OP-142 Gene Expression Profile of Nonfunctional Human Pancreatic Islets: Predictors of Transplant Failure?

The Effect of Donor Factors on Human Islet Yield and Their in Vivo Function

Background

A major problem in the islet field is the high selectivity exercised in accepting cadaveric pancreas for islet isolation. This practice is based on experience that indicates that islet yield and posttransplant function are related to donor demographics and injury mechanisms.

Objective

To examine factors influencing islets recovery and in vivo function with emphasis on donor-related factors.

Methods

Islets were isolated from 99 human donor pancreata, and islet yield was reported as islet equivalent per gram pancreatic tissue. Donor, procurement, and isolation factors were collected for each isolation and correlation statistics were performed between these variables and islet yield.

Results

Results indicated a differential effect of enzyme mixes on yield with 2 Collagenase P digestion most suitable for increased ischemic time ( R2 = 0.1; P < .08), Liberase with small donor pancreas size and elevated preprocurement glucose ( R2 = 0.15; P < .02), and Serva with female donors ( R2 = 0.17; P < .06). Islets from 29 isolations were further tested by transplantation under the kidney capsule of immune-deficient NOD-SCID mice. Although all 29 preparations had acceptable in vitro perfusion parameters indicating viability, only 19 functioned in vivo with serum levels of insulin >5 U/mL and C peptide >1.5 ng/mL. No significant differences in donor, procurement, and isolation factors were evident between the islet preparations that functioned in vivo and those that were nonfunctional.

Conclusions

These data demonstrate that although yield is affected by a variety of donor factors and enzyme mixes, these factors do not affect islet in vivo function.

Characterization of a nanogland for the autotransplantation of human pancreatic islets

Despite the clinical success of pancreatic islet transplantation, graft function is frequently lost over time due to islet dispersion, lack of neovascularization, and loss of physiological architecture. To address these problems, islet encapsulation strategies including scaffolds and devices have been developed, which produced encouraging results in preclinical models. However, islet loss from such architectures could represent a significant limitation to clinical use. Here, we developed and characterized a novel islet encapsulation silicon device, the NanoGland, to overcome islet loss, while providing a physiological-like environment for long-term islet viability and revascularization. NanoGlands, microfabricated with a channel size ranging from 3.6 nm to 60 μm, were mathematically modeled to predict the kinetics of the response of encapsulated islets to glucose stimuli, based on different channel sizes, and to rationally select membranes for further testing. The model was validated in vitro using static and perifusion testing, during which insulin secretion and functionality were demonstrated for over 30-days. In vitro testing also showed 70-83% enhanced islet retention as compared to porous scaffolds, here simulated through a 200 μm channel membrane. Finally, evidence of in vivo viability of human islets subcutaneously transplanted within NanoGlands was shown in mice for over 120 days. In this context, mouse endothelial cell infiltration suggesting neovascularization from the host were identified in the retrieved grafts. The NanoGland represents a novel, promising approach for the autotransplantation of human islets.

Mathematical models of diabetes progression

Few attempts have been made to model mathematically the progression of type 2 diabetes. A realistic representation of the long-term physiological adaptation to developing insulin resistance is necessary for effectively designing clinical trials and evaluating diabetes prevention or disease modification therapies. Writing a good model for diabetes progression is difficult because the long time span of the disease makes experimental verification of modeling hypotheses extremely awkward. In this context, it is of primary importance that the assumptions underlying the model equations properly reflect established physiology and that the mathematical formulation of the model give rise only to physically plausible behavior of the solutions. In the present work, a model of the pancreatic islet compensation is formulated, its physiological assumptions are presented, some fundamental qualitative characteristics of its solutions are established, the numerical values assigned to its parameters are extensively discussed (also with reference to available cross-sectional epidemiologic data), and its performance over the span of a lifetime is simulated under various conditions, including worsening insulin resistance and primary replication defects. The differences with respect to two previously proposed models of diabetes progression are highlighted, and therefore, the model is proposed as a realistic, robust description of the evolution of the compensation of the glucose-insulin system in healthy and diabetic individuals. Model simulations can be run from the authors' web page.

Glucose-stimulated increment in oxygen consumption rate as a standardized test of human islet quality

Standardized assessment of islet quality is imperative for clinical islet transplantation. We have previously shown that the increment in oxygen consumption rate stimulated by glucose (DeltaOCR(glc)) can predict in vivo efficacy of islet transplantation in mice. To further evaluate the approach, we studied three factors: islet specificity, islet composition and agreement between results obtained by different groups. Equivalent perifusion systems were set up at the City of Hope and the University of Washington and the values of DeltaOCR(glc) obtained at both institutions were compared. Islet specificity was determined by comparing DeltaOCR(glc) in islet and nonislet tissue. The DeltaOCR(glc) ranged from 0.01 to 0.19 nmol/min/100 islets (n = 14), a wide range in islet quality, but the values obtained by the two centers were similar. The contribution from nonislet impurities was negligible (DeltaOCR(glc) was 0.12 nmol/min/100 islets vs. 0.007 nmol/min/100 nonislet clusters). The DeltaOCR(glc) was statistically independent of percent beta cells, demonstrating that DeltaOCR(glc) is governed more by islet quality than by islet composition. The DeltaOCR(glc), but not the absolute level of OCR, was predictive of reversal of hyperglycemia in diabetic mice. These demonstrations lay the foundation for testing DeltaOCR(glc) as a measurement of islet quality for human islet transplantation.

Examination of Gene Expression Profile of Functional Human Pancreatic Islets After 2-Week Culture

Islet transplant faces significant challenges, mainly because of the high incidence of primary nonfunction of transplanted islets. Protocol modifications to improve the rate of islet function have included changes in pancreatic preservation and the introduction of short-term culture. Islet culture for 48 to 72 hours has become a standard part of most successful protocols for clinical islet transplantation. We have previously reported gene expression profiles associated with human pancreatic islet function. The aim of this study was to determine the change in gene expression profiles of functional islets after 2 weeks of culture in Memphis-serum free media. Human islets from four isolations were maintained in culture for 14 days in Memphis-serum free media. RNA was extracted from 10000 IEQ for analysis of the gene expression profiles using high-density Affymetrix U133A GeneChips and Genespring software. Islet function was assessed by measurements of human C-peptide at days 7 and 14 posttransplant into NOD-SCID mice. Human C-peptide levels were determined by radioimmunoassay. Our preliminary data showed that genes related to functionality, such as those directed toward insulin processing and secretion, did not vary over 14 days of culture, while genes related to exocrine pancreas and organ architecture and immune-associated genes decreased over time. The ability to maintain islets in culture is an important step toward the development of islet tissue repositories, as well as toward screening human islet preparations for additional pathogens.

Assessment of islet function following islet and pancreas transplantation

Pancreas and islet transplant recipients are monitored using various metabolic and imaging methods. The inaccessibility of the transplanted whole pancreas and of the isolated islets poses specific problems (eg, all assessment techniques are indirect). Although successful pancreas transplantation typically restores normal glucose homeostasis, islet transplantation into the liver does not completely normalize islet hormone secretion and glucose metabolism. Development of better testing strategies, such as direct islet imaging, will significantly advance the field.