**10. Reduced penetrance and pleiotropic effects**

Incomplete penetrance and variable expressivity are discussed greatly in the literature in an attempt to explain the complexity of the biological phenomena involved in ASD (Zhao et al., 2007; State & Levtt, 2011; Coleman & Colbert, 2012).

While some genomic regions seem to play very important roles in predisposition for autism and others not so much, there are some for which there is much divergence in published da‐ ta. For example, the variations that involve 15q11-13, 16p11.2 and 22q11.2 are among the most commonly found CNVs in patients with autism, however the resulting behavioral phe‐ notype varies from normal to strongly affected (Cook & Scherer, 2008; Adzhubei et al., 2011; Devlin & Scherer, 2012). The explanation for the recurrence of rearrangements in these re‐ gions is that they are flanked by low repetition segments containing 99% of the identity, which due to their genomic structure predisposes them to non-allelic homologous recombi‐ nation and a high mutation rate (Cottrell et al., 2011). However, another four regions in which large deletions are observed (2q22.1, 3p26.3, 4q12 and 14q23) are not flanked by un‐ stable elements and also have a strong potential for involvement in the etiology of autism (Nicholas et al., 2007; Cottrell et al., 2011; Griswold et al, 2011).

With the recent clinical and research use of high resolution array-based strategies for charac‐ terizing CNVs in children with developmental disabilities, there has been a significant in‐ crease in the diagnosis of interstitial 15q duplications, including small, atypical duplications. The association between ASD and 15q duplications has been described in many case reports since the 1980s and sparked interest in 15q11-13 as a candidate region in autism linkage studies. Prevalence estimates of proximal 15q duplications in cohorts of patients with autism range from 0.5% to 3%, depending on ascertainment criteria and methods used to identify the duplications (Simon et al., 2010). This region contains receptors of an inhibitory neuro‐ transmitter, gamma-aminobuytric acid (GABA) that acts on a complex channel receptor mainly permeable to chloride anions to reduce neuronal excitability. GABA signaling is also established before glutamatergic transmission suggesting that GABA is the principal excita‐ tory transmitter during early development (Takayama & Inoue, 2004; Ben-Ari et al., 2007).

X-linked *DDX53-PTCHD1* locus, of an enrichment of CNVs disrupting functional gene sets involved in cellular proliferation, projection and motility, and GTPase/Ras signaling, (Pinto et al., 2010). Another pathway with many of these variations is that related to the develop‐ ment of synapses, and maintenance and motility of neurons, the NLGN-NRXN-SHANK

But, CNVs have been regarded as "rare variants" within the panorama of the etiology of ASD and it is important to emphasize the care that should be taken in associating them in the etiology, especially those not previously described. Accuracy is required in the study of these changes during the formation of control groups. On the other hand, when it comes to autism, on finding the same CNVs in carriers and controls is not a reason to dismiss their participation in the phenotype, given the complexity of factors that may be acting jointly

Incomplete penetrance and variable expressivity are discussed greatly in the literature in an attempt to explain the complexity of the biological phenomena involved in ASD (Zhao et al.,

While some genomic regions seem to play very important roles in predisposition for autism and others not so much, there are some for which there is much divergence in published da‐ ta. For example, the variations that involve 15q11-13, 16p11.2 and 22q11.2 are among the most commonly found CNVs in patients with autism, however the resulting behavioral phe‐ notype varies from normal to strongly affected (Cook & Scherer, 2008; Adzhubei et al., 2011; Devlin & Scherer, 2012). The explanation for the recurrence of rearrangements in these re‐ gions is that they are flanked by low repetition segments containing 99% of the identity, which due to their genomic structure predisposes them to non-allelic homologous recombi‐ nation and a high mutation rate (Cottrell et al., 2011). However, another four regions in which large deletions are observed (2q22.1, 3p26.3, 4q12 and 14q23) are not flanked by un‐ stable elements and also have a strong potential for involvement in the etiology of autism

With the recent clinical and research use of high resolution array-based strategies for charac‐ terizing CNVs in children with developmental disabilities, there has been a significant in‐ crease in the diagnosis of interstitial 15q duplications, including small, atypical duplications. The association between ASD and 15q duplications has been described in many case reports since the 1980s and sparked interest in 15q11-13 as a candidate region in autism linkage studies. Prevalence estimates of proximal 15q duplications in cohorts of patients with autism range from 0.5% to 3%, depending on ascertainment criteria and methods used to identify the duplications (Simon et al., 2010). This region contains receptors of an inhibitory neuro‐ transmitter, gamma-aminobuytric acid (GABA) that acts on a complex channel receptor mainly permeable to chloride anions to reduce neuronal excitability. GABA signaling is also

pathway (Bourgeron, 2009).

and in combination in different cases.

226 Recent Advances in Autism Spectrum Disorders - Volume I

**10. Reduced penetrance and pleiotropic effects**

(Nicholas et al., 2007; Cottrell et al., 2011; Griswold et al, 2011).

2007; State & Levtt, 2011; Coleman & Colbert, 2012).

In multiplex families, i.e. those with more than one affected individual, different individuals have similar microdeletions or microduplications and exhibit autistic, broad or normal phe‐ notypes (Fernandez et al., 2010; Kumar et al., 2008). The microduplication of 15q11-13, which seems to have a major impact on the expression of ASD, is a good example of differ‐ ent effects in carriers. For example, Veltman et al. (2005) reported a female proband referred for evaluation of a possible ASD. Genetic analyses indicated that the proband, her father and one of her sisters, carried a paternally derived interstitial duplication involving 15q11-13. The proband showed evidence of ASD (PDD-NOS), borderline mental retardation, mild hypotonia and joint laxity. Her father and her sister were of normal intelligence and neither was thought to have an ASD, although speech/language difficulties and some autis‐ tic type behaviors were reported to have been present early in the development of the sister.

In studies of idiopathic cases, the most common change is the 16p11.2 microduplication/ microdeletion, comprising a region of around 600 Kb. Microduplication of 16p11.2 is present in autistic and schizophrenic individuals but is also seen in normal individuals. This varia‐ tion is also observed in cases of ASD with and without dysmorphic signs (Fernandez et al., 2010; Shinawi et al., 2010). Microdeletions of 16p11.2 seem to be more penetrating (around 100%) and are associated with the presence of major dysmorphic signs, while microduplica‐ tions have reduced penetrance (~50%) and are associated with minor dysmorphic signs (Ra‐ mocki & Zoghbi, 2008; Bourgeron, 2009; Fernandez et al., 2010).

One possible explanation for these cases, as already mentioned in previous sections, is that probably there are other genetic components that are contributing to the disease in these families that, depending on how they are associated in individuals, result in different phe‐ notypes (Griswold et al., 2012). But it is also very likely that different CNVs display different penetrance depending on the sensitivity of the affected gene to dosage (number of copies), the function of the gene and the affected region.

Endophenotypes (or intermediate phenotypes) are defined as heritable traits that form a causal link between genes and observable symptoms. Brain based endophenotypes offer several important advantages over clinical phenotypes in the search for pleiotropic genes in ASD. They provide insight in the causal chains of action leading from gene to symptom ex‐ pression, they aid in forming etiologically more homogeneous subgroups of patients (Gould & Gottesman, 2006).

In addition to the variability of simple and more complex genetic mechanisms proposed in the expression of the ASD phenotype, environmental factors via epigenetic mechanisms seem to be very important.
