Sunday 3 November 2002
SFN 2002 - Day 1
by Rabiya Tuma
In the opening public lecture of the 32nd annual meeting of the Society for Neuroscience, Carla Shatz not only described how the visual system is established, she discussed her latest revelation: Molecules best known in immune recognition are involved in building the neural connection system.
Many scientists liken neural connections in the brain to their electrical counterparts in a computer. But Shatz, who specializes at Harvard Medical School in the development of mammalian visual systems, showed that connections in the mammalian brain are much more flexible than the hardwired links in computer chips.
During eye development, signals are relayed from the retina to the lateral geniculate nucleus (LGN), and then on to the visual cortex of the brain, where the information is processed. Initially the connections from one eye are randomly intermixed with those from the other eye in the LGN and at the cortex. But early in development - before the eye even opens or functions - neighboring groups of retinal neurons fire in waves, without any input signal. These waves of autoimpulses from the neighboring retinal neurons seem to strengthen some connections in the LGN and cortex, while weakening others.
The end result is that the signals from one eye are interspersed in a layered pattern with inputs from the other. Like the black and white stripes of a zebra, the stripes of input from each eye are juxtaposed, but not intermixed, in the mature brain.
Shatz's group has shown previously that if they block the autoimpulses from the retina to the LGN, then the neat layering of the normal adult pattern fails to form, and the immature intermixed pattern remains. Thus the researchers conclude that the adult pattern is the result of simultaneous weakening of some connections and reinforcement of others. But the exact molecules involved in the process have remained largely unknown.
Now Shatz and colleagues have made the surprising discovery that proteins in the major histocompatibility complex class I (MHCI)are directly involved in the process. Previously, MHC molecules were thought to be in the sole purview of the immune system, where they are responsible for presenting foreign antigens to T-cells and inducing cellular immune response. But when Shatz's group used microarrays to compare mRNA from LGN neurons that had normal inputs from the retina to ones that had the connections blocked, they found the MHCI RNA was present only in the normal LGN.
Shatz's group also has light microscope data indicating that the MHCI proteins are present at the synapse in several regions of both the developing brain and the adult brain.
Her current hypothesis is that the MHCI proteins are the "anti-glue" that allow the inappropriate early connections to be broken down. Imagine the MHC1 protein sitting on in the membrane of the post-synaptic neuron, she suggests, and interacting with a protein complex on the presynaptic neuron. Shatz speculates that the interaction initiates a signaling cascade that somehow reduces the strength of the connection, eventually leading to its complete elimination.
But why would the brain leave such important connections to chance, rather than establishing them in a genetically hardwired manner? "Nature sets up the basic connections," said Shatz. "But the adaptive process that follows allows us to adapt to and learn from our experiences."
As for the presence of immune-system proteins in the brain, Guy McKhann,
a neurologist specializing in brain function and development, told BioMedNet
News that with Shatz's work, "this observation has suddenly gone from phenomenology
© Elsevier Science Limited 2002