Bacteria May Be Early Signal of Oral Cancer

Elevated levels of three bacterial species in human saliva may be diagnostic indicators for the most common form of oral cancer, oral squamous cell carcinoma, according to a study by researchers from the Forsyth Institute, the Dana–Farber Cancer Institute, and HSDM. The study was the first of its kind to conduct a comprehensive examination of the microbiota of oral soft tissues, and the scientists hope their results will lead to a new method of detecting oral cancer with an early noninvasive diagnostic test using saliva. Results of the study appeared in the July 7 Journal of Translational Medicine.

Courtesy Donna Mager
Heightened levels in human saliva of three species of bacteria, Prevotella melaninogenica (left), Capnocytophaga gingivalis (top right), and Streptococcus mitis (bottom right) may be an early indicator of oral squamous cell carcinoma.

Oral cancer is diagnosed in nearly 30,000 Americans annually and has one of the lowest five-year survival rates of any cancer, according to lead researcher Donna Mager. “Although we have improved our treatment and the quality of life of people with oral cancer, if we don’t catch oral cancer at its early stages, the death rate will be just about what it was decades ago,” she said. Patients in early stages of the cancer are generally asymptomatic, making early detection a significant challenge. In response to the need for better detection efforts, Mager, an HSDM research associate in oral medicine, infection, and immunity at Forsyth, decided to investigate whether the microbiota of the saliva or oral soft tissues could provide diagnostic clues to the disease.

The investigation, which began in 2000 and involved patients from the Harvard–Partners hospitals, examined 40 of the most common oral bacteria in the saliva of 228 cancer-free subjects and compared the subjects’ samples to 45 oral cancer patients’ saliva. DNA counts per milliliter of saliva were tallied for each bacterial species and averaged across the subjects in the two groups. Counts of Capnocytophaga gingivalis, Prevotella melaninogenica, and Streptococcus mitis were elevated in the oral cancer subjects and found to have diagnostic power in 80 percent of the oral cancer cases reviewed. “We found that there appeared to be a threshold above which these three bacteria rose, and once they got above that threshold, they became diagnostically sensitive and specific indicators for the presence of an oral cancer lesion,” Mager said. The team then matched the 45 oral cancer subjects and controls by age, gender, and smoking history, getting similar results. In the future, the researchers hope to look at different populations, both inside and outside the country, to obtain more diverse data.

Oral cancer has been linked to alcohol and smoking in 75 percent of all cases. But incidence in young people is increasing worldwide and many of these individuals have no known risk factors.

Currently the best way to detect cancer lesions, Mager said, is for every adult to be given an oral cancer exam once a year. Still, these screenings are inefficient for detecting oral cancer in large populations. So a simple and inexpensive diagnostic tool, such as examining saliva for specific bacteria, is much needed. The National Institute of Dental and Craniofacial Research funded this study.

—Rachel Patzer

Step Taken Toward $1,000 Personal Genome

The theoretical price of having one’s personal genome sequenced has fallen from the prohibitive $20 million dollars to about $2.2 million, and the goal is to reduce the amount further—to about $1,000—to make individualized prevention and treatment realistic.

The sharp drop is due to a new DNA sequencing technology developed by MD–PhD student Jay Shendure, PhD student Gregory Porreca, professor George Church, and their colleagues, and appearing in the Sept. 9 Science. The team sequenced the E. coli bacterial genome at a fraction of the cost of conventional sequencing using off-the-shelf instruments and chemical reagents. Their technology appears to be even more accurate and less costly than a commercial DNA decoding technology reported earlier that week. The group is now scaling up their technology to resequence the human genome.

The Church group’s technology is based on converting a widely available and relatively inexpensive microscope with a digital camera for use in a rapid automated sequencing process that does not involve the much slower electrophoresis, a mainstay of the conventional Sanger sequencing method.

“Meeting the challenge of the $1,000 human genome requires a significant paradigm shift in our underlying approach to the DNA polymer,” write the scientists. Their goal of tackling the human genome will require converting several microscopes to function as sequencing instruments, as well as further miniaturization and process refinement.

Their current technique calls for replicating thousands of DNA fragments attached to one-micron beads, allowing for high signal density in a small area that is still large enough to be resolved through inexpensive optics. One of four fluorescent dyes corresponding to the four DNA bases binds at a specific location on the genetic sequence, depending on which DNA base is present. The fragment then shines with one of the four colors, revealing the identity of the base. Recording the color data from multiple passes over the same sequences, a camera documents the results and routes them to computers that reinterpret the data as a linear sequence of base pairs.

In their study, the researchers matched the sequence information against a reference genome, finding genetic variation in the bacterial DNA that had evolved in the lab.

“The cost of $1,000 for a human genome should allow prioritization of detailed diagnostics and therapeutics, as is already happening with cancer,” said Church, an HMS professor of genetics who also heads the Lipper Center for Computational Genetics at the Medical School.

—Robert Neal
and Judith Montminy

Fat Cell Protein Seen to Cause Insulin Resistance

Researchers from Beth Israel Deaconess Medical Center are the first to find that a vitamin-carrier protein secreted by mouse adipose cells promotes insulin resistance, one of the most common risk factors for diabetes. Results of their study, which appear in the July 21 issue of Nature, may offer insight into potential new drug therapies for humans suffering from diabetes, obesity, and metabolic syndrome.

The protein, retinol-binding protein-4 (RBP4), was thought to have only one role in the body—to deliver vitamin A to tissues. But the scientists, led by Barbara Kahn, HMS professor of medicine and chief of the Division of Endocrinology, Diabetes, and Metabolism at BID, with senior postdoctoral students Qin Yang and Tim Graham, found that the carrier protein also appeared to play a causal role in diabetes. Though previous studies revealed elevated RBP4 in diabetes patients, researchers had not investigated whether the protein was playing a causal role in the disease, until now.

Earlier work in Kahn’s lab involved understanding the paradox surrounding a specific glucose transporter, GLUT4. The expression of GLUT4 is decreased in adipocytes of both humans and mice in insulin--resistant states, but not in muscle cells. To decipher the mechanism of GLUT4, the researchers made transgenic mice, one group overexpressing GLUT4 and the other underexpressing the transporter. “Interestingly, overexpression of GLUT4 in adipose tissue resulted in enhanced glucose tolerance and insulin sensitivity,” Kahn said. “In contrast, mice with markedly reduced GLUT4 expression in adipose tissue but normal GLUT4 expression in muscle are insulin resistant and have an increased risk for overt diabetes.” Muscle from these mice is resistant to the action of insulin in vivo, but responds normally ex vivo. These results suggested that GLUT4 in adipose tissues affects whole-body insulin sensitivity, possibly through a protein secreted from the fat tissue.

The scientists conducted a global gene expression analysis to search for genes whose expression was altered in adipose tissue that also had a primary genetic alteration in GLUT4 expression. The RBP4 gene emerged from this work. The researchers then showed that elevating levels of RBP4 in mice through increased gene expression or injection of the purified RBP4 protein causes insulin resistance. They also found that a decrease in RBP4 in insulin-resistant states ameliorates the condition.

The researchers hope to further characterize the role of RBP4 as a cause for insulin resistance and the mechanisms by which it does this. Several key questions remain. One is whether dietary vitamin A affects insulin sensitivity and whether the combination of vitamin A with a high-fat diet may play a role in promoting diabetes and obesity. “The stereotypical Western diet generally contains a high proportion of fat and vitamin A,” Kahn said. “We are investigating this potential link right now.”

—Rachel Patzer

top