The background genetic effect of the genes underlying the broad autism phenotype as a unifying feature in gene × gene and gene × environment causal mechanisms in autism

Al l a ut ho rs a bi de b y th e As so ci ati on fo r M ed ic al E th ic s ( AM E) e th ic al ru le s o f d isc lo su re . Abstract Introduction This model examines the role of proposed broader autism phenotype candidate genes and unfavourable pre-, periand neonatal factors and environmental hazards associated with risk for early disruption of brain development, organisation of neural circuitry and increased risk for autism. A number of designated autism susceptibility genes may be more robustly characterised as broader autism phenotype candidate genes. This review proposes five broader autism phenotype candidate genes (SLC6A4, COMT, CNTNAP2, MET, FOXP2) for further review. Phenotypes result from the expression of an organism’s genes as well as the influence of environmental factors and the interactions between the two. Broader autism phenotype candidate genes are pleotropic and include common heritable polymorphisms associated with general population risk for variable genotype–phenotype expressions that are frequently seen in autism, unaffected family members and in the general population. The general population broader autism phenotype candidate genes and genotype–phenotype expressions include social cognition and personality features, immune deficiencies, fine and gross motor incoordination, developmental language impairment, eating disorders, depression, anxiety and panic disorders, sensory processing impairments, obsessive compulsive behaviours, diabetes, gastrointestinal disorders, irritable bowel syndrome and repetitive behaviours that cluster within affected and unaffected family members and that are continuously and variably distributed throughout the general population. The independent broader autism phenotype component part is always reliant on other gene mutations inherited and/or de novo, environmental risk factors and epigenetic events acting alone or in concert that are involved in the transition to strictly defined autism. The general population risk for autism in developmentally compromised or at-risk individuals associated with a specific pre-, perior neonatal insult is calculated at about 7%. This review discusses the background genetic effect of the genes underlying the broad autism phenotype as a unifying feature in gene × gene and gene × environment causal mechanisms in autism. Conclusion Identifying unfavourable pre-, periand neonatal risk factors and environmental hazards associated with the severe developmental disorders (autism, intellectual disability, attention deficit hyperactivity disorder and schizophrenia) should be a high priority that might lead to more effective prevention strategies for these debilitating developmental conditions.


Introduction
Autism is behaviourally defined by the Diagnostic and Statistical Manual, Fourth Edition, Text Revision 1 and involves a constellation of behavioural symptoms related to social cognition personality features, delayed, absent or unusual communication styles and stereotypical or obsessional and repetitive behaviours that fall under the umbrella category of the pervasive developmental disorders (PDD).The three major sub-categories are autistic disorder, PDD-Not Otherwise Specified and Asperger syndrome.In this review all sub-categories within the PDD's are referred to as autism.Individuals diagnosed with autism and their unaffected family members may share many subtle qualitative similarities in social-cognitive personality features, language, sensory processing and repetitive behaviours that are thought to mirror some aspects of autism features but that are not associated with a debilitating neurodevelopmental condition.The broader autism phenotype (BAP) and associated features are continuously and variably distributed throughout the general population 2,3 .
An emerging consensus suggests that the broad spectrum of developmental disorders, including autism, may involve an early, long-lasting disruption of brain development that interferes with pre-programmed neuronal migration and is disruptive of normal synaptic connectivity.Neuropathological and imaging studies have consistently reported microscopic and macroscopic structural anomalies in the brain in autism 4,5 .All authors contributed to the conception, design, and preparation of the manuscript, as well as read and approved the final manuscript.

Genetic and environmental influences have been invoked but autism
All authors abide by the Association for Medical Ethics (AME) ethical rules of disclosure.
For citation purposes: Jensen RA.The background genetic effect of the genes underlying the broad autism phenotype as a unifying feature in gene x gene and gene x environment causal mechanisms in autism.OA Autism 2013 May 01;1(2):11.
aetiology has proven to be elusive 6 .
Heterogeneity and non-specificity in genetic, phenotypic, neuropathological, neuroimaging and environmental studies have emerged as a significant obstacle in advancing the understanding of the biological complexity and mechanisms underlying autism aetiology 7,8 .Two competing conceptual models for autism aetiology have been in place for decades.One model views autism aetiology primarily as a heritable genetic condition and the competing model views autism as primarily involving environmentallyinduced structural anomalies in the brain.Both models associate autism with atypical brain development, however no single rare large genetic deletion or duplication, common genetic variant, unfavourable pre-, peri-or neonatal event or environmental hazard is predictive of autism.Autism is likely to involve multiple risk factors of small effect that in aggregate increases total risk in any individual case [9][10][11] .
Rutter 12 has proposed a two-hit model and the existence of BAP candidate genes: "In other words, what is required for autism 'proper' to develop are the susceptibility genes and some other risk factor that could be either genetic or environmental in origin.The implication, if it is a two hit process is that the genes underlying the broader autism phenotype may not be exactly the same as those involved in the transition to the handicapping disorder." The long-lasting disruption of early brain development is involved in the transition to a broad spectrum of diverse neurodevelopmental conditions including autism, ADHD, schizophrenia and intellectual disability (primary mechanism).The common heritable genetic polymorphisms underlying the BAP component part (secondary mechanism) is a background genetic effect that when present is determinative of a developmental trajectory towards an autism diagnosis.In early childhood during the long-lasting rapid disruption of brain development and organisation of brain circuitry occurs, an array of BAP genotype-phenotype expressions may interactively manifest themselves at the extremes in strictly diagnosed autism.The aim of this review was to discuss the background genetic effect of the genes underlying BAP as a unifying feature in gene × gene and gene × environment causal mechanisms in autism.

The BAP candidate genes
An increasing number of autism candidate genes have been proposed and common heritable gene polymorphisms that are distributed throughout the general population may be more robustly characterised as BAP candidate genes.BAP candidate genes harbour within them subunits of genetic variants associated with autism risk.BAP candidate genes are pleotropic.Phenotypes result from the expression of an organism's genes as well as the influence of environmental factors and the interactions between the two.BAP candidate genes should include regions that harbour common genetic variants associated with general population risk for variable genotype-phenotypic expressions.
General population genotype-phenotype expre ssions in BAP genes include unique social cognition and personality type features, immune deficiencies, fine and gross motor incoordination, developmental language impairment, eating disorders, depression, anxiety and panic disorders, bipolar disorder, sensory processing impairments, obsessive compulsive behaviours, diabetes, gastrointestinal disorders, irritable bowel syndrome and repetitive behaviours that cluster within strictly diagnosed autism, unaffected first-degree relatives in the families and in the general population.
The common genetic variants in the BAP candidate genes are heritable and very common in the general population.For example, the MET (7q31) gene region harbours a polymorphism, the MET promoter variant rs1858830 allele 'C', which is present in 47% of the general population and is associated with immune function, gastrointestinal repair, neuronal growth and development 13 .One of the first autism candidate genes proposed is the serotonin transporter gene SLC64A (5-HTT), which maps to chromosome 17q 11,12 .Multiple studies have implicated common genetic polymorphisms within the COMT gene (22q11) region as an autism candidate gene.One of the earliest target-rich regions containing autism candidate genes was mapped to chromosome 7 and include MET (7q31), CNTNAP2 (7q35) and FOXP2 (7q35).Extreme genotypephenotype expressions in the BAP genes that harbour common polymorphisms may involve gene by environment or epigenetic mechanisms given how widely these polymorphisms are distributed throughout the general population.General population genotype-phenotype expressions in BAP proposed candidate genes are listed in Table 1.
In early childhood during the rapid and long-lasting disruption of early brain development, severe symptoms begin to emerge and an array of independent BAP component parts may be variably expressed at the extremes in strictly diagnosed autism.The BAP candidate genes all involve common polymorphisms associated with general population risk for obsessive compulsive disorder.In strictly diagnosed autism in early childhood, severe obsessive compulsive disorder-like symptoms may emerge including repetitively checking things, counting things, hoarding behaviours, lining up objects and amassing collections of objects.When these behaviours are disrupted severe anxiety and panic attacks are commonly observed 42 .Sensory impairments are seen and Licensee OA Publishing London 2013.Creative Commons Attribution License (CC-BY) Competing interests: none declared.Conflict of interests: none declared.
All authors contributed to the conception, design, and preparation of the manuscript, as well as read and approved the final manuscript.
All authors abide by the Association for Medical Ethics (AME) ethical rules of disclosure.
For citation purposes: Jensen RA.The background genetic effect of the genes underlying the broad autism phenotype as a unifying feature in gene x gene and gene x environment causal mechanisms in autism.OA Autism 2013 May 01;1(2):11.
may include unusual responses to touch, light, sound, heat and pain 43 .Idiosyncratic feeding behaviours are frequently observed perhaps related to general population risk for eating disorders 24 .Medical complications such as severe constipation, gastrointestinal problems, irritable bowel syndrome and immune deficiencies occur more frequently than general population norms.Impairments in gross and fine motor skills are a commonplace 44 .
The phenomenon of transient echolalia appears to be universal in verbal young children diagnosed with autism 45 .Echolalia has been seen in other conditions that feature aberrant neuronal circuitry including patients with Alzheimer disease, brain tumour and stroke [46][47][48] .

General population prevalence of the BAP and at the extremes
The BAP social cognition features concept by child psychiatry has its origins in Leo Kanner's 42 1943 classic article ' Autistic Disturbance of Affective Contact'.He described the parents as follows: "One other fact stands out prominently.In the whole group, there are very few really warm hearted fathers and mothers.For the most part, the parents, grandparents and collaterals are persons strongly preoccupied with abstractions of a scientific, literary, or artistic nature, and limited in genuine interest in people." A number of groups have measured the general population prevalence of BAP social cognition, communication and repetitive traits by devising a check list of questionnaires including the Childhood Asperger Syndrome Test, the Autism Spectrum Questionnaire and the Social Reciprocity Scale.These checklists do not contain any biological symptoms that may cluster in affected and unaffected family members and in the general population that may also be associated with increased risk for disruption of early brain development.
Genetic epidemiology has measured the prevalence of BAP traits in the general population by recruiting thousands of volunteer samples often taken from twin registries with research questionnaires to be completed by parents or teachers or by self report.The studies have consistently reported that the prevalence of BAP traits are moderately to highly heritable and extend very broadly throughout the general population.Hoekstra et al. 49 measured the BAP using Autism-Spectrum Quotient scores from volunteers recruited Of the 374 students (males = 118, females = 256) completing the test, 18 scored below the cut-off of 30 (males = 11.9%,females = 2.9%).About 20% of unaffected siblings have a history of language delay far greater than the 7.4% of general population children with a history of language delay 53,54 .Ronald et al. 55 selected the top 5% highest BAP scorers and reported high heritability that showed modest phenotypic and genetic overlap between three measures: social impairments, communication impairments and restricted repetitive and interests.
Happe and colleagues 56 reported that "Around 10% of all children showed only social impairment, only communicative difficulties or only rigid and repetitive interests and behaviour, and these problems appeared to be at a level of severity comparable to that found in children with diagnosed autism in our sample." The Rutter hypothesis was demonstrated in a gene × gene (G × G) causal mechanism, Down syndrome with autism, featuring the interplay between a de novo genetic mutation and a background BAP genetic effect that follows a developmental trajectory to an autism diagnosis.Ghaziuddin 57,58 compared a group of Down syndrome individuals and their first-degree relatives (parents and siblings) with or without a diagnosis of autism.In Down syndrome with autism there was an excess of first-degree relatives who met the description of BAP features compared to first-degree relatives in Down syndrome without autism who did not.The Down syndrome mutation and autism was not present in first-degree relatives, parents and siblings, and the genes underlying the BAP component part are independent of and are a background genetic effect secondary to the disruption of early brain development in Down syndrome and the transition to autism as predicted by the Rutter hypothesis.

Unfavourable pre-, peri-and neonatal factors in autism: gene × environment (G × E) causal mechanism
Several design methodologies examining obstetrical and environmental hazards associated with autism risk have been in place for decades.The most common design method is selecting autism diagnosed only individuals and largely unaffected sibling, neurotypical and/or national statistic controls.This method has produced consistent evidence that unfavourable events in the pre-, periand neonatal period may or may not be associated with autism risk.The data have been difficult to analyse as the factors representing possible risk for autism are not specific to autism and may represent various forms of pathological processes and developmental problems.No single unfavourable factor or unifying feature stands out as representing high autism risk 8 .
A second less frequently used design method was serendipitously introduced by Chess 59 in her 1971 seminal article ' Autism in Children with Congenital Rubella'.In the aftermath of the last rubella epidemic that took place in 1960s in the US, 243 children diagnosed with congenital rubella syndrome were placed under her care at the New York University Medical Centre.Of the 243 children, a psychiatric diagnosis of autism was reported in 10 children and a psychiatric diagnosis of atypical autism in a further eight children which represented 7.4% of the total group, consistent with cut-off estimates for highscoring BAP social cognition trait prevalence in the general population.A comprehensive detailed description of the behavioural disturbances in two of the boys under Chess's care diagnosed with autism was presented.The behavioural disturbances observed were indistinguishable from the very detailed behavioural disturbances of the 11 children described by Kanner 42 .
The estimates for the prevalence of autism risk represented by the BAP highest scorers by various cut-offs in the general population of between 5% and 10% have been reported by Plomin's group.The cut-offs can be applied to the variable diagnostic outcomes within studies based on the Chess design method.The design selects all individuals identified with a specific unfavourable pre-, peri-or neonatal event and reports the prevalence of autism within the group.The Rutter hypothesis would expect striking heterogeneity in outcomes and that the prevalence of autism within the studies would be consistent with and test the reliability of the general population autism risk estimates based on high BAP scoring cut-off ranges.Exclusionary criteria are studies with less than 20 participants, retrospective studies based on parental recall, studies that compared autism diagnosed samples only with various control groups and studies that used autism screening tools only (Table 2).

Discussion
Of the 17 studies listed only one, infantile spasms, was an outlier for autism risk with a 30% Licensee OA Publishing London 2013.Creative Commons Attribution License (CC-BY) Competing interests: none declared.Conflict of interests: none declared.
All authors contributed to the conception, design, and preparation of the manuscript, as well as read and approved the final manuscript.
All authors abide by the Association for Medical Ethics (AME) ethical rules of disclosure.
For citation purposes: Jensen RA.The background genetic effect of the genes underlying the broad autism phenotype as a unifying feature in gene x gene and gene x environment causal mechanisms in autism.OA Autism 2013 May 01;1(2):11.
prevalence (6/20).The results of 16 of the 17 studies were predicted by the estimates of autism risk associated with a specific pre-, peri-or neonatal unfavourable factor in the general population of developmentally compromised or at-risk individuals of around 7% (the highest BAP scorers) and by the Rutter hypothesis.Six of the syndromes listed (Table 2) have such a hugely massive environmental effect that preventative measures can or have been put in place.Prevention measures are in place for congenital rubella syndrome and thalidomide embryopathy with the development of an effective rubella vaccine and the ban on the use of thalidomide in obstetrics.Foetal alcohol syndrome, valproate acid syndrome, perinatal cocaine exposure syndrome and post institutional autistic syndrome are all preventable syndromes with a massive environmental effect and high autism risk.
The prevalence of autism in many genetically determined syndromes is frequently much greater than the approximate 7% that the Rutter hypothesis would predict.Moss and Howlin 76 have presented good evidence that in the genetically determined syndromes autism symptoms are associated with intellectual disability and cautioned against over interpreting the superficial similarities between autism and the behavioural phenotypes seen in most genetically determined syndromes.The authors also found that many genetically determined syndromes have their own unique pattern of superficial autism symptoms.
Twin studies in autism where one or both twins met diagnostic criteria for strictly defined or broadly defined autism have produced monozygotic twin concordance rates as high as 92% in broadly defined autism.Studies of general population twins have reported high BAP concordance rates and heritability estimates in general population twins, dependent on arbitrary cutoffs.Classical twin study design heritability estimates cannot control for the high rates of de novo mutations in autism therefore autism twin study heritability estimates are inflated.Common gene polymorphisms in candidate BAP genes are inherited therefore BAP concordance rates and heritability estimates may be more informative with respect to heritability estimates and general population autism risk.The implication within this model is that what the twin studies are reflecting may not be the heritability of autism at all but rather the concordance rate and heritability estimates of the independent BAP component part.The independent BAP component part can be characterised as the 'missing heritability' in autism 77 .
The background genetic effect of the genes underlying the broad autism phenotype as a unifying feature in gene × gene and gene × environment causal mechanisms in autism RA Jensen* Diagnosis Advancements *Corresponding author Email: rajensen088@aol.com6100 N. Brookline Ave, #27, Oklahoma City, OK 73112, USA Licensee OA Publishing London 2013.Creative Commons Attribution License (CC-BY) Competing interests: none declared.Conflict of interests: none declared.

Table 1 BAP candidate genes: genotype-phenotype expressions
41Licensee OA Publishing London 2013.Creative Commons Attribution License (CC-BY)Competing interests: none declared.Conflict of interests: none declared.All authors contributed to the conception, design, and preparation of the manuscript, as well as read and approved the final manuscript.All authors abide by the Association for Medical Ethics (AME) ethical rules of disclosure.For citation purposes: Jensen RA.The background genetic effect of the genes underlying the broad autism phenotype as a unifying feature in gene x gene and gene x environment causal mechanisms in autism.OA Autism 2013 May 01;1(2):11. 52reported on the selfadministered Empathy Quotient test, thought to measure the prevalence of a putative BAP trait, empathy, taken by undergraduate students.A score below 30 represents the cut-off that best differentiates the presence of this putative BAP trait from controls.

Table 2 Autism prevalence identified with specific pre-, peri-and neonatal unfavourable factors in the general population Description ASD (%) Total diagnosed ASD Total in group Study design Author (s)
*, pooled results of two post institutional autistic syndrome papers