Protein patterns of normal human colonic crypts, isolated from different regions of the large intestine, and several colorectal cancer cell lines were compared using two-dimensional electrophoresis gels (2-DE). As detected by intrinsic radiolabeling and Coomassie Brilliant Blue staining, the protein patterns for normal crypts isolated from the ascending, and descending, regions of the colon and the rectum, were almost (> 95%) identical. While 75-80% of the protein spots from normal crypts and the colorectal cancer cell line (LIM 1863), a cell line that grows as organoids and differentiates spontaneously into crypt-like structures in vitro, can be matched, the relative expression levels of a large number of proteins differ. At least two protein spots (undetectable in the protein pattern from normal cells), proteins a (M(r) approximately 18,000, pI 6.7-6.9) and b (M(r) approximately 24,000, pI 5.9-6.0), were detected in the 2-DE gel protein pattern in the three cell lines LIM 1863, LIM 1215 and LIM 1899. The identity of these proteins is not yet known and further studies are required before they can be considered as potential colon tumor markers. Approximately 60% of the cellular proteins from LIM 1215 cells, a colon carcinoma cell line that exhibits many properties associated with columnar cells, can be matched with LIM 1863 cells. The results presented here represent an initial phase in our efforts to develop a comprehensive protein database for normal human colon cells and several colorectal cancer cell lines. While our initial protein identification relied on microsequencing methodologies, we are presently evaluating peptide-mass fingerprinting, utilizing capillary reversed-phase high-performance liquid chromatography (RP-HPLC) and electrospray mass spectrometry, as a means for rapid identification of proteins at subpicomole levels. Using this approach, protein #3 (M(r) approximately 66,000, pI 6.2) was identified as heat shock protein 60 from as few as seven tryptic peptide masses when they were screened against the molecular weight search (MOWSE) peptide-mass database.