The insufficient consistency of enzyme blends is one of the main reasons for the low standardization of enzymatic isolation of human islets. Although manufacturers succeeded to provide purified products consisting solely of collagenase class I and II (CI and CII), and a complementary protease, such as neutral protease or thermolysin, these enzyme blends are still characterized by a lot-dependent variability (1). Previous retrospective analysis proposed that the efficiency of enzyme blends is mainly determined by the integrity of CI required to initiate effective cleavage of collagen fibers within the periinsular extracellular matrix (2–4). This notion is supported by the observation that degradation of CI is associated with the loss of at least one of the total two collagen-binding domains resulting in an isoform of 100 kDa that has a specific collagen-degrading activity of only 10% compared with intact (115 kDa) isomers (5). Because no prospective isolation study has been performed yet, the present investigation was aimed to clarify whether intact (CI-115) and truncated (CI-100) CI isomers are capable of participating in the release of islets from human pancreatic tissue. All enzymes were provided as a modular system by Serva/Nordmark (Heidelberg/Uetersen, Germany). The raw Clostridium histolyticum product was purified and chromatographically separated into fractions of tryptic-like activity, neutral protease, CI-115, and CI-100 that were recombined with purified CII at a CII-to-CI ratio of 0.7 (Fig. 1). Characterization by reversed-phase high-performance liquid chromatography and the Wunsch assay demonstrated a purity of 92%, 93%, and 89% for Cl-115, Cl-100, and CII, and an amidolytic activity of 0.072, 0.074, and 8.5 4-phenylazobenzyloxycarbonyl-l- prolyl-l-leucylglycyl-prolyl-d-arginine U/ mg, respectively. All fractionated classes were produced from one single batch.FIGURE 1.: Mono-Q high-performance liquid chromatography analysis of the collagenase class I isomers of 100 kDa (class I-100, top) and 115 kDa (class l-115, middle), and collagenase class II (bottom) performed at a wavelength of 280 nm.The blends were supplemented with 1.5 U dimethyl-casein/g of neutral protease and 2.6 U benzoyl-l-arginine-ethyl-ester/g of tryptic-like activity, dissolved in 1.5 mL/g of Hank's balanced salt solution, and intraductally injected into research grade pancreata using a CII activity of 25 U 4-phenylazobenzyloxycarbonyl-l-prolyl-l-leucylglycyl-prolyl-d-arginine/g.islets were isolated as previously described (6). After culture for 2 to 3 days at 37°C, islets were assessed for insulin release during static glucose incubation (2 vs. 20 mM) and for viability (Syto 13/ethidium bromide). Donor age (57.8±4.5 vs. 58.9±3.8), body mass index (23.1±1.2 vs. 24.9±0.4), pancreas weight (60.4±8.6 g vs. 74.1 ±14.1 g), and cold ischemia time (18.4±4.1 hr vs. 14.7±3.3 hr) did not differ between CI-115 (n=8) and CI-100 (n=8) (mean±SE, not significant by Mann-Whitney U test). The proportion of male-to-female donors (3:5 vs. 5:3) and histidine-tryptophan- ketoglutarate to University of Wisconsin solution perfusion (3:5 vs. 2:6) were approximately the same in pancreata processed by means of CI-115 and CI-100 (not significant by Fisher's exact test). No significant differences were observed between CI-115 and CI-100 regarding recirculation time (28.0±2.2 min vs. 24.0±2.3 min), percentage of undigested tissue (8.1±1.1% vs. 12.5±2.0%) and embedded islets (15.0±2.2% vs. 15.3± 5.2%), purity (52.8±4.9% vs. 48.6± 8.9%), and islet yield per gram (2340±320 islet equivalent/g vs. 3010±640 islet equivalent/g) equivalent to a postpurification recovery of 50.7±9.1% and 64.6±7.8%, respectively. Islet morphology and size distribution were similar in both the experimental groups (data not shown). Islet survival (67.1%±7.5% vs. 64.6%±5.9%) and purity (48.5%±4.7% vs. 41.7%± 8.7%) determined postculture were also comparable. Quality assessment indicated no detrimental effect of CI-100 on islet viability (74.7±1.3 vs. 78.1±2.8), insulin content (331±103 μU/ng DNA vs. 255±93 μU/ng DNA), or glucose stimulation index (2.5±0.5 vs. 3.9±0.7) when compared with CI-115. The rapid release of islets from within acinar tissue using effective enzymes is of significant importance to recover islets from the harmful environment present during pancreas digestion to preserve morphologic and metabolic integrity. CI-100 seems to have the same capacity to dissociate human pancreatic tissue and to release islets as CI-115. This is in contrast to previous retrospective analysis (2, 3). Nevertheless, we cannot exclude that a potential deficiency of CI-100 to cleave collagen fibers in the perinsular extracellular matrix is compensated by complementary proteases, such as neutral protease, as demonstrated in rat islet isolation (7). A similar auxiliary function can be discussed for tryptic-like activity identified as clostripain (E.C. 3.4.22.8), which efficiently reduced digestion time in rat and human islet isolation (8). Most importantly, CI-100 was not associated with potential islet-harmful side effects as observed for other low-molecular-weight collagenase isomers that have been regularly removed from final collagenase NB1 preparations (9). This is supported by the observation that in vitro function, viability, or survival postculture was not reduced after CI-100 utilization. In conclusion, the present data indicate that the presence of collagenase CI-100 does not prevent successful islet isolations from the human pancreas. Whether this finding is related to a potential compensatory function of noncollagenolytic proteases has to be clarified in future studies using a model of marginal enzyme activities. Heide Brandhorst1 Sana Asif1 Karin Andersson1 Johanna Mönch2 Olaf Friedrich3 Nicole Rämsch-Günther2 Christian Rämsch3 Melanie Steffens2 Jörg Lambrecht3 Thomas Schräder3 Manfred Kurfürst3 Helene H. Andersson4 Marie Felldin5 Aksel Foss6 Kaija Salmela7 Annika Tibell8 Gunnar Tufveson9 Olle Korsgren1 Daniel Brandhorst1 1 Department of Oncology, Radiology, and Clinical Immunology Uppsala University Uppsala, Sweden 2 Serva Electrophoresis GmbH Uetersen, Germany 3 Nordmark Arzneimittel GmbH & Co. KG Uetersen, Germany 4 Department of Nephrology and Transplantation University Hospital Malmö, Sweden 5 Department of Transplantation University Hospital Gothenburg, Sweden 6 Division of Surgery, Section for Transplantation Oslo University Hospital Rikshospitalet, Oslo, Norway 7 Division of Transplantation, Surgical Hospital Helsinki University Helsinki, Finland 8 Division of Transplantation Surgery, CLINTEC Karolinska Institute Stockholm, Sweden 9 Department of Surgical Sciences, Division of Transplantation Surgery Uppsala University Hospital Uppsala, Sweden

BACKGROUND:Previous studies indicated different roles of collagenase class I, class II and neutral protease in the enzymatic islet release from pancreatic tissue. Because no information has been available, this study was aimed to investigate the isolation efficiency of different ratios between collagenase class II and I (C-ratio) in the rat pancreas serving as model for the human pancreas without being restricted by the large variability observed in human donors.METHODS:Rat pancreata were digested using a marginal neutral protease activity and 20 PZ-U of purified collagenase classes recombined to create a C-ratio of 0.5, 1.0, or 1.5. Collagenase efficiency was evaluated in terms of isolation outcome and posttransplantation function in diabetic nude mice.RESULTS:The highest yield of freshly isolated islets was obtained using a C-ratio of 1.0. Purity and fragmentation of freshly isolated islets were not influenced by the C-ratio. After 24-hr culture performed for quality assessment, a marginal but significant reduction of viability was observed in islets isolated by means of a C-ratio of 0.5 and 1.5. Islet in vitro and posttransplantation function revealed no negative effect mediated by different C-ratios.CONCLUSIONS:The present study demonstrates that the C-ratio is of significant relevance for the outcome after enzymatic rat islet isolation. The data indicate further that purified collagenase class I or class II does not damage islet tissue even if used in excess. The present study can serve as a start for subsequent experiments in the human pancreas.

OBJECTIVE:The utilization of purified enzyme blends consisting of collagenase class I (CI) and II (CII) and neutral protease is an essential step for clinical islet isolation. Previous studies suggested that the use of enzyme lots containing degraded CI reduced islet release from human pancreata. The present study sought to assess the effect of degraded collagenase on islet function in vitro and posttransplantation.MATERIALS AND METHODS:Crude collagenase was chromatographically separated into CI, CII, and a mixture of degraded CI and CII isomers. Subsequently, classes were recombined to obtain a CII/CI ratio of 0.5. Rat islets were isolated utilizing neutral protease and 20 units of recombined collagenase containing either intact (Ci) or degraded isomers (Cd).RESULTS:Digestion time was reduced utilizing Cd (P < .001). The highest islet yield and lowest islet fragmentation were obtained with Ci (P < .01). Utilization of Cd corresponded to a reduction in viability and in vitro function (NS). Islet transplantation reversed hyperglycemia in diabetic nude mice, but revealed an absence of weight gain in recipients receiving islets isolated using Cd (P < .01).CONCLUSION:This study suggested that islet function posttransplantation is affected by degraded collagenase isomers. This finding has to be considered for the purification process of collagenase.

Background. Biologic characteristics of enzyme products used for islet isolation are critical for the success of islet transplantation. In particular, lot-to-lot variability significantly affects the yields of the isolation procedure. In this study, we have evaluated a new enzyme preparation in which highly purified collagenase can be blended with separately supplied neutral protease in a predetermined ratio.Methods. We compared the results of human-islet isolations performed either with Collagenase NB1 supplemented with Neutral Protease NB (group I, n = 9) or with Liberase (group II, n = 9).Results. Endotoxin contents of the enzyme preparations were lower in group I. Total islet yields were similar in both groups, but islet equivalents per gram of pancreas was higher in group I (4,020 +/- 1,240 vs. 2,360 +/- 1,350; P < 0.05). Islet morphology was improved in group I with significantly higher proportion of free and intact islets (71 +/- 9% vs. 52 +/- 14%; P < 0.01). In vitro function was improved and apoptosis rate was lower in group I.Conclusions. This new enzyme blend was as efficient as Liberase in terms of islet yields and showed improvements in islet morphology, viability, and in vitro function. The possibility to control lot-to-lot variability, and the low endotoxin contents make Collagenase NB1 a promising product for human-islet isolation.

Transient receptor potential proteins (TRP) are supposed to participate in the formation of store‐operated Ca2+ influx channels by co‐assembly. However, little is known which domains facilitate the interaction of subunits. Contribution of the N‐terminal coiled‐coil domain and ankyrin‐like repeats and the putative pore region of the mouse TRP1β (mTRP1β) variant to the formation of functional cation channels were analyzed following overexpression in HEK293 (human embryonic kidney) cells. MTRP1β expressing cells exhibited enhanced Ca2+ influx and enhanced whole‐cell membrane currents compared to mTRP1β deletion mutants. Using a yeast two‐hybrid assay only the coiled‐coil domain facilitated homodimerization of the N‐terminus. These results suggest that the N‐terminus of mTRP1β is required for structural organization thus forming functional channels.