Noninvasive diagnosis of recurrent autoimmune type 1 diabetes after islet cell transplantation

Abstract

Islet cell transplantation is curative therapy for patients with complicated autoimmune type 1 diabetes (T1D). We report the diagnostic potential of circulating transplant islet-specific exosomes to noninvasively distinguish pancreatic β cell injury secondary to recurrent autoimmunity vs immunologic rejection. A T1D patient with hypoglycemic unawareness underwent islet transplantation and maintained normoglycemia until posttransplant day 1098 before requiring exogenous insulin. Plasma analysis showed decreased donor islet exosome quantities on day 1001, before hyperglycemia onset. This drop in islet exosome quantity signified islet injury, but did not distinguish injury type. However, analysis of purified transplant islet exosome cargoes showed decrease in insulin-containing exosomes, but not glucagon-containing exosomes, indicating selective destruction of transplanted β cells secondary to recurrent T1D autoimmunity. Furthermore, donor islet exosome cargo analysis showed time-specific increase in islet autoantigen, glutamic acid decarboxylase 65 (GAD65), implicated in T1D autoimmunity. Time-matched analysis of plasma transplant islet exosomes in 3 control subjects undergoing islet cell transplantation failed to show changes in islet exosome quantities or intraexosomal cargo expression of insulin, glucagon, and GAD65. This is the first report of noninvasive diagnosis of recurrent autoimmunity after islet cell transplantation, suggesting that transplant tissue exosome platform may serve as a biomarker in islet transplant diagnostics.

1 |. INTRODUCTION

Islet cell transplantation is performed in patients with autoimmune type 1 diabetes (T1D) with hypoglycemic unawareness or glycemic lability. Currently, glucose monitoring and C-peptide-based assays are utilized for monitoring islet allograft function,¹ but these noninvasive platforms lack the required time-sensitivity and accuracy to diagnose early injury to the transplanted islet mass. Furthermore, transplant islet injury in these patients can occur secondary to immunologic rejection or recurrent autoimmunity.^2,3 Since both these types of injury clinically manifest as hyperglycemia, currently, liver biopsy-based histological analysis of allo-islet clusters is the only method of determining the cause of recurrent diabetes after islet cell transplantation. There are only a few case studies reporting recurrent autoimmunity in islet allografts,^2,3 where biopsy is rarely performed due to the poor reliability of liver biopsy to include islet tissue necessary for pathologic evaluation,⁴ and due to the invasive nature of liver biopsy.

Exosomes are circulating extracellular microvesicles that carry cell-specific protein and RNA cargoes reflecting the conditional status of their tissue counterparts.^5–11 We previously reported the diagnostic potential of transplant tissue-specific exosome profiling from recipient circulation as a novel, noninvasive biomarker for acute rejection.^12–14 The translational potential of this platform was validated in patients undergoing allogeneic islet cell transplantation.¹² Transplant islet exosomes were shown to carry islet endocrine hormone mRNAs and proteins, insulin, glucagon, and somatostatin, as part of their intraexosomal cargoes, suggesting that islet exosomes were cell specific, representing the endocrine cellular constituents of the pancreatic α, β, and δ cells, respectively. Therefore, we hypothesized that transplant islet exosome cargo analysis would not only enable detection of islet allograft injury, but also the type of islet allograft injury–immunologic rejection vs recurrent T1D autoimmunity. Here, we report the results of the first noninvasive diagnosis of recurrent T1D autoimmunity in an islet transplant recipient utilizing a transplant tissue exosome platform.

2 |. CASE REPORT

A 56-year-old man (Patient A) with a 44-year history of T1D complicated by hypoglycemic unawareness underwent intraportal infusion of single donor allogeneic islet cells according to the CIT07 protocol.¹⁴ He maintained long-term normoglycemia on tacrolimus and sirolimus immunosuppression (Figure 1A). Monitoring for autoantibodies against glutamic acid decarboxylase-65 (GAD65), tyrosine phosphatase islet cell antigen (ICA)-512/IA-2, insulin, and for alloantibodies against class I and II human leukocyte antigens was performed every 3–6 months posttransplant as previously described.¹⁵ At 18 months posttransplant, the patient developed pneumocystis pneumonia, requiring change in immunosuppression from sirolimus to mycophenolic acid and low-dose tacrolimus. At 24 months (733 days), the allograft β cell secretory capacity as assessed by glucose-potentiation of arginine-induced insulin secretion demonstrated stable functional β cell mass compared to the 12-month (363 day) time point, supporting the establishment of a sufficient reserve capacity for insulin secretion by the graft to resist metabolic exhaustion and the absence of nonimmunologic islet β cell loss; assessments of alloantibody and autoantibody remained negative. At 36 months (1098 days), the patient developed recurrent diabetes. Glucose-potentiated arginine testing showed a 58% reduction in p cell secretory capacity (Figure 1B). Autoantibody analysis showed a new GAD65 autoantibody at 33 months (1001 days), which was not present on earlier assessments; testing for islet donor-specific alloantibodies remained negative. Furthermore, none of the control Patients B-D developed new or increasing titer of autoantibodies against GAD65, ICA-512/IA-2, or insulin, nor developed a positive panel-reactive alloantibody screen during follow-up. Although suggested by the development of a new and progressive titer of GAD65 autoantibody in the absence of any alloantibody reactivity, establishing a diagnosis of recurrent autoimmune diabetes currently requires demonstration of selective destruction of islet β cells along with preservation of islet α cells by histologic examination of transplanted tissue biopsy as previously reported for whole pancreas and islet cell transplantation.^1–3

FIGURE 1.

Glycemic control and recurrence of diabetes in patient A over posttransplant follow-up is shown. A, Exogenous insulin requirement, hemoglobin A1c, and autoantibody GAD65 titers are shown. After maintaining normoglycemia post-islet cell transplantation, patient A developed GAD65 autoantibody at 33 months (1001 d); and at 36 months (1098 d), fasting hyperglycemia was noted, requiring resumption of insulin therapy. B, β cell secretory capacity derived from the acute insulin response to arginine stimulation under hyperglycemic clamp conditions provided a measure of the functional allogeneic islet mass yearly after transplant. In Patient A, β cell secretory capacity was near-normal and stable at 12 and 24 months, but declined by 58% at 36 months, when the patient developed recurrent diabetes. GAD65, glutamic acid decarboxylase 65

We previously reported plasma transplant islet exosome quantities in Patients A-D over the follow-up period.¹² For the sake of ease of interpretation, transplant islet exosome signal and GAD65 autoantibody titer in Patient A for follow-up time points are shown (Figure 2). In Patient A, an initial increase in donor islet exosome signal was noted posttransplant similar to Patients B-D, but over follow-up there was a drop in the signal at measured time points, with the donor islet exosome signal on day 1001 decreasing to lower than the mean signal value in Patients B-D.¹² This signal change in Patient A occurred before onset of recurrent diabetes, exogenous insulin requirement, low C peptide levels, and changes in glucose-potentiated arginine studies. Although the signal decreased in this patient, it did not drop to pretransplant/background signal level, suggesting that there was still some persistent release of exosomes by the islet allograft. However, the transplant islet exosome signal in time-matched samples in Patients B-D, represented by the mean signal values, remained stable during follow-up.¹²

FIGURE 2.

Change in circulating transplant islet exosome signal correlates with rise in GAD65 antibody titer. A, Nanoparticle detector analysis for transplant islet-specific exosomes (TISE) signal in Patient A over the follow-up is shown. Blue represents total plasma exosomes, and the red curve represents TISE signal characterized using anti-donor HLA-specific antibody quantum dot on the fluorescence mode. Compared to the pretransplant sample, TISE were seen at 60 minutes post-islet cell transplantation, and detectable over long-term follow-up. Day 1001 sample showed a decrease in TISE signal to below time-matched islet transplant controls with long-term insulin independence (not shown). B, Antibody titer to islet autoantigen, GAD65, in Patient A is shown. The first increase in GAD65 antibody was noted from day 1001 sample, at the same time point when TISE signal drop was noted in Patient A compared to time-matched controls. GAD65, glutamic acid decarboxylase 65

Quantitative analysis of circulating islet exosomes enabled noninvasive diagnosis of islet allograft injury, but it does not help distinguish the type of islet injury (immunologic rejection vs recurrent T1D autoimmunity). The persistent low level of islet exosome signal in Patient A after onset of recurrent diabetes suggests that there was residual allo-islet mass responsible for maintaining the sustained circulating islet exosome output, thus supporting a diagnosis of recurrent T1D over immunologic rejection, as the latter would eliminate all transplanted tissue. Our recent work demonstrated that transplant islet exosomes carry endocrine hormone proteins and mRNAs as part of their cargoes.¹² Thus, if islet exosome subpopulations are specific to their islet endocrine cell of origin, we proposed that with recurrent T1D there would be a selective decrease in insulin-containing exosomes, with minimal to no changes in the glucagon-containing exosome subpopulation, as the latter hormone is specifically released by islet α cells. If the cause was immunologic rejection, then there would be a corresponding decrease in both insulin and glucagon signals within the transplanted islet exosome subpopulation, with preserved ratio of expression of the 2 hormones.

We enriched donor islet exosomes from recipient fasting plasma samples over the follow-up period in the 4 patients and analyzed for insulin and glucagon mRNA content by reverse transcription polymerase chain reaction (Figure 3A–D). In Patients B-D, insulin and glucagon mRNAs were detected in all posttransplant samples, with preserved ratio of expression during the entire follow-up (Figure 3B–D). In Patient A, preserved insulin-to-glucagon mRNA expression was noted on the day 56 time point, but selective decrease in insulin mRNA was noted as early as day 455. The insulin mRNA signal decreased further on later time points, suggesting progressive and selective destruction of allo-islet β cells. Therefore, before the onset of recurrent diabetes and changes in clinical markers for β cell monitoring, there was time sensitive, selective decrease in insulin-containing islet exosomes without any reduction in glucagon-containing islet exosomes. Furthermore, intraexosomal cargo analysis for insulin mRNA was more time sensitive for β cell injury than quantitative changes in transplant islet exosome signal. Taken together, these findings indicate that transplant islet exosome cargo analysis may enable early diagnosis of β cell-specific injury and noninvasive classification of recurrent autoimmune from alloimmune islet injury.

FIGURE 3.

Intraexosomal mRNA cargo of donor islet-specific exosomes reflects endocrine cell-specific constituents of the functional transplanted islet mass. Circulating donor islet-specific exosomes were enriched from recipient plasma in Patients A, B, C, and D using anti-donor HLA I-specific antibody conjugated beads. Reverse transcription polymerase chain reaction analysis for insulin and glucagon mRNA cargoes in transplant islet exosomes at follow-up time points is shown for all 4 patients (A-D). In all cases, IgG isotype beads (negative control) from posttransplant samples and pretransplant samples did not show any enrichment of insulin or glucagon in the samples. In Patients B, C, and D, insulin and glucagon mRNAs were detected in all posttransplant samples, with similar ratios of expression between these 2 endocrine markers, suggesting stable β and α cell function, respectively, in the transplanted islets. In Patient A, day 56 sample showed high expression of insulin and glucagon mRNAs, but days 455, 1001, and 1197 samples showed progressive loss of insulin expression without change in glucagon expression, demonstrating selective loss of β cell exosome output secondary to selective β cell destruction in the transplanted islets with preserved transplant islet α cell function. β-actin mRNA control is also shown

In support of the above findings, selective upregulation of GAD65 autoantibody levels, but not alloantibody, were noted starting at the 33-month (day 1001) time point in Patient A (Figure 2B). Because alloexosomes may have immunoregulatory roles, we assessed whether islet exosomes carried GAD65 autoantigen as part of its cargo in a time- and condition-specific manner. In Patient A, GAD65 mRNA expression was seen in the donor islet exosome subpopulation at time points before onset of recurrent diabetes and before upregulation of circulating anti-GAD65 autoantibody levels (Figure 4A), inversely correlating with intraexosomal insulin mRNA levels. In Patients B and C, there was no detection of GAD65 mRNA at all tested time points (Figure 4A); insufficient sample remained from Patient D for analysis. These findings were further validated by Western blot analysis of islet exosomes, where selective and time-specific expression of GAD65 protein was seen in Patient A, but not in Patients B and C (Figure 4B). These findings further support the idea that the cause of recurrent diabetes in Patient A was due to T1D autoimmunity recurrence. Collectively, this suggested that islet exosomes may carry autoantigens implicated in T1D autoimmunity, and exosomes may be one of the methods of autoantigen trafficking and presentation. Intraexosomal cargo analysis for autoantigens may further improve the diagnostic accuracy of the transplant exosome platform for noninvasive monitoring of β cell mass in islet transplant recipients. (All experiments are performed according to methods described in the Supporting Information.)

FIGURE 4.

Recurrent diabetes in Patient A is associated with upregulation of islet self-antigen GAD65 in circulating transplant islet exosomes. Because exosomes are implicated in immune regulation, donor islet-specific exosome cargo was analyzed for expression of T1D self-antigen, GAD65, using anti-donor HLA I-specific antibody conjugated beads. A, Reverse transcription polymerase chain reaction for GAD65 mRNA expression showed its absence in controls, Patients B and C, but selective upregulation was seen in days 455 and 1197 samples in Patient A. This patient developed GAD65 autoantibody on day 1001, suggesting that exosome-based expression of selfantigens may precede development of T1D autoantibody. B, Western blot analysis of donor islet-specific intraexosomal cargo also showed selective upregulation of GAD65 protein in Patient A at day 455 time point, but was undetectable in Patient B. Patient C showed trace to undetectable levels of GAD65 protein. GAD65, glutamic acid decarboxylase 65; T1D, type 1 diabetes

3 |. DISCUSSION

We report the biomarker potential of transplanted islet exosome profiling for noninvasive diagnosis of β cell autoimmunity as the cause of recurrent diabetes in a T1D recipient 1098 days after islet cell transplantation. Currently, histological evaluation of the allo-islet tissue obtained via surgical biopsy showing selective β cell injury with preservation of islet architecture, including α cells, is the only described method for distinguishing recurrent autoimmunity from immunologic rejection.^2,3 Exosomes are circulating tissue-specific extracellular microvesicles with protein and RNA cargoes specific to their cell of origin, with expression profiles reflecting conditional changes imposed on their source tissue. In islet transplantation, donor islet exosomes would carry islet endocrine hormone proteins and mRNAs as their intraexosomal cargo. In this context, recurrent T1D autoimmunity would lead to specific changes in β cell exosomes, with preservation of the α cell exosome contribution into the total exosome pool from peripheral blood, whereas rejection would lead to changes in both the β cell and α cell exosome subpopulations. The findings in this study strongly support the diagnosis of recurrent autoimmunity in the index patient. Time- and patient-specific quantitative changes in donor islet exosomes, along with changes in cargoes of islet exosome subpopulations specific to their cellular constituents of the islet cluster, were noted. These findings suggest that plasma islet exosome profiles are dynamic. Furthermore, the time- and patient-specific detection of GAD65 mRNA and protein in transplant islet exosomes in Patient A before detection of serum anti-GAD65 autoantibody and onset of recurrent diabetes suggests that transplant islet exosomes may play a role in immune interactions and autoantigen presentation.

This dynamic and tissue-specific nature of islet exosome response that can be tracked from a peripheral blood sample might enable time-sensitive detection of β cell injury in islet transplant recipients before clinical manifestation. In support of this idea, in Patient A we noted a >20% drop in the fasting plasma islet exosome signal at time points >50 days before requirement for exogenous insulin therapy, while hemoglobin A_1c, fasting glucose, and C peptide values were still normal.¹¹ This transplant islet exosome platform may serve as an adjunct to current clinical standards for monitoring the islet allograft in patients undergoing islet cell transplantation. If so, such a biomarker may provide early diagnosis of islet injury and help distinguish the type of injury (recurrent autoimmunity or immunologic rejection), and thus provide an earlier window for therapeutic interventions.

In summary, we report the first case of noninvasive diagnosis of recurrent autoimmunity in a type 1 diabetic patient after allogeneic islet cell transplantation. Future studies may help in better understanding the biomarker potential of transplant islet exosome platform.

4 |. STUDY LIMITATIONS

This was a retrospective analysis, and the limited availability of plasma samples from islet transplant patients (250–500 μL) precluded more detailed analysis looking for time- and patient-specific changes in microRNA cargoes of transplant islet exosomes. Further analyses may provide additional mechanistic insights and biomarkers to further improve the diagnostic accuracy and time-sensitivity of the proposed platform. The present study was strengthened by the concurrent assessment of in vivo functional islet β cell mass and serologic evidence for recurrent autoimmunity that help to exclude nonimmunologic islet β cell loss as can occur with excessive metabolic demand imposed on an islet graft as an explanation for the selective loss of β cell exosomes.¹⁶ However, the discrimination of nonimmunologic from autoimmune β cell loss may be more challenging without correlative markers of autoimmunity. The present study did not compare the exosome approach to other methods of detecting active cellular autoimmunity.¹⁷ While further validation of the exosome platform should include autoreactive T cell assays, the simple sample collection and processing requirements for assessing plasma exosomes may allow for future broader application in clinical settings.

Abbreviations:

GAD65

glutamic acid decarboxylase 65

T1D

type 1 diabetes