9 November 2010. Two new reports reveal new facets of the interaction between neurexin and neuroligin—cell adhesion molecules implicated in schizophrenia and other brain disorders. One study, published in Cell on October 15, devises a way to directly visualize binding between neurexin and neuroligin, and finds that the two are brought together during development and in response to synaptic activity. The other study, published in Neuron on September 23, finds that neurexin-neuroligin binding has a role beyond synapses, regulating the formation of dendrites.
Discoveries of microdeletions within the neurexin gene in schizophrenia and autism, and of mutations in neuroligin genes in autism, have prompted a view of these disorders as "synaptopathies"—stemming from faulty wiring and signaling between neurons (Südhof, 2008). Consistent with a synaptic role, neurexin is located on the tip of an axon where it can bind to neuroligin, anchored on the receiving end of a neuron—forming a kind of trans-synaptic bridge across the synaptic cleft. Although in vitro studies show that the neurexin-neuroligin interaction induces synapses, knocking out these genes in mouse models reveals no deficit in synapse number, but rather in synapse function. This suggests that these proteins are involved in making the synapse fully operational, maybe by recruiting the protein machinery needed on both the pre-synaptic and post-synaptic sides (Chubykin et al., 2007). The picture is further complicated by other, potentially redundant binding partners (see SRF related news story).
Rather than overexpressing or knocking down neurexin or neuroligin, it would help to have more subtle ways to probe this partnership. Both studies make use of interaction-deficient mutant proteins, and develop new ways to visualize the results. Alice Ting at the Massachusetts Institute of Technology invented a method called biotin labeling of intercellular contacts (BLINC) that can rapidly visualize contact between two proteins, in this case, neurexin and neuroligin 1. Kurt Haas at the University of British Columbia (UBC) studied the interaction while imaging dendrite formation in vivo. Together, the studies reveal the interaction between neurexin and neuroligin to be a highly dynamic one with multiple roles in wiring up the brain.
First author Amar Thyagarajan and Ting made use of an enzyme from E. coli called biotin ligase (BirA), which can add a large biotin molecule to a small acceptor peptide (AP) when BirA and AP get close enough to interact in the presence of biotin molecules. The researchers introduced neurexin tagged with BirA into one pool of neurons, inserted neuroligin 1 tagged with AP into a separate pool of neurons, and cultured the neurons together. At day 16 in culture, when neurons are in the midst of synapse development, the researchers added a biotin-providing molecule for 15 minutes, and then stained with fluorescent streptavidin to visualize the biotin molecules. This gave robust BLINC signals that turned up where expected, colocalizing with several other pre- and post-synaptic markers. When interaction-deficient mutants of neurexin or neuroligin were used, no BLINC signal appeared.
They then used BLINC to see how neurexin-neuroligin interactions evolve during synaptic development. Clusters of BLINC signals grew as cultures matured and neurons connected—as though, when binding to one another, neurexin-neuroligin proteins aggregate with each other. This growth could be stalled by blocking electrical activity through NMDA receptors. The BLINC signals also grew in response to acute synaptic activity, with the intensity of a single cluster increasing over seven times. This growth of neurexin-neuroligin complexes stemmed from changes in trafficking of these molecules, increasing the insertion of new neurexin and neuroligin proteins into the cell membrane and slowing their internalization back into the cell. Neurexin-neuroligin complex growth also spurred AMPA receptor insertion, a hallmark of synapse maturity. The researchers noted that, because BLINC signals appeared at the earliest stage they looked at, and because they overlapped with Bassoon, an early marker of pre-synaptic development, neurexin-neuroligin complexes could be one of the earliest events leading to a mature, stable synapse.
In their Neuron paper, first author Simon Xuan Chen and colleagues at UBC explored the role of neurexin and neuroligin in forming dendrites during early brain development. This interest stemmed from a "synaptotropic" hypothesis for dendrite formation, which proposes that synapse formation itself somehow shores up the post-synaptic cell membrane to enable dendrite growth. Because binding between neurexin and neuroligin bridges the gap between pre- and post-synaptic neurons, and because they are involved in synapse building, they are good candidates for stabilizing a budding dendrite.
Using two-photon time-lapse imaging, the team watched dendrite formation unfold in 3D in intact Xenopus laevis brain. This revealed a highly dynamic picture of dendrite formation, with the dendritic shaft of a post-synaptic neuron extending and retracting processes called filopodia, as though searching for an axonal input. When a filopodium contacts an axon tip, it stabilizes, leaving a newborn dendrite; when a filopodium doesn't make contact after some time, it is eliminated.
The researchers established that blocking the neurexin-neuroligin interaction tilted this picture away from stabilization and toward elimination. Using a soluble form of neurexin to prevent endogenous neurexin from binding to neuroligin, they were able to eliminate nearly twice as many dendritic filopodia as in control dendrites. This resulted in a 35 percent decrease in filopodia density, but also an increase in filopodia motility—as though the new processes intensified their search for a neurexin-carrying axon. Disrupting the neurexin-neuroligin interaction with a neuroligin mutant that could not bind neurexin produced similar results, as did knocking down endogenous neuroligin 1 expression. In contrast, overexpression of neuroligin 1 led to hyperstabilization: motility was reduced and fewer filopodia were eliminated, leading to an increase in filopodia density and complex branching.
The researchers found that both ends of neuroligin 1 contribute to dendrite stabilization by using a neuroligin 1 mutant that cannot bind its intracellular binding partner—the synaptic scaffolding protein PSD-95—but that it can bind neurexin just fine. The researchers propose that upon initial contact, the neurexin-neuroligin interaction transiently stabilizes a growing dendrite long enough to keep it from being eliminated, but that the assembly of post-synaptic protein complexes are required for the dendrite to stick around for the longer term. This first stage of stabilization occurs independent of activity, whereas the second stage requires activity through NMDA receptors.
After four days, the hyperstabilizing effects of overexpressing neuroligin 1 led to dendritic arbors that were spatially restricted yet complex—bush-like instead of tree-like—whereas knockdown of neuroligin 1 resulted in stunted arbors that were small with fewer branch points. Dendrite patterning can ultimately effect how signals are computed and shuttled around the brain, and abnormalities have been noted in schizophrenia and autism (Jan et al., 2010).
Together, the two studies delve deeper into the function of this intriguing partnership and come up with a more complex picture of synapse biology. Neurexin and neuroligin proteins have not only given researchers an entry point for understanding the multi-step process of synapse building, but also a framework for understanding—and potentially developing therapies for—brain disorders like schizophrenia.—Michele Solis.
Thyagarajan A, Ting AY. Imaging Activity-Dependent Regulation of Neurexin-Neuroligin Interactions Using trans-Synaptic Enzymatic Biotinylation. Cell. 2010 Oct 29;143: 456-69. Abstract
Chen SX, Tari PK, She K, Haas K. Neurexin-neuroligin cell adhesion complexes contribute to synaptotropic dendritogenesis via growth stabilization mechanisms in vivo. Neuron. 2010 Sep 23; 67: 967-83. Abstract