Quantitative Trait Locus Analysis, Nonverbal Communication and Autism: A research paper review

Contributors: Inzamam-ul Hoque Kabya, Saoda Tasneem, Ashik Ahmed Akash , Naushin Shaira Ananna

Title: Quantitative Trait Locus Analysis of Nonverbal Communication in Autism Spectrum Disorders. 

Abstract:  QTL analysis of NVC in order to find out the regions of the chromosomes that are most significantly responsible for non-verbal quantitative traits in people with ASD. 

Keyword: Autism, Non-verbal Communication, QTL Analysis

• Autism: Autism is a genetic disorder. ASD has 3 deficits. Autism is diagnosed when the 3 deficits are in the severe stage.
• Non-Verbal Communication: Non-Verbal Communication is the nonlinguistic transmission of information through visual, auditory, tactile, and kinestic channels. It includes the use of visual cues such as body language, distance, and physical environments of voice and of touch.  It includes facial expression, tone and pitch of voice, gestures displayed through body language, and the physical distance between communicators. 
• QTL Analysis: QTL is a locus that correlates with the variation of a quantitative trait in the phenotype of a population of organisms. A region of DNA is associated with a particular phenotypic trait, which varies in degree. Those are found in different chromosomes. 

Introduction: ASD (Autism Spectrum Disorder) has deficits in three behavioral domains:

1. Social Interaction 
2. Communication 
3. Restrictive & repetitive behavior 

The disorders under ASD has at least one of the three behavioral deficits. Autism is the most severe form of ASD as it has deficits in all of these three domains. There are some common predisposing genes in the disorders under ASD which create chromosomal heterogeneity. 

According to ADI-R (Autism Diagnostic Interview-Revised), ASD children with repetitive behavior is nine times more likely to have a parent with the obsessive-compulsive disorder than ASD children without this behavioral problem. There is strong evidence of genetic liability for having ASD but the responsible gene for ASD is not specified yet. 

QTL is an alternative mapping strategy that decreases the heterogeneity among genetically complex disorders. The method tests each chromosome region for a relationship between the degree of genotype sharing and the similarity of that quantitative trait in affected relative pairs. QTL analysis is a linkage of genotypic data and phenotypic data. NVC is a subdomain of ASD which reflects the severity of other impairment and also a useful trait for ASD gene mapping. It is a good candidate for QTL Analysis as verbal communication skills can only measure for high functioning ASD such as Asperger’s but NVC can be measured for all the subtypes of ASD which is potentially more informative. 

Research Objective: To find out the significant responsible region in the chromosome of ASD people using QTL Analysis through which Genotype and Trait data of a complete sample can be analyzed.

Method: The method consists of five steps. 

• Study Sample: AGRE (Autism genetic resource exchange) ascertained nuclear families consists of at least two siblings with ASD. The ADI-R, a semi-structured interview given by trained testers to caregivers of autistic suspected children. ADI-R questions focus on assessing possible deficits in the three classical domains of autism: verbal/NVC, social development/play, and repetitive stereotyped interests/behaviors. The ADI-R algorithm is based on criteria in the ICD-10 and the Diagnostic and Statistical Manual of Mental Disorders, 4th Edition (DSM-IV) which is used to diagnose a subject with autism. Suspected ASD people are classified into two groups as Autism and Not quite Autism. Individuals are diagnosed with autism if their scores meet the criteria within the three behavioral domains and their onset of symptoms can be noticed before 36 months of age and less severely affected subjects were classified as not-quite-autism (NQA) based on AGRE criteria.

A total sample of 1132 ASD people with NVC data was classified for NVC QTL analysis by defining NVC scores. NVC scores were analyzed to measure the whole range of the NVC distribution and to test the association of NVC and verbal impairment. Some individuals with monozygotic twins and nonidiopathic autism such as fragile X syndrome and families not having two or more siblings with NVC scores and genotypes were excluded from the QTL and OSA analysis. Among the remaining 228 nuclear families, 203families had two affected siblings, 23 had three affected siblings, and two had four affected siblings, resulting in 284 sibling pairs or 483 individuals with ASD. 

• NVC Trait Distribution: NVC score is defined by the summation of responses (ranges from 0-3) of 7 items (29-33, 63 & 65) from ADI-R. Every response ranges from 0 to 3 where 0 means no observable deficit and higher value 3 correspond to greater impairment. An age depending scoring criterion was used according to the ADI-R questions where 2 types of responses were labeled as ‘CURRENT’ which described the subject’s impairment at the time of interview and ‘MOST ABNORMAL AT 4–5’ which described their most extreme impairment at any time point between ages 4 and 5. The response to ‘CURRENT’ was used for the children under 4 years old and ‘MOST ABNORMAL AT 4–5’ was used for those beyond that age range. ADI-R items are used to define NVC score because it describes a broader set of skills and can be applied on people of different age groups. 228 QTL families having the ASD sibpairs were categorized according to their levels of verbal impairment to find out the relation between NVC and their verbal ability as ADI-R question ‘overall level of language’ (Item 19). The response of this item ranges from 0 to 2 where 0 denotes 3 or more words containing comprehensible phrases, 1 denotes the absence of three-word phrases but the presence of speech on a daily basis with at least five distinct words, and 2 denotes the presence of fewer than five words and/or speech not used on a daily basis.

• Genotyping: A genome scan of 408 microsatellite markers was conducted in the parents and children of 228 families using the PREST software. When the Mendelian error was detected in a family for a marker, the families were being excluded from the analysis of that specific marker. 

• Trait Distribution and Correlation:

     Softwares that are used to find out the NVC score: 

• • R and SAS Software Package – used for statistical analysis. 
  • The Anderson–Darling statistic – used to measure the normality of trait distribution.
  • The Kendall-Tau statistic – used to estimate correlations in sibling trait and ADI-R domains. 

• The nonparametric Jonckheere–Terpstra statistic – used to measure the relationship between NVC score and the severity of language deficit. 

• Genetic Analysis: NVC QTL analysis was conducted by using the non-parametric linkage static of the Genehunter 2.1 software. OSA package Software was used to tests whether the siblings with the highest NVC score share excess allele-sharing. QTL having lod score greater than 3.0 is likely to harbor a gene(s) contributing to ASD through deficits in NVC. 

Result: 

In a result, there are four tables.

• Table 2 mainly represents the NVC distribution between the classification of verbal ability and the total sample. Here the mean and median of NVC increase with the impairment level.
• Table 3 shows that for those who have a higher NVC score, their impairment level was also high. The NVC score = 0.38 was highest, so the impairment of that sibling pairs was 2 which was highest. Even when that was taken with the individual sample, the result was the same. NVC score is significantly correlated with social, repetitive, and restrictive behavior and the other domains of ADI-R.
• In table 4, five chromosomal regions were taken by QTL analysis from 284 sibling pairs whose z score was more than 2.33. OSA conduction was occurred upon the ASD families with those 5 regions to understand whether there is any linkage between NVC score and allele sharing. 
• The regions whose low score was more than 3, were included in table 5 as a significant lod score. It is the region where the lod score will increase if the allele sharing increases.

By doing allele-sharing analysis on 175 families, it is seen that in the region 8q23-24 lodscore increases from 1.6 to 3.4. In the same pattern in 70 families, in the region 16p12-13 lodscore increases from 0.3 to 3.8. The other 3 regions also contribute in NVC variation but do not affect in a particular direction.

Discussion:

From the above analysis, it is understood that, in genetically complex disorders like ASD, QTL identification is a useful strategy for predisposing gene mapping. 

The NVC QTL which is identified by this study were previously highlighted by other studies of ASD linkage through the Lander and Kruglyak criteria. According to their criteria, if (p<0.05) it’s possible’ and if (p<0.0007) then its ‘suggestive’ level.

Most significant QTL 1p13-q12 which was ‘suggestive’ and ‘possible’ for sample 17 and 139 sibships in the two previous independent studies.

In the region, 4q21-25 and 8q23-24, ascertained ‘possible’ level and both linked to ASD at the ‘suggestive’ level of 381 pairs.

At 16p12-13, a QTL found in ‘possible’ level. It resulted in linkage in ‘suggestive’ level in 83 ASD sib pairs and in ‘possible’ level linkage in 151 sibships.

These findings of ASD linkage support using the QTL approach and shows the importance of the chromosome region in ASD. At QTL 7q35 the ‘age at first word’ and NVC was found in the same location. It is found that by ordering the families from least to the greatest time of ‘age at the first word’, the allele sharing increased. Current NVC QTL study shares, a study of ‘age at the first word’ and NVC correlated at 0.21 in 117 sibling pairs.

Conclusion:

NVC deficits is an important subdomain and familial feature of ASD that is likely to be genetic and result from variation at 8q23–24 and 16p12–13. The modest linkage analysis should be considered with caution as replication of these results in independent ASD samples will be critical for establishing the validity of the QTL and OSA approach to understanding the etiology of genetically complex disorders and for identifying the genes contributing to the risk for developing ASD through deficits in NVC.