November 15, 2005
Synaptic connections need nurturing to retain their structure and keep outsiders at bay
The ability of the brain to transmit and process information requires a lifelong commitment to maintaining the integrity of synapses--the special connections that permit the passage of nerve impulses from one nerve cell to another, according to investigators at St. Jude Children's Research Hospital and colleagues in Hokkaido University School of Medicine (Japan). A report on this work appears in the November 15 issue of Nature Neuroscience.
This long-term commitment requires proteins called synaptotrophins, the prototype of which is Cbln1, to maintain countless millions of synapses in good working order, the researchers said. In the absence of such proteins, the synapses weaken and eventually fall apart. This not only compromises nerve transmission, but also provides the opportunity for other nerves to extend their axons toward these faltering synapses and make inappropriate connections that further disrupt brain function.
The idea that synapses require maintenance factors in the adult is not new, although the identification of specific substances that contribute to this process has proven elusive, Morgan said. Therefore, the researchers used a laboratory model of the cerebellum to identify proteins that maintain a specific set of synaptic connections. Using this model they found that Cbln1 maintains correct synaptic connections after they have been established.
The St. Jude team showed that Cbln1 maintains the connection between two types of nerves--Purkinje cells and granule cells. Purkinje cells are large nerves that are aligned like dominos across the upper part of the cerebellum and send information to other parts of the brain. Parallel fibers are the axons of granule cells and they form synapses with the dendrites of Purkinje cells. The investigators showed that granule cells release Cbln1 from the ends of their axons--the parallel fibers--in order to maintain their synaptic connections with these dendrites. Dendrites are threadlike branches on nerves that conduct incoming impulses from synapses to the body of the nerve cell.
When the investigators studied the electrical activity at the synapse between granule and Purkinje cells in models that lacked the gene for cbln1 (cbln1-/-), they found that the signals in Purkinje cells stimulated by granule cells were consistently smaller than normal. In addition, the number of synapses between parallel fibers and Purkinje cells in cbln1-/- models was markedly reduced compared to cbln+/+ models. This was because the endings of the axons from granule cells (the presynapse) progressively detached from Purkinje cells. However, the specialized regions of Purkinje cell dendrites that participate in the synapse (postsynaptic spines), were still present in normal numbers. But these unoccupied spines led to a second synaptic abnormality in the cbln1-/- model: the pattern of synapses between the Purkinje cells and another type of nerve, called the climbing fibers, was abnormal. Normally, only one climbing fiber makes a synapse with a particular Purkinje cell. But in the cbln1-/- models many climbing fibers had grown into the area of the cerebellum where Purkinje cells normally form these single synapses with parallel fibers; and these climbing fibers had established many synaptic contacts on each Purkinje cell at sites previously occupied by parallel fibers.
"Nerves are territorial and respond to signals that keep other types of nerves out of their area of the brain unless they have legitimate business there," Morgan said. "Without proteins like Cbln1 in the brain, some nerves extend their axons into neighboring territories, usurping abandoned postsynaptic sites and disrupting normal function."
The researchers also found that Cbnl1 plays a critical role in maintaining the molecular mechanisms that underlie long-term depression (LTD) at the parallel fiber-Purkinje cell synapse. LTD, an experience-dependent modification in the response of a nerve to stimulation, is viewed as a molecular form of memory. The molecular machinery that produces LTD is thought to be located in Purkinje cells. This suggests that Cbln1 released by granule cells must somehow influence LTD in Purkinje cells, Morgan said. The authors propose this occurs through a protein in the postsynaptic spine (dendrites) of Purkinje cells called the orphan delta-2 glutamate receptor. They note that models that lack delta-2 glutamate receptor in Purkinje cells are nearly identical to models that lack Cbln1 in granule cells. This suggests that delta-2 glutamate receptor somehow mediates the action of secreted Cbln1. As both delta-2 glutamate receptor and Cbln1 are members of larger families of proteins that have distinct patterns of expression throughout the brain, this type of interaction might provide a way to ensure that the correct synaptic contacts form between different sets of nerve cells.
"Our findings are a significant step in our goal of understanding how the brain maintains its synaptic integrity," Morgan said. "Moreover, these findings open the possibility that disruption of synaptotrophin function could play a role in the development of neurological and psychiatric disorders. Therefore, these proteins and the pathways through which they function might represent potential targets for therapeutic intervention in neurological and psychiatric diseases."
Other authors of the paper include Hirokazu Hirai (currently at Japan Science and Technology Agency, Kanazawa), Zhen Pang (currently at Roche Palo Alto, Palo Alto, Calif.), Dashi Bao, Leyi Li, Jennifer Parris, and Yongqi Rong (St. Jude); Taisuke Miyazaki, Eriko Miura, and Masahiko Watanabe (Hokkaido University School of Medicine, Sapporo, Japan) and Michisuke Yuzaki (currently at Keio University School of Medicine, Tokyo, Japan).
On the World Wide Web: