Combination of MS Protein Identification and Bioassay of Chromatographic Fractions to Identify Biologically Active Substances from Complex Protein Sources

Purification of biologically active proteins from complex biological sources is a difficult task, usually requiring large amounts of sample and many separation steps. We found an active substance in a serum response element-dependent luciferase reporter gene bioassay in interstitial cystitis urine that we attempted to purify with column chromatography and the bioassay. With anion-exchange Mono Q and C₄ reversed-phase columns, apparently sharp active peaks were obtained. However, more than 20 kinds of proteins were identified from the active fractions with MS, indicating that the purification was not complete. As further purification was difficult, we chose a candidate molecule by means of studying the correlation between MS protein identification scores and bioassay responses of chromatographic fractions near the active peaks. As a result, epidermal growth factor (EGF) was nominated as a candidate molecule among the identified proteins because the elution profile of EGF was consistent with that of the bioassay, and the correlation coefficient of EGF between MS protein identification scores and bioassay responses was the highest among all the identified proteins. With recombinant EGF and anti-EGF and anti-EGF receptor antibodies, EGF was confirmed to be the desired substance in interstitial cystitis urine. This approach required only 20 ml of urine sample and two column chromatographic steps. The combination of MS protein identification and bioassay of chromatographic fractions may be useful for identifying biologically active substances from complex protein sources.Purification and identification of biologically active proteins existing in minute amounts from biological sources such as urine is still a difficult task (1). It requires a large volume of the sample and many separation steps for purification (2, 3). Nevertheless the recent progress of MS has dramatically changed protein analysis (4). With MS, smaller protein samples can be used than with classical protein identification methods such as N-terminal peptide sequencing.Interstitial cystitis (IC)¹ is a chronic inflammatory disease characterized by frequency and urgency and/or severe pelvic pain (5). The International Continence Society also selected the term “painful bladder syndrome” for IC (6). The quality of life of IC patients is extremely low because of their severe symptoms. The pathogenesis of IC is unclear, and effective treatments have not been established. To elucidate the mechanism of IC pathogenesis, we attempted to find characteristic proteins in IC urine using proteomics techniques and have already reported active neutrophil elastase as an IC urinary marker (7). We had also performed gene expression analysis of IC bladder tissues using GeneChip technology and found that mRNA expression of GPR18, a member of the G-protein-coupled receptors, was higher in IC bladder than in the control.² We tried to confirm whether GPR18 endogenous ligand existed in IC urine by using a bioassay with GPR18 transfectant cells.In the present study, the existence of an active substance in IC urine was suggested in the bioassay using the serum response element (SRE)-dependent luciferase reporter gene with the stable recombinant HEK293 cell line expressing GPR18. We thought that the response was derived from GPR18 and tried to purify the active substance from a small volume of IC urine using chromatographic techniques. Among the many proteins identified from partially purified samples, we clearly nominated epidermal growth factor (EGF) as a candidate molecule judging from the correlation between MS protein identification and the bioassay of chromatographic fractions. With recombinant EGF and anti-EGF antibody, EGF was confirmed to be the desired substance found in IC urine. The complete inhibition of the bioassay response by anti-EGF receptor antibody also indicated that the response was based on the EGF receptor, not GPR18, suggesting that GPR18 overexpression enhanced the EGF signal via the endogenous EGF receptor of the HEK293 cell line.