ZO-1 is a 210- to 220-kDa peripheral membrane protein associated with the cytoplasmic surface of the epithelial tight junction. Because ZO-1 may interact with other unidentified tight junction proteins, we have looked for other polypeptides that bind to ZO-1. A 160-kDa polypeptide was identified that coimmunoprecipitates with ZO-1 from detergent extracts of metabolically labeled Madin-Darby canine kidney (MDCK) cells. This polypeptide appears to be distinct from ZO-1, rather than a degradation product, by several criteria. It lacks ZO-1 epitopes recognized by both monoclonal antibodies and a polyclonal serum to ZO-1, since it is not detectable in immunoblots of either whole cell extracts or ZO-1 immunoprecipitates. Also, it exhibits a peptide map different from that of ZO-1 on one-dimensional "Cleveland gels." Moreover, because the kinetics of appearance of newly synthesized 160-kDa polypeptide in anti-ZO-1 immunoprecipitates is much slower than that of ZO-1, its presence in immunoprecipitates cannot be simply explained by degradation of ZO-1 during cell lysis. Like ZO-1, the 160-kDa polypeptide seems to be a cytoplasmic peripheral membrane protein. It cannot be labeled by two different cell surface labeling reagents. It can be extracted from the membrane by high salt concentration in the absence of detergents. As expected for a protein complex, the 160-kDa polypeptide and ZO-1 turn over with similar kinetics. We propose that the 160-kDa polypeptide is a component of the tight junction.