Autism is a neurodevelopmental disorder characterized by deficits in social communication and interaction, impaired language development, and stereotyped repetitive behaviors. The broader category of autism spectrum disorders (ASDs) includes individuals with subsets of these symptoms (Geschwind and Levitt, 2007). The prevalence of autism in the USA has increased since the inception of the diagnosis in the 1980s – to about 1 in every 110 children. The majority of this rise is attributed to greater screening efforts and expanded diagnostic criteria (Centers for Disease Control and Prevention, 2009).
The study of autism is challenging because of the heterogeneous phenotypic presentation of the disease and the complexity of its inheritance, and no primary treatment exists. Although the genetic heritability is high – with parental susceptibility genes contributing to an estimated 80% of cases – the polymorphic nature of autism has made identification of individual genes difficult (Freitag, 2007). Unraveling this genetic component is crucial to fully understanding ASDs.
Investigations of familial autism have yielded several candidate genes, including neuroligin-3 and -4, which are associated with ASDs in a subset of family pedigrees (Jamain et al., 2003; Laumonnier et al., 2004; Yan et al., 2005). Neuroligins are post-synaptic cell adhesion molecules that bind to presynaptic neurexins and function in the maturation, stability and maintenance of synapses (Sudhof, 2008). This function of neuroligins is consistent with numerous lines of evidence supporting a role for neuronal connections, synaptogenesis and plasticity in autism (Levy et al., 2009).
Discovering a role for neuroligins in autism had immediate implications for developing animal models of the disease. Most previously existing mouse models of autism represent co-morbid genetic disorders such as Rett syndrome or fragile X syndrome, with limited utility to address ASDs specifically. The mouse neuroligin-3-knockout model for autism generated excitement as the first monogenic animal model of autism (Radyushkin et al., 2009). A neuroligin-1-knockout mouse soon followed (Blundell et al., 2010). Both sets of mice lack an obvious autistic phenotype, but display endophenotypes reminiscent of ASDs. Although complex animals such as mice have the best potential to recapitulate complex human behavioral phenotypes, organisms such as Caenorhabditis elegans, Drosophila melanogaster and Saccharomyces cerevisiae provide powerful, tractable systems with which to elucidate the fundamental genetic and mechanistic causes of disease.
In a recent issue of DMM, Hunter et al. report behavioral phenotypes consistent with ASDs following removal of the neuroligin homolog, nlg-1, from C. elegans (Hunter et al., 2010). Through cDNA sequencing and reverse-transcriptase polymerase chain reaction (RT-PCR) they demonstrate that the nlg-1 transcript undergoes extensive alternative splicing [figure 1 in Hunter et al. (Hunter et al., 2010)]. The structure of the longest predicted neuroligin isoform is similar to that of mammalian neuroligins, containing a large extracellular cholinesterase-like domain, a type 1 transmembrane domain, and a small intracellular domain terminating in a PDZ-binding motif [figure 2 in Hunter et al. (Hunter et al., 2010)].
Like its mammalian homologs, C. elegans nlg-1 is neuronally expressed and localizes to postsynaptic regions. The authors define the expression pattern and subcellular localization of this neuroligin using transcriptional and translational reporters. An nlg-1-promoter-driven transcriptional reporter was expressed in a subset of ventral nerve cord, head and body neurons – roughly 45 of the 302 present in adult worms [figure 3 in Hunter et al. (Hunter et al., 2010)]. Furthermore, a fusion protein comprising neuroligin and yellow fluorescent protein (YFP) localized to nerve ring and ventral nerve cord synapses throughout development, with specific juxtaposition to presynaptic markers in adult animals [figure 4 in Hunter et al. (Hunter et al., 2010)]. The neuroligin expression pattern in C. elegans suggests that it is important to neuronal function.
Knockout of nlg-1 produces phenotypes consistent with ASDs. To explore the function of nlg-1, the effect of two independent mutant nlg-1 alleles on worm development and function were examined [figures 1 and 2 in Hunter et al. (Hunter et al., 2010)]. Nervous-system formation and synaptic function were grossly intact in homozygous nlg-1 mutants, which were superficially wild type in appearance, development and behavior. However, upon further inspection, some phenotypes were uncovered in the mutant worms. Although nlg-1-deficient animals responded normally to most chemical cues, their behavior was altered in the presence of the aversive chemical 1-octanol or when presented with two conflicting chemosensory inputs [figure 5 in Hunter et al. (Hunter et al., 2010)]. Mutants of nlg-1 also completely lacked any thermal response and had altered locomotion after being removed from their food source [figure 5 in Hunter et al. (Hunter et al., 2010)]. Importantly, phenotypes in mutant animals were rescued by transgenic nlg-1 expression [figure 5 in Hunter et al. (Hunter et al., 2010)]. The sensory defects present in nlg-1-mutant worms are intriguing, as ASD patients often have difficulties processing or integrating information (American Psychiatric Association, 1994).
The metabolism of nlg-1-mutant worms was also altered. Basal levels of oxidative damage to proteins were significantly higher in mutants than in wild-type animals [figure 6 in Hunter et al. (Hunter et al., 2010)]. Animals deficient in neuroligin were thus more sensitive than their wild-type counterparts to agents like paraquat, copper and mercury, which cause or exacerbate oxidative stress. Again, transgenic expression of nlg-1 was able to rescue all mutant phenotypes [figure 6 in Hunter et al. (Hunter et al., 2010)]. Notably, individuals with autism also present biomarkers associated with oxidative stress (Chauhan and Chauhan, 2006).
This C. elegans model is a monogenic model of autism corresponding to orthologous genes in the human, mouse and worm, with synergistic potential to study ASDs. C. elegans provides a tractable and well-characterized system in which to study autism. Nematodes are unique in that neurodevelopment is characterized at the cellular level and one can observe the developmental, behavioral and molecular phenotypes simultaneously. Future research would benefit from a more rigorous examination of the structure and function of neuroligin in C. elegans synapses and monitoring neurogenesis in order to define the developmental and molecular mechanisms of disease pathogenesis. In addition to elucidating how a neuroligin deficit contributes to autism, this C. elegans model presents the opportunity to study other genetic complexities of autism. The relative ease of nematode genetics enables researchers to uncover other potential autism-linked genes by screening for suppressors and enhancers of the phenotypes described by Hunter et al. This model will also allow exploration of how polymorphisms modulate nlg-1-mutant phenotypes by making equivalent nlg-1 mutations in different available C. elegans genetic backgrounds. The worm phenotype is subtle, similar to mouse models of autism and consistent with the ambiguity of many autism characteristics in humans. The sensory deficits present in the neuroligin-deficient nematode and mouse complement each other and mimic facets of human ASD behavior.
C. elegans models are amenable to high-throughput analysis, especially with technological advances such as the Complex Object Parametric Analyzer and Sorter and microfluidic manipulation (Pulak, 2006; Rohde et al., 2007). Recently, a C. elegans model of dystonia was used to screen for small molecules that affect movement, and the resulting compounds were validated in mouse and human cells (Cao et al., 2010). The development of a high-throughput behavioral screen in neuroligin-deficient worms would allow for further study of gene-gene and gene-environment interactions, and for the screening of compounds that might possibly ameliorate nlg-1-mutant phenotypes and thus provide leads towards potential autism treatments.
This C. elegans model is a viable starting place to examine a variety of questions relevant to autism. If mutation of nlg-1 increases oxidative damage, does the oxidative stress cause the behavioral changes in the mutant worms? What gene expression changes are induced by mutant forms of nlg-1 and could these transcriptional alterations explain the behavior of the mutant worms? Despite the need for further characterization, the work of Hunter et al. represents an important addition to the field of autism research. Even if this particular mutant is not validated as an autism model, information will be gained about neuroligins and their role in development – knowledge with direct applicability to ASD research and treatment.