Marginal Transplantation Dose Study for the Treatment of Insulin-independent Type 1 Diabetes After Allogeneic Islet Transplantation in Macaca Fascicularis Monkeys


 Many groups are working to improve the results of clinical allogeneic islet transplantation in a primate model. However, few studies have focused on the optimal islet dose for achieving normal glycemia without exogenous insulin after transplantation in primate models or on the relationship between rejection and islet amyloid polypeptide (IAPP) expression. We evaluated the dose (group 1/10,000, group 2/20,000, and group 3/>25,000 islet equivalents (IEQ)/kg) needed to achieve normal glycemia without exogenous insulin after transplantation using eleven cynomolgus monkeys, and we analyzed the characteristics exhibited in the islets after transplantation. Group 1 (n2) failed to control blood glucose level, despite injection with the highest dose of exogenous insulin, and group 2 (n5) achieved unstable control, with a high insulin requirement. However, group 3 (n4) achieved normal glycemia without exogenous insulin and maintained it for more than 60 days. Immunohistochemistry results from staining islets found in liver biopsies indicated that as the number of transplanted islets decreased, the amount of IAPP accumulation within the islets increased, which accelerated immune cell infiltration. In conclusion, the marginal transplantation dose for achieving a normal glycemia without exogenous insulin in our cynomolgus monkey model was >25,000 IEQ/kg, and the accumulation of IAPP early after transplantation, which depends on the transplanted islet dose, can be considered one factor in rejection.


Introduction
Islet transplantation is a feasible therapeutic option for patients with type 1 diabetes mellitus (T1DM) that can offer optimal glycemic control and prevent severe hypoglycemia. The Edmonton protocol demonstrated that transplanting more than 10,000 islet equivalents per kilogram of body weight (IEQ/kg) cured T1DM patients, although all patients required islets from two deceased-donor pancreases 1 . Clinical islet transplantation faces several remaining challenges, including a shortage of human donors. The current technique for islet isolation cannot obtain the maximum number of islets available from each pancreas donor 2,3 . After intra-portal infusion, a considerable number of islets is lost to an instant bloodmediated in ammatory reaction [4][5][6] . In addition, immunosuppressive drugs applied after islet transplantation can be diabetogenic 7,8 . To overcome those clinical obstacles, researchers have used nonhuman primates (NHPs) as surrogates for humans because of their evolutionary proximity, immune systems similar that of humans, and larger body sizes and longer lifespans, compared with other experimental animals [9][10][11][12][13] .
When studying allogeneic islet transplantation using NHPs, the number of islets required for insulin independence is important information for which limited evidence is available. Furthermore, the environment of transplanted islets and how it changes according to the infused islet dose should also be clari ed. Several studies of allogeneic islet transplantation in primates reported various results, depending on the choice of immunosuppressants. Studies using JAK3 inhibitors (3,900 to 12,500 IEQ/kg) 14 or anti-CD40 (9,361 to 32,387 IEQ/kg) 15,16 immunosuppression after allogeneic islet transplantation achieved normal blood glucose levels with or without a signi cant reduction in the amount of exogenous insulin required. In the results of a study that transplanted 20,000 IEQ/kg allogeneic islets, the group that received co-transplantation with bone marrow-derived spheroids (to improve transplant prognosis), achieved normal glucose levels without exogenous insulin and maintained it for a period. However, the control group that received only islets did not achieve a normal blood glucose level without exogenous insulin 17 . In our previous study, the surface manipulation of islets to improve the results of allogeneic islet transplantation (9,000 to 20,000 IEQ/kg) also did not produce insulin-independent blood glucose level normalization in the control groups (10,000 and 20,000 IEQ/kg) 18 . However, a study using an immunosuppressive protocol similar to the one we used in this study reported achieving a normal blood glucose level without exogenous insulin in the group that received 2-30,000 IEQ/kg 19 .
Meanwhile, various results have been made in determining the cause of T1DM, and islet amyloid polypeptide (IAPP) is one possibility. IAPP is well known for its role in type 2 diabetes. It is produced with insulin in the beta cells through the endoplasmic reticulum and Golgi and is an oligomer that exhibits intracellular toxicity, ultimately increasing apoptosis and destroying beta cells [20][21][22][23][24] . For this reason, the opinion that IAPP is the main cause of type 2 diabetes is highly convincing. IAPP has also received attention as a cause of beta-cell dysfunction in T1DM [25][26][27] , and some studies have reported a relationship between T1DM and IAPP [28][29][30][31] . However, within our knowledge, no studies have examined the correlation between IAPP and the dose of transplanted islets.
Our aim in the present study was to establish the optimal allogeneic islet transplantation dose to normalize the blood glucose level without requiring exogenous insulin in cynomolgus monkeys and to explain the correlation between the dose of transplanted islets and the IAPP expressed in those islets.

Results
Transplant-recipient monkeys. Eleven monkeys using an immunosuppressant regime successfully received allogeneic islets from eleven donor monkeys. Eight monkeys (A, B, C, D, H, I, J, and K) underwent rabbit anti-thymocyte globulin (ATG, Genzyme, Cambridge, MA) induction alone, and 3 monkeys (E, F, and G) in group 2 underwent combined anti-CD20 monoclonal antibody, rituximab (RTX, MabThera, Roche Pharma, Schweiz), and ATG induction (Supplementary Figure S1). The mean age and body weight at the time of transplantation were 50.5 (41-63) months and 3.7 (2.4-4.6) kg (Table 1). The mean dose of transplanted islets was 10,000 (group 1, n = 2), 19,900 (19,500-20,000) (group 2, n = 5), and 26,095 (25,000-27,000) (group 3, n = 4) IEQ/kg ( Fig. 1 and Table 1). Their levels of serum C-peptide were 2.1 ± 9.5 (monkey A) and 0.1 ± 0.0 (monkey B) ( Fig. 2A and B). Because of this, the experiment with monkey B was terminated 60 days after transplantation. Because monkey A continued to show high blood glucose levels despite exogenous insulin usage and had low serum C-peptide levels, as seen in monkey B, we terminated the experiment 87 days after transplantation ( Fig. 2A). Between one and two months after transplantation (POD 30-60), the ve monkeys in group 2 required 1.1 ± 0.1 (monkey C), 1.3 ± 0.1 (monkey D), 2.5 ± 0.1 (monkey E), 0.5 ± 0.0 (monkey F), and 1.2 ± 0.1 (monkey G) IU/kg/day of exogenous insulin to maintain of blood glucose levels near 200mg/dL. In the same period, their levels of serum C-peptide were 1.4 ± 0.1 (monkey C), 2.4 ± 0.2 (monkey D), 1.4 ± 0.4 (monkey E), 3.0 ± 0.2 (monkey F), and 1.8 ± 0.2 (monkey G) ng/ml (Fig. 2C, D, E, F, and G). 47 days after transplantation, monkey E expired unexpectedly and without signi cant changes in the relevant biological parameters. The group 3 monkeys required either no exogenous insulin (monkeys H and K) or a very low dose (monkey I used 0.2 IU/kg/day at 43 days after transplantation, and monkey J used 0.1 IU/kg/day at 36 and 46 days after transplantation) to maintain blood glucose levels below 100 mg/dL. Their levels of serum C-peptide were 3.3 ± 0.4 (monkey H), 2.2 ± 0.4 (monkey I), 2.5 ± 0.2 (monkey J), and 3.4 ± 0.3 (monkey K) ng/ml (Fig. 2H, I, J and K). Those results were maintained for three months (POD 90) after transplantation in each subject except monkey K. Interestingly, monkey K, who received a lobectomy (whole segment 2, left lateral lobe) at POD 60, showed dramatic changes in insulin independence, serum C-peptide level (decreased), and demand for exogenous insulin to maintain glucose level control after lobectomy (Fig. 2K). Intravenous glucose tolerance testing (IVGTT) 1 month after transplantation showed that the group 3 monkeys had signi cantly better blood glucose levels (Fig. 3A) and serum C-peptide levels ( Fig. 3B) than the monkeys in the other groups. Those results indicate that a dose of at least 25,000 IEQ/kg is required to achieve normal blood glucose levels without exogenous insulin after transplantation.
Liver biopsy results after transplantation. One month after transplantation, group 1 was unavailable for biopsy because of health problems caused by uncontrolled blood glucose levels. In the same period, the group 2 and 3 monkeys were biopsied, and 8 and 5 liver tissue samples were obtained, respectively. Two months after transplantation, 3 liver tissue samples were obtained from monkey A. Monkey B was terminated in this period, in accordance with our criteria, and 49 liver tissue samples (whole segment 2, left lateral) were obtained. In the same period, 8 and 9 liver tissue samples were biopsied from groups 2 and 3, respectively. No islets were found in the tissues biopsied from group 1 two months after transplantation. One month after transplantation, 11 and 17 islets were found in samples from groups 2 and 3, respectively. Two months after transplantation, 59 and 39 islets were found in samples from groups 2 and 3, respectively. Information from the liver biopsies performed one and two months after transplantation is summarized in Table 2.
These results indicate that as the amount of IAPP expressed within the islets increased, especially early after transplantation, the number of dysfunctional islets increased, and the level of IAPP expression increased as the number of transplanted islets decreased.

Discussion
Our present study demonstrates that > 25,000 IEQ/kg is the islet dose su cient to achieve an exogenous insulin-independent normal blood glucose level after allogeneic islet transplantation when using a recent clinical immunosuppression protocol. The group that received 20,000 IEQ/kg required exogenous insulin to maintain blood glucose levels close to the reference value. This result is consistent with previous studies. The group that received 10,000 IEQ/kg had to be terminated from the experiment 2 months after transplantation to meet our criteria because of their high level of exogenous insulin demand after transplantation and unregulated blood glucose levels despite supplemental insulin. Upon termination, we found no functional islets in the biopsied liver. According to our previous report, segment 2 accounts for about 20% of the islets transplanted to the entire liver 32 . Two months after transplantation, one monkey (K) in group 3, which received > 25,000 IEQ/kg, showed indicators of diabetes (blood glucose level, serum C-peptide level, and insulin requirement) and regressed similarly to the monkeys in group 2 after we resected the whole of segment 2. That suggests that > 25,000 IEQ/kg is a marginal dose for achieving a normal blood glucose level after transplantation without exogenous insulin. However, that result is from only one monkey, so more detailed dose studies between 20,000 and > 25,000 IEQ/kg are needed.
ATG and rituximab are the immunosuppressive therapies used for T and B cell-regulation in clinical transplantation settings, and they are also used in allogeneic islet transplantation in NHPs 14,[16][17][18][19]26 . In our three monkeys that used both ATG and rituximab, there were no differences in the factors related to the normalization of blood glucose level, especially the B cell in ltration affected by rituximab. This result is presumed to have been affected by our method of inducing diabetes, in which we removed the pancreas and spleen together, as previously described 18,33 .
The characteristics of the islets found in the liver biopsy after transplantation differed slightly in each group, but the size, insulin expression level, and glucagon expression level differences were not signi cant 32,34 . On the other hand, the IAPP expression level within the islets and CD3 + T cell in ltration seemed to correlate with each other. In diabetic monkeys, it is di cult to explain clearly why doses above 25,000 IEQ/kg were able to achieve normalization of blood glucose levels after transplantation without exogenous insulin. However, the expression level of IAPP within the islets correlates with the in ltration rate of CD3 + T cells doses between the 20,000 IEQ/kg and > 25,000 IEQ/kg groups. We thus expect that the mechanism of IAPP production occurs alongside the insulin production process in beta cells, allowing higher IAPP levels to accumulate in the group that received a lower dose of islet cells under the same diabetic condition. When an insu cient number of islets is transplanted into a diabetic monkey, each cell must produce a larger amount of insulin to normalize the blood glucose level, which would allow a larger amount of IAPP to accumulate within the islets compared with the islets of monkeys that received a su cient dose of transplanted islets. This expectation is consistent with our results, in which the islets of monkeys that received a smaller number of islets showed higher levels of IAPP accumulation than those under the same diabetic condition that received more islets (Fig. 6A and B).
On the other hand, several studies have reported that IAPP activates T cells [28][29][30][31] . Based on those results, we predicted that the increased IAPP accumulation found in the group that received an insu cient dose of islets would attract T cells into the transplanted islets. Our results are consistent with that expectation. Group 2, which received an insu cient dose of islets (20,000 IEQ/kg), showed a higher level of T cell in ltration than group 3 (> 25,000 IEQ/kg) both 1 and 2 months after transplantation (Fig. 5D). However, in group 1, which received the smallest number of islets (10,000 IEQ/kg), no islets were found two months after transplantation (Fig. 5). Interestingly, the phenomenon of T cell in ltration in the results from the biopsied islets after transplantation showed that as the expression of IAPP increased, the population of scar islets that did not function also increased, which accelerated T-cell in ltration into the islets (Fig. 6C). Thus, a hyperglycemic environment accelerates the accumulation of IAPP into transplanted islets, especially early after transplantation, and IAPP accumulation within an islet negatively affects the results after transplantation 26 .
One limitation of our present study is the number of biopsied islets containing histological information.
The probability of nding islets in liver tissue obtained through a single edge biopsy is low. The excision of large amounts of the liver to ensure successful detection of islets would affect the health of the monkeys and could thus negatively affect the stability of both the blood glucose level and the study as a whole. Therefore, the number of islets that could be found in the biopsies was small, making statistically reliable results di cult to obtain. In our previous study, which we conducted around the same time as this study, the optimal liver biopsy site for successful islet detection was segment 1. Most of the biopsies in this study were performed in segment 2, which had less than half the density of islets per liver area found in segment 1 32 . To overcome that limitation, future studies should conduct biopsies in a site with a high success rate.
In conclusion, the marginal transplantation dose to achieve a normal blood glucose level without the need for exogenous insulin in a cynomolgus monkey model was > 2,5000 IEQ/kg, and an accumulation of IAPP, which varied with the transplanted islet dose early after transplantation, can be considered a factor of rejection.

Materials And Methods
Animals and animal care. Eleven cynomolgus monkeys (Macaca fascicularis) were supplied by Orient Bio Co. Ltd (Seongnam, Korea). The monkeys were isolated in individual cages. Water was given ad libitum, and biscuits were supplied twice a day (Certi ed Primate Diet 5048, LabDiet, St. Louis, MO, USA). Fresh fruits, vegetables, and nuts were also provided twice a day. The inhabited environment was maintained at a temperature of 25 ± 2°C, 40-60% humidity, 1-5 mmHg positive pressure air conditioning, and 300 lux illuminance alternating with darkness every 12 hours. All the procedures related to infection screening, housing, handling, care, and treatment in this study were performed as previously described 33 . This study was approved by the Institutional Animal Care and Use Committee of Orient Bio Laboratories (ORIENT-IACUC-14195), and the experiments were performed in accordance with the relevant guidelines and regulations. All methods were carried out in compliance with the ARRIVE guidelines.
Induction and management of type 1 diabetes mellitus. Pancreatectomy and the induction, con rmation, and maintenance of T1DM in cynomolgus monkeys were performed as previously described 33 . Brie y, the donor monkey's pancreas was removed through subtotal (> 70% of the pancreas) or total pancreatectomy. The removed pancreas was used for islet isolation, and the donor monkey whose pancreas was removed became a recipient monkey after an injection of 60-80 mg/kg of streptozotocin (Sigma, St Louis, MO, USA). T1DM was diagnosed when the following criteria were satis ed: 1) sustained hyperglycemia (blood glucose level > 250 mg/dl), 2) fasting C-peptide level < 0.5 ng/ml, and 3) decrease in stimulated C-peptide response in the IVGTT. After the onset of diabetes and islet transplantation, the blood glucose level was monitored 2 to 4 times daily, and exogenous insulin (glargine: Lantus; Sano -Aventis, Bridgewater, NJ, USA, and glulisine: Apidra, Sano -Aventis) were used to maintain blood glucose levels < 200 mg/kg to protect the animals from hyperglycemia (Fig. 2).
Isolation and culture of islets. Resected partial or complete pancreases from the donor monkeys were processed as previously described to isolate the islets 33 . Brie y, the islets were isolated from the pancreas using the modi ed Ricordi method with collagenase MTF C/T (Roche, Indianapolis, IN). The discontinuous Ficoll density gradient method was used to purify the islets. The puri ed islets were cultured with CMRL-1066 supplemented medium (Corning, NY, USA) supplied with 10% fetal bovine serum (Gibco-Thermo Fisher Scienti c, Waltham, MA, USA) and 1% antibiotics (Gibco-Thermo Fisher Scienti c) in a humidi ed 5% CO 2 atmosphere at 37°C.
Allogeneic islet transplantation and perioperative management. Initially, we planned for the follow-up time after transplantation to be 3 months (90 days). In fact, we stopped the experiment early for some monkeys in accordance with humanitarian treatment because the affected animals endured the following conditions for two continuous months after transplantation: 1) blood glucose level > 200mg/dL, 2) exogenous insulin usage > 3 IU/kg/day, and 3) serum C-peptide levels < 0.5 ng/ml. The transplantation of cultured islets and perioperative management were performed as previously described 18,33 . Brie y, at least 30 days after con rmation of T1DM, 11 monkeys underwent islet transplantation via the intraportal injection of cultured islets. According to the injected IEQ/kg, the monkeys were divided into three groups: group 1 (10,000 IEQ/kg, n = 2), group 2 (20,000 IEQ/kg, n = 5), and group 3 (> 25,000 IEQ/kg, n = 4) ( Table 1). After inducing general anesthesia, we performed central venous access port insertion via the right internal jugular vein. Through this central line, all monkeys received ATG four times at 12-hour intervals to a cumulative dose of 20 mg/kg as induction immunosuppression. RTX injections at a dose of 375 mg/m 2 were added in three monkeys (E, F, G) in group 2 as combination induction immunosuppression. Laparotomy began with an upper midline incision. After a self-retractor was applied, the portal vein was isolated. Islet-mixed heparin (75 IU/kg) was infused through an 18-gauge angiocatheter inserted into the portal vein. After islet infusion was nished, the angiocatheter was removed, and the puncture site was closed with 6 − 0 Prolene sutures. After each monkey awakened from anesthesia, it was returned to its cage. The immunosuppression schedule for pre-and posttransplantation is summarized in Supplementary Figure S1.
Postoperative management. Postoperative management were performed as previously described 18 . Brie y, the monkeys received oral FK506 (Tacrolimus, Prograf, Astellas Pharma Europe Ltd, Addlestone, UK) and mycophenolate mofetil (Cellcept, Roche Pharmaceuticals AG, Basel, Switzerland) as maintenance immunosuppressive drugs. To prevent in ammatory events, etanercept was given on the day of the transplant (day 0) and on days 3 and 6. A subcutaneous injection of anakinra (Kineret™, Swedish Orphan Biovitrum, Stockholm, Sweden) was also given daily from days 0 to 7 (Supplementary Figure S1). One month after transplantation, IVGTT was performed after 12 hours of fasting. After sedation with ketamine, two blood samples were drawn for C-peptide and blood glucose measurements. Then, 0.5 g/kg of dextrose was given intravenously, and blood samples were drawn 1, 3, 5, 7, and 10 min thereafter. Blood samples were also drawn at 15,20,25,30, and 60 min to measure the glucose disappearance rate. Serum C-peptide was measured using a radioimmunoassay kit developed for human plasma (C-Peptide IRMA kit; IMMUNOTHECH, Beckman Coulter Inc., Prague, Czech Republic), which shows 90% cross-reactivity with plasma from cynomolgus monkeys. The acute C-peptide response was calculated as the difference between the mean C-peptide after glucose infusion and C-peptide at baseline ( Figs. 2 and 3).
Liver biopsy and immunohistochemical staining. Liver biopsies and the immunohistochemical analyses were conducted following our previously published criteria 32 . Brie y, to observe and investigate the environment of the grafted islets after transplantation, a protocol liver biopsy was performed one month (30 days) and two months (60 days) after transplantation, dependent on each monkey's health condition.
Monkey K of group 3 received a left lateral lobectomy 2 months after islet transplantation to assess the continuation of insulin independence after the removal of a considerable volume of islets from the grafted liver. The monkeys underwent a laparotomy to expose the liver, and then segment 1 or 2 of the liver was excised. The biopsied liver tissues were xed in 10% neutral buffered formalin for 24h and then embedded in para n. Each para n block was sliced into 30 para n slides that were 4 µm in thickness in a total of three sections, as previously described 32 to produce 10 slides/section (Fig. 1). The slides produced in each section were stained for target proteins in the following order: hematoxylin and eosin (1st slide), anti-insulin from abcam (ab6995) (2nd slide), anti-CD3 from DAKO (A0452) (3rd slide), anti-CD20 from DAKO (M0755) (4th slide), anti-amyloid oligomers from abcam (ab126892) (5th slide), and anti-glucagon from abcam (ab92517) (6th slide). After depara nization and heat retrieval of the epitope, each slide was stained with the target protein for the 1st antibody. To visualize of each target protein as brown, 3, 3'-diaminobenzidine tetrahydrochloride staining was performed using a DAKO EnVision system (DAKO, Santa Clara, CA, USA) according to the manufacturer's instructions (Fig. 4).
Analysis of immunohistochemically stained images. The immunohistochemistry slide images were analyzed as previously described 32 . Brie y, an image le of each stained slide was acquired using a ScanScope AT slide scanner (Leica Biosystems, Wetzlar, Germany) and Aperio ScanScope software (Leica Biosystems). Using the Aperio Positive Pixel Count algorithm (version 9.1) in the Aperio ImageScope program (version 12.1.0.5029; Leica Biosystems), we obtained the islet diameter (µm), islet area (µm 2 ), and insulin-, glucagon-, CD3-, CD20-, and IAPP-positive areas within the total islet area (%) (Fig. 5). Because of the non-uniform shape and size of the islets and the change in the position of the islets on the slides from a continuous para n section, the number of islets for which we could analyze all the characteristics was small. To compensate for that, we analyzed the statistical signi cance of the characteristics between islets containing all the characteristics (complete data set) and islets containing only some of the characteristics (incomplete date set). As a result, we con rmed that there were no statistically signi cance differences in the remaining characteristics except for the size of the islets and the beta-cell expression rate within the islet area of group 2 at 2 months. Therefore, we used the incomplete data set (Supplementary Table S1).
Statistical analysis. The IVGTT results were analyzed using one-way ANOVA with Bonferroni`s multiple comparison post hoc test in GraphPad Prism version 5.00 (GraphPad Software, San Diego CA, USA) ( Fig. 3A and B). The characteristics of the biopsied islets were analyzed using unpaired t testing with GraphPad Prism version 5.00 (GraphPad Software) (Fig. 5, Fig. 6B, and Supplementary Table S1). All data are presented as means ± standard error of the mean (SEM).

Declarations Data Availability
The datasets generated during and/or analyzed for the current study are available from the corresponding author upon reasonable request.  Representative results from the immunohistochemical analysis of islets found in liver biopsies conducted one month after transplantation. Liver biopsies were performed in group 1 (a) -(e), group 2 (f) -(j), and group 3 (k) -(o). (a) -(e) In Group 1, no liver biopsies were performed due to the subjects' poor health caused by high blood glucose levels (ND, not detected). Insulin (f and k) and glucagon (g and i) were expressed in high amounts within the islets of groups 2 and 3. Small numbers of CD3+ T cells (arrowed) (h and m) and CD20+ B cells (i and n) were detected in the islet areas of both groups 2 and 3, but they were higher in group 2. IAPP (j and o) was slightly expressed in all parts of the islets. Brown indicates the target protein in each result (scale bars = 100 μm).