Quantitative assessment of collagenase blends for human islet isolation

Background. The variability in collagenase blends has been speculated as the single most important determinant of the success or failure in isolated islet yields in clinical islet transplantation. Examination of the formulation and potency of the widely used Liberase HI enzyme blend will uncover possible sources of imprecision. Methods. High performance liquid chromatography (HPLC) and kinetic measurements of collagenase and protease activity were used to assess potency. Between four and nine clinical lots were assessed for various parameters such as relative formulation of collagenase isoforms, and recovered collagenase and protease potencies postreconstitution. Results. Six vials from a single typical lot had a mean enzyme content of 489 +/- 62.5 mg (mean +/- SEM; range 398 - 610 mg). The mean recovered collagenase activity was 2235 +/- 310 Wimsch units (WU)/vial (range 1794-2968 WU/vial). The percent coefficients of variation for collagenase and protease activity in these vials were 17.4%, and 13.4%, respectively. The increase in the presence of the collagenase Ib (CIb) isoform detected by HPLC analysis was related to the chronological order of the date of manufacture. The CIb isoform was found to have a reduced specific activity compared to intact collagenase I (CI) (3.8 +/- 1.2 WU/mg vs. 2.1 +/- 0.7 WU/mg, P<0.05). The presence of CIb was related to reduced islet yields in twelve human isolations studied. Conclusions. Variation in potency was observed between, and within lots of Liberase HI in this study. Differences in relative collagenase isoform composition may also affect the stability and potency characteristics of these blends.