5th Congress Autism-Europe
Articulos / Proceeding
Autism-Spain

Association studies in Autism

J.P. MUH, J. Hérault, E. Petit, C. Barthélémy

INSERM U 316 - CHRU Bretonneau 37044 TOURS Cedex (France)

Family studies and epidemiological data suggest that genetic susceptibility factors are involved in infantile autism. These etiological factors could be considered especially with regard to behavioral symptomatology originally described by Kanner and by other specialists in autistic syndrome associated with neurological disorders and certain genetic diseases.

Indeed, infantile autism is characterized by an incapacity for verbal and non verbal communication and by a deficit in emotional relationships.

The pattern observed may result from different cerebral structures, mainly those which are involved in the control of behavior and those related to regulation of perception and association.

Thus, we can divide the pathophysiology of autistic syndrome into pre, peri and postnatal environmental factors and genetic causes.

In view of the overlaps in clinical patterns and the two possible origins I have mentioned it seems that the genetic forms of autism result from mutations in genes controlling brain development and that environmental and genetic causes are able to damage the same cerebral function areas involved in the control of behavior and in particular the regulation of perception and association.

As reported by Folstein and Rutter:

"Some evidence suggests that quite often it is not autism itself that is inherited but rather some genetic abnormality of language or sociability that interacts with other factors to produce autism"

Genetic causes

Epidemiologic studies of relatives and siblings since 1991 have reported that the overall recurrence risk is between 3.3 % and 14.5 % (mean around 8 %). A family aggregation study by Jorde et al determined that the familiar aggregation and likelihood of genetic basis was 22 times higher in the autistic than in the control group.

Nevertheless, the greatest care must be taken in the interpretation of results of genetic investigations.

The genetic involvement in the etiology of autism could be divided into two parts :

- first: the arguments for genetic factors in idiopathic Kanner syndrome,

- second : the report that characterized genetic disorders could be associated with autism.

In this talk, I will focus on some of the genetic disorders and molecular markers which have been reported to he linked with autism.

Slide 1:

There are two methodological approaches to identify genetic influences on autism. The first, clinical, begins with the study of the behaviour of a child using DSM III R and DSM IV criteria, and continues through the study of brain functions by means of electrophysiological and imaging techniques to, possibly, genetic investigations (using epidemiological and familiar studies). The other, a theoretical approach, begins with the gene, through the brain and then to the behaviour. To my knowledge the latter is an approach not yet fully in use in the exploration of infantile autism.

Slide 2 :

Autistic syndrome was described in 1943 by Kanner and Asperger. The genetic origin of autism was suggested by Kanner in 1954. The arguments for this origin were the prevalence of 4 - 5 for 10,000 births. There is a higher rate if we consider autism with associated abnormalities. The sex ratio is 3 boys for one girl. Among siblings, the prevalence of autism is fifty times greater than in the general population. Moreover 15% of sisters, brothers and other relations of autistic patients present language problems. All these epidemiological studies support the arguments for a genetic origin of autism.

Slide 3:

The reports of twin studies are significant. Individual cases of monozygotic and dizygotic twins were published before 1977. The incidence of concordance was 83% for monozygotic and 40% for dizygotic twins. The reports of Folstein and Rutter, Ritvo and Steffenburg since 1977 have shown a concordance of 82% - 95.6% for monozygotic and a concordance of 10% - 23.5% for dizygotic twins. The last data from Bolton and Rutter (1990) confirmed this proportion for MZ and DZ.

Slide 4:

If we consider the cases with associated abnormalities, we can quote phakomatosis (tuberous sclerosis and neurofibromatosis) and other chromosomal syndromes. However, I want to draw your attention to fragile-X syndrome.

Slide 5:

Opinions differ over the interpretation of the association of fragile-X syndrome and autism. Different authors suggest that fragile-X syndrome does not exist in autistic boys. Others assert that there is an association between fragile-X syndrome and autistic syndrome in both boys and girls.

If we consider the proportion of fragile-X among autistic males, the mean incidence is about 8%. The last review by Fisch of 21 recent studies revealed a proportion of 6.4 fragile-X to 1,006 autistic males. The most recent methods of molecular diagnosis described Mandel's team enable us to obtain more accurate epidemiology data.

Slide 6:

Investigation with X chromosome DNA probes was expected to reveal different lesions to fra-X but Crowe's data obtained with probes for the Xq 26-28 region and for the short arm were inconclusive, although the team recognized that deletions may have been present but undetected by their probes.


Slide 7:

We report here the study of different biallelic markers of chromosome X. We found a significant Chi Square for the locus DXS 287 in Xq22-23.

Slide 8:

MRX genes for non specific MR.

Recent extensive research has supported the classification of X linked mental retardation genes. From 1988 to 1992, 13 were genes described. The main location of MRX genes is on the pericentromeric region of the short arm of the X chromosome (X p 11.21 to X p 22.12). Recently in our laboratory new localizations were described in families now identified as MRX 14 and 15 and 16.


Slide 9:

We studied a population of 145 autistic children (aged 2-12 years), diagnosed according to revised DSM III. Using more precise methods to assess the association of the CNS development process with behaviour disorders, we measured urine and blood dopamine, methoxyamines, serotonin and hydroxyvanillic acid, because such neurochemical abnormalities could in part explain the behavioral syndrome.

Slide 10:

We observed differences in the urine levels of dopamine, hydroxyvanillic acid and serotonin in about half of autistic subjects. The main difference is that the levels for the youngest are higher. During maturation, from the second to eighth year, the pattern of production of these components remains higher for autistic children than for controls.


Slide 11:

This is the reason why we performed association studies with markers involved with the metabolism of neurotransmitters. The results were not significant. Similarly the use of markers of receptor genes were inconclusive. However, we performed a controlled association study for the gene of DBH (chromosome localisation 9q34), the gene of TH (11p 15.5), the closely located HRAS 1 gene and the gene of TPH (11p15).

Slide 12:

The population studied comprised 70 autistic patients and 70 normal matched controls. On this slide we have the results of the association study with markers for candidate genes in the 11p15.5 region (close to the locus of TH gene) which was significant for the allele distribution of B2 and B3 of the Hras gene.


Slide 13:

Among other causes of autistic syndrome there are prenatal causes. Firstly, there is the expression and activity of genes involved in cerebral development. One group regulates metabolic functions and differentiation. A second group consists of genes such as homeotic genes which are specifically responsible for neurogenesis and growth of the brain.

Our interest in the study of homeotic genes is the result of the data published by various authors in neurophysiology and imaging.

To explain behavioral pathology several authors have suggested disturbances in the connections of the brainstem cerebellum and the limbic system. The cerebellum is known to control motility and regulation of perception.

Histological and imaging studies of the cerebellum have shown abnormalities such as hypoplasia of the cerebellar vermis. This supports the possibility of mutations and/or abnormal functions of genes involved in the development of the CNS.

Slide 14:

On this slide, you could consider the report of different cerebellar abnormalities. Imaging studies using magnetic resonance and computerized tomography have shown reduced size brain parenchyma particularly in the vermis. This probably corresponds to the autopsy studies which have demonstrated abnormal size and/or loss of cellular density in these areas.

We know that homeobox genes govern CNS segmentation and neural cell differentiation, acting as DNA-binding proteins with the control of regulatory expression and/or function. In humans the main localization of expression of EN1 an EN2 is the region of the pons cerebelli and cerebellum.

Slide 15:

The human homologs of Drosophila engrailed gene and of mouse homeogenes En1 and En2 have been characterized and localized by LOGAN and JOYNER. The EN1 gene was mapped to chromosome 2.2p13-p21. In our study the probe used was MIPG (including homeobox).

The EN2 gene was localized to chromosome 7.7q32-qter. The probes used were MP4 (upstream from homeobox) and MPS (including homeobox). The probes were kindly provided by Ms LOGAN.

Slide 16:

With the MP4 probe showing a Pvull polymorphism, significant differences in the allele frequencies between the two populations were found. With the MP5 probe showing a SstI polymorphism no difference appeared. These results are preliminary and we shall continue this study to demonstrate its potential involvement in the pathophysiology of autism.

Conclusion:

I want to emphasize three points:

FIRST:

We have frequently observed that there is an association between MR and autism (about 80 %). Familial X-linked mental retardation (MR) has been reported by several teams.

Statistical studies demonstrated that a predominance of MR males exist: the same findings could appear in autism. This fact could explain the controversy about the relationship between autistic and fra-X syndromes.

The male predominance in MR and autistic children could be explained

by the advent of lethal conditions in the female foetus

or by transmission of undefined X-linked traits through generations.

Is autism a fra-X syndrome or a X-linked non fra-X disorder or an autosomal disease ?

The answer may be found by molecular genetics.

SECOND :

With regard to our results, it could be suggested that:

These results raise the possibility that there is a relationship between a chromosome 11 p 15-5 locus and the gene(s) for autistic syndrome and for other mental disorders. Van Tol et al. (1992) suggest that polymorphic variations in DRD4 gene in the human population underlie individual susceptibility and could be involved in the appearance of cognitive disorders during brain development.

In the future, genetic linkage studies will permit us to establish a broader definition of the autistic phenotype including cognitive and social deficits with autism.

THIRD:

Recent data in our association study (Petit et al. 1995) with markers of engrailed homeogenes have demonstrated that allele 2 of the MP4 probes is significantly associated with the autistic population.

The result of this association study may be enhanced by cerebral imaging data.

The combination of these approaches reinforces the hypothesis that infantile autism may be a multigenic disease.