Long tandem repeats as a form of genomic copy number variation: structure and length polymorphism of a chromosome 5p repeat in control and schizophrenia populations

Abstract

Objectives: Genomic copy number variations (CNVs) are a major form of variation in the human genome and play an etiologic role in several neuropsychiatric diseases. Tandem repeats, particularly with long (>50 bp) repeat units, are a relatively common yet underexplored type of CNV that may significantly contribute to human genomic variation and disease risk. We therefore carried out a pilot experiment to explore the potential role of long tandem repeats as risk factors in psychiatric disorders.

Methods: A bacterial artificial chromosome-based array comparative genomic hybridization (aCGH) platform was used to examine CNVs in genomic DNA from 34 probands with schizophrenia or schizoaffective disorder.

Results: The aCGH screen detected an apparent deletion on 5p15.1 in two probands, caused by the presence in each proband of two low copy number (short) alleles of a tandem repeat that ranges in length from fewer than 10 to greater than 50 3.4 kb units in the population examined. Short alleles partially segregate with schizophrenia in a small number of families, though linkage was not significant. An association study showed no significant difference in repeat length between 406 schizophrenia cases and 392 controls.

Conclusion: Although we did not demonstrate a relationship between the 5p15.1 repeat and schizophrenia, our results illustrate that long tandem repeats represent an intriguing type of genetic variation that have not been studied in earlier connection with psychiatric illness. aCGH can detect a small subset of these repeats, but systematic investigation will require the development of specific arrays and improved analytical methods.

Figure 1

BAC array results demonstrating diminished log2 signal ratio on chromosome 5p15.1 in DNA from study participant S56. The Y axis depicts the log2 ratio of hybridization signal compared to pooled controls. The X axis depicts location in megabases along chromosome 5. The arrow identifies the loci defined by BACs RP11-88L18 and RP11–91D21.

Figure 2

Genomic structure of 5p15.1. A. Representation of 200 kb of chromosome 5p15.1 from the human reference sequence (hg 18 build 36) illustrating the relative location and orientation of 26 repeat units. The units colored white have > 99% identity; the units colored grey have identity ranging from 88% to 93%. Shown are BACs detecting the copy number change, the qPCR sites used to define the copy number change (telomeric qPCR begins at 17,521,414 bp; centromeric qPCR ends at 17, 718,335 bp and is located in a spliced EST-BM682321) and the HindIII restriction sites (17,571,914 and 17,648,982 respectively) used in pulse field experiments. Small arrows in the repeat units depict their orientation. B. Conservation pattern across species is shown for the central 19 tandem units (http://genome.ucsc.edu). C. Representation of a single 3434 bp repeat. The figure depicts the relative position of the qPCR product used to measure copy number, the PCR product used to detect transcription, Genscan gene prediction, human ESTs, and the conservation across the repeat unit (http://genome.ucsc.edu). D. Transcript expression of the conserved region of 5p15.1 tandem repeat units. Shown is RT-PCR performed with cDNA obtained from frontal cortex of 2 individuals. The same result was obtained using cDNA from whole brain of 3 individuals and hippocampus of 1 individual. Ladder is in lane 1 with units in base pairs. A band of the predicted 249 base pairs located in the central conserved region of the repeat unit is shown in lanes 2 and 4. The region was amplified in the RT+ samples but not in the corresponding RT- samples or the water blank.

Figure 3

The 5p15.1 tandem repeat length is variable. PFGE of HindIII restriction fragments containing the repeat from probands and controls, detected with a repeat specific probe. S prefix = study probands, GM prefix = controls. Ladder shown in kilobases.

Figure 4

Stability of repeat length inheritance and partial segregation of repeat length with disease in four small pedigrees. Bands with fewer than 22 repeat units are indicated by arrows for each pedigree. Size markers are shown in kilobases. Individual S276 carries a short allele but was not affected at last contact.

Figure 5

qPCR is a valid measure of combined allele length. Approximate repeat length as determined by PFGE for each allele is strongly correlated with normalized qPCR results for the combined alleles. r² = 0.82. The increasing scatter at longer repeat lengths reflects limitations of PFGE resolution.

Figure 6

Distribution of 5p15.1 repeat length in 406 patients and 392 controls. Relative repeat length, as indicated by qPCR, binned as shown on the X-axis. Y axis indicates overall frequency of lengths in each bin.