DnaJ, an Escherichia coli Hsp40 protein composed of 376 amino acid residues, is a chaperone with thioldisulfide oxidoreductase activity. We present here for the first time a small angle x-ray scattering study of intact DnaJ and a truncated version, DnaJ (1-330), in solution. The molecular weight of DnaJ and DnaJ (1-330) determined by both small angle x-ray scattering and size-exclusion chromatography provide direct evidence that DnaJ is a homodimer and DnaJ (1-330) is a monomer. The restored models show that DnaJ is a distorted -shaped dimeric molecule with the C terminus of each subunit forming the central part of the , whereas DnaJ (1-330) exists as a monomer. This indicates that the deletion of the C-terminal 46 residues of DnaJ impairs the association sites, although it does not cause significant conformational changes. Biochemical studies reveal that DnaJ (1-330), while fully retaining its thiol-disulfide oxidoreductase activity, is structurally less stable, and its peptide binding capacity is severely impaired relative to that of the intact molecule. Together, our results reveal that the C-terminal (331-376) residues are directly involved in dimerization, and the dimeric structure of DnaJ is necessary for its chaperone activity but not required for the thiol-disulfide oxidoreductase activity.Hsp40 proteins collaborate specifically with Hsp70 proteins and some other factors to participate in a wide range of cellular processes essential to cell survival, such as assisting the folding, assembly, disassembly, and translocation of newly synthesized proteins, by suppressing aggregation or mediating degradation of misfolded proteins (1, 2). The Hsp40 family proteins are divided into three subtypes (3) (see Scheme 1). They all contain a J domain responsible for binding to the ATPase domain and stimulating the ATPase activity of Hsp70. Type I Hsp40 proteins, such as Escherichia coli DnaJ, yeast Yjd1, and human Hdj-2, have a Gly/Phe-rich region, which has been suggested to assist J domain interactions with Hsp70 (4). The Gly/Phe-rich region is followed by a conserved cysteine-rich region forming a zinc finger domain and a poorly conserved C-terminal domain. The peptide binding site(s) has been found to locate within these two domains (5-7). Type II Hsp40 proteins, such as yeast Sis1 and mammalian Hdj-1, also contain the J, a Gly/Phe-rich region, and C-terminal domains but without the zinc finger domain (8, 9). Both type I and type II Hsp40 proteins act as molecular chaperones to bind and deliver nonnative proteins to Hsp70 (7, 8, 10), but they are not functionally equivalent (11). Type III Hsp40 proteins, such as yeast YJL162c and virus T antigen, contain only the J domain, and they do not function as molecular chaperones (3).In addition to the DnaK-dependent co-chaperone activity described above, DnaJ also exerts autonomous DnaK-independent chaperone activity (9, 12, 13), which is characterized by its ability to bind with unfolded proteins and prevent them from irreversible aggregation (10, 11). Moreover, DnaJ has ...