ity D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y M

onas

h U

nive

rsity

on

10/3

0/14

For

pers

onal

use

onl

y.

AN I N VIVO MODEL FOR THE STUDY O F IDDM DEVELOPMENT 101

.- P C 6 u)

._ L

C

3 ul C

._ -

.-

120 - 1 80

60

40

20

0

7 12 23 37 40 90 173

Days after transplantation

FIGURE 6 Graft insulin staining in percentage of total graft area from allografted BB-DP rats.

100

90

80

70

60

50

40

30

20

10

0 7 10 12 17

Days after transplantation

MHC Class I1 m ED1 04%-TCR

FIGURE 7(a)

Aut

oim

mun

ity D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y M

onas

h U

nive

rsity

on

10/3

0/14

For

pers

onal

use

onl

y.

2500 ~

7

No. cells/mm2 transdant

T

10 12 Days after transplantation

17

E4 MHC class I I H MHC class I 0 CD4 0 ED1 a 4 - T C R 0 CD8

FlGURE 7(b)

FIGURE 7 (a) Percentage of infiltrated islets in sifu WK rats killed on days 7, 10, 12 and 17 after allografting with W F islets (n=6) . (b) Number of infiltrating cells/mni' in the grafts from W K rats allografted with W F islets, killed on days 7, 10, 12 and 17 after allografting with W F islets ( n = 6) . # p < 0.01 comparing day 10 to day 7. * p < 0.01 comparing day 12 to day 7.

B

T

T T

7 10 12 17

D a y s after transplantatan

FIGURE 8 comparing day 7 to 12. # p=O.01 comparing day 12 to 17.

Graft insulin staining in percentage of total graft area from WK rats allografted with W F islets. *p=0.05

Aut

oim

mun

ity D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y M

onas

h U

nive

rsity

on

10/3

0/14

For

pers

onal

use

onl

y.

A N I N Vf VO MODEL FOR THE STUDY OF IDDM DEVELOPMENT

3500

3ooo

2500

1500

500

0 BB-WK WF-WK

Days 7 and 12 after transplantation

loo, I

90 I 80

70

60

50

40

30

20

10

0

80-WK WF-WK

Days 7 and 12 after transplantation

( b)

103

FIGURE 9 (a ) Number of infiltrating cells/mni' in the WK rats allografted with either BB-DP islets ( n = 6 ) or W F islets ( P I = 12). killed on days 7 and 12 after grafting. * p = O . O 2 . (b) Graft insulin staining in percentage of total graft area from WK rats allografted with either BB-DP or WF islets. * p = 0.01.

Aut

oim

mun

ity D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y M

onas

h U

nive

rsity

on

10/3

0/14

For

pers

onal

use

onl

y.

104 U.B. CHRISTENSEN r t d

in situ from syngrafted BB-DP rats at onset of diabetes ( p < 0.01) (Figs. 7(a) and 3(a)).

Allogeneic Transplantation of BB Islets to WK Rats

Examination of the islets in .vim was not performed. Interestingly, BB-DP islet allografts showed signif- icantly lower number of MHC class IT, CD4, CD8 and @TCR positive cells (Fig. 9(a)) and signifi- cantly higher insulin staining (Fig. 9(b)) compared to allografted MHC compatible (RTI “) WF islets at days 7 and 12.

DISCUSSION

In this study we have used a transplantation model of neonatal syngeneic BB-DP rat islets grafted to 30 day old BB-DP rats. By placing 200 islets in one graft the islets will with time form a syncytium or “macroislet”. We hypothesized that the graft would be infiltrated in a “synchronized” fashion, in contrast to the physically separated islets in .situ and infiltrated in a way that reflects the events taking place in a single islet during the destructive pathogenetic process leading to IDDM.

We found a strong positive correlation between the mononuclear cell infiltration and islet cell destruction in the graft and islets in situ in the BB-DP rats grafted with islets syngrafts. Further we found that allografting WK islets to BB-DP rats resulted in inhibition of diabetes and insulitis.

A heavy infiltration of the WF islet allografts grafted to WK rats was seen without infiltration of the islets in sifu. Thus, islet allografting to normal rats did not sensitize the recipients against islets in s i t u .

The results from the BB-DP syngrafts indicate that syngrafts are infiltrated with the same kinetics and by the same cell types as each islet in situ regardless of the time of retrieval of the graft and that the graft represents a model of the infiltration and its development in the islets in situ. The syn- chronization between destruction of islets in s i t i r

and in the graft has previously been In that study, however, rats were grafted with either MHC incompatible islets or islets from diabetes resistant BB-rats to investigate if MHC incompat- ible islets would be susceptible to immune attack. The rats were grafted at a higher age (90 120 days), and killed either 21 days after grafting or at onset of disease. Islets were pre-treated to resist allograft attack, and the rats treated with antilym- phocytic serum immediately after grafting. The synchronization of the destruction between islets in situ and in a graft was confirmed by Prowse et a/.[231 in a similar study, using MHC incompatible islet donors pretreated with cyclophosphamide. In none of these studies the natural history of graft infiltration was studied. In a study in prediabetic NOD mice[241 syn-, allo-, and xenografted islets were transplanted to the same prediabetic mouse to investigate the degree of rejection of the grafts. No unambiguous correlation was found between the infiltration in the syngraft and in the pancreatic islets, and no report was made concerning the effects of the grafting on diabetes incidence, or whether transplanting both syn-, allo-, and xeno- grafts to the same mouse would affect the infiltra- tion of the grafts or islets in situ.

In our study, grafting of 200 BB-DP islets to BB- DP rats did not alter diabetes incidence. Four hundred islets were not sufficient to reverse diabetes,’251 and even though external prophylactic treatment of insulin to BB-DP rats reduced incidence of diabetes,[*‘] transplantation of islets to normal rats resulted in a reduction of [%cell mass in the graft and a normal P-cell mass in the host pancreatic islets.[271 This indicates that the grafted islets represents a functional reserve and do not “take over” the control of the insulin response.

Graft insulin content was determined as percent- age of insulin positive cells per total graft area, implying that a few insulin positive cells in a small graft would be assessed as having a high percentage of graft insulin content. This was seen in the WF islets grafted to WK rats, where the insulin content fell concurrently with an increase in mononuclear cells from day 7 to day 12. At day 17 only a few

Aut

oim

mun

ity D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y M

onas

h U

nive

rsity

on

10/3

0/14

For

pers

onal

use

onl

y.

AN I N VIVO MODEL FOR THE STUDY OF IDDM DEVELOPMENT 105

insulin positive cells were identifiable, but since the graft had decreased significantly in size ( p < 0.05) the insulin score was quite high. Within each group and between the groups the area of grafts varied in size dependent upon whether the section was from the middle of the graft or from one of the edges, but apart from allografted WK rats on day 17, none of the variations reached statistical significance. Inter- estingly, only 5% of the allografted BB-DP rats developed diabetes, eventhough the cellular infil- trate was quantitatively and qualitatively alike in the allo- and syngrafted BB-DP rats until onset of diabetes in the syngrafted BB-DP rats and that the insulin content in the WK islet allografts were preserved. We do not have an explanation for this interesting observation, which is based on an obser- vation of n = 3 in each group. Curative transplanta- tion by grafting 1000- 1200 allogeneic islets from MHC mismatched donors to diabetic BB-DP rats resulted in prevention of diabetes,"'] whereas grafts consisting of 40 MHC mismatched islets to predia- betic BB-DP rats were destroyed.[231 This indicates that the amount of allografted islets transplanted may be of importance for the graft survival and prevention. Possibly allografting just after weaning induces a tolerant state towards p-cells since Hegre et u ~ . [ ~ ' I found that grafting of 100-200 MHC mismatched neonatal islets to aglucosuric 46-58 day old BB-DP rats resulted in glucosuria, and insulitis in the host islets but no mononuclear cell infiltration in the allograft. Since we did not co- transplant a non-islet allograft it could be specu- lated that the prevention seen is caused by a non-islet specific effect of allografting. However, grafting of MHC mismatched islets and pituitary tissue to prediabetic BB-DP rats rendered the islets infiltrated and the pituitary tissue unaffected.'*31 Further studies are needed to elucidate possible therapeutic perspectives of these observations.

Allografting of neonatal BB-DP rat islets to WK rats resulted in a significantly lower number of infiltrating cells in the grafts compared to WF islets allografted to WK rats. 1ti vi fro it has been shown that islets isolated from 8 day old BB-DP rats are significantly less vulnerable to cytotoxic attack from

spleen mononuclear cells than islets from age matched WF rats.["' Further, KO et ~ 1 . ' ~ ' ~ demon- strated that expression of a membrane-bound islet cell-specific 38 kDa antigen was not detected before the BB rats were 30 day old, whereas W F rats and diabetes resistant rats express the antigen at birth. This suggests that neonatal BB-DP islets might be less antigenic compared to neonatal islets from other rat strains.

In the isolated islets used for grafting we found MHC class I1 positive cells after the 5 day culture period. Analysis of the islets prior to grafting was made to investigate if mononuclear cells were at all present after 5 days of culture and if present, whether the amount of cells would be different in WF and BB-DP islets and then affect the total number of mononuclear cells counted in the graft. But since the number of positive cells was so small it did not affect the total mononuclear cell count. We also estimated the number of MHC class I1 positive cells in whole islets from 30 days old BB- DP and WF rat. The area of at least 14 islets in situ from 6 WF and BB-DP rats was measured and the number of infiltrating MHC class I1 positive cells counted and pooled. The number of MHC class I1 positive cells within one mm3 was calculated to 7800 in BB-DP rats and 9700 in W F rat islets. One mm3 approximately equals 1065 islets"51 resulting in approximately 8 MHC class I1 positive cells in islets from BB-DP rats and 10 in islets from W F rats in accordance with previous observation^.[^" The presence of MHC class I1 positive cells in isolated adult islets has been described p r e v i ~ u s l y . " ~ ~ ~ ' ~ Judged from the shape of the cells they were most likely dendritic cells and/or macr~phages.'"~ Ten to thirty MHC class I1 cells/islet were present in freshly isolated islets, less after four days of culture."*] Interestingly, Ulrich et ~ 1 . ' ~ ~ ' found that islets from Dark Agouti rats (RTI") contained 3 times less MHC class I1 positive cells compared to Lewis (RTI') and CAP rats (RTI'). In accordance with our findings Pipeleers e f ul.'"] found no islet passenger lymphocytes after 4 days of culture.

N o significant increase was seen in the cellular infiltrate in grafts in prediabetic syngrafted BB-DP

Aut

oim

mun

ity D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y M

onas

h U

nive

rsity

on

10/3

0/14

For

pers

onal

use

onl

y.

I06 U.B. CHRISTENSEN P/ u/

rats, but we did observe an interesting trend towards an increase in number of infiltrating cells on day 37 after transplantation. The time-point for this increase correlates with an increase in free radical activity measured by an increase in expired pentane in BB-DP rats before diabetes onset.[341

The study has shown that grafting of 200 isolated neonatal BB-DP islets to 30 day old BB-DP recip- ients did not alter time of onset or incidence of diabetes. The kinetics and the types of cells seen in the cellular infiltration in the syngeneic grafts mirror the percentage of infiltrated islets in .situ in BB-DP rats. The insulin content in the syngeneic BB-DP grafts was inversely correlated to the mononuclear cell infiltration in the graft. A similar correlation was not found for WK islets allografted to BB-DP rats, in which allografting seemed to prevent insu- litis and diabetes. The syngeneically transplanted BB-DP islets form a “macroislet” and constitute a unique model for molecular examinations in v i vo of the processes taking place in the islets either during the destruction of islets in the pathogenetic process leading to IDDM or during investigations of intervention modalities.

Acknowledgements

The skillful technical assistance by Ms. E. Schjerning and Ms. H. Foght, Steno Diabetes Center (Gentofte, Denmark), Ms. S. Day, Department of Pathology, Division of Immunology, (Cambridge) and by Mr. Troels Momberg-J0rgensen and his staff at the Biological Department of Novo Nordisk A/S (Gentofte, Denmark) is greatly appreciated. J.N. is the recipient of a research grant (#1921125), U.B.C. and H.U.A. of a postdoctoral fellowship from the Juvenile Diabetes Foundation International N.Y. and U.B.C. of a research grant from The Danish Research Council. This work was supported by grants from the Danish Diabetes Association (U.B.C., H.U.A., T.M.-P.T.S), The Bernhard and Marie Klein Foundation (U.B.C., H.U.A.), The Poul and Erna Sehested Hansen Foundation (U.B.C.), The King Christian d. X Foundation (U.B.C.). Dr. L. O’Reilly, Ph.D., (Department of

Pathology, Division of Immunology, (Cambridge) is warmly thanked for inspiring discussions.

Refkvences

[ I ] Ckpts, W. Pathologic anatomy of the pancreas in juvcnilc diabetes mellitus. Diuhete.u 1965. 14: 619 - 3 3 .

[2] Foulis, A.K., Liddle, C.N., Farquharson, M.A., Rich- mond. J.A. and Weir, R.S. The histopathology of thc pancreas in type I (insulin-dependent) diabetes mellitus: a 25-year review ofdeaths in patients under 20 years ofage in the United Kingdom. Diuhrtologio 1986. 29: 267-274.

[3] Juncker, K.. Egeberg, J., Kromann, H. and Nerup, J. An autopsy study of the islets of langerhans in acute-onset juvenile diabetes nicllitus. A c / u p t r / h . inicrohiol. sc,uritl. 1977. 85: 699-706.

[4] Hanenberg, H., Kolh-Bachofen, V., Kantwerk-Funke, G. and Kolb. H . Macrophage infiltration precedes and is a prerequisite for lymphocytic insulitis in pancreatic islets of pre-diabetic RB-rats. Diuhcrologiu 1989. 32: 126 132.

[ S ] Nakhooda. A.F., Like,A.A.. Chappel, C.I.. Wei, C.-N. and Marliss, E.B. The spontaneously diabetic Wistar rat (the BB rat). Diuhrtologitr 1978. 14: 199-207.

[6] Bonc, A.J., Walker, K., Varcy, A.-M., Cooke, A. and Baird, J.D. Effect of cyclosporin on pancreatic events and development of diabetes in BB/Edinburgh rats. Ditrhr,tca.\ 1990. 39: 508- 14.

[7] Walker, R.. Bone, A.J., Cooke, A. and Baird, J.D. Distinc macrophage subpopulation in pancreas of prediabetic BB/ E rats. Diuhews 1988. 37: 1301-1304.

[8] O’Reilly, L.A., Hutchings,P.R.,Crocker, P.R.,Simpson, E., Lund, T., Kioussis, D.. Takei, F.. Baird, J . and Cookc. A. Characterization of pancreatic islet cell infiltrats in NOD mice. Effcct of cell transfer and transgene cxprcssion. Eirr. J . / / 7 ? f I ~ Z / f 7 0 / . 1991. 21: 1171-1 180.

[9] Fujita, T., Yui, R. Histological changes in thc pancrcatic islets of NOD mice. In Tarui S. , Tochino Y., Nonaka K. ed. / t i .s ir / i / rs utid / j p c I t / i c r /w /c~s . Academic Press, Tyoko. 19x6. 3 5 - SO.

[lo] Makino, S., Kuninioto, D., Muraoka, Y., Mizushima. Y.. Katagiri, K. and Tochimo, Y. Breeding of a nonobese diabetic strain ofmice. E.\-p. h i m . 1980. 29: I 13.

[ I I ] Lambert, E.F., Signore, A,, Gale. E.A.M. and Pouilli, P. Lessons from the NOD niouse for the pathogenesis and iiiiiiiunotherapy of human Type 1 (insulin dcpendent) diabetes mellitus. Diuhrtologirr 1989. 32: 7?3- 708.

[I21 Brunsted. J.. Nielsen, J.H. and Lernmark, A. The Hagedom Study Group. Isolation of islets from mice and rats. Mclhotls in Diuherc2.s Rr.sc.urch. Larner Johl SL eds. New York, Wiley. 19x4. 1: 131-138.

[I31 Korsgren, 0.. Jansson. L., Sandlcr, S. and Andersson, A. Hyperglycemia-induced B ccll toxicity. The fate of pancreatic islets transplanted to mice dependent on their genetic background. J . Clin. f i rvrsi . 1990. 86: 2161-6X.

[I41 Butlcr, L., Gubcrski, D.L. and Like. A.A.: Gcnctic nnulysis of the BB/W diabetic rat. Ctrn. .I . G m , / . Cj,to/. 1983. 25: I IS.

[ IS] Fukunioto. T., McMaster, W.R. and Williams. A.F. Mouse monoclonal antibodies against rat major histocon- patihility antigens. Two la antigens and cxpression of la and class I antigens in rat thymus. Eur. J . / f J f f ? f f / i l f / ~ / . 1982. 12: 23 7 - 243.

Aut

oim

mun

ity D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y M

onas

h U

nive

rsity

on

10/3

0/14

For

pers

onal

use

onl

y.

AN I N VIVO MODEL FOR THE STUDY OF IDDM DEVELOPMENT 107

[I61 McMaster. W.R. and Williams, A.F. Identification of la glycoproteins in the rat thymus and purification from rat spleen. Eur. J . Inimunol. 1979. 9: 426-433.

[I71 Hunig, T., Wallne, H.J., Hartley, J.K., Lawetzky, A. and Tiefenthaler, G. A monoclonal antibody to a constant determinant of the rat antigen receptor that induces T cell activation. Differential reactivity with subset of immature and mature T lymphocytes. J . E.xp. Med. 1989. 169: 73-86.

[I81 Parnes, J.R. Molecular biology and function o f C D 4 and CD8. Adv. Imrnuwol. 1989. 44: 265-311.

[I91 Brideau, F.J.,Carter, P.B., McMaster, W.R., Mason, D.W. and Williams, A.F. Two subsets of rat T lymphocytes defined with monoclonal antibodies. Eur. J . Ininiunol. 1980. 10: 609-615.

[20] Barclay, A.N. The localization of populations of lympho- cytes defined by monoclonal antibodies in rat lymphoid tissues. Iriimnnologq' 198 I . 42: 596- 600.

[21] Torres-Nagel, N., Kraus, E., Brown, M.N., Tiefenthaler, G., Mitnacht, R., Williams, A.F. and Hunig, T. Differential thymus dependence of rat CD8 isoform expression. Eur. J . Inmiuriol. 1992. 22: 2841 -2848.

[22] Werninger. E.J. and Like, A.A. Immune attack on pancreatic islet transplants in the spontaneously diabetic biobreeding/worcester (BB/W) rat is not M H C restricted. Journul of lniniunologj~ 1985. 134: 2383-86.

[23] Prowse, S.J., BEllgrau, D. and Lafferty, K.J. Islet allografts are destroyed by disease occurrence in the spontaneously diabetic BB rat. Diuhrtrs 1986. 35: 110--14.

[24] Mandel, T.E., Koulmanda, M., Loudovaris, T. and Bacelj, A. Islet grafts in NOD mice: A comparison of iso-, allo-, and pig xenografts. Trunspluntution Proctwlings 1989. 21: 3813-38 14.

[25] Derek, W.R., Gray, P.M. and Morris, P.J. The effect of hyperglycemia an isolated rodent islets transplanted to the kidney capsule site. Trunspluntution 1986. 41: 699-703.

[26] Bertransd, S., Paepe, M. de, Vigeant, C. and Yale, J-F. Prevention of adoptive transfer in BB rats by prophylactic insulin treatment. Diuhete.v 1992. 41: 1273-77.

[27] Montana, E., Bonner-Weir, S. and Weir, G.C. Beta cell and growth after syngeneic islet cell transplantation in normal and streptozocin diabetic C57BL/6 mice. J . Clin. Invest. 1993. 9: 780-787.

[28] Woerhle, M., Markman, J.F., Beyer, K., Naji, A, , Bretzel, R.G. and Feeerlin, K. The influence of the implantation site (kidney capsule vs portal vein) on islet survival. Hormone und Metuholism Resrurcli 1990. supl. I . 25: 163-5.

[29] Hegre.0.D.. Enriques, A.J., Ketchum, R.J., Weishaus,A.J. and Serie, J.R. Islet transplantation in spontaneously diabetic BBjW or rats. Diuhetcs 1989. 38: 1148-54.

[30] Ekblond, A. , Schou, M. and Buschard, K. Cytotoxicity towards neonatal versus adult BB rat pancreatic islet cells. Autoimmunity 1995. 20: 93-98.

[31] KO, I.Y., Jun, H.S., Kim, G.S. and Yoon, J.W. Studies on autoimmunity for initiation of beta-cell destruction X- Delayed expression of a membrane-bound islet cell-specific 38 kDa autoantigen that precedes insulitis and diabetes in the diabetes-prone BB rat. Diuhetologia 1994. 37: 460-465.

[32] Pipeleers, D.G., Pipeleers-Marichal, M., Hannaert, J.-C., Berghmans, M., In't Veld, P.A., Rozing, J., Van De Winkel, M. and Gepts, W. Transplantation of purified islet cells in the diabetic rats. Diuhetes 1991. 40: 908-19.

[33] Ulrichs, K. and Muller-Ruchholtz, W. M H C class I1 antigen expression on the various cells of normal and activated isolated pancreatic islets. Dicrgnostic~ hnmunology 1985. 3: 47-55.

[34] Pitkanen, O.M., Martin, J.M., Hallman, M., Alerblom, H.K., Sariola, H. and Andersson, S.M. Free radical activity during development of insulin-dependent diabetes mellitus in the rat. L(/e Sciences 1991. 50: 3 3 5 ~ ~ 339.

[35] Haist, R.E. and Pugh, E.J. Volume measurement of the islets of Langerhans and the effects of age and fasting. Am. J . Phq'siol. 1948. 152: 36-41.

Aut

oim

mun

ity D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y M

onas

h U

nive

rsity

on

10/3

0/14

For

pers

onal

use

onl

y.