Schizophrenia Research Forum - A Catalyst for Creative Thinking

Live Discussion: Neurexin—From Basic Biology to Cognitive Diseases


Thomas Südhof

Steven Clapcote

On 23 March 2011, SRF held a live online discussion about neurexin with Thomas Südhof of Stanford University and Steven Clapcote of the University of Leeds.

From its humble beginnings as the identified target of a toxin in black widow venom, to its current status as suspected culprit in some cases of schizophrenia and autism, neurexin fascinates basic scientists and clinicians alike.

Anchored in the tip of an axon, neurexin binds to molecules on the receiving end of the synapse. This interaction across the synaptic cleft assembles the working parts of synapses and influences how they work. Because miniscule changes in synapse function can cascade into substantial changes in how signals are routed through the brain, mutations of neurexin and its associated molecules can aid us in understanding the basics of communication across the synapse, and what goes wrong in brain disorders like schizophrenia.

Using Clapcote's recent review article (Reichelt et al., 2011) as a starting point, Clapcote and Südhof guided a discussion of this multifaceted molecule that boasts thousands of isoforms and multiple binding partners. How might neurexin participate in specifying the different synapse types in the brain, and how might it shape our understanding of brain disorders like autism and schizophrenia as "synaptopathies" stemming from dysfunctional synapses?

Suggested Reading: Reichelt AC, Rodgers RJ, Clapcote SJ. The role of neurexins in schizophrenia and autistic spectrum disorder. Neuropharmacology. 2011 Jan 22. Abstract

View Transcript of Live Discussion — Posted 13 April 2011

View Comments By:
Tristram Lett, James L. Kennedy — Posted 22 March 2011



Transcript

Attendees/Participants

Nick Brandon. Pfizer Neuroscience
Steven Clapcote, University of Leeds
Felipe Espinosa, UT Southwestern Medical Center
Kathy Evans, University of Edinburgh
Kevin Jones, Children’s National Medical Institute
John Gray, University of California San Francisco
Hakon Heimer, Schizophrenia Research Forum
Tristram Lett, Centre for Addiction and Mental Health, Toronto
Deborah Levy, McLean Hospital
Aaron Morley
Eric Norstrom, University of Chicago
Seth Shipman, University of California San Francisco
Anna Smith
Michele Solis, Schizophrenia Research Forum
Pippa Thomson, University of Edinburgh
Rosie Walker, University of Edinburgh

Note: Transcript has been edited for clarity and accuracy.


Hakon Heimer
Let's start off by having all participants in the "room" introduce themselves. Please type your name and affiliation or institution, and you might like to mention in a sentence your main project or interest in neurexin, since neurexin is a relatively new point of focus. I'm Hakon Heimer, editor of the Schizophrenia Research Forum.

Michele Solis
I'm Michele Solis, science writer for the Schizophrenia Research Forum.

Kevin Jones
Kevin Jones, postdoc, Children's National Medical Institute.

Pippa Thomson
Pippa Thomson, University of Edinburgh.

Steve Clapcote
I'm Steve Clapcote, lecturer at Leeds University in the U.K.

Kathy Evans
Kathy Evans, University of Edinburgh. We identified differential expression of Nrxn1 and Nrxn3 in the DISC1 L100P model of psychiatric illness.

Felipe Espinosa
I am a senior research scientist at UT Southwestern Medical Center, Dallas.

Tristram Lett
Tristram Lett, Centre for Addiction and Mental Health, Toronto.

Hakon Heimer
If you haven't chatted before, you'll see that many things can happen at once in the chatroom. So, while the rest of you introduce yourselves, I would like to introduce and thank our chat leaders. In our usual informal spirit, I won't list their lengthy achievements, but just say that Steve Clapcote is at the University of Leeds, where he works on the molecular biology and genetics of mental disorders. Tom Südhof is at Stanford University, and his lab focuses on the formation of synapses and also on pre-synaptic aspects of neurotransmitter release.

Eric Norstrom
Eric Norstrom, postdoc, Department of Neurobiology, University of Chicago. Typically Alzheimer's research. I'm interested in neurexin as an influencer of cognitive disease states.

Seth Shipman
Seth Shipman, graduate student, UCSF.

Felipe Espinosa
I've been studying corticostriatal neurotransmission in neuroligin 1 mutants in the lab of Craig Powell.

Hakon Heimer
I thank Steve and Tom for spending their lunch/dinner hours chatting with us, and also Michele Solis, SRF writer, for making this happen. I'll now turn the floor over to Steve.

Steve Clapcote
Thanks, Hakon. I spoke with Hakon and Michele last Friday to try to get some structure for our one-hour discussion. We came up with a rough plan that started with the association of Nrxn1 with autism and schizophrenia. Do we have any human geneticists in the chatroom who contributed to this work?

Tristram Lett
Jim Kennedy is here with Tris Lett in Toronto.

Steve Clapcote
From my reading of the literature, Nrxn1α has been implicated in a subset of autism, mental retardation, and schizophrenia cases, based on the finding of deletions affecting the coding sequence of Nrxn1α but not β.

Michele Solis
There is also a little evidence for duplications within Nrxn1α associated with autism and schizophrenia.

Steve Clapcote
What is most interesting about the Nrxn1α CNV findings are that the associated phenotypes are so varied, from autism that is diagnosed before three years of age to schizophrenia, which is a young adult-onset disorder. Presumably, there are other as yet unidentified genetic and environmental factors that are influencing the manifestation of the phenotype.

Deborah Levy
The same seems to be true of most, if not all, of the rare variants associated with schizophrenia.

Steve Clapcote
Deborah, yes, I notice that SHANK3 CNVs are also associated with both autism and schizophrenia.

Deborah Levy
So what accounts for the variable expressivity and age of onset?

Steve Clapcote
Deborah, regarding that variable expressivity, I don't have any data to support my answer. However, it's possible that all Nrxn1α deletion carriers show a cognitive endophenotype that's not always manifested as illness. The original DISC1 family is like this—all carriers of the translocation showed an information processing deficit, but only 60 percent were mentally ill.

Deborah Levy
Steve, interesting. The information processing deficit is more penetrant, possibly because it is more benign.

Tristram Lett
Regarding the genetics, it is notable that the GWAS of schizophrenia and autism did not find much for neurexin, presumably since they only looked at certain SNPs and could only assess common variants of those SNPs.

Kathy Evans
Tristram, probably phenotypic heterogeneity also causes problems in GWAS of these conditions.

Michele Solis
Tristram, what is your sense for how carefully probed the neurexin area is in various GWAS?

Tristram Lett
Michele, there will be regions of this large gene that were not fully assessed, and resequencing will cover the whole gene in detail in the future.

Steve Clapcote
Another recent finding that connects Nrxn1 with schizophrenia is Kathy Evans's finding that the expression of Nrxn1 is dysregulated in a DISC1 mutant mouse model of schizophrenia. Kathy, maybe you could tell us more about that work?

Kathy Evans
Steve, we carried out whole-genome expression microarray analysis comparing adult hippocampal expression in the DISC1 L100P mouse model of schizophrenia and found differential expression of Nrxn1 and 3. We then showed differential expression of these genes throughout development. The biggest differences were at E18 and P7.

Pippa Thomson
All, Kirov et al., 2008, reported a de novo duplication of the Nrxn1 interactor APBA2 in schizophrenia.

Hakon Heimer
Steve, since we got a late start—I suggest throwing out your questions on the methods/technologies/mice/flies, etc., while the genetics discussion continues separately. As discussion leader, you get to play speedchess with everyone. ::blink

Steve Clapcote
Kathy, can you tell us something about the difficulty you had in getting Nrxn1 antibodies to work. This is an issue for those of us new to the field.

Kathy Evans
Steve, we tried two different antibodies for both Nrxn1 and 3; none of the four antibodies worked.

Steve Clapcote
Tom, you've done more work than the rest of us combined on neurexins. Is there a reliable Nrxn1α antibody available? Preferably commercially?

Tom Südhof
Steve, the only neurexin antibodies that we have tested that are commercially available and work are from Synaptic Systems GmbH.

Nick Brandon
Hakon, it's Nick Brandon here. Reading out of general interest but indulged by recent DISC1 link.

Hakon Heimer
Nick Brandon, what does the DISC1-Nrxn1 reported link trigger in your mind in terms of possible interactions?

Nick Brandon
Hakon, there is a need to understand the mechanism behind the change reported by Kathy. We are starting to get a better idea on the role of DISC1 at synapse, but all studies have not looked at neurexin biology. There are some key experiments to be done here. We are still working in the area.

Steve Clapcote
Kathy Evans's finding that the expression of Nrxn1 and 3 were both dysregulated suggests that Nrxn3 could be an autism or schizophrenia risk factor as well. The evidence for Nrxn3 from human genetic studies so far isn't convincing, though.

Kathy Evans
Steve, Nrxn3 is also differentially expressed in the cell lines from individuals with the translocation that disrupts DISC1.

Steve Clapcote
The Nrxn1α KO mouse that Tom Südhof's lab created is a potentially powerful tool. Craig Powell and Tom have published a preliminary behavioral and electrophysiological study of hom KO mice. These mice are available from JAX, although they've also got Nrxn2 and Nrxn3 KO'd as well, so one has to backcross.

Steve Clapcote
Following on from the PNAS paper that Tom and Craig Powell published about two years ago, has anymore work been done on the Nrxn1 KO mice?

Michele Solis
And what to make of the difference between neurexin1α knockouts in mice and Drosophila? Mice show no change in synapse number, but flies do.

Tom Südhof
Michele, the Drosophila studies do not show that synapse numbers are altered; they show that the organization of synapses is altered. Please note that the relation between synapse numbers and synaptic activity is complex, and depends on the type of synapses, i.e., differs between brain regions, neuromuscular junctions, etc. Finally, Drosophila doesn't have β-neurexins....

Hakon Heimer
All, we're moving more into the biology and neurexin role in synapse development, but certainly if there are anymore thoughts/questions on the genetics or modeling/methods, throw those out.

Pippa Thomson
All, apart from the synapse, is there any evidence for Nrxn1 at nodes of Ranvier?

Michele Solis
Tom, I'm wondering what your take is on the new binding partners for neurexin that have been discovered in the past year or so (LRRTM2s, Cbln, GABAARs). Do the multiple types of interaction suggest ways of specifying different kinds of synapses, or redundancy?

Tom Südhof
My hunch is that neurexins will interact with many post-synaptic binding partners to mediate a large number of trans-synaptic signals that are regulated among others by alternative splicing of neurexins, and that all of the recently reported interactions are important.

Michele Solis
Tom, do you think that multiple interactions may well cooperate at the same synapse? And how to explore the role of each interaction separately? This seems like a real challenge.

Tom Südhof
Some interactions are mutually exclusive; others are compatible with each other. The way to analyze the role of the various interactions will be to do the hard work, and to create conditional mice carrying various mutations—there are no shortcuts, and there is no fast easy way of looking at these proteins. It will take a few years, and I wish more labs would work on this to accelerate the process, since the evidence that neurexins are central regulators of everything synaptic is becoming stronger.

Michele Solis
I guess the bright side—for me—is that these numerous interactions may be specific to certain synapses or regions of the brain. This provides a way to reconcile how deficits in a brainwide synaptic protein can alter only certain brain circuits in disease. How do others think about this?

Felipe Espinosa
Michele and Tom, I agree with your point that specific interactions of splice variants of neurexin1 with diverse proteins may vary among different synapses within a circuit or in different neuronal networks, and that this may explain neural network-specific symptoms. In fact, animal models start showing exactly that (as an example, see Gibson et al., 2008).

Steve Clapcote
In the DISC1 field, we got lucky by finding some point mutations that seem to differentially affect the binding sites of interactor proteins. It's a bit of a shot in the dark approach, but some interesting findings are coming out from those mice now. Maybe it is worth screening an ENU mutant archive for Nrxn1 mutations?

Nick Brandon
Tom, Steve, this may have been asked, but have you looked at DISC1 levels or location in any neurexin mice?

Steve Clapcote
Nick, I've only just got a Nrxn1 KO colony established. However, it would be an easy experiment to do, now what I know from Tom where I can get a reliable Nrxn1α antibody. Alternatively, I could use RT-PCR like Kathy did. Kathy, apart from finding different expression, have you seen any other changes in neurexins in the DISC1 mice?

Kathy Evans
Steve, the lack of antibodies stopped us doing this so far. My understanding is that Nrxn1 is synaptic activity dependent; therefore, if DISC1 is mutant, this may affect the synapse and thus Nrxn, and maybe expression of many other genes. I don't think we can call this an interaction at this stage. It could also be due to an impact of DISC1 on transcription or some sort of compensatory response.

Michele Solis
Tom, Steve, your thoughts on what neurexin biology tells us about schizophrenia and autism?

Steve Clapcote
Michele, the finding of Nrxn1 deletions in both disorders adds to the evidence that their etiology is closer than was previously thought.

Tom Südhof
I would interpret the human genetics data, not only those on neurexin1α, as telling us that schizophrenia and autism are, at least in part, synaptic diseases, and that any serious drug development targeting these diseases needs to target synapses, and can only be based on an understanding of synapses.

Hakon Heimer
All, we're now at the end of our official time—so time for final thoughts or grand sweeping statements.

Michele Solis
All, as a non-clinician, I would say the deficits in synaptic function—and not synapse number—in neurexin deletions seem more promising to me from a therapeutic standpoint. Wouldn't it be easier to fix problems in existing synapses, rather than having to rewire them?

Steve Clapcote
Yes, I agree with that. Seth Grant's work on synaptic proteins is pointing in the same direction. I've heard Seth say that there are some existing drugs that could help synaptic diseases, but I'm not sure which he was referring to.

Michele Solis
It seems a real challenge to get drugs that will work specifically on the synapses that are ailing, and leave alone the ones that are not. But I agree with Tom; only a detailed understanding of how synapses work will lead us there.

Felipe Espinosa
All, regarding treatment, James Bibb's lab at UT Southwestern Medical Center has a very interesting approach to the problem of targeting a protein that is known to interact with many others and avoiding side effects. Their approach is to specifically target the interaction of their protein of interest with another partner they believe is directly involved in the pathology, sparing the interactions with all the other partners.

Steve Clapcote
Felipe, it would be useful to find out more about that approach—it may be useful for studies on DISC1 as well.

Felipe Espinosa
Actually, I think this is pretty new stuff, so this may not be published yet. But check his lab webpage.

Hakon Heimer
I will thank Tom, Michele, and Steve for leading this discussion! For those of you who have to go to lunch and dinner (Steve!), we thank you. However, the rest of you can stay as long as you like to discuss.

Steve Clapcote
Hakon, thanks for the invite. I'm new to neurexin research, and assisted my postdoc to write a review article as a way to educate myself. I'm going to work on the Nrxn1α KO mice that Tom made. I'd be happy to send brains to anyone interested in collaborating.

Tom Südhof
Thanks—Tom

Michele Solis
Steve, sounds like a great project. It'll be fun to track your results. Tom is right; more people do need to work on this.

Steve Clapcote
Kathy, have you ever heard of that antibody company that Tom spoke of? I've got a Santa Cruz Nrxn1α antibody, but I'm sure it'll be crap.

Kathy Evans
Steve, I haven't but will be looking it up. Have a horrible feeling that one of our duff antibodies was from SC.

Eric Norstrom
Steve, Synaptic Systems. I've had good luck with many of their antibodies.

Steve Clapcote
Eric, thanks for that. I don't have much grant money, so tend to use SC, as they're cheap. However, it's a false economy because half the time they don't work!

Eric Norstrom
Steve, ha ha! So true!

Steve Clapcote
Kathy, incidentally, Tom has also made a Nrxn1α transgenic overexpressor mouse. I keep meaning to ask him whether he's still got these mice because they might be useful for your research. He may also have the original cDNA clone as well. Hakon, I'm signing off now. Thanks once again for the invite.

Michele Solis
Bye, Steve. Thanks!

Hakon Heimer
Thanks to everyone. I'll go for coffee now, too.

Michele Solis
Thanks everyone; signing off from Seattle.

Felipe Espinosa
I am taking off to do experiments. Thanks to organizers and leaders.

Comments on Online Discussion
Comment by:  Tristram LettJames L. Kennedy (SRF Advisor) (Disclosure)
Submitted 22 March 2011
Posted 22 March 2011

Accumulating evidence suggests that heterozygous deletions in the neurexin-1 (NRXN1) gene pose a significant risk to autism spectrum disorders and schizophrenia. In contrast, no common variant in NRXN1 has been associated with either disorder, suggesting that the risk created by these markers is limited. However, this does not exclude the possibility that common variants could be important in alternate phenotypes. For instance, a genetic change that does not result in truncated or dysfunctional protein, but likely in aberrant gene regulation at critical times of development, may have greater relevance in a more homogeneous subgroup of schizophrenia patients.

Recent evidence suggests that NRXN1 modulates post-synaptic differentiation of glutamatergic synapses potentially through neuroligin or leucine-rich repeat transmembrane proteins. Considering some antipsychotic medications’ potential action on glutamate regulation, dysregulation of NRXN1 by common variants (or haplotypes) could potentially alter antipsychotic efficacy. We have found that common alleles in NRXN1 are indeed related to the phenotype of clozapine response in schizophrenia (Souza et al., 2010). With the characterization of these genotype-intermediate phenotype relationships, it may become more clear as to how disruption of NRXN1 regulation may underpin the neuropathology of schizophrenia, autism spectrum disorders, and mental retardation. Therefore, we advocate additional examination of common variants in NRXN1, particularly using more specific subphenotypes (e.g., clozapine response or non-response).

References:

Souza RP, Meltzer HY, Lieberman JA, Le Foll B, Kennedy JL. Influence of neurexin 1 (NRXN1) polymorphisms in clozapine response. Hum Psychopharmacol. 2010 Nov;25(7-8):582-5. doi: 10.1002/hup.1146. Abstract

View all comments by Tristram Lett
View all comments by James L. Kennedy