High-resolution microarray analysis has facilitated the identification of multiple new microdeletion/microduplication syndromes in individuals with idiopathic mental retardation and congenital anomalies [10–13]. Because it is not reliant on clinical suspicion of a known syndrome, microarray analysis can identify the underlying genomic lesion in affected individuals for whom a diagnosis has not been possible – because the rarity of the causative chromosomal abnormality may prevent accurate reporting of index cases in the literature or because the variability or mildness of the phenotype escapes characterization.
Among individuals with microdeletions of 3q29, the phenotype varies widely, with mild to moderate mental retardation/developmental delay, microcephaly, and mild dysmorphic features (including high nasal bridge and short philtrum) the only features common to the majority. The lack of an easily discernible clinical phenotype complicates diagnosis and may explain why so few individuals with this microdeletion have been identified in the literature. However, the presence of autistic features in several individuals [1, 2] suggests that a deletion of 3q29 may be considered in individuals with autism. Like that of the 3q29 microdeletion, the phenotype of the reciprocal duplication varied, with mental retardation the only common feature among the three individuals for whom clinical information was available. The careful clinical examination of additional individuals with these chromosome abnormalities are needed to further refine the phenotypes.
Although microduplications have been predicted to occur with equal frequency as microdeletions that are flanked by LCRs, fewer have been identified. Those that have been found tend to result in milder phenotypes than microdeletions, which may explain in part why microduplication syndromes appear to be less common than their counterparts. Of the 19 microduplications of 3q29 identified by our laboratory, only five were the reciprocal duplication of the common-sized deletion of the same region; the remaining 14 microduplications flanked, overlapped or were smaller than the common deletion region. These results suggest that other mechanisms in addition to NAHR mediate rearrangements of 3q29. A similar situation occurs in rare rearrangements of 17p11.2 associated with Smith-Magenis syndrome, in which the breakpoints do not fall within the paired segmental duplications that flank most deletions of this region or appear to be mediated by known genomic architectural features .
Only two of 10 individuals with microduplications of 3q29 for whom parental DNA was available for testing had de novo abnormalities. Typically, abnormalities inherited from phenotypically normal parents are considered benign copy-number variants. However, subclinical phenotypes not appreciated in a parent may be manifested more severely in an affected child; the most common example is the deletion of 22q11.2 . Therefore, the clinical significance of the reciprocal 3q29 microduplication is still unclear. The three-generation family with five individuals with reciprocal 3q29 microduplications recently described by Lisi and colleagues  would seem to confirm this duplication is a clinically relevant abnormality. However, the presence of a ~169 kb deletion of 4q13.1 in all affected individuals, and absent in unaffected family members, suggests that the 3q29 microduplication may not be causative of the affected individuals' phenotypes. Nonetheless, given the rate at which de novo genomic rearrangements occur in the human population, a combination of CNVs at the same or different loci, inherited from parents in whom the single variation was insufficient to cause disease, might be necessary to produce a phenotype. A similar hypothesis has been proposed for thrombocytopenia-absent radius (TAR) syndrome; the authors of a recent study in which 30 individuals with TAR syndrome had a microdeletion of 1q21.1, 75% of which were inherited from a normal parent, suggest the presence of a genetic modifier in addition to the 1q21.1 microdeletion was necessary to cause the disease phenotype . High-resolution whole-genome screening of a large cohort of affected individuals is required to determine whether another abnormality is present elsewhere in the genome.
High-resolution breakpoint mapping using BAC and oligo-based microarrays identified common and unique breakpoints in both the 3q29 microdeletion and microduplication cohorts suggesting that other mechanisms in addition to nonallelic homologous recombination may play a role in mediating these rearrangements. This study demonstrates that array CGH is especially suited to identify chromosome abnormalities in individuals with unclear or variable presentations.