Abstract
Optimal human islet isolation requires the delivery of bacterial collagenase to the pancreatic islet–exocrine interface. However, we have previously demonstrated the presence of collagenase within human islets immediately following intraductal collagenase administration. This potentially has significant implications for patient safety. The present study aimed to determine if collagenase becomes internalized into islets during the isolation procedure and if it remains within the islet postisolation. Islet samples were taken at various stages throughout 14 clinical human islet isolations: during digest collection, following University of Wisconsin solution incubation, immediately postisolation, and after 24 h of culture. Samples were embedded in agar, cryosectioned, and then assessed by immunolabeling for collagenase and insulin. Immunoreactivity for collagenase was not observed in isolated islets in any preparation. Collagenase labeling was detected in one sample taken at the digest collection phase in one islet preparation only. No collagenase-specific labeling was seen in islets sampled at any of the other time points in any of the 14 islet preparations. Collagenase that enters islets during intraductal administration is washed out of the islets during the collection phase of the isolation process and thus does not remain in islets after isolation. This observation alleviates some of the important safety concerns that collagenase remains within islet grafts.
Introduction
Optimal human islet isolation requires the delivery of bacterial collagenase to the pancreatic islet–exocrine interface in order to digest the extracellular matrix and release intact islets. However, we have previously demonstrated the presence of collagenase within human islets immediately following intraductal collagenase administration using standard, widely used techniques (10). This could lead to internalization of enzyme into the cells of the islet, as has been shown in a study using fluorescein isothiocyanate (FITC)-conjugated Liberase and mouse islets (3). This in turn could cause overdigestion of islets resulting in islet loss but, more significantly, could have major implications for patient safety if such islets are transplanted. In addition, residual collagenase could impair graft function and may result in the instant blood-mediated inflammatory reaction (IBMIR) or other inflammatory or immunological reactions in the recipient. Furthermore, the transplantation of islets with intracellular collagenase has implications for islet integrity and islet graft survival.
Tissue dissociation enzymes for islet isolation have been extensively examined to optimize isolation efficiency (1,2,5,6,8,11). However, whether these enzymes are present within islet grafts following islet transplantation remains unknown. Therefore, the aims of this study were to determine if collagenase persists in islet spaces during the isolation procedure and remains within the islet postisolation.
Materials and Methods
Islet Isolation and Sampling Technique
With appropriate consent and ethical approval, human pancreases (n = 14) were recovered from multiorgan donors (for donor and islet isolation characteristics, see Table 1). Islets were isolated by intraductal enzyme infusion using collagenase NB1 (with activity between 2101 and 2965 PZ-U) and neutral protease NB (50 DMC-U, Serva, Heidelberg, Germany). The enzyme mixture was administered into the pancreatic duct by manual syringe loading. Injection speed was adjusted, aiming to maintain pressures of 80 mmHg for 5 min, slowly increasing to 180 mmHg for the subsequent 10 min, as stated in the Edmonton Protocol (14). Pressures were monitored using a single pressure monitoring transducer (Smiths Medical, Lancs, UK) and pressure monitor (Hewlett Packard, UK). Enzyme infusion was followed by automated digestion and continuous density gradient purification, as described previously (4,13). Tissue samples were taken from the body of the pancreas immediately after collagenase infusion and snap frozen in liquid nitrogen. Islet samples were fixed in 4% paraformaldehyde (Sigma, Poole, UK) at various stages throughout the isolation process: during digest collection, following 1 h of University of Wisconsin (UW) solution incubation, immediately postisolation, and then after 24 h of culture in Connaught Medical Research Laboratories (CMRL)-1066 medium (PAA Laboratories, Pasching, Austria) at 37°C. A total of 14 islet preparations were used in the study (Table 1), although samples were not collected from each stage in all preparations.
Donor and Islet Isolation Characteristics
BMI, body mass index; CIT, cold ischemia time; IEQ, islet equivalent.
Distension quality score: the quality of pancreas distension following enzyme infusion was observed macroscopically and scored from 1 to 4: 1 = excellent distension, 2 = good distension, 3 = satisfactory distension, 4 = poor distension.
Islet Processing and Immunolabeling
The fixed digest material and islets were stained with 2% eosin, in order to aid their identification, and embedded in a warm solution of 2% agar in phosphate-buffered saline followed by cooling in a refrigerator. Frozen sections (8 μm thick) of the collagenase-infused pancreas samples and the fixed digest material and islets in agar were cut onto coated slides. Double immunofluorescent labeling for collagenase and insulin was performed. Sections were incubated with 10% normal goat serum (Sigma) for 30 min, followed by primary antibodies, overnight at 4°C: polyclonal rabbit anti-collagenase (1:1,000; Cortex Biochem, San Leandro, CA, USA) and polyclonal guinea pig anti-insulin (1:100; Sigma). Sections were then incubated with secondary antibodies for 1 h at room temperature, Alexa Fluor 488 goat anti-rabbit (1:500; Invitrogen, Paisley, UK) and Texas red-conjugated goat anti-guinea pig (1:50; Sigma). Negative control sections were incubated with normal rabbit IgG in place of the collagenase antibody. Images were captured using a Zeiss LSM510 confocal laser scanning microscope (Carl Zeiss, Jena, Germany).
Exposure of Isolated Islets and Frozen Tissue Sections to Collagenase
To confirm immunoidentification of collagenase in islets, in vitro exposure to the enzyme was used. Following 24 h of culture postisolation, islets were divided into two groups: one group was treated with Serva collagenase and neutral protease (1.15 and 0.14 mg/ml, respectively, in Hank's balanced salt solution, HBSS), and the second group was treated with HBSS alone for 1 h at 37°C. The islets were washed with HBSS and fixed in 4% paraformaldehyde, then sectioned, and immunolabeled as above.
To ensure we could detect both intact and potentially degraded collagenase using the collagenase antibody, Serva collagenase and neutral protease (concentrations as above) were incubated at conditions chosen to mimic that of the isolation process: 20 min at 4°C, followed by 30 min at 37°C and 60 min at room temperature (referred to from now on as “preincubated enzyme”). Frozen pancreas sections were then incubated with either fresh or preincubated enzyme for 5 min, fixed, and then labeled for collagenase and insulin as described above.
Results
Collagenase labeling was widespread throughout the exocrine pancreas as well as being present around the periphery of and within the islets in samples of pancreas taken immediately following intraductal collagenase administration (Fig. 1a), confirming in this study what we have already reported previously (10).
(a) Collagenase labeling (green) is seen peripheral to as well as within the islet (red, insulin labeling of β-cells) in samples of collagenase-infused pancreas. (b) Collagenase labeling is detected adjacent to islet cells in samples taken during digest collection in only 1 of 14 islet preparations. No collagenase labeling was seen in samples taken after 1 h of University of Wisconsin (UW) solution incubation (c), from the final preparation following digest purification (d), and after overnight culture of isolated islets (e). Isolated islets exposed to collagenase for 1 h at 37°C were positive for collagenase labeling (f). Collagenase labeling is widespread throughout the pancreas sections, adjacent to islet cells, with no difference in labeling seen between pancreas sections treated with fresh (g) or preincubated (h) collagenase. Scale bars: 50 μm.
Samples from 14 islet preparations were analyzed for immunolabeling; collagenase labeling was detected in one sample taken from the first pool of digest material during the collection phase, in one preparation only (Fig. 1b). No collagenase-specific labeling was seen in islets sampled at any of the other time points during and after the islet isolation process (Fig. 1c–e, Table 2). Islets labeled with a control, species-matched antibody showed no specific fluorescence.
Collagenase-Specific Labeling of Samples Taken During the Islet Isolation Process and After Islet Culture
A total of 14 islet preparations were studied, although samples were not collected from each stage in all preparations. UW, University of Wisconsin.
As a positive control for collagenase labeling, isolated islets were exposed to collagenase for 1 h at 37°C and then immunolabeled for collagenase. Collagenase labeling in the extracellular spaces was detected (Fig. 1f), indicating that the enzyme can bind to substrates in islets. No labeling was seen in islets exposed to HBSS alone.
In an additional positive control, collagenase labeling was found in pancreas sections treated with both fresh and preincubated enzyme, suggesting that the collagenase antibody is able to recognize intact (Fig. 1g) and potentially degraded collagenase (Fig. 1h).
Discussion
We have previously shown that intraductally administered collagenase is able to enter extracellular spaces of human islets (10). This finding had potentially significant implications for patient safety, especially as over recent years there have been considerable concerns over the fact that mammalian neural tissue had been used during the manufacture of the collagenase Liberase. This study was therefore an important one, aimed at determining whether collagenase persists within islets during and after the isolation process.
Our results indicate that collagenase is not present within islets following isolation. No collagenase was detected in islets sampled at and subsequent to the UW solution incubation stage, which suggests that collagenase is washed out of the islets during the dilution and collection phases of the isolation process. This helps to resolve some of the safety concerns that collagenase may still be present within islet grafts. However, the presence of collagenase within the extracellular spaces of islets immediately following intraductal administration may cause damage to islets during the digestion process and therefore have a negative impact on islet integrity and islet graft survival. During the digestion process, collagenase will be active for a prolonged period of time. Dissociation followed by dilution and collection can take up to 2 h, and although a decrease in temperature and substantial dilution will help to reduce collagenase activity and concentration, respectively, islet integrity could be compromised, particularly in those islets that are liberated later on during the digestion process, and so have been exposed to collagenase at temperatures at which it is active for longer.
Both our study and that of Balamurugan et al. (3) found that collagenase enters human islets if exposed to it postisolation; however, our results suggest that, even if this does happen in the pancreas immediately following enzyme infusion, any collagenase that may have bound will have been washed away by the time at which the islets are collected after pancreas digestion. Conversely, collagenase has been found to remain in isolated mouse islets when they are isolated using fluorescein-labeled Liberase (3). The discrepancy between these results and those from our study using human islets may be due to differences in the islet isolation procedure between the mouse and human pancreas, in addition to the variation in the structure of the islet–exocrine interface and differences in islet morphology between human and mouse pancreas (7,9,12).
Although we only investigated the presence of Serva collagenase NB1 within human islets, due to its exclusive use for clinical islet isolation in the UK, it is unlikely that there would be significant differences in results using other brands of commercially available collagenase that are produced in a similar way from Clostridium histolyticum and are delivered into the human pancreas in the same manner.
In summary, the present study shows that collagenase is not present in isolated islets. Therefore, safety concerns that bacterial products may still be present in transplanted islets and are able to contribute to IBMIR, inflammatory, or immunological reactions in the recipient are negated.
Footnotes
Acknowledgments
This study was supported in part by grants from the Diabetes Research and Wellness Foundation (S.E.C., S.J.H., and P.R.V.J.) and the National Institute for Health Research Biomedical Research Centre, Oxford (P.R.V.J.). This work was carried out in the Nuffield Department of Surgical Sciences and the Oxford Centre for Diabetes, Endocrinology, and Metabolism, both at the University of Oxford. The authors would like to thank the members of the Oxford Human Islet Isolation Team for their excellent work in the isolation facility and Professor Patrik Rorsman for use of the confocal microscope. S.E.C., S.J.H., D.W.R.G., and P.R.V.J. participated in research design. S.E.C., S.J.H., A.C., D.W.R.G., and P.R.V.J. participated in the writing of the manuscript. S.E.C. and S.J.H. participated in the performance of the research. A.C. and P.R.V.J. contributed reagents and analytical tools. S.E.C. and A.C. participated in data analysis. The authors declare no conflict of interest.