SAN DIEGO - New research is overturning the old dogma that the nervous system and the immune system are separated by a Chinese wall. Indeed, the opposite may be true. Some major histocompatibility complex or MHC class 1 molecules, currently conceptualized as immune system proteins, are expressed specifically on neurons. They help regulate plasticity both during development and in response to injury, and animals that lack them have some behavioral symptoms that also are seen in humans with schizophrenia and autism.

So-called classical MHC genes are involved in presenting antigens to T cells, allowing the immune system to fight off invaders. But roughly a decade ago, MHC class I molecules were discovered on neurons in the lab of Carla Shatz, then at University of California at Berkeley and now at Harvard University.

When asked how her lab discovered those proteins, Shatz, who spoke at a symposium on "The New Neuroimmunology: Immune Proteins in Synapse Formation, Plasticity and Repair" at the Society for Neuroscience annual meeting in San Diego on Sunday, as well as at a press conference afterward, told reporters that "they found us."

Shatz added, "we were looking for molecules that were turned up or down during experience." And among the molecules that her group's screens turned up were MHC class I proteins. Shatz' group also found that MHC class I proteins appear to act as a molecular brake of sorts, limiting and tuning the connections that form between neurons during development.

Shatz's discovery opened the gates for a number of discoveries on the roles of MHC class I proteins in neural plasticity, both developmentally and in response to injury.

Shatz's erstwhile postdoc Lisa Boulanger, who is now at University of California at San Diego and chaired the symposium, reported on work from her laboratory showing that in mice, MHC class I proteins appear to be involved in determining the strength of neural connections.

Neurons form connections, and weaken or strengthen them, according to environmental stimuli. Boulanger and her colleagues showed that certain stimuli that weaken the connections between two normal neurons will strengthen those between two cells lacking MHC class I proteins.

While such strengthening might be positive under certain circumstances, the inability to weaken unnecessary or even counterproductive connections can be problematic, and may be one of the neural underpinnings of autism. Separate studies showed that autistic children have larger brains than normal controls early in development, which might be due to an inability to get rid of inappropriate connections.

Boulanger cautioned that important work still needs to be done to nail down any role of MHC class 1 proteins in autism. It is impossible to tell whether an animal has the equivalent of a complex disorder like schizophrenia or autism, and her laboratory is only now beginning to study blood and tissue samples from autistic and schizophrenic patients to see whether they show abnormalities in MHC class 1 proteins.

Still, animal data are consistent with such an association. Autistics also appear to have trouble ignoring unimportant stimuli, and that, too, fits with characteristics of neurons lacking MHC class 1. Boulanger and her colleagues found that such mice did not show a neural adaptation known as prepulse inhibition, which reflects the ability to filter inappropriate stimuli.

At the Karolinska Institute in Stockholm, Sweden, Staffan Cullheim is studying another facet of MHC class I function: how the proteins affect the response to injury. After an injury, motor neurons specifically get rid of excitatory connections, possibly as a way to focus on healing without being distracted by electrical input they cannot pass on. Staffan found that motor neurons lacking MHC class I proteins pruned many more synaptic connections than wild-type ones. They also did not show a preference for pruning excitatory over inhibitory connections, suggesting that such pruning probably was ineffective in getting them some peace and quiet.

In some ways, the findings of Cullheim's group are the opposite of what's seen during development. While a lack of MHC leads to too little synaptic pruning during development, it leads to too much pruning after injury. Boulanger said that such apparent contradictions highlight how much remains to be learned about MHC class I molecules. Of the more than 70 members that make up the MHC class I family, most "are nonclassical and have no known immune function," Boulanger said. "That implies to me that the main function of this gene family . . . remains to be discovered."

The 37th annual meeting of the Society for Neuroscience runs through Wednesday. In clinical developments reported at the meeting:

• Advanced Cell Technology Inc., of Los Angeles, and the Casey Eye Institute at Oregon Health & Science University reported results of a study using ACT's embryonic stem cell-derived RPE cells in the RCS rat, a validated animal model of retinal disease, showing that the RPE cells demonstrated a statistically significant therapeutic effect compared to controls, with a maximum efficacious effect at a medium dose level. Histological assessment showed integration of the human RPE cells into the rodent's RPE layer without migration into the retina. Those results indicated that ES-derived RPE cells might provide an effective donor cell source to rescue photoreceptors in conditions such as age-related macular degeneration, where RPE function is compromised. ACT intends to present those data to the FDA when it files for an investigational new drug application for its RPE therapy.

• Avigen Inc., of Alameda, Calif., disclosed first-in-patient top-line Phase IIa data for its investigational neuropathic pain product, AV411 (ibudilast). Summary results showed AV411 was safe and well tolerated in the study at doses up to 80 mg/day, also indicating a favorable dose-to-blood plasma level relationship. Although there appeared to be no difference in overall mean visual analogue scale (VAS) scores between active and placebo arms, a pharmacokinetic assessment indicated an encouraging correlation between AV411 plasma levels and the number of patients who reported a decrease in pain scores as assessed by VAS. There was a trend toward decreased opioid use that correlated with increased AV411 dose, as well. Data were presented at the 10th International Conference on the Mechanisms and Treatment of Neuropathic Pain held in Salt Lake City. Avigen also presented data from a Phase I study with AV650 (tolperisone) that showed the compound for the treatment of disabling neuromuscular spasticity and spasm to be well tolerated with no evidence of sedation.

• Pipex Pharmaceuticals Inc., of Ann Arbor, Mich., offered preliminary results using its lead anti-copper drug candidate, oral Coprexa (tetrathiomolybdate) in the most commonly utilized transgenic preclinical murine model of Alzheimer's disease. Data showed a significant reduction in central nervous system copper (p < 0.05) and a 40 percent reduction in insoluble amyloid beta (p < 0.05). The results need to be replicated and expanded before proceeding to clinical trials, the firm noted.

• Sangamo BioSciences Inc., of Richmond, Calif., presented in vivo data showing that treatment with a zinc finger DNA-binding protein therapeutic targeting TrkA, a validated pain receptor, significantly reduced the perception of pain in a mouse model of bone-cancer pain.