Use of array CGH in the evaluation of dysmorphology, malformations, developmental delay, and idiopathic mental retardation
Introduction
The World Health Organization defines mental retardation as a condition of arrested or incomplete development of the mind, especially characterized by impairment of skills that are manifested during the developmental period and which contribute to the overall level of intelligence (i.e. cognitive, language, motor, and social abilities; http://www.who.int/mental_health/media/en/69.pdf). The term developmental delay is often used before age five, and the delay can involve motor function, cognitive ability, language or combinations thereof. Determining whether the disabilities are associated with malformations or multiple congenital anomalies and/or dysmorphic features can be helpful because it sometimes can suggest a syndromic clinical diagnosis to a skilled clinician and will guide the selection of diagnostic testing. Mental retardation is estimated to affect 2–3% of the population [1, 2].
Genetic abnormalities are the most common identifiable cause of developmental delay–mental retardation. However, there is considerable discrepancy in estimates as to the percentage of cases of developmental delay–mental retardation in which an etiologic diagnosis can be established (as reviewed by Moog [3]). The report of a National Institutes of Health (NIH) Consensus Conference held in 1995 concluded “that a diagnosis or cause of the mental retardation can be identified in 40–60% of all patients undergoing evaluation” [4], although one study reported a specific genetic or syndrome diagnosis in only 19.9% of cases [5].
Various new discoveries and methods such as sequencing of X-linked genes causing mental retardation and array comparative genomic hybridization (CGH) currently allow for the identification of a very specific molecular cause for developmental delay–mental retardation, at least 10% more frequently than was the case at the time of the NIH Consensus Conference 12 years ago. Practice guidelines for the cytogenetic evaluation of patients with developmental delay–mental retardation available from 2005 recommend high-resolution chromosome banding, individual fluorescent in situ hybridization (FISH) tests, and subtelomeric analysis in various circumstances [6], but these will clearly need to be revised with the advent of array CGH.
In this review, we focus on recent advances in research and diagnostic application of array CGH in the evaluation of patients with idiopathic mental retardation.
Section snippets
Array CGH
Cytogenetic diagnostic tools have improved over the past 30 years, with higher resolution chromosomal banding and FISH being landmark advances. Over the past 7–8 years, diagnosis of developmental delay–mental retardation has been improved by the use of multi-subtelomeric FISH and specific amplification methods such as multiplex ligation-dependent probe amplification (MLPA) to search primarily for subtelomeric imbalances. These tests and the earlier development of metaphase chromosome CGH served
Detection rates in studies of mental retardation
The frequency with which chromosome abnormalities and/or genomic rearrangements are detected in patients with developmental delay–mental retardation is higher with the presence of malformations or dysmorphic features and with more severe retardation. Based on a literature review, van Karnebeek et al. [10] calculated the mean yield of chromosome aberration as 9.5%. Shevell et al. [11] estimated the detection rate of visible chromosome abnormalities in patients with mental retardation as 3.7%.
Genome architecture and copy-number variation
The use of array CGH to analyze the genomes of normal humans led to the discovery of extensive genomic rearrangements ranging in size from kilobases to megabases and not recognizable by high-resolution chromosomal banding. These changes have been called copy-number variations (CNVs) [26•, 27•]. Recently, a tiling-path BAC array was used, together with an oligonucleotide array (Affymetrix 500K) that genotypes 500 000 single nucleotide polymorphisms (SNP) to compare the genomes of 270 individuals
Identification of new mental retardation genes and novel deletion and duplication syndromes
Based on the knowledge of higher-order human genome architecture, Sharp et al. [20••, 32•] used bioinformatic analysis of segmental duplications in the genome to design a microarray with 2007 BAC clones specific for 130 genomic regions that were found to be flanked by directly oriented low-copy repeats (LCRs). Using this microarray, the authors investigated 290 individuals with mental retardation. They identified 16 pathogenic rearrangements including five novel imbalances on chromosomes
Use of array CGH in evaluation of developmental delay–mental retardation
We identified thirteen studies performing array CGH in 40 or more patients with idiopathic developmental delay–mental retardation (Table 1). These include one report using Affymetrix 100K oligonucleotide arrays, and twelve studies using BAC arrays. Of the studies using BAC arrays, five used BACs at approximately 1 Mb intervals across the genome (∼3000 BACs), one used an array targeted for rearrangement-prone genomic regions (∼2000 BACs), two used a tiling array with 32 447 BACs, and three used
Advantages and disadvantages of array CGH over alternative methods
Array CGH detects virtually all unbalanced chromosome abnormalities that can be detected by banded karyoptye analysis. In practice, array CGH often detects abnormalities that could or should have been detected by karyotype but which were reported as being normal by prior chromosome analysis. In addition, array CGH is far superior to alternative methods for detecting genomic abnormalities that are clearly below the resolution of karyotype detection. Typically, multiple probes per subtelomeric
BAC versus oligonucleotide arrays
Oligonucleotide arrays have important advantages compared with BAC arrays, which are cumbersome, expensive, and time-consuming for printing. Many of the genomic gains or losses can be smaller than the BACs or overlap the BAC only partially, in which cases the abnormality can be missed or give only weak evidence of a gain or loss of copy number. In the case of a weak but real indication of gain or loss on the array, the abnormality may not be detected by FISH using the same BAC as a probe,
Tiling versus targeted arrays in the clinic
The question of whether targeted or tiling arrays should be used in routine clinical diagnosis has generated vigorous and healthy debate [56, 57]. Targeted arrays typically cover the chromosomal telomeres and known regions of deletion and/or duplication syndromes (Figure 4). Clinical versions often have excluded regions of CNV by trial and elimination. The advantages of these arrays are that the great majority of well known genomic disorders can be diagnosed with minimal interference from CNVs
Prenatal use of array CGH
Array CGH can be applied to the analysis of prenatal samples to detect conditions associated with developmental delay–mental retardation [43, 59••, 60, 61]. There are suggestions that prenatal use of array CGH will become widespread [62], and there have been many strong statements in opposition to prenatal use of array CGH [63, 64]. We view the availability of array CGH as a major advance in prenatal diagnosis. The ability to detect disorders as being severe or more severe than Down syndrome is
Array CGH and autism
A very wide range of chromosomal abnormalities have been reported in autism, as reviewed by Vorstman et al. [67•]. The diversity of involved regions makes it clear that ordering single locus FISH tests is impractical. The most common abnormalities in the Vorstman et al. review were deletions of 2qter, deletions of 22qter, and duplications of the PWS/AS region at 15q11-q13. Also, two separate reports using array CGH found these three abnormalities to be the most common [68•, 69••]. The 15q11-q13
Conclusions
In summary, multiple recent studies have shown that array CGH is a powerful and efficient method for diagnosis and research of mental retardation. Array CGH has proven to have a 10–15% overall rate of detection of genomic abnormalities — mainly interstitial deletions and duplications — in patients with mental retardation. Commercially available genome-wide high-resolution microarrays with more than 300 000 oligonucleotide probes will probably replace the BAC arrays in clinical laboratories.
Update
Recent work has demonstrated that de novo deletions and duplications are a more frequent cause of autism than previously appreciated [69••].
Disclosure statement
The Department of Molecular and Human Genetics at Baylor College of Medicine (BCM) offers extensive genetic laboratory testing and derives revenue from this activity. Array CGH is offered in collaboration with Athena Diagnostics. In addition, BCM, but not the authors personally, owns stock in Spectral Genomics, and ALB served on the Scientific Advisory Board of Spectral Genomics in the past. Spectral Genomics sells microarrays for CGH to LabCorp, Mayo, Arup, academic laboratories, and others.
References and recommended reading
Papers of particular interest, published within the period of review, have been highlighted as:
• of special interest
•• of outstanding interest
Acknowledgements
We appreciate the critical reviews of JR Lupski. We thank SW Cheung for providing Figure 1, Figure 2. This work has been generously supported by the Baylor College of Medicine Mental Retardation Research Center (HD 2406407). We apologize to colleagues and the authors of relevant papers that could not be cited owing to space limitations.
References (69)
American College of Medical Genetics guideline on the cytogenetic evaluation of the individual with developmental delay or mental retardation
Genet Med
(2005)- et al.
The use of telomere probes to investigate submicroscopic rearrangements associated with mental retardation
Curr Opin Genet Dev
(2003) - et al.
Diagnostic genome profiling in mental retardation
Am J Hum Genet
(2005) - et al.
Large-scale copy number polymorphism in the human genome
Science
(2004) - et al.
Recurrent infections, hypotonia, and mental retardation caused by duplication of MECP2 and adjacent region in Xq28
Pediatrics
(2006) - et al.
BAC to the future! or oligonucleotides: a perspective for micro array comparative genomic hybridization (array CGH)
Nucleic Acids Res
(2006) - et al.
Application of array-based comparative genomic hybridization to clinical diagnostics
J Mol Diagn
(2006) - et al.
Diagnostic genome profiling: unbiased whole genome or targeted analysis?
J Mol Diagn
(2006) - et al.
A tiling resolution DNA microarray with complete coverage of the human genome
Nat Genet
(2004) Fetal genetic testing: Screen test
Nature
(2005)