Gene Therapy for Pain

NIH scientists are attacking chronic pain with a novel form of gene therapy that targets the spinal cord. Though still in the animal testing stage, this approach has overcome one of the major obstacles to gene therapy as a way to manipulate spinal cord function. Rather than injecting genes directly into a localized area of the spinal cord, the pain-relieving gene is introduced into the sheath of tissue that surrounds the cord. From that strategic location, the gene can pump out its product and bathe many nerves, thus extending the range of its pain-numbing effect. The investigators hope that this simplified approach can be used to generate a variety of products in the tissue surrounding nerves, including factors that could stimulate new nerve growth.

The study, carried out by scientists from the National Institute of Dental and Craniofacial Research (NIDCR) and the University of Pennsylvania, was reported in the May 1 issue of Human Gene Therapy.

"We are totally pumped up that this approach is working in an animal model," said Dr. Mike Iadarola, chief of NIDCR's Neuronal Gene Expression Unit. "The animal studies have shown us that genes are readily taken up by the connective tissue cells that surround the central nervous system. So, given the right gene, our approach has application to a broad range conditions, from pain control to spinal cord injury and disorders like multiple sclerosis and Parkinson's disease."

In the study, investigators used an adenovirus--similar to a cold virus-- to deliver the beta-endorphin gene to the rat spinal cord. The virus particles were injected into the spinal fluid, where they were readily taken up by the protective sheath of connective tissue, called the pia mater, which surrounds the cord. Within 24 hours the sheath cells began secreting beta-endorphin, one of the body's natural sedatives for alleviating pain.

"The incredible simplicity and relative noninvasiveness of this approach provides a new frame of reference for gene therapy of the nervous system," said co-author, Dr. Alan Finegold, previously with NIDCR's Pain and Neurosensory Mechanisms Branch and now in the private sector.

The spinal cord was selected as the target for the beta-endorphin gene, not only because of its ease of access, but also because it is the first processing point for relaying pain signals to the brain, and pain can be effectively controlled at this location. The idea was to have beta-endorphin block pain signals before they reached the brain, where pain perception occurs.

The researchers observed that beta-endorphin levels in the spinal fluid increased nearly 10-fold following a single injection of virus. Cellular analysis confirmed that sheath cells, not spinal cord neurons, were the source of beta-endorphin.

To determine if the method had a therapeutic effect, the investigators used the rat "hindpaw" model for evaluating pain response. It is based on the time that elapses before a rat voluntarily pulls its paw away from a heat lamp. The system allows a rat to be tested for the normal pain response in one paw and the so-called "hyperalgesic" response in the other paw, which has been inflamed by injection of an irritant. This latter type of super-sensitive pain results in rapid paw withdrawal and is used as a model for the chronic pain of cancer or arthritis.

The rats responded to beta-endorphin by exhibiting a delayed response in pulling the inflamed paw away from the heat source, a sign that the hyperalgesic pain sensation was reduced. An added bonus was the observation that the non-inflamed paw had a normal withdrawal response. This points not only to a lack of toxicity from the treatment procedure but also to a selective therapeutic effect for beta-endorphin. The results are similar to a person getting relief from chronic cancer pain, yet not losing the normal sense of feeling to react to painful stimuli. The reason for this distinction is not completely understood, but scientists feel that inflammation may help activate receptors on the affected nerves, making them more responsive to the blocking effect of beta-endorphin.

Development of this novel method evolved from some preliminary trial and error testing. Initial attempts were aimed at injecting virus directly into neural tissues. "We discovered early-on that brain and spinal cord were not a hospitable environment for direct injection of virus," said Dr. Iadarola. "There are physical barriers that prevent the virus from infiltrating the space between the neurons, keeping any beneficial effects very localized. We shifted our approach to the spinal fluid, which we thought would be an excellent medium to expose a wide swath of neurons to the therapeutic virus."

What they observed however, was the protective sheath of connective tissues that coats the spinal cord acted like a sponge, soaking up the virus and preventing direct contact with nerve tissue. What initially appeared as an obstacle turned out to be the makings of a new approach for gene therapy to the nervous system. Although the nerve cells could not be made to effectively take up the gene, they wound up being exposed to beta-endorphin that was produced by neighboring sheath cells.

As with other studies that have used adenoviruses to deliver genes, the effects of beta-endorphin were not permanent. Production peaked after 3 - 7 days and tailed off dramatically by day 15. However, the investigators are optimistic that improvements in vector design will result in a single injection that provides long-term gene expression, not only of beta-endorphin, but genes to treat a variety of spinal cord and brain disorders.

Working with Drs. Iadarola and Finegold was Dr. Andrew Mannes from the University of Pennsylvania, Department of Anesthesiology.