How to use image analysis for islet counting

Probing the characteristics and biofunctional effects of disease-affected cells and drug response via machine learning applications

Drug-induced transformations in disease characteristics at the cellular and molecular level offers the opportunity to predict and evaluate the efficacy of pharmaceutical ingredients whilst enabling the optimal design of new and improved drugs with enhanced pharmacokinetics and pharmacodynamics. Machine learning is a promising in-silico tool used to simulate cells with specific disease properties and to determine their response toward drug uptake. Differences in the properties of normal and infected cells, including biophysical, biochemical and physiological characteristics, plays a key role in developing fundamental cellular probing platforms for machine learning applications. Cellular features can be extracted periodically from both the drug treated, infected, and normal cells via image segmentations in order to probe dynamic differences in cell behavior. Cellular segmentation can be evaluated to reflect the levels of drug effect on a distinct cell or group of cells via probability scoring. This article provides an account for the use of machine learning methods to probe differences in the biophysical, biochemical and physiological characteristics of infected cells in response to pharmacokinetics uptake of drug ingredients for application in cancer, diabetes and neurodegenerative disease therapies.

The Flaws and Future of Islet Volume Measurements

When working with isolated islet preparations, measuring the volume of tissue is not a trivial matter. Islets come in a large range of sizes and are often contaminated with exocrine tissue. Many factors complicate the procedure, and yet knowledge of the islet volume is essential for predicting the success of an islet transplant or comparing experimental groups in the laboratory. In 1990, Ricordi presented the islet equivalency (IEQ), defined as one IEQ equaling a single spherical islet of 150 μm in diameter. The method for estimating IEQ was developed by visualizing islets in a microscope, estimating their diameter in 50 μm categories and calculating a total volume for the preparation. Shortly after its introduction, the IEQ was adopted as the standard method for islet volume measurements. It has helped to advance research in the field by providing a useful tool improving the reproducibility of islet research and eventually the success of clinical islet transplants. However, the accuracy of the IEQ method has been questioned for years and many alternatives have been proposed, but none have been able to replace the widespread use of the IEQ. This article reviews the history of the IEQ, and discusses the benefits and failings of the measurement. A thorough evaluation of alternatives for estimating islet volume is provided along with the steps needed to uniformly move to an improved method of islet volume estimation. The lessons learned from islet researchers may serve as a guide for other fields of regenerative medicine as cell clusters become a more attractive therapeutic option.

Fully Automated Islet Cell Counter (ICC) for the Assessment of Islet Mass, Purity, and Size Distribution by Digital Image Analysis

For isolated pancreatic islet cell preparations, it is important to be able to reliably assess their mass and quality, and for clinical applications, it is part of the regulatory requirement. Accurate assessment, however, is difficult because islets are spheroid-like cell aggregates of different sizes (<50 to 500 μm) resulting in possible thousandfold differences between the mass contribution of individual particles. The current standard manual counting method that uses size-based group classification is known to be error prone and operator dependent. Digital image analysis (DIA)-based methods can provide less subjective, more reproducible, and better-documented islet cell mass (IEQ) estimates; however, so far, none has become widely accepted or used. Here we present results obtained using a compact, self-contained islet cell counter (ICC3) that includes both the hardware and software needed for automated islet counting and requires minimal operator training and input; hence, it can be easily adapted at any center and could provide a convenient standardized cGMP-compliant IEQ assessment. Using cross-validated sample counting, we found that for most human islet cell preparations, ICC3 provides islet mass (IEQ) estimates that correlate well with those obtained by trained operators using the current manual SOP method ( r² = 0.78, slope = 1.02). Variability and reproducibility are also improved compared to the manual method, and most of the remaining variability (CV = 8.9%) results from the rearrangement of the islet particles due to movement of the sample between counts. Characterization of the size distribution is also important, and the present digitally collected data allow more detailed analysis and coverage of a wider size range. We found again that for human islet cell preparations, a Weibull distribution function provides good description of the particle size.

Digital Image Analysis to Assess Quantity and Morphological Quality of Isolated Pancreatic Islets

Quantity and quality assessment of human pancreatic islets are essential processes to define a safe and potent quality product used for clinical transplantation. The conventional method of manual assessment has been used in the field for longer than two decades. The high degree of variability in product quantity and lack of archival imaging records of the product for verification are two major disadvantages of using the manual method for quantity and quality assessment of human pancreatic islets. Investigators have developed promising new methods for technical improvement. In this study, we briefly review the published methods and highlight the advantages of digital imaging analysis (DIA) when compared to the manual method. The application of DIA reduces measurement variability and increases the precision of islet equivalent (IEQ) determination for batch analysis. It produces images that can be archived for retrospective analysis and validation, and the data can be transmitted electronically for off-site analysis. These features are important for quality pancreatic islet assessment and are consistent with FDA requirements of current good manufacturing practice for clinical islet transplantation.

Development of a peptide-functionalized imaging nanoprobe for the targeting of (FXYD2)γa as a highly specific biomarker of pancreatic beta cells

Diabetes is characterized by a progressive decline of the pancreatic beta cell mass (BCM), which is responsible for insufficient insulin secretion and hyperglycaemia. There are currently no reliable methods to measure non-invasively the BCM in diabetic patients. Our work describes a phage display-derived peptide (P88) that is highly specific to (FXYD2)γa expressed by human beta cells and is proposed as a molecular vector for the development of functionalized imaging probes. P88 does not bind to the exocrine pancreas and is able to detect down to ~156 human pancreatic islets/mm(3) in vitro after conjugation to ultra-small particles of iron oxide (USPIO), as proven by the R2 measured on MR images. For in vivo evaluation, MRI studies were carried out on nude mice bearing Capan-2 tumours that also express (FXYD2)γa. A strong negative contrast was obtained subsequent to the injection of USPIO-P88, but not in negative controls. On human histological sections, USPIO-P88 seems to be specific to pancreatic beta cells, but not to duodenum, stomach or kidney tissues. USPIO-P88 thus represents a novel and promising tool for monitoring pancreatic BCM in diabetic patients. The quantitative correlation between BCM and R2 remains to be demonstrated in vivo, but the T2 mapping and the black pixel estimation after USPIO-P88 injection could provide important information for the future pancreatic BCM evaluation by MRI.

A Simple Method to Replace Islet Equivalents for Volume Quantification of Human Islets

Human islets come in a variety of sizes and shapes, and the total volume of islets used for research or clinical transplants must be estimated in a manner that is simple and valid. Islet equivalent (IEQ) measurements are the standard estimate of islet volume. We published a new method (the Kansas method) for estimating rat islet volume using cell numbers that was reliable and valid. Here we modified the method for human islets. We measured the dimensions of isolated human islets showing that they are not spherical and became less so in larger islets, with an average smallest/largest diameter ratio of 0.73 in large islets and 0.85 in small islets. Human islets were individually loaded into 96-well plates, dissociated into single cells, and the total cell number per islet determined with computer-assisted cytometry. Based on the counted cell number per islet, a regression model was created to convert islet diameter to cell number with a high R(2) value (0.99). Separate regression equations for male and female donors or young and old donors were not significantly different than the pooled data and did not improve the regression values. There was an inverse correlation between the cell number per IEQ and islet size. The Kansas method was validated with ATP/cell and cell viability data. Compared to the actual cell count, conventional IEQ measurements overestimated tissue volume of large islets by nearly double. Examples of differences in results obtained from the same data sets normalized to IEQ or the Kansas method included viability and insulin secretion concentrations. The implications of the error associated with the current IEQ method of volume estimation are discussed.

A replacement for islet equivalents with improved reliability and validity

Islet equivalent (IE), the standard estimate of isolated islet volume, is an essential measure to determine the amount of transplanted islet tissue in the clinic and is used in research laboratories to normalize results, yet it is based on the false assumption that all islets are spherical. Here, we developed and tested a new easy-to-use method to quantify islet volume with greater accuracy. Isolated rat islets were dissociated into single cells, and the total cell number per islet was determined by using computer-assisted cytometry. Based on the cell number per islet, we created a regression model to convert islet diameter to cell number with a high R2 value (0.8) and good validity and reliability with the same model applicable to young and old rats and males or females. Conventional IE measurements overestimated the tissue volume of islets. To compare results obtained using IE or our new method, we compared Glut2 protein levels determined by Western Blot and proinsulin content via ELISA between small (diameter≤100 μm) and large (diameter≥200 μm) islets. When normalized by IE, large islets showed significantly lower Glut2 level and proinsulin content. However, when normalized by cell number, large and small islets had no difference in Glut2 levels, but large islets contained more proinsulin. In conclusion, normalizing islet volume by IE overestimated the tissue volume, which may lead to erroneous results. Normalizing by cell number is a more accurate method to quantify tissue amounts used in islet transplantation and research.

Quantitative histopathological assessment of retardation of islets of langerhans degeneration in rosiglitazone-dosed obese ZDF rats using combined insulin and collagens (I and III) immunohistochemistry with automated image analysis and statistical modeling

Islets of Langerhans represent a heterogeneous population in insulin resistant and diabetic animals and humans as histological appearances and function vary substantially. Mathematical representation that reflects this morphological diversity will assist in assessment of degeneration and regeneration, enabling comparisons between species, strains, and experimental investigations. Our investigative approach used a model of islet degeneration in diabetic male obese Zucker Diabetic Fatty (ZDF) rats and evaluated its prevention using rosiglitazone treatment. Immunohistochemical staining (insulin and collagens I/III) with automated image analysis reliably measured numbers, area, clustering, and staining intensity of β-cells and degree of islet fibrosis. Finite mixture mathematical modeling for the joint probability distribution of seven islet parameters to represent islet numerical data variation provided an automatic procedure for islet category allocations as normal or abnormal. Allocations for obese ZDF controls and rosiglitazone-treated animals were significantly different, with no significant difference between the latter and lean ZDF controls, indicative of differences within islet populations of individual animals, between lean and obese rat strains and following drug treatment. Islet morphology showed clear association with mathematical characterization. Information on islet morphology obtained by histopathological assessment of single pancreatic tissue sections was confirmed by this method showing drug-induced retardation of islet of Langerhans degeneration.

Quantification of the islet product: presentation of a standardized current good manufacturing practices compliant system with minimal variability

BACKGROUND

Accurate islet quantification has proven difficult to standardize in a good manufacturing practices (GMP) approved manner.

METHODS

The influence of assessment variables from both manual and computer-assisted digital image analysis (DIA) methods were compared using calibrated, standardized microspheres or islets alone. Additionally, a mixture of microspheres and exocrine tissue was used to evaluate the variability of both the current, internationally recognized, manual method and a novel GMP-friendly purity- and volume-based method (PV) evaluated by DIA in a semiclosed, culture bag system.

RESULTS

Computer-assisted DIA recorded known microsphere size distribution and quantities accurately. By using DIA to evaluate islets, the interindividual manually evaluated percent coefficients of variation (CV%; n=14) were reduced by almost half for both islet equivalents (IEs; 31% vs. 17%, P=0.002) and purity (20% vs. 13%, P=0.033). The microsphere pool mixed with exocrine tissue did not differ from expected IE with either method. However, manual IE resulted in a total CV% of 44.3% and a range spanning 258 k IE, whereas PV resulted in CV% of 10.7% and range of 60 k IE. Purity CV% for each method were similar approximating 10.5% and differed from expected by +7% for the manual method and +3% for PV.

CONCLUSION

The variability of standard counting methods for islet samples and clinical quantities of microspheres mixed with exocrine tissue were reduced with DIA. They were reduced even further by use of a semiclosed bag system compared with standard manual counting, thereby facilitating the standardization of islet evaluation according to GMP standards.

Reduced expression of PDX-1 is associated with decreased beta cell function in chronic pancreatitis

OBJECTIVES

The present study was conducted to monitor the expression of pancreas and duodenal homeobox gene (PDX-1) for assessing beta-cell function in islets from patients with chronic pancreatitis (CP).

METHODS

Islets isolated from the pancreata of 40 surgical patients categorized as control group, patients with mild CP, and patients with advanced CP were assessed for their yield, size, and glucose-stimulated insulin secretion. Expressions of genes coding for PDX-1, insulin, and glucagon were simultaneously monitored by reverse transcription polymerase chain reaction and confirmed by immunohistochemistry.

RESULTS

In comparison with the control group (2673 +/- 592 islet equivalents [IEq]/g), islet yield did not differ much in the patients with mild CP (2344 +/- 738 IEq/g) but was significantly reduced (P < 0.0001) in the patients with advanced CP (731 +/- 167 IEq/g). Although the marginal decrease in islet size observed in the patients with mild CP was not significantly different from that observed in the control group, there was a 58% decrease observed in the patients with advanced CP that was also accompanied by a significant reduction in beta-cell mass (P < 0.05). The expression of insulin and PDX-1 genes, but not of glucagon, was significantly reduced in the patients with advanced CP as confirmed by immunohistochemistry. Islets obtained from the patients with advanced CP retained 53% glucose-stimulated insulin secretion function in comparison with those of the control group.

CONCLUSION

The results indicate that beta-cell dysfunction during progression of CP correlates with the decrease in PDX-1 gene expression.

Validation of methodologies for quantifying isolated human islets: an Islet Cell Resources study

BACKGROUND

Quantification of islet mass is a crucial criterion for defining the quality of the islet product ensuring a potent islet transplant when used as a therapeutic intervention for select patients with type I diabetes.

METHODS

This multi-center study involved all eight member institutions of the National Institutes of Health-supported Islet Cell Resources Consortium. The study was designed to validate the standard counting procedure for quantifying isolated, dithizone-stained human islets as a reliable methodology by ascertaining the accuracy, repeatability (intra-observer variability), and intermediate precision (inter-observer variability). The secondary aim of the study was to evaluate a new software-assisted digital image analysis method as a supplement for islet quantification.

RESULTS

The study demonstrated the accuracy, repeatability and intermediate precision of the standard counting procedure for isolated human islets. This study also demonstrated that software-assisted digital image analysis as a supplemental method for islet quantification was more accurate and consistent than the standard manual counting method.

CONCLUSIONS

Standard counting procedures for enumerating isolated stained human islets is a valid methodology, but computer-assisted digital image analysis assessment of islet mass has the added benefit of providing a permanent record of the isolated islet product being evaluated that improves quality assurance operations of current good manufacturing practice.

The current status of islet transplantation and its perspectives

Transplantation of human pancreatic isolated islets can restore beta-cell function but it requires chronic immunosuppression. The outcome of islet transplantation mainly depends on both the quality of islet preparations, and the survival of the graft. The quality of islet preparations can be evaluated by the results of isolation, which determines the chance to achieve insulin independence. The survival of islet grafts is reflected by the amount of engrafted functional tissue that maintains metabolic control. Immunosuppressive therapy prevents the immunological rejection of grafts, but impairs their function and impedes their regenerative capacity. Therefore, the selection of high quality islet preparations and the reduction of toxic effects of immunosuppressive regimens might dramatically improve the outcomes. The application of stem cell therapy in islet transplantation may contribute to a better understanding of the mechanisms responsible for tissue homeostasis and immune tolerance. Xenogeneic islets may serve as an unlimited source if immune tolerance can be achieved. This may be a strategy to enable a substantial improvement in function while overcoming potentially deleterious risks.