Researchers Develop Method to Identify Proteins in the Human Serum Proteome
In the June 2004 issue of Clinical Proteomics, Drs. King C. Chan, David A. Lucas, and colleagues at SAIC-Frederick, in the Laboratory of Proteomics
and Analytical Technologies at NCI-Frederick, report on a new method for isolating and identifying proteins found in trace quantities in blood. These proteins are potential
biomarkers to alert clinicians of certain diseases, including ovarian, breast, and prostate cancer.
The researchers crafted a multistep
procedure for separating blood proteins derived from serum. Prior efforts to identify these proteins, collectively
known as the serum proteome,
came up short mainly because the separation steps meant to reduce amounts of large, highly abundant proteins caused a simultaneous loss of smaller, low-abundance proteins. The authors avoided this problem by using multiple separation and fractionation
steps, based on the size, electric charge, and other chemical properties that differ between proteins,
to produce samples that were then analyzed by mass spectrometry, a high-throughput technique for identifying individual proteins.
"Our investigation resulted in the identification of 1,444 proteins in serum," said coauthor Dr. Thomas Conrads, director of the Mass Spectrometry
Center at NCI-Frederick. "The proteins identified by earlier research overlapped only slightly with those characterized by our group," said Conrads. "This emphasizes the wide scope and complexity of the human serum proteome, which has been estimated to contain more than 10,000 proteins."
The authors created a publicly available
database of the human serum proteome (available at http://bpp.nci.nih.gov) to serve as a resource for other researchers
Similar Risk of Lung Cancer Among Male and Female Smokers
Results from two large cohorts indicate
that men and women with comparable
smoking histories have similar
risks of developing lung cancer. Data analysis from the Nurses' Health Study of women and the Health Professional Follow-Up Study of men found no evidence for a greater risk of lung cancer among women who smoke, even though some previous case-controlled studies have suggested
that women are at greater risk.
In a study in the June 2 Journal of the National Cancer Institute, an international
group of researchers led by Dr. Diane Feskanich of Brigham and Women's Hospital in Boston, directly compared lung cancer incidence rates using data from 60,296 women and 25,397 men, aged 40 to 79, who were current or former smokers. Findings from the two cohort studies (with common ages and follow-up periods) do not support a greater risk of lung cancer for women.
An accompanying editorial charts the history of lung cancer research for the past century, and notes that early studies tended to show lower risks of lung cancer among women smokers - mainly because of the lag in women's uptake of smoking, women's lower average cigarette consumption, and other factors. Although some case-controlled studies in the 1990s indicated that women may be more at risk for lung cancer, the authors agree that the "clear picture that emerges from the cohort studies is that women do not have higher rates of smoking-induced lung cancer than men."
Tumor Supressor Gene Analysis May Yield New Targeted Therapies
Dr. Zhenghe Wang and colleagues from the Sidney Kimmel Cancer Center at Johns Hopkins University have sequenced the entire gene family that codes for key cellular signaling proteins, known as tyrosine phosphatases,
from human cancers. According
to the results of their study, published
in the May 21 issue of Science, the scientists found mutations that affected over a quarter of colorectal cancers, as well as a smaller subset of lung, breast, and gastric cancers.
Though targeted therapies toward protein tyrosine kinases - such as the epidermal growth factor receptor (EGFR) - have been directly linked to tumorigenesis, there has not been as much investigation into tyrosine phosphatases, which directly regulate the activity of kinases and the downstream
proteins in their signaling pathways.
In this study, researchers looked at all 87 members of the phosphatase gene family in 18 colorectal cancers, and identified 6 genes that were specifically
mutated in tumors. They then sequenced these 6 genes from an additional
157 colorectal cancers and identified
77 different mutations, which in total were found in 26 percent of the tumors. Further examination of a subset
of these mutations revealed that they reduce the function of the phosphatase proteins for which they code, thereby hampering the ability of the proteins to regulate cellular functions
such as growth, differentiation, death, and tissue invasion.
Risk Prediction Models Workshop Sets Goals
Estimating absolute risk of cancer can have profound implications for targeted
prevention strategies and clinical decision-making. On May 20, more than 100 experts met in Washington, D.C., for a workshop about cancer risk prediction models. "This interdisciplinary
workshop broke ground by bringing
together the cancer risk prediction modeling community for the first time and helping identify the research steps needed to move this field forward," noted Dr. Andrew Freedman, workshop cochair from NCI's Division of Cancer Control and Population Sciences. Other cosponsors were NCI's Division of Cancer
Epidemiology and Genetics (DCEG) and Office of Women's Health.
The workshop included four sessions on risk prediction models: applications,
development and implementation,
evaluation and validation, and predicting germline mutation carrier
status. DCEG's Dr. Ruth Pfeiffer, workshop cochair noted, "After intensive
discussions, model developers and clinicians reached the consensus that model performance should be judged in the context of specific applications, and further methodological
research is needed to develop criteria for model assessment.
"Priorities for future research include identifying cancer sites for which new risk prediction models are useful, finding ways to improve current and future cancer risk prediction models by incorporating new clinical and biological markers, and providing data resources and study populations for modeling and validation.
|