Proteomics-based identification of DEAD-box protein 48 as a novel autoantigen, a prospective serum marker for pancreatic cancer

Patients with cancer frequently develop autoantibodies, and the identification of panels of tumor autoantigens may have utility in early cancer diagnosis and immunotherapy. This study aims to exploit the autoantibody repertoire in pancreatic cancer and identify the possible serum marker for pancreatic cancer. Sera from 55 newly diagnosed patients with pancreatic cancer and 52 healthy controls were analyzed for antibody-based reactivity against Hep-2, a human larynx epithelioma, cancer cell line, with one-dimensional immunoblot assay. From this analysis, we observed a prominent hand with a molecular weight of 47 kDa in 63.64% (35/55) patients, while in only 1.9% normal group (1/52). Using immunoblot analysis after two-dimensional electrophoresis combined with liquid chromatography-electrospray ionization tandem mass spectrometry, this target antigen was identified as DEAD-box protein 48 (DDX48). BLAST analysis showed that it was highly similar to eukaryotic initiation factor 4A and might play a role in pre-mRNA processing. An enzyme-linked immunosorbent assay was performed using recombinant, purified DDX48 as an antigen to detect anti-DDX48 autoantibodies in sera. Reactivity was observed in 20 of 60 (33.33%) pancreatic cancer patients, 3 of 30 (10.00%) colorectal cancer patients, 2 of 30 (6.67%) gastric cancer patients, 2 of 30 (6.67%) hepatocellular cancer patients, while none of the 20 chronic pancreatitis patients, 30 lung cancer patients, and 60 normal individuals. Together, these results demonstrate that the detection of autoantibodies to DDX48 may have clinical utility for the improved diagnosis of pancreatic cancer. (c) 2005 Elsevier Inc. All rights reserved.