The transmembrane forms of all immunoglobulin (Ig) classes are associated with two glycoproteins, mb-1 and B29, that are crucial for signal transduction following antigen binding to the Ig molecule. We have investigated the transcription and protein expression of mb-1 and B29 genes during B-cell development. Sixty human continuous cell lines (35 B-lineage, 11 T-lineage, 11 myeloid-lineage and three non-hematopoietic) and 75 hematopoietic malignancies (55 B-lineage, 12 T-lineage and eight myeloid-lineage), were tested for RNA expression by Northern blotting experiments with the mb-1 pRA3 cDNA probe, and a newly isolated B29 cDNA probe. Protein expression was analyzed by immunofluorescence microscopy of cytocentrifuge preparations, which were labeled with the anti-mb-1 HM57 monoclonal antibody (mAb) and an anti-B29 polyclonal antiserum, directed against intracellular epitopes of these polypeptides. Except for two early precursor B-cell lines, mb-1 and B29 transcripts and proteins were detected in all B-cell lines and B-cell malignancies, i.e. from immature to more mature B cells, irrespective of their Ig class expression. Transcription of mb-1 genes seems to be down-regulated at the plasma cell stage, because no mb-1 transcripts and mb-1 proteins could be detected in the four plasma cell lines and two plasma cell leukemias tested. B29 transcripts were detectable in these cell samples, but low levels of B29 proteins were only detected in one plasma cell line. The HM57 mAb gave strong labeling on fresh cytocentrifuge preparations of all B-cell samples, and this mb-1 protein expression appeared to be B-cell specific. We therefore conclude that the HM57 mAb is well suited for the detection of the mb-1 molecule as a pan-B-cell marker for the diagnosis of immature and mature B-cell malignancies. The expression pattern of the mb-1 protein is comparable to that of the CD19 and CD22 antigens, but has the advantage of being B-lineage specific. Although B29 protein expression was restricted to B-lineage cells, the anti-B29 antiserum is less suitable for diagnosis of B-cell malignancies, because of the variable and generally weak signals on cytocentrifuge preparations.