- Email: [email protected]
Transplantation of discordant pig islet xenografts in diabetic rats
Diabetes Research and Clinical Practice, 15 (1992) c-: 1992 Elsevier
Science Publishers
197-204
197
B.V. All rights reserved 0168-8227/92/$05.00
DIABET 00603
Transplantation
of discordant pig islet xenografts in diabetic rats
Wah Jun Tze, Joseph Tai and Sze-Shuen Departments of Pediatrics and Patholoa.
Cheung
University of British Columbia, Vancouver, British Columbia. Canada
(Received 8 April 1991) (Revision accepted 4 October 1991)
Summary Xenotransplantation of pig islets under the kidney capsule (KC) of diabetic rats was performed. Natural preformed AC1 rat anti-pig leukocytotoxicity, leukoagglutination and hemagglutination antibody titers ranged from Neat-l : 16, 1 : 8-1 : 32 and 1 : 128-1: 256, respectively (n = 14). Normal AC1 sera were non-toxic to pig islets during short term incubation. Pig islet xenograft survival times in the nonimmunosuppressed AC1 rats, AC1 rats immunosuppressed with antithymocyte serum (ATS) or cyclosporin A were 3.8 k 0.4 (mean + SE; n = 5), 10.4 & 0.7 (n = 13) and 6.0 k 1.0 (n = 5) days, respectively. Pig islets implanted in the abdominal testis of AC1 recipients immunosuppressed with 5 doses ATS survived for a mean of 6.4 ? 1.O days (n = 7). The mean K rate following an intravenous glucose tolerance test (IVGTT) in AC1 rats 1 week after transplantation with pig islet under the KC was 2.2 k 0.4 (n = 10) compared to that of 2.91 + 0.30 found in normal control rats (n = 8). Peak insulin at 1 min was 60.1 + 3.9 pU/ml (n = 4). Histological and immunohistochemical examination showed that the xenograft from recipients treated with 5 doses of ATS still contained well-preserved islet tissue with many insulinand glucagon-containing cells on the day of graft removal when blood glucose had returned to hyperglycemic level. Both CD4 and CD8 positive cells were in the vicinity of the graft tissue. This study has shown that hyperacute rejection of islet xenograft did not occur in this discordant xenograft model despite the presence of preformed anti-pig antibodies in the rat recipients. Further islet xenotransplantation using this discordant model would be useful in the design of an effective immunosuppressive regimen. Key words: Pig; Rat; Islet; Xenotransplantation;
Correspondence
to: W.J. Tze, Depts. of Pediatrics
and Patholof British Columbia, 4480 Oak Street, Vancouver, B.C., Canada V6H 3V4. Abbreviations: ATS, antithymocyte serum; BSA-HBSS, Hanks’ balanced salt with 0.4% bovine serum albumin; IVGTT, intravenous glucose tolerance test; KC, kidney capsule; K rate, glucose disappearance rate; STZ, streptozotocin.
Streptozotocin;
Diabetes
Introduction
ogy, University
Islet allotransplantation into the diabetic rat has been shown to normalize the carbohydrate metabolism as well as preventing secondary diabetic complications in the recipients [ 1,2]. Although promising results have been achieved by islet allo-
198
transplantation into diabetic animals, a corresponding approach to human therapy is hampered by the limited supply of human islet tissue. Islet xenograft is an appealing alternative to islet allograft because this approach would enable the use of an unlimited supply of large animal donor tissue for transplantation in human diabetics. In spite of the potential, islet xenograft has been the subject of relatively few studies. The majority of the islet xenotransplantations were performed between donor-recipient of closely related species such as rat to mouse or hamster to mouse [ 3,4]. Relatively few islet xenotransplantation studies were reported between widely disparate combinations of animals [5]. Pig islet is a potential donor tissue for xenotransplantation in humans. Earlier, Ricordi reported xenotransplantation of adult pig islet in diabetic mice [6]. Simeonovic et al. [ 71 demonstrated that xenografts of fetal pig proislets reversed STZ-induced diabetes and achieved long-term normoglycemia > 139-168 days in CD4-depleted recipient mice. Presently, there is little information on the metabolic function of adult pig islet xenograft in diabetic rat recipients. The present study assessed the natural anti-pig antibodies in the AC1 rats and the survival time and histology of adult pig islet xenografts in the discordant STZ-induced diabetic recipients.
Materials and Methods Animals
Male AC1 rats (Harlan SD, Indianapolis, IN) of 200-225 g body weight were used as recipients. Diabetes was induced by a single intravenous injection of 55 mg/kg body weight of STZ and they were considered as diabetic when their blood glucose (BG) was >400 mg/dl for at least 10 days. Preparation of pig islets
Pig pancreatic islets were isolated from sow pancreata obtained from a local abattoir. The warm and cold ischemic times were < 20 min and 2 h,
respectively. Pig pancreatic islets were isolated with a modified technique of Gray et al. [ 81. The pancreas was perfused with Type XI collagenase (0.75-1.0 mg/ml, 1 ml/g tissue; Sigma Chemical Co. St. Louis, MO) in Hanks’ balanced salt solution supplemented with l”/ bovine serum albumin and digested for between 15-20 min at 37 oC. The digested gland was then placed in cold Hanks’ balanced salt solution with 0.4% bovine serum albumin (BSA-HBSS) and chopped with fine scissors and teased with blunt forceps to release the islets. The islets and non-islet fragments were washed with BSA-HBSS and sedimented at unit gravity repeatedly to remove the exocrine enzymes. After overnight culture at room temperature, islets were purified by multilayered Percoll gradient (SG 1.020, 1.045 and 1.068). Islet tissue was resuspended in Percoll with SG 1.068 and then overlayered with Percoll with SG 1.045 and 1.020. After 20 min of equilibration at unit gravity, islets were collected from the layer with SG 1.045. Their identity was confirmed by staining with dithizone [ 91. The contaminant consisted mainly of ductal tissue. For an in vitro functional study, 100 islets (150-250 pm in diameter) hand-picked from the Percoll purified preparation were tested for their response to glucose stimulation in a perifusion system which consisted of a 13 mm Swinnex filter unit. BSAHBSS at 37°C was introduced into the chamber at 1 ml/min with a peristaltic pump. Glucose concentrations of the BSA-HBSS were 100 mg/dl between O-60 min, 121-180 min and 450 mg/dl between 6 l-120 min. Samples collected were stored at - 20 ‘C until insulin assay. Xenotransplantation and immunosuppression
Between 1000-4000 pig islets (150-250 pm in diameter; 1 g sedimentation volume of 4000 islets, equivalent to = 50-80 ~1 in a 50 U insulin syringe) were transplanted into each STZinduced diabetic rat under the KC or into the abdominal testis [ lo]. For immunosuppression with ATS, 1 ml of the antiserum prepared in our laboratory [ 111, was given intraperitoneally immediately after islet transplantation and daily for
199
the 4 subsequent days. Cyclosporin A was given subcutaneously at 30 mg/kg/day for 3 consecutive days following transplantation [ 121. Histology and immunohistochemical
study
BG levels were checked
daily until the day of removal of the kidney containing the graft, that is, when BG of the recipient rose above 200 mg/dl on 2 consecutive days. Equal halves were snap frozen in LN2 or fixed in Bouin’s fixative. Paraffin sections were stained for insulin, glucagon and somatostatin with immunoperoxidase staining and frozen (ABC Staining Kits, Vectastain) sections for cellular markers with monoclonal antibodies against leukocyte common antigen (0X1), class II antigen (0X6), CD4 (W3/25) and CD8 (0X8) (Cedarlane Lab., Hornby, Ontario). IVGTT (1 g/kg body weight) was performed on 8 normal rats, 5 diabetic rats and 10 diabetic AC1 rats with functioning pig islet xenograft under the KC 7 days after transplantation. During IVGTT, water was allowed but not food. The rates of glucose disappearance (K rate) were calculated. Insulin was determined by radioimmunoassay using pig insulin as standard (Pharmacia Diagnostics, Montreal, Quebec). Determination of preformed
rat anti-pig antibodies
Natural occurring leukocytotoxicity, leukoagglutination and hemagglutination titers in the sera of normal AC1 rats were assessed on pig WBC and RBC [ 131. All rat sera were heat inactivated for 30 min at 56°C. For assessing the toxic effect of normal rat sera on pig islets, groups of 20 pig islets were placed in wells of 96-well microtiter plates in the presence of 100 ~1 of rat serum. They were examined after 1 h of incubation at room temperature. Rabbit complement (100 ,uI; 1 : 8 dilution, Low-Tox Complement, Cedarlane Lab, Hornby, Ontario) was added for another hour and then the islets were re-examined. Results The pig islets used for the transplantation are less compact than those of either humans or rats. The
major contaminant of the preparation is the ductal elements. Fig. 1 shows that all preparations of the isolated pig islets secreted basal insulin. However, there is variation among the different islet preparations in their response to glucose stimulation. Table 1 shows that natural rat anti-pig antibodies were detected in all normal AC1 animals before transplantation (n = 14). These antibodies were non-toxic to the pig islets during short-term incubation. Table 2 and Fig. 2 show the functional period
p
I
50
-
8
40
-
i 2
30
-
$E
20
-
B
10
LOW
-HIQHl
LOW 60
I
I
T
I
/
I
I
1
I
H
zl 5
‘a.lI
t
01 0
’
20
’
40
’
60
’
80
’
100
’
120
’
140
’
160
I
180
’
200
Minutes
Fig. 1. Insulin secretion kinetics of groups of 100 pig islets during perifusion with BSA-HBSS containing 100 mg/dl 450 mg/dl glucose 121-180 min) and (O-60 min, (61-120 min; mean f SE; n = 10). TABLE 1 Leukocytotoxicity, leukoagglutination, hemagglutination and islet toxicity titers of normal AC1 rat sera against pig cells Antibody titers* Leukocytotoxicity Leukoagglutination Hemagglutination Islet toxicity
N (9)#; I:8 (3); 1:16 (2) I:8 (I); 1:16 (11); 1:32 (2) 1:128 (1); I:256 (13)
* For leukocytotoxicity titers, the end point was the last well showing 50% lysis. For leuko- and hemagglutination assays titer was taken as the last well showing agglutination. # The number in the bracket indicates the sample size. + Minimal cytotoxic effect was observed for normal AC1 sera. Positive control sera from 4 AC1 rats sensitized with pig cells showed lysis of up to 507, of the islets at I:4 dilution.
200 TABLE 2 Functional
period of pig islet xenograft in diabetic AC1 rats
KC” KC” AT,’ KC KC’ KC“ ~_ ._~~~.
1
2 3 4 5 6 _
Immunosuppression
Survival days
(n) 5 13 7 6 6 7
None ATS x ATS x CsA x None ATS x
3,3,4,4,5 5,7,8,10 x 3,10b,ll 44557 7 , , 7 . 10,lO 4.5,5,5,6,11 < 2 days <2 days
Number
Transplant site
Group
5 5 3 5
Mean + SE’
3.8 > 10.4 6.4 6.0
x 2,12b,13,14,14b
f f + +
0.4 0.7 1.0 1.0
.I Transplanted with > 4000 Percoll purified islets under the kidney capsule (KC) or abdominal testis (AT); n = number of recipients. h Graft functional at time of death. ’ Transplanted with > 4000 Ficoll purified islets under the kidney capsule (KC). ” Transplanted with ~2000 Percoll purified islets under the kidney capsule. ’ Group 2 vs 1, P < 0.003, calculated by the Mann-Whitney U test; Group 3 vs I, P < 0.04; Group 2 vs 3, P < 0.03.
g
500
-
\, E
400
-
E: 300 Z a E 200
-
x
E @I
-
100
-
0' -2
1
I
I
I
I
L
0
2
4
0
8
10
I
12
I
14
I 16
was 6.4 t 1.0 days (Group 3). With cyclosporin A as immunosuppressant, the functional period was 6.0 + 1.0 days (Group 4). Group 5 recipients each received x 4000 pig islets purified by Ficoll gradient in the KC. Only transient lowering of BG for <2 days was observed in all animals. In the recipients of ~2000 islets (Group 6) the lowering of BG was transient for < 2 days. Fig. 3 shows that there is no correlation
DAY
Fig. 2. Non-fasting serum glucose (mean F SE, n = 13) of diabetic AC1 rats transplanted with Percoll purified pig islets under the kidney capsule, immunosuppressed with ATS ( 1 ml/day, day O-4).
of xenotransplanted pig islets in diabetic rats. BG levels returned to normoglycemic state in diabetic recipients of islets under the KC in l-2 days. In the non-immunosuppressed recipients transplanted with > 4000 pig islets, the survival period was 3.8 + 0.4 days (Group 1). Significantly prolonged function of the graft with serum glucose levels maintained at < 200 mg/dl for more than 10 days was seen in lo/13 rats that received 5 doses of ATS after the islet xenograft (Group 2). When similar numbers of islets were transplanted into the abdominal testis of diabetic recipients given 5 doses of ATS, graft survival
.
30 f F
.
.
3
20 I 10
1
.
a
l
.
.
I
l
.
.
.
I
Fig. 3. Duration of pig islet xenograft function in diabetic serum hemagglutination (o ) and AC1 rats and leukoagglutination titers (0) at the time of pig islet xenotransplantation.
201
immunosuppression the graft was mostly fibrotic and only occasional insulin-containing cells were seen by day 5 after transplantation.
Discussion 80
I
-20
0
20
40
60
60
100
120
140
Fig. 4. Serum glucose (mean + SE; A) and insulin (B) levels following IVGTT (1 g/kg body weight) in overnight fasted normal AC1 rats (o-----o; n = 8); diabetic AC1 rats (A-----A ; n = 7), diabetic AC1 rats with functioning pig islet xenograft under the kidney capsule, treated with 5 doses of ATS, 1 week after transplantation (V-----V; n = 10 and 4 for glucose and insulin), and 1 week after rejection of pig islet xenograft (A-----A ; n = 4).
between the pig islet functional period and the anti-pig hemagglutination and leukoagglutination titers in the AC1 recipients at the time of transplantation. Fig. 4 shows that successful pig islet xenograft resulted in a significant improvement in the glucose disappearance as measured by the K rate, from ~0.3 (n = 5) in diabetic rats to 2.2 f 0.4 (mean + SE; n = 10) in rats with functioning pig xenograft in the KC. The K rate for normal controls was 2.9 k 0.3 (n = 8). Significant levels of insulin were present in the AC1 recipients with functional pig islet xenograft under the KC and this returned to basal levels present in the diabetic controls following graft rejection. Fig. 5 shows that islet tissue at the time of the graft removal was easily recognizable in the recipients treated with 5 doses of ATS. In some sections, both CD4 + and CD8 + cells are numerous in the vicinity of the graft whereas in others they are totally absent. One common feature is the fibrosis seen surrounding the graft tissue. In contrast, in recipients without any
Use of xenogeneic tissue for transplantation presents a solution for the shortage of donor islets in clinical transplantation. Pig islet is a potential donor candidate because of the similarity in molecular structure and biological action of pig and human insulin. Therefore once the islet isolation is perfected an unlimited tissue supply would become available for preclinical testing. However, despite recent reports on the successful islet isolation from adult pigs and subsequent transplantation in mice [ 14-161, comparable studies in rats are lacking. Our results demonstrated that pig islets isolated from sows were functional in vivo in the amelioration of diabetes in AC1 rat recipients. Even though extended function was not observed, the results would suggest that pig islet xenograft is a feasible approach as ATS alone can improve the graft survival time to a significantly longer duration than that of untreated controls. Even though abdominal testis has been shown to be an effective immunoprivileged site for islet transplantation in both allograft and xenograft models [ 17,101, our results with the pig-rat xenograft model were not so favourable. The immunosuppressive effect of ATS was superior to that of the cyclosporin A regimen used, which was very effective in prolonging islet allotransplantation [ 121. With a combination of immunosuppressive agents such as FK506 and/or islet pretreatment by UV irradiation or low temperature prior to transplantation, it is likely that extended survival could be achieved. Pig islets do not survive well when cultured in media suitable for rat and human islets. Therefore, improvement in the culture condition is needed before transplantation studies with lowtemperature cultured islets become feasible. IVGTT in rats with functional pig xenograft yielded an improved K rate over the diabetic controls.
Fig. 5. Histological appearance of pig islet xenograft under the kidney capsule of a diabetic AC1 rat immunosuppressed with ATS, with normogiycemia, 14 days after transplantation. No lymphoid cell infiltration is obvious and the islets are well vascularized (magnification x 660).
Insulin data confirmed that the pig islets responded to glucose stimulation. Furthermore, negligible levels of insulin were detected in the same recipients following islet graft rejection. The present study also demonstrated that the quantity of pig islets required to achieve normoglycemia in diabetic rat recipients is much more than that with rat and human islets. In the preliminary transplantations, diabetic rats transplanted with 1000 and 2000 islets had transient lowering of BG and only those given ~4000 islets had a normalization of BG. The reason for this is likely to be multifactorial. One reason could be the lower functional activity of pig islets and/or porcine insulin in rats. The pig islets vary in size and were less compact than rat or human islets. The
fact that 2000 pig islets have a 1 g sedimentation volume of x 30 ~1, equivalent to that of 1500 rat islets that we used for our previous successful islet allotransplantation studies, would support the view that sufficient pig islets were transplanted. As a large number of adult pig islets are required for a successful rat xenotransplant, the purity of islet tissue for transplantation becomes a critical issue. Inclusion of excessive non-endocrine tissue in the graft would result in a poor environment for the transplanted islets, especially in the kidney capsule, and increases the chance of cellular death due to non-immunologic mechanisms. Pig islets can be purified by several gradient centrifugation procedures available including Ficoll and Percoll. However, we favour the Percoll over the
203
Ficoll gradient. Percoll solutions used were isoosmotic whereas Ficoll solutions at the SG used for gradient centrifugation were hyperosmotic. Also batches of Ficoll have been found to contain ingredients toxic to islets [ 181. In one experiment, none of the rats transplanted with adequate numbers of Ficoll purified islets in the KC had normoglycemia for > 2 days despite the fact that the islet preparation was pure and viable at the time of transplantation. Some subtle changes to the pig islets is likely to have occurred during the purification process and when compounded by the trauma exerted on them during the transplantation procedure resulted in the death of a large number of endocrine cells, thus resulting in the graft being non-functional. Since pig islets are very fragile and fragment easily during the isolation and purification process, further improvements of the currently available methods are needed to overcome this obstacle. In the morphological and immunohistochemical studies, CD4 + and CD8 + lymphoid cells were observed as early as day 3 following islet transplantation in rats without immunosuppression. In diabetic rat recipients given ATS for 5 days, at the time of the removal of the graft, the islets were still well preserved and stained positively for insulin and glucagon. The fact that the host lymphoid cells were not detected within the islets would suggest that direct contact of the host cells with the graft tissue was not essential for the killing of the target cells. The fact that ATS significantly prolonged the graft survival would indicate the importance of T lymphocytes in the rejection process. As CD4 + and CD8 + cells were present at the graft area shortly after transplantation, their relative contribution towards target killing remains to be delineated. In the rat-mouse and fetal pig proislet-mouse xenotransplantation models there was an indication that CD4 + cells were responsible for the rejection process [ 7,19,20]. Depletion of CD4 + T cells resulted in prolonged xenograft survival. In the 3 rats that died on days 10, 12 and 14, the animals had a BG below 100 mg/dl. Histological examination of kidneys revealed normal appearing islets at the
graft site, with little lymphoid cell infiltration. This finding suggests that these 3 islet xenografts were not rejected and could have functioned for a much longer period of time. The cause of death was likely the result of complications of ATS injections. Preformed antibodies present in the xenograft recipients in most instances contribute towards the hyperacute rejection of the graft [ 51. In the present study the role played by these preformed antibodies in the rejection process is unclear as no immediate loss of islet graft function was observed even in recipients without immunosuppression. Also, we failed to observe any correlation between the functional period of the graft and natural antibody titers of the pig islet recipients. Further studies are needed to define the participation of these antibodies in the rejection of the long-term functioning xenograft. The present finding that prolongation of survival of discordant transplants of pig islets in diabetic rats can be achieved without continuous immunosuppression is encouraging. The discordant xenotransplantation model would be useful for testing the efficacy of new immunosuppressive agents.
Acknowledgements This work was supported by the Medical Research Council of Canada and the Lam Foundation. We thank Dr. N. Bissada and Mr. W. Yep for technical assistance, and Ms. G. Chiew for secretarial help.
References 1 Tze, W.J. and Tai, J. (1988) Allotransplantation of dispersed single pancreatic endocrine cells in diabetic rats. Diabetes 37, 383-392. 2 Sima, A.A.F., Zang, W.-X., Tze, W.J., Tai, J. and Nathaniel, V. (1988) Diabetic neuropathy in the streptozotocin diabetic rat and the effect of allogeneic islet cell transplantation. A morphometric analysis. Diabetes 37, 1129-1136.
204 3 Bobzien, B., Yasunami, Y., Majercik, M., Lacy, P.E. and Davie, J.M. (1983) Intratesticular transplants of islet xenografts (rat to mouse). Diabetes 32, 213-216. 4 Sullivan, F.P., Ricordi, C., Hauptfeld, V. and Lacy, P.E. (1987) Effect oflow-temperature culture and site of transplantation on hamster islet xenograft survival (hamster to mouse). Transplantation 44, 465-468. 5 Auchincloss, H., Jr. (1988) Xenogeneic transplantation. A review. Transplantation 46, I-20. 6 Ricordi, C. and Lacy, P.E. (1987) Renal subcapsular xenotransplantation of purified porcine islets. Transplantation 44, 721-723. 7 Simeonovic, C.J., Ceredig, R. and Wilson, J.D. (1989) Reversal of diabetes in CD4 + T cell-depleted mice by xenotransplantation of pig proislets. Transplant. Proc. 21, 3811-3812. 8 Gray, D.W.R., McShane, P., Grant, A. and Morris, P.J. (1984) A method for isolation of islets of Langerhans from the human pancreas. Diabetes 33, 1055-1061. 9 Latif, Z.A., Noel, J. and Alejandro, R. (1988) A simple method of staining fresh and cultured islets. Transplantation 45, 827-830. 10 Selawry, H.P., Whittington, K.B. and Forster, H.G. (1988) Intratesticular islet xenograft survival in relation to tissue cyclosporine levels. Am. J. Med. Sci. 294, 497-502. 11 Davis, R.C., Cooperband, S.R. and Mannick, J.A. (1969) Preparation and in vitro assay of effective and ineffective antilymphocyte sera. Surgery 66, 58-64. 12 James. R.F., Lake, S.P., Chamberlain, J. et al. (1989)
13
14
15
16 17
18
19
20
Gamma irradiation of isolated rat islets pretransplantation produces indefinite allografi survival in cyclosporine-treated recipients. Transplantation 47,929-933. Shons, A.R. and Najarian, J.S. (1974) Xenograft rejection mechanisms in man. Trans. Am. Sot. Artif. Int. Organs 20, 562-568. Ricordi, C., Finke, E.H. and Lacy, P.E. (1986) A method for the mass isolation of islets from adult pig pancreas. Diabetes 35, 649-653. Calafiore, R., Calcinaro, F., Basta, G. et al. (1990) A method for the massive separation of highly purified adult porcine islets of Langerhans. Metabolism 39, 175-181. Ricordi, C., Socci, C., Davalli, A.M. et al. (1990) Isolation of the elusive pig islet. Surgery 107, 688-694. Bellgrau, D. and Selawry, H.P. (1990) Cyclosporineinduced tolerance to intratesticular islet xenografts. Transplantation 50, 654-657. Vonen, B., Florholmen, J., Giaever, A.K. and Burhol, P.G. (1987) A methodological study of discontinuous Percoll-gradient separation of pancreatic islets from rats. Stand. J. Clin. Lab. Invest. 47, 415-420. Kawai, M., Gotoh, M., Monden, M. et al. (1989) Effect of L3T4 and Lyt2 monoclonal antibodies on islet xenograft (rat to mouse) rejection. Transplant. Proc. 21, 2709-2710. Simeonovic, C.J., Wilson, J.D. and Ceredig, R. (1990) Antibody-induced rejection of pig proislet xenografts in CD4 + T cell-depleted diabetic mice. Transplantation 50, 657-662.
Science Publishers
197-204
197
B.V. All rights reserved 0168-8227/92/$05.00
DIABET 00603
Transplantation
of discordant pig islet xenografts in diabetic rats
Wah Jun Tze, Joseph Tai and Sze-Shuen Departments of Pediatrics and Patholoa.
Cheung
University of British Columbia, Vancouver, British Columbia. Canada
(Received 8 April 1991) (Revision accepted 4 October 1991)
Summary Xenotransplantation of pig islets under the kidney capsule (KC) of diabetic rats was performed. Natural preformed AC1 rat anti-pig leukocytotoxicity, leukoagglutination and hemagglutination antibody titers ranged from Neat-l : 16, 1 : 8-1 : 32 and 1 : 128-1: 256, respectively (n = 14). Normal AC1 sera were non-toxic to pig islets during short term incubation. Pig islet xenograft survival times in the nonimmunosuppressed AC1 rats, AC1 rats immunosuppressed with antithymocyte serum (ATS) or cyclosporin A were 3.8 k 0.4 (mean + SE; n = 5), 10.4 & 0.7 (n = 13) and 6.0 k 1.0 (n = 5) days, respectively. Pig islets implanted in the abdominal testis of AC1 recipients immunosuppressed with 5 doses ATS survived for a mean of 6.4 ? 1.O days (n = 7). The mean K rate following an intravenous glucose tolerance test (IVGTT) in AC1 rats 1 week after transplantation with pig islet under the KC was 2.2 k 0.4 (n = 10) compared to that of 2.91 + 0.30 found in normal control rats (n = 8). Peak insulin at 1 min was 60.1 + 3.9 pU/ml (n = 4). Histological and immunohistochemical examination showed that the xenograft from recipients treated with 5 doses of ATS still contained well-preserved islet tissue with many insulinand glucagon-containing cells on the day of graft removal when blood glucose had returned to hyperglycemic level. Both CD4 and CD8 positive cells were in the vicinity of the graft tissue. This study has shown that hyperacute rejection of islet xenograft did not occur in this discordant xenograft model despite the presence of preformed anti-pig antibodies in the rat recipients. Further islet xenotransplantation using this discordant model would be useful in the design of an effective immunosuppressive regimen. Key words: Pig; Rat; Islet; Xenotransplantation;
Correspondence
to: W.J. Tze, Depts. of Pediatrics
and Patholof British Columbia, 4480 Oak Street, Vancouver, B.C., Canada V6H 3V4. Abbreviations: ATS, antithymocyte serum; BSA-HBSS, Hanks’ balanced salt with 0.4% bovine serum albumin; IVGTT, intravenous glucose tolerance test; KC, kidney capsule; K rate, glucose disappearance rate; STZ, streptozotocin.
Streptozotocin;
Diabetes
Introduction
ogy, University
Islet allotransplantation into the diabetic rat has been shown to normalize the carbohydrate metabolism as well as preventing secondary diabetic complications in the recipients [ 1,2]. Although promising results have been achieved by islet allo-
198
transplantation into diabetic animals, a corresponding approach to human therapy is hampered by the limited supply of human islet tissue. Islet xenograft is an appealing alternative to islet allograft because this approach would enable the use of an unlimited supply of large animal donor tissue for transplantation in human diabetics. In spite of the potential, islet xenograft has been the subject of relatively few studies. The majority of the islet xenotransplantations were performed between donor-recipient of closely related species such as rat to mouse or hamster to mouse [ 3,4]. Relatively few islet xenotransplantation studies were reported between widely disparate combinations of animals [5]. Pig islet is a potential donor tissue for xenotransplantation in humans. Earlier, Ricordi reported xenotransplantation of adult pig islet in diabetic mice [6]. Simeonovic et al. [ 71 demonstrated that xenografts of fetal pig proislets reversed STZ-induced diabetes and achieved long-term normoglycemia > 139-168 days in CD4-depleted recipient mice. Presently, there is little information on the metabolic function of adult pig islet xenograft in diabetic rat recipients. The present study assessed the natural anti-pig antibodies in the AC1 rats and the survival time and histology of adult pig islet xenografts in the discordant STZ-induced diabetic recipients.
Materials and Methods Animals
Male AC1 rats (Harlan SD, Indianapolis, IN) of 200-225 g body weight were used as recipients. Diabetes was induced by a single intravenous injection of 55 mg/kg body weight of STZ and they were considered as diabetic when their blood glucose (BG) was >400 mg/dl for at least 10 days. Preparation of pig islets
Pig pancreatic islets were isolated from sow pancreata obtained from a local abattoir. The warm and cold ischemic times were < 20 min and 2 h,
respectively. Pig pancreatic islets were isolated with a modified technique of Gray et al. [ 81. The pancreas was perfused with Type XI collagenase (0.75-1.0 mg/ml, 1 ml/g tissue; Sigma Chemical Co. St. Louis, MO) in Hanks’ balanced salt solution supplemented with l”/ bovine serum albumin and digested for between 15-20 min at 37 oC. The digested gland was then placed in cold Hanks’ balanced salt solution with 0.4% bovine serum albumin (BSA-HBSS) and chopped with fine scissors and teased with blunt forceps to release the islets. The islets and non-islet fragments were washed with BSA-HBSS and sedimented at unit gravity repeatedly to remove the exocrine enzymes. After overnight culture at room temperature, islets were purified by multilayered Percoll gradient (SG 1.020, 1.045 and 1.068). Islet tissue was resuspended in Percoll with SG 1.068 and then overlayered with Percoll with SG 1.045 and 1.020. After 20 min of equilibration at unit gravity, islets were collected from the layer with SG 1.045. Their identity was confirmed by staining with dithizone [ 91. The contaminant consisted mainly of ductal tissue. For an in vitro functional study, 100 islets (150-250 pm in diameter) hand-picked from the Percoll purified preparation were tested for their response to glucose stimulation in a perifusion system which consisted of a 13 mm Swinnex filter unit. BSAHBSS at 37°C was introduced into the chamber at 1 ml/min with a peristaltic pump. Glucose concentrations of the BSA-HBSS were 100 mg/dl between O-60 min, 121-180 min and 450 mg/dl between 6 l-120 min. Samples collected were stored at - 20 ‘C until insulin assay. Xenotransplantation and immunosuppression
Between 1000-4000 pig islets (150-250 pm in diameter; 1 g sedimentation volume of 4000 islets, equivalent to = 50-80 ~1 in a 50 U insulin syringe) were transplanted into each STZinduced diabetic rat under the KC or into the abdominal testis [ lo]. For immunosuppression with ATS, 1 ml of the antiserum prepared in our laboratory [ 111, was given intraperitoneally immediately after islet transplantation and daily for
199
the 4 subsequent days. Cyclosporin A was given subcutaneously at 30 mg/kg/day for 3 consecutive days following transplantation [ 121. Histology and immunohistochemical
study
BG levels were checked
daily until the day of removal of the kidney containing the graft, that is, when BG of the recipient rose above 200 mg/dl on 2 consecutive days. Equal halves were snap frozen in LN2 or fixed in Bouin’s fixative. Paraffin sections were stained for insulin, glucagon and somatostatin with immunoperoxidase staining and frozen (ABC Staining Kits, Vectastain) sections for cellular markers with monoclonal antibodies against leukocyte common antigen (0X1), class II antigen (0X6), CD4 (W3/25) and CD8 (0X8) (Cedarlane Lab., Hornby, Ontario). IVGTT (1 g/kg body weight) was performed on 8 normal rats, 5 diabetic rats and 10 diabetic AC1 rats with functioning pig islet xenograft under the KC 7 days after transplantation. During IVGTT, water was allowed but not food. The rates of glucose disappearance (K rate) were calculated. Insulin was determined by radioimmunoassay using pig insulin as standard (Pharmacia Diagnostics, Montreal, Quebec). Determination of preformed
rat anti-pig antibodies
Natural occurring leukocytotoxicity, leukoagglutination and hemagglutination titers in the sera of normal AC1 rats were assessed on pig WBC and RBC [ 131. All rat sera were heat inactivated for 30 min at 56°C. For assessing the toxic effect of normal rat sera on pig islets, groups of 20 pig islets were placed in wells of 96-well microtiter plates in the presence of 100 ~1 of rat serum. They were examined after 1 h of incubation at room temperature. Rabbit complement (100 ,uI; 1 : 8 dilution, Low-Tox Complement, Cedarlane Lab, Hornby, Ontario) was added for another hour and then the islets were re-examined. Results The pig islets used for the transplantation are less compact than those of either humans or rats. The
major contaminant of the preparation is the ductal elements. Fig. 1 shows that all preparations of the isolated pig islets secreted basal insulin. However, there is variation among the different islet preparations in their response to glucose stimulation. Table 1 shows that natural rat anti-pig antibodies were detected in all normal AC1 animals before transplantation (n = 14). These antibodies were non-toxic to the pig islets during short-term incubation. Table 2 and Fig. 2 show the functional period
p
I
50
-
8
40
-
i 2
30
-
$E
20
-
B
10
LOW
-HIQHl
LOW 60
I
I
T
I
/
I
I
1
I
H
zl 5
‘a.lI
t
01 0
’
20
’
40
’
60
’
80
’
100
’
120
’
140
’
160
I
180
’
200
Minutes
Fig. 1. Insulin secretion kinetics of groups of 100 pig islets during perifusion with BSA-HBSS containing 100 mg/dl 450 mg/dl glucose 121-180 min) and (O-60 min, (61-120 min; mean f SE; n = 10). TABLE 1 Leukocytotoxicity, leukoagglutination, hemagglutination and islet toxicity titers of normal AC1 rat sera against pig cells Antibody titers* Leukocytotoxicity Leukoagglutination Hemagglutination Islet toxicity
N (9)#; I:8 (3); 1:16 (2) I:8 (I); 1:16 (11); 1:32 (2) 1:128 (1); I:256 (13)
* For leukocytotoxicity titers, the end point was the last well showing 50% lysis. For leuko- and hemagglutination assays titer was taken as the last well showing agglutination. # The number in the bracket indicates the sample size. + Minimal cytotoxic effect was observed for normal AC1 sera. Positive control sera from 4 AC1 rats sensitized with pig cells showed lysis of up to 507, of the islets at I:4 dilution.
200 TABLE 2 Functional
period of pig islet xenograft in diabetic AC1 rats
KC” KC” AT,’ KC KC’ KC“ ~_ ._~~~.
1
2 3 4 5 6 _
Immunosuppression
Survival days
(n) 5 13 7 6 6 7
None ATS x ATS x CsA x None ATS x
3,3,4,4,5 5,7,8,10 x 3,10b,ll 44557 7 , , 7 . 10,lO 4.5,5,5,6,11 < 2 days <2 days
Number
Transplant site
Group
5 5 3 5
Mean + SE’
3.8 > 10.4 6.4 6.0
x 2,12b,13,14,14b
f f + +
0.4 0.7 1.0 1.0
.I Transplanted with > 4000 Percoll purified islets under the kidney capsule (KC) or abdominal testis (AT); n = number of recipients. h Graft functional at time of death. ’ Transplanted with > 4000 Ficoll purified islets under the kidney capsule (KC). ” Transplanted with ~2000 Percoll purified islets under the kidney capsule. ’ Group 2 vs 1, P < 0.003, calculated by the Mann-Whitney U test; Group 3 vs I, P < 0.04; Group 2 vs 3, P < 0.03.
g
500
-
\, E
400
-
E: 300 Z a E 200
-
x
E @I
-
100
-
0' -2
1
I
I
I
I
L
0
2
4
0
8
10
I
12
I
14
I 16
was 6.4 t 1.0 days (Group 3). With cyclosporin A as immunosuppressant, the functional period was 6.0 + 1.0 days (Group 4). Group 5 recipients each received x 4000 pig islets purified by Ficoll gradient in the KC. Only transient lowering of BG for <2 days was observed in all animals. In the recipients of ~2000 islets (Group 6) the lowering of BG was transient for < 2 days. Fig. 3 shows that there is no correlation
DAY
Fig. 2. Non-fasting serum glucose (mean F SE, n = 13) of diabetic AC1 rats transplanted with Percoll purified pig islets under the kidney capsule, immunosuppressed with ATS ( 1 ml/day, day O-4).
of xenotransplanted pig islets in diabetic rats. BG levels returned to normoglycemic state in diabetic recipients of islets under the KC in l-2 days. In the non-immunosuppressed recipients transplanted with > 4000 pig islets, the survival period was 3.8 + 0.4 days (Group 1). Significantly prolonged function of the graft with serum glucose levels maintained at < 200 mg/dl for more than 10 days was seen in lo/13 rats that received 5 doses of ATS after the islet xenograft (Group 2). When similar numbers of islets were transplanted into the abdominal testis of diabetic recipients given 5 doses of ATS, graft survival
.
30 f F
.
.
3
20 I 10
1
.
a
l
.
.
I
l
.
.
.
I
Fig. 3. Duration of pig islet xenograft function in diabetic serum hemagglutination (o ) and AC1 rats and leukoagglutination titers (0) at the time of pig islet xenotransplantation.
201
immunosuppression the graft was mostly fibrotic and only occasional insulin-containing cells were seen by day 5 after transplantation.
Discussion 80
I
-20
0
20
40
60
60
100
120
140
Fig. 4. Serum glucose (mean + SE; A) and insulin (B) levels following IVGTT (1 g/kg body weight) in overnight fasted normal AC1 rats (o-----o; n = 8); diabetic AC1 rats (A-----A ; n = 7), diabetic AC1 rats with functioning pig islet xenograft under the kidney capsule, treated with 5 doses of ATS, 1 week after transplantation (V-----V; n = 10 and 4 for glucose and insulin), and 1 week after rejection of pig islet xenograft (A-----A ; n = 4).
between the pig islet functional period and the anti-pig hemagglutination and leukoagglutination titers in the AC1 recipients at the time of transplantation. Fig. 4 shows that successful pig islet xenograft resulted in a significant improvement in the glucose disappearance as measured by the K rate, from ~0.3 (n = 5) in diabetic rats to 2.2 f 0.4 (mean + SE; n = 10) in rats with functioning pig xenograft in the KC. The K rate for normal controls was 2.9 k 0.3 (n = 8). Significant levels of insulin were present in the AC1 recipients with functional pig islet xenograft under the KC and this returned to basal levels present in the diabetic controls following graft rejection. Fig. 5 shows that islet tissue at the time of the graft removal was easily recognizable in the recipients treated with 5 doses of ATS. In some sections, both CD4 + and CD8 + cells are numerous in the vicinity of the graft whereas in others they are totally absent. One common feature is the fibrosis seen surrounding the graft tissue. In contrast, in recipients without any
Use of xenogeneic tissue for transplantation presents a solution for the shortage of donor islets in clinical transplantation. Pig islet is a potential donor candidate because of the similarity in molecular structure and biological action of pig and human insulin. Therefore once the islet isolation is perfected an unlimited tissue supply would become available for preclinical testing. However, despite recent reports on the successful islet isolation from adult pigs and subsequent transplantation in mice [ 14-161, comparable studies in rats are lacking. Our results demonstrated that pig islets isolated from sows were functional in vivo in the amelioration of diabetes in AC1 rat recipients. Even though extended function was not observed, the results would suggest that pig islet xenograft is a feasible approach as ATS alone can improve the graft survival time to a significantly longer duration than that of untreated controls. Even though abdominal testis has been shown to be an effective immunoprivileged site for islet transplantation in both allograft and xenograft models [ 17,101, our results with the pig-rat xenograft model were not so favourable. The immunosuppressive effect of ATS was superior to that of the cyclosporin A regimen used, which was very effective in prolonging islet allotransplantation [ 121. With a combination of immunosuppressive agents such as FK506 and/or islet pretreatment by UV irradiation or low temperature prior to transplantation, it is likely that extended survival could be achieved. Pig islets do not survive well when cultured in media suitable for rat and human islets. Therefore, improvement in the culture condition is needed before transplantation studies with lowtemperature cultured islets become feasible. IVGTT in rats with functional pig xenograft yielded an improved K rate over the diabetic controls.
Fig. 5. Histological appearance of pig islet xenograft under the kidney capsule of a diabetic AC1 rat immunosuppressed with ATS, with normogiycemia, 14 days after transplantation. No lymphoid cell infiltration is obvious and the islets are well vascularized (magnification x 660).
Insulin data confirmed that the pig islets responded to glucose stimulation. Furthermore, negligible levels of insulin were detected in the same recipients following islet graft rejection. The present study also demonstrated that the quantity of pig islets required to achieve normoglycemia in diabetic rat recipients is much more than that with rat and human islets. In the preliminary transplantations, diabetic rats transplanted with 1000 and 2000 islets had transient lowering of BG and only those given ~4000 islets had a normalization of BG. The reason for this is likely to be multifactorial. One reason could be the lower functional activity of pig islets and/or porcine insulin in rats. The pig islets vary in size and were less compact than rat or human islets. The
fact that 2000 pig islets have a 1 g sedimentation volume of x 30 ~1, equivalent to that of 1500 rat islets that we used for our previous successful islet allotransplantation studies, would support the view that sufficient pig islets were transplanted. As a large number of adult pig islets are required for a successful rat xenotransplant, the purity of islet tissue for transplantation becomes a critical issue. Inclusion of excessive non-endocrine tissue in the graft would result in a poor environment for the transplanted islets, especially in the kidney capsule, and increases the chance of cellular death due to non-immunologic mechanisms. Pig islets can be purified by several gradient centrifugation procedures available including Ficoll and Percoll. However, we favour the Percoll over the
203
Ficoll gradient. Percoll solutions used were isoosmotic whereas Ficoll solutions at the SG used for gradient centrifugation were hyperosmotic. Also batches of Ficoll have been found to contain ingredients toxic to islets [ 181. In one experiment, none of the rats transplanted with adequate numbers of Ficoll purified islets in the KC had normoglycemia for > 2 days despite the fact that the islet preparation was pure and viable at the time of transplantation. Some subtle changes to the pig islets is likely to have occurred during the purification process and when compounded by the trauma exerted on them during the transplantation procedure resulted in the death of a large number of endocrine cells, thus resulting in the graft being non-functional. Since pig islets are very fragile and fragment easily during the isolation and purification process, further improvements of the currently available methods are needed to overcome this obstacle. In the morphological and immunohistochemical studies, CD4 + and CD8 + lymphoid cells were observed as early as day 3 following islet transplantation in rats without immunosuppression. In diabetic rat recipients given ATS for 5 days, at the time of the removal of the graft, the islets were still well preserved and stained positively for insulin and glucagon. The fact that the host lymphoid cells were not detected within the islets would suggest that direct contact of the host cells with the graft tissue was not essential for the killing of the target cells. The fact that ATS significantly prolonged the graft survival would indicate the importance of T lymphocytes in the rejection process. As CD4 + and CD8 + cells were present at the graft area shortly after transplantation, their relative contribution towards target killing remains to be delineated. In the rat-mouse and fetal pig proislet-mouse xenotransplantation models there was an indication that CD4 + cells were responsible for the rejection process [ 7,19,20]. Depletion of CD4 + T cells resulted in prolonged xenograft survival. In the 3 rats that died on days 10, 12 and 14, the animals had a BG below 100 mg/dl. Histological examination of kidneys revealed normal appearing islets at the
graft site, with little lymphoid cell infiltration. This finding suggests that these 3 islet xenografts were not rejected and could have functioned for a much longer period of time. The cause of death was likely the result of complications of ATS injections. Preformed antibodies present in the xenograft recipients in most instances contribute towards the hyperacute rejection of the graft [ 51. In the present study the role played by these preformed antibodies in the rejection process is unclear as no immediate loss of islet graft function was observed even in recipients without immunosuppression. Also, we failed to observe any correlation between the functional period of the graft and natural antibody titers of the pig islet recipients. Further studies are needed to define the participation of these antibodies in the rejection of the long-term functioning xenograft. The present finding that prolongation of survival of discordant transplants of pig islets in diabetic rats can be achieved without continuous immunosuppression is encouraging. The discordant xenotransplantation model would be useful for testing the efficacy of new immunosuppressive agents.
Acknowledgements This work was supported by the Medical Research Council of Canada and the Lam Foundation. We thank Dr. N. Bissada and Mr. W. Yep for technical assistance, and Ms. G. Chiew for secretarial help.
References 1 Tze, W.J. and Tai, J. (1988) Allotransplantation of dispersed single pancreatic endocrine cells in diabetic rats. Diabetes 37, 383-392. 2 Sima, A.A.F., Zang, W.-X., Tze, W.J., Tai, J. and Nathaniel, V. (1988) Diabetic neuropathy in the streptozotocin diabetic rat and the effect of allogeneic islet cell transplantation. A morphometric analysis. Diabetes 37, 1129-1136.
204 3 Bobzien, B., Yasunami, Y., Majercik, M., Lacy, P.E. and Davie, J.M. (1983) Intratesticular transplants of islet xenografts (rat to mouse). Diabetes 32, 213-216. 4 Sullivan, F.P., Ricordi, C., Hauptfeld, V. and Lacy, P.E. (1987) Effect oflow-temperature culture and site of transplantation on hamster islet xenograft survival (hamster to mouse). Transplantation 44, 465-468. 5 Auchincloss, H., Jr. (1988) Xenogeneic transplantation. A review. Transplantation 46, I-20. 6 Ricordi, C. and Lacy, P.E. (1987) Renal subcapsular xenotransplantation of purified porcine islets. Transplantation 44, 721-723. 7 Simeonovic, C.J., Ceredig, R. and Wilson, J.D. (1989) Reversal of diabetes in CD4 + T cell-depleted mice by xenotransplantation of pig proislets. Transplant. Proc. 21, 3811-3812. 8 Gray, D.W.R., McShane, P., Grant, A. and Morris, P.J. (1984) A method for isolation of islets of Langerhans from the human pancreas. Diabetes 33, 1055-1061. 9 Latif, Z.A., Noel, J. and Alejandro, R. (1988) A simple method of staining fresh and cultured islets. Transplantation 45, 827-830. 10 Selawry, H.P., Whittington, K.B. and Forster, H.G. (1988) Intratesticular islet xenograft survival in relation to tissue cyclosporine levels. Am. J. Med. Sci. 294, 497-502. 11 Davis, R.C., Cooperband, S.R. and Mannick, J.A. (1969) Preparation and in vitro assay of effective and ineffective antilymphocyte sera. Surgery 66, 58-64. 12 James. R.F., Lake, S.P., Chamberlain, J. et al. (1989)
13
14
15
16 17
18
19
20
Gamma irradiation of isolated rat islets pretransplantation produces indefinite allografi survival in cyclosporine-treated recipients. Transplantation 47,929-933. Shons, A.R. and Najarian, J.S. (1974) Xenograft rejection mechanisms in man. Trans. Am. Sot. Artif. Int. Organs 20, 562-568. Ricordi, C., Finke, E.H. and Lacy, P.E. (1986) A method for the mass isolation of islets from adult pig pancreas. Diabetes 35, 649-653. Calafiore, R., Calcinaro, F., Basta, G. et al. (1990) A method for the massive separation of highly purified adult porcine islets of Langerhans. Metabolism 39, 175-181. Ricordi, C., Socci, C., Davalli, A.M. et al. (1990) Isolation of the elusive pig islet. Surgery 107, 688-694. Bellgrau, D. and Selawry, H.P. (1990) Cyclosporineinduced tolerance to intratesticular islet xenografts. Transplantation 50, 654-657. Vonen, B., Florholmen, J., Giaever, A.K. and Burhol, P.G. (1987) A methodological study of discontinuous Percoll-gradient separation of pancreatic islets from rats. Stand. J. Clin. Lab. Invest. 47, 415-420. Kawai, M., Gotoh, M., Monden, M. et al. (1989) Effect of L3T4 and Lyt2 monoclonal antibodies on islet xenograft (rat to mouse) rejection. Transplant. Proc. 21, 2709-2710. Simeonovic, C.J., Wilson, J.D. and Ceredig, R. (1990) Antibody-induced rejection of pig proislet xenografts in CD4 + T cell-depleted diabetic mice. Transplantation 50, 657-662.