5 January 2010. A synapse-promoting protein called LRRTM2 has been identified as a ligand for neurexin, a strong candidate susceptibility gene for schizophrenia, autism, and other neuropsychiatric disorders (see SRF related news story). This newly discovered LRRTM2-neurexin interaction, described in two reports published in Neuron on December 24, adds to the list of diverse partnerships between cell adhesion molecules that bridge across neurons to regulate how synapses are established.
The two studies—one from Thomas Südhof and colleagues at Stanford University and one from Anirvan Ghosh's group at the University of California, San Diego—come swiftly after the initial identification of LRRTM (that's leucine-rich repeat transmembrane) proteins as stimulators of synapse formation in cell culture by Anne Marie Craig's group at the University of British Columbia (Linhoff et al., 2009). LRRTMs comprise a family of four proteins found in neurons that span the cell membrane and protrude into the extracellular space with a long stretch of leucines, a structure predisposed to interacting with other proteins. A variant in the gene for the LRRTM1 type has been linked to schizophrenia (Francks et al., 2007 and Ludwig et al., 2009), and although the new findings concern LRRTM2 specifically, they reiterate the theme that problems with synapse assembly and maintenance can lead to psychiatric disorders (Südhof, 2008).
Using complementary and sometimes overlapping methods, the new studies place LRRTM2 on the receiving, postsynaptic end of a developing synapse. When LRRTM2 binds the extracellular part of neurexin located in a contacting axon, this triggers excitatory synapse formation, as documented by increases in synapse density, size, glutamate receptor aggregation, and synaptic currents. Although similar to the synapse-promoting bridge formed between neuroligins and neurexins, LRRTMs are unrelated to neuroligins, and as Südhof's group found, somewhat more discerning: LRRTM2 bound only those neurexin proteins that lacked a specific splice site, while neuroligins bind neurexins both with and without this site. This difference underscores a variety of possible interactions between cell adhesion molecules, which can contribute to specifying the diverse synapse types in the brain.
Similar to Craig's study, first author Jaewon Ko and colleagues in Südhof's lab initially probed the synapse-promoting properties of LRRTM2 by seeing if they could trick a neuron into forming an artificial synapse with a non-neuronal cell made to express LRRTM2. After two days, the LRRTM2-containing cells sported synapse-like junctions that stained positively for presynaptic markers of vesicles containing glutamate, but not the inhibitory neurotransmitter GABA. This indicated that LRRTM2 induced specifically excitatory synapses. To verify this with real synapses, they overexpressed LRRTM2 in cultured neurons and also found an increase in excitatory, but not inhibitory, synapse density and size.
Next, the researchers looked for LRRTM2's presumed partner in triggering synapse formation. Using affinity chromatography, the researchers trolled a mixture of brain proteins with LRRTM2 to trap those that could stick tightly to LRRTM2. They pulled out both α and β forms of neurexin, a known presynaptic protein. Neurexins bound LRRTM2 tightly, even when LRRTM2 was on the cell surface, indicating a direct interaction between the two proteins involving their extracellular ends. Further analysis revealed the sensitivity of LRRTM2 to splice site #4 on both forms of neurexin. The LRRTM2-neurexin interaction was critical to artificial synapse formation because adding free-floating neurexin—which interferes with the access of axon-tethered neurexins to LRRTM2—substantially reduced synapse formation.
A functional focus
The study by Ghosh, first author Joris de Wit, and colleagues also begins with an artificial synapse assay, but once the researchers found that LRRTM2 induced more synapse-like structures than the other members of the LRRTM family, they were quickly tinkering with endogenous LRRTM2 in hippocampal neurons. When the researchers introduced a silencing RNA molecule directed to LRRTM2 to lower endogenous levels of LRRTM2 in cultured neurons, they found a 40 percent decrease in synapse density in only excitatory synapses; inhibitory synapse formation was unchanged.
In addition to inducing presynaptic specializations, LRRTM2 also seemed to organize the post-synaptic side of things. Knocking down LRRTM2 levels in neurons led to a 33 percent decrease in the density of GluR1 protein—a subunit of the AMPA type of glutamate receptor—on the cell surface, and the extracellular portion of LRRTM2 colocalized to several other glutamate receptor subunits introduced into non-neuronal cells. Meanwhile, the intracellular portion of LRRTM2 colocalized to PSD-95, a synaptic scaffolding protein found exclusively in excitatory synapses, suggesting that LRRTM2 assembles post-synaptic elements both above and below the cell membrane. Then came an acid test: Could decreasing LRRTM2 tamper with excitatory synapse function in vivo? To address this, the researchers injected a lentivirus containing the LRRTM2-lowering RNA molecule into the hippocampus of young rats. A week or so later, hippocampal slices were cut and simultaneous recordings of infected and uninfected dentate gyrus neurons showed that the infected ones—with less LRRTM2—had a 58 percent reduction in synaptic currents flowing through AMPA receptors when compared to uninfected cells with normal amounts of LRRTM2; a 54 percent reduction was measured for NMDA currents. These results argue that loss of LRRTM2 decreases glutamatergic neurotransmission in vivo.
The group then turned their sights to identifying a partner for LRRTM2, whose extracellular end they determined to be crucial for forming synapses. They, too, found that neurexins readily bound LRRTM2, and further experiments winnowed down LRRTM2's binding partners to the neurexin 1α and 1β isoforms. The study also concluded by showing that neurexin was critical to the synapse-forming abilities of LRRTM2 in the artificial synapse assay, using a technique different from Südhof's study: when an RNA construct that lowered neurexin 1 expression was introduced into the neurons, the non-neuronal cells expressing LRRTM2 could no longer induce synapses.
The different approaches in these two studies converge onto a picture of a new partnership between cell adhesion molecules: when post-synaptic LRRTM2 binds to presynaptic neurexin, excitatory synapses emerge. The exact nature of this synapse-promoting ability remains unclear, however. Rather than instigating the initial contact between two neurons, the LRRTM2-neurexin interaction could be critical for organizing or stabilizing transient synapses that are initially formed by another independent mechanism. What is clear is that there are multiple cell adhesion pathways operating in parallel between neurons, and these may specify the wide range of synapse types in the brain, or even provide redundancy. The complexities of this synaptic apparatus also point to more ways to disrupt—as well as potentially fix—synapse formations that may go awry in brain disorders like schizophrenia and autism.—Michele Solis.
Ko J, Fuccillo MV, Malenka RC, Südhof TC. LRRTM2 functions as a neurexin ligand in promoting excitatory synapse formation. Neuron. 2009 Dec 24; 64: 791-798.
de Wit J, Sylwestrak E, O'Sullivan ML, Otto S, Tiglio K, Savas JN, Yates JR, Comoletti D, Taylor P, Ghosh A. LRRTM2 interacts with neurexin1 and regulates excitatory synapse formation. Neuron. 2009 Dec 24; 64: 799-806.