Pediatric Cancer
Varicella Zoster
Technology Transfer
The Short White Coat
Research Briefs
Around & About

Researchers in the Columbia Genome Center have constructed the world’s first carbohydrate chip, an array of 100 different sugars attached to a chemically modified glass slide. The technology may be used someday to design a single chip, with thousands of carbohydrate antigens, that could help clinicians diagnose many common infectious diseases from a few microliters of blood. The method to make the prototype carbohydrate chips, which would detect antibodies to microbial pathogens in the blood, was published in the March Nature Biotechnology.

DNA and proteins get most of the microarray attention, but when antibodies recognize a foreign antigen in the body the antigen could also be a sugar chain. Since a microbe may have its own unique sugar on its surface, a single array containing many sugar structures would be able to detect antibodies generated to a wide range of infections, the researchers say. Sugar arrays also could be used in basic research to understand how protein-carbohydrate interactions help cells recognize each other.

The production of carbohydrate arrays was delayed until now by difficulty in getting the sugars to stick to a slide. The Columbia researchers, led by Dr. Denong Wang, head of functional genomics in the Genome Center, overcame the problem. They printed different sugar solutions directly to a nitrocellulose-coated slide with a machine designed to make DNA arrays.

After extensive washing, most of sugars stayed in their spots, as seen when the fluorescently tagged sugars were viewed with a scanner. More importantly, the researchers also found the immobilized sugars retained their ability to bind to other molecules. Antibodies that detect different carbohydrate structures in standard antigen binding assays (ELISA) also detected the sugar structures on the array.

With room for 20,000 different spots of sugar, the array can hold the signature sugars from nearly all common pathogens. “We’re working hard to identify the sugar that is unique for each microorganism, including anthrax and pneumococcus,” Dr. Wang says. “Our goal is to push the chip into clinical situations as soon as possible.”

— Susan Conova

In the mid-1990s, Columbia Health Sciences investigators identified a group of genes that inhibit skin cancer as part of an effort to understand the genetic processes underlying uncontrolled cell growth and cell lineage determination (cellular differentiation). When active, these melanoma genes can revert diverse cancerous cells to a state of normal growth and can even promote cell death in tumor cells.

Now the researchers report the protein product of one of these genes, mda-5 (melanoma differentiation associated gene-5), has unusual properties: Based on structure and initial functional studies this gene may code for a protein that can unwind RNA and also cause programmed cell death, or apoptosis. This gene appears to be the first example where these two functions are linked in a single protein.

The research, led by Dr. Paul B. Fisher, professor of clinical pathology, director of neuro-oncology research, and the Michael and Stella Chernow Urological Cancer Research Scientist, appeared in the Jan. 22 issue of the Proceedings of the National Academy of Sciences.

Dr. Fisher, the senior author, and his colleagues report the mda-5 gene may code for a protein that is an RNA helicase, which can unwind a cell’s RNA. The MDA-5 protein also contains a caspase recruitment domain, an important component in programmed cell death and, if operational, may give this protein the ability to target a cell for destruction.

With its apparently unique properties, the mda-5 gene might be able to act as a sensor of certain types of viral infection, such as from RNA viruses like influenza and HIV, and could elicit apoptosis, preventing the propagation of RNA viruses or viruses that contain double-stranded intermediates. Being able to kill virally infected cells selectively could lead to new ways to treat viral infection and stop viruses before they can cause damage, Dr. Fisher says. He hypothesizes that a drug could be developed to turn on the mda-5 gene or mimic its activity to prevent virus replication.

Dr. Dong-chul Kang, lead author of the paper and associate research scientist; Dr. Rahul V. Gopalkrishnan, associate research scientist; Qingping Wu, research technician; Dr. Eckhard Jankowsky, formerly a postdoctoral research fellow at Columbia and currently an assistant professor of biochemistry at Case Western Reserve University; and Dr. Anna Marie Pyle, professor of biochemistry and molecular biophysics and an investigator in the Howard Hughes Medical Institute, also participated in this research. Dr. Fisher and his colleagues are in the process of patenting their discovery.

— Matthew Dougherty

Scanning electron micrographs of human pelvis bone biopsies showing normal, above, and osteoporotic trabecular bone, below. (Copyright 2001, David W. Dempster, Ph.D.)

Bone density measurements are key to diagnosing osteoporosis, but those numbers do not indicate how bone is being lost.

Two Columbia researchers have developed a computer bone loss model that shows the different ways the microstructure of trabecular bone is stripped of its strength. Trabecular bone, which has a mesh-like structure and is found in the hip, spine, and wrist, is the bone type most vulnerable to osteoporosis because it tends to degrade faster than other bone types.

Dr. X. Edward Guo, associate professor of biomedical engineering, and Chi Hyun Kim, a biomedical engineering graduate student, have developed a 3-D computer simulation of trabecular bone loss and thinning to better understand how strength and stiffness are affected by osteoporosis. The model also simulates how osteoporosis medications increase bone strength by making the connective parts of the mesh structure or the struts thicker.

The results of their studies suggest that physicians need to analyze more than bone density when assessing osteoporosis. Two patients, for example, may have similar bone densities but one patient may have bone where individual struts have thinned while the other patient may have lost those bone struts, a more serious condition.

Using the model, the researchers showed that bone strut loss was much more detrimental to the strength of bone than strut thinning. While it had been known that bone loss decreased the strength of bone, the modeling showed how much bone loss affects bone strength. They quantified the loss of strength to the bone due to bone loss and found the amount was far greater than had been measured before.

The study results were published in the February issue of Bone.

— Matthew Dougherty