By David N. Leff

When dueling was in flower, a century or two ago, one favored technology for slaying your adversary was derringers at 20 paces. Your body's immune defenses, it now turns out, practice a similar form of mutual assured destruction.

This cellular mano-a-mano fight to the death pits two cytokines against each other - tumor necrosis factor (TNF-alpha) and insulin-like growth factor I (IGF-I). When injury or inflammation attacks neurons, for example, these two peptides descend on the affected cells to duke it out between programmed cell death (apoptosis) - TNF-alpha's specialty - and survival, courtesy of the IGF-I hormone.

The blood-brain barrier bars these two molecules from entry into the central nervous system, which however is beset by numerous cerebral insults, from ischemic stroke to AIDS dementia to the range of neurodegenerative diseases that afflict the elderly.

"TNF-alpha is a hot drug," observed immunophysiologist Keith Kelley, a professor at the University of Illinois in Urbana, and president of the PsychoNeuroImmunology Research Society. "Antagonists of the TNF receptor, and of TNF itself," he pointed out, "have now been approved for treatment in rheumatoid arthritis and Crohn's disease. But no one's looked at the brain, because the only way you can get this stuff into the brain is by drilling a hole in the skull.

"The point is," Kelley continued, "everyone believes that the mechanism of action, and at least the cell-death function of TNF and its receptor, is direct killing of the cell. I think we've shown, for the first time, that maybe a more important pathophysiological role of TNF is not direct killing, because it takes greater than a nanogram per milliliter of TNF-alpha to kill normal cells."

Kelley's novel spin on TNF's death duty appears in the current Proceedings of the National Academy of Sciences (PNAS), dated Aug. 17, 1999. Its title is, "A new mechanism of neurodegeneration: A proinflammatory cytokine inhibits receptor signaling by a survival peptide."

Foes Fight Joined At The Hip

"I would say," Kelley told BioWorld Today, "that apoptotic factors like TNF cause cytotoxic cell death. We know that survival factors, like IGF-I on the neurons, cause life. But the real question is - take stroke, cerebral ischemia, for example, where endogenous production drives both these proteins up at the site of injury. Then who's going to win? Clearly, the proinflammatory cytokine, TNF, which goes up in response to injury, wins. You get a dead neuron, and a leg that doesn't move - all these devastating effects of stroke.

"On the other hand," he went on, "if you add IGF-I, and it wins this little teeter-totter feedback loop, then the neuron might survive. And I think most folks today believe that we've either got to get more neuronal survival factors, such as IGF, into the brain, or find some way to increase the efficacy of such drugs.

"Their view is that we've got TNF on the left hand and IGF on the right hand, so to save the neuron we've either got to inhibit TNF or increase IGF," Kelley continued. "What we're saying in PNAS is that it might be just as effective to improve the efficacy of IGF by blocking TNF. It means that you could add as much IGF-I as you wanted to the cells, but in the presence of TNF-alpha, which is silently inhibiting this IGF signal, all the IGF in the world isn't going to work, because the TNF at the site of the brain lesion, in stroke, is going to inhibit the survival of IGF-I's survival enzymes, such as PI3 kinase."

To de-puzzle the mode of action between TNF-alpha and IGF-I, the co-authors chose cerebellar granule cells as their model brain neurons. "There's a whole layer of granule cells in the cerebellum," explained the paper's lead author, M.D.-Ph.D. student Homer Venters, "and there's a similar layer in the hippocampus, which is involved in learning and memory."

Brain's Granule Cells - In Vitro Animal Model

"The granular layer in the cerebellum," he told BioWorld Today, "has a huge number of cells. In fact there are more neurons in that layer than in the whole cortex. Something like 1011 or 1012 cells. One reason that they're a good model for us to use as a primary cell culture," Venters went on, "is that it's 95 percent one type of cell. When we observe an effect, it's an effect that's particular to this type of neuron, not an interaction between two different types of cells."

The team first showed that cell death induced in normal mouse cerebellar granule cells could be markedly inhibited, in a dose-related manner, by treatment with recombinant IGF-I. Increased activity of the IGF-I-related protective enzyme, PI3, paced this effect.

"We asked how those neurons lived and died," Kelley recounted, "and found that IGF-I caused them to live, but that if we pretreated them with TNF, IGF-I's survival effect didn't work any more. So the good guy was IGF and the bad guy TNF. But that didn't work the way we expected it to. We expected that TNF would just kill 'em. In reality, very low concentrations - 10 picograms per ml - merely blocked the ability of IGF to save the cell.

"Now we are trying to convince ourselves that in important neurodegenerative diseases, where both these endogenous proteins go up in the brain, we're on the right track by looking at how TNF inhibits the ability of IGF to save neurons. With that concept, what we're trying to do is first of all, understand which TNF receptor is responsible for that; we don't know yet. That's where we're headed.

"We've seen the damaging effect of stroke on the quality of peoples' lives," Kelley observed, "so I hope our findings would encourage drug companies to target ways of getting inhibitors - antagonists of TNF - into the brain. "I think," he concluded, "it's a new way of looking at the world."