Selective enterocyte transplantation is a potential alternative to small intestinal transplantation for treatment of the short bowel syndrome. Our aim was to compare chelation and enzymatic methods of isolating enterocytes with respect to initial cell yield and characteristics and in vitro growth. Enterocytes were harvested from adjacent ileal segments in 35 rabbits using chelation with EDTA and warm trypsinization. Determinations were made of initial viability by trypan blue exclusion, cell yield, and proportion of intact crypts. Cells (5 x 10(6)) were seeded in growth media into culture vessels. Cell attachment was estimated by measuring cells liberated by Dispase at 24 hr. In vitro growth was assessed at 14 and 28 days. Although total cell yield (15.0 +/- 9.9 vs 12.5 +/- 8.3 x 10(6) cells/cm) and intact crypts were similar, the trypsin technique resulted in cells with higher initial viability (86 +/- 7 vs 71 +/- 18%, P < 0.05). Cell attachment (7 +/- 8 vs 8 +/- 4%) and enterocyte disaccharidase activity were similar using both techniques. While the number of epithelial cell growth foci was not significantly different at 14 days, there was significantly greater surface coverage on both plain (44 +/- 20% vs 1 +/- 0%, P < 0.05) and Matrigel-coated (80 +/- 14 vs 16 +/- 25%, P < .05) vessels at 28 days with trypsin-isolated cells. Trypsinization resulted in a cell population which had a higher percentage of viable cells but a similar proportion of intact crypts and differentiated cells compared to those resulting from the chelation technique. Trypsin-liberated cells had greater capacity for in vitro growth particularly on Matrigel-coated surfaces.(ABSTRACT TRUNCATED AT 250 WORDS)