Using probes based on the human protein sequence, we cloned a cDNA from a human melanoma library (Summers et al 1996), thus confirming the amino acid sequence identity between platelet-derived EPF and Cpn10. within 24 hours of fertilization in all mammalian species tested and that persisted for at least the first half of gestation (Morton et al 1987). Subsequent studies demonstrated that EPF is not confined to pregnancy; activity can H-1152 dihydrochloride be detected in serum during the processes of normal tissue renewal, such as liver regeneration (Quinn et al 1994), as well as in pathologic situations, such as the development of cancer (Quinn 1991;Quinn and Morton 1992). It can be demonstrated in vitro that the appearance of EPF in the extracellular compartment closely parallels cellular growth, and that the induction of growth arrest or differentiation causes the rapid disappearance of EPF (Quinn et al 1990). Before our identification of EPF as a homologue of chaperonin 10 (Cpn10;Cavanagh and Morton 1994), we raised monoclonal antibodies to that protein partially purified from a medium conditioned by the human choriocarcinoma cell line BeWo (Athanasas et al 1989;Quinn et al 1990). These antibodies were invaluable neutralizing agents and enabled us to determine that EPF is not only closely associated with cellular growth but is also required for normal embryonic development (Athanasas et al 1989;Athanasas-Platsis et al 1991), establishment and maintenance of tumors (Quinn and Morton 1992), and normal liver regeneration (Quinn et al 1994). As low-affinity immunoglobulinM antibodies , however, their utility was limited. Furthermore, antibody production by the hybridomas was unstable (Quinn et al 1990). These characteristics were considered to result from the essential growth-regulatory properties of EPF. Hybridomas, like the parent myeloma, both produce EPF and require it for growth; hence, the only hybrid H-1152 dihydrochloride cells capable of surviving production of these EPF-neutralizing agents were those producing the least effective antibodies. Once the EPF amino acid sequence was established, we undertook production of polyclonal antibodies to synthetic peptides that corresponded with different parts of this sequence to produce a more robust array of reagents. The need for such tools was made even more compelling by the actual identity of the EPF protein sequence. Seventy percent of the amino acid sequence of the molecule isolated from human platelets was determined, and except for a single residue now known to represent a species difference, this sequence was found to be identical to that of rat Cpn10 (Hartmann et al 1992). Cpn10 functions in eubacteria and within eukaryote mitochondria and plastids as an accessory molecule to chaperonin 60 (Ellis and van der Vies 1991). TheEscherichia coliforms of these molecules (the products of the GroE operon) are known as GroES and GroEL, respectively (Zeilstra-Ryalls et al 1991). Further studies determined that human plateletderived EPF and rat mitochondrial Cpn10 were functionally interchangeable in vitro, but GroES did not exhibit activity in the EPF bioassay (Cavanagh and Morton 1994). Using probes based on the human protein sequence, we cloned a cDNA from a human melanoma library (Summers et al 1996), thus confirming the amino acid sequence identity between platelet-derived EPF and Cpn10. This apparent identity between an H-1152 dihydrochloride extracellular molecule with growth-regulatory and immunomodulatory properties and a molecular chaperone targeted to intracellular organelles raises unprecedented regulatory and mechanistic questions. In this study, we describe the production, assessment, and application of antibodies to selected epitopes of Cpn10 that will be useful Vegfa tools in the search for answers to such questions. == RESULTS AND DISCUSSION == == Antibody response to Cpn10-derived synthetic peptides displays an unusual pattern == During the initial attempts at antibody production, short synthetic peptides corresponding with the N-terminal sequence of Cpn10 (residues 111) and an internal sequence (residues 3344) were each conjugated to ovalbumin and administered according to a standard immunization schedule (Johnstone and Thorpe 1996). Rabbits did not respond to booster doses of H-1152 dihydrochloride antigen, and with time, specific antibody production declined (data not shown). With a.