We have investigated a potential relationship between expression of beta 1-integrins and adhesiveness of the beta 2-integrin LFA-1 (alpha L beta 2, CD11a/CD18). By an approach of random mutagenesis and selection we established clones from the human acute lymphatic leukemia cell line HPB-ALL with (i) constitutively active LFA-1 and (ii) with no apparent integrin-beta 1 cell surface expression. Thirty-seven of 42 clones selected for activated LFA-1 were found to have lost apparent integrin-beta 1 expression. Conversely, 7 of 21 clones selected for lack of beta 1 expression were found to have activated LFA-1. Since this pointed toward a possible coupling between beta 1 expression and LFA-1 activity, we further analyzed at which level beta 1 expression was blocked. We focused on one clone, HAP4, with activated LFA-1 and no detectable beta 1 cell surface expression and found, surprisingly, that it expressed wild-type levels of beta 1 mRNA and, in Western blots of whole cell lysates, apparently normal levels of beta 1 protein. However, in addition to beta 1 of the expected molecular weight, HAP4 expressed a unique 48-kDa band recognized by the polyclonal anti-beta 1 antiserum. Immunoprecipitation experiments revealed that the epitope recognized by the anti-beta 1 antibody 4B4 was hidden or lost. The alpha 4-chain was found in its precursor form but it did not associate with any beta-chain, and it was not processed to its mature form. Instead alpha 4-chains were eventually degraded. Taken together this showed that beta 1-chains were produced but not properly processed in HAP4. From this we propose that HAP4 is deficient in a gene product required both for proper beta 1 folding and for repression of LFA-1 adhesiveness.