A Penalized Spline Based Method for Detecting the DNA Copy Number Alteration in an Array-CGH Experiment

Abstract Cytogenetics and comparative genomic hybridization (CGH) have been used to identify large genomic amplifications and deletions in all subtypes of acute myeloid leukemia (AML). Up to 15–20% of AML patients have a normal karyotype at diagnosis. While cytogenetic abnormalities confer important prognostic information for patients with AML, there remain differences in the therapeutic response and outcome among patients with the same cytogenetic profile, implying that other, more subtle, genetic abnormalities may exist. We hypothesized that a subset of AML patients with normal cytogenetics may contain genomic DNA copy number changes that are too small to be detected using standard cytogenetic techniques. To address this possibility, we used high-resolution bacterial artificial chromosome (BAC) array CGH technology to examine 31 AML patients with normal cytogenetics. The BAC arrays contain 2,464 BAC clones spotted in triplicate on glass slides, and provide a 1 Mb resolution of the entire human genome. Technical generation of the arrays, hybridization parameters, and analysis were similar to that reported for murine BAC array CGH (Nat Genet. 2001 Dec;29(4):459–64). The 31 AML samples included 4 M0, 8 M1, 10 M2, and 9 M4 patients. Array CGH experiments were performed using 500 nanograms of Cyanine 5 labeled genomic DNA from unmanipulated AML bone marrow, mixed with an equal amount of control DNA (a pool of DNA from 4 cancer-free individuals) labeled with Cyanine 3. Using the human 1 Mb BAC arrays, we identified amplifications and deletions from multiple samples that were confirmed with G-banding cytogenetics [del(7)(q31), del(7)(p11.2), +8, del(11)(q13q23), +21, add(21)(q22), −X, −Y, +Y]. In addition, BAC arrays robustly detected copy number alterations that were identified in as few as 4/21 metaphases. We identified 5/31 (16%) patients with normal cytogenetics that contained altered genomic DNA copy numbers using BAC array CGH. Copy number changes were confirmed for several of these genomic loci using a dye-swap experiment, where the AML DNA was labeled with Cyanine 3, and the control DNA with Cyanine 5. Two of the 5 patients with abnormalities detected using array CGH would be reclassified from “intermediate” to “unfavorable” cytogenetics [del(7)(q31.31q34), add(11)(q23.3qter), and 17(p12pter)]. These results suggest that a subset of AML patients with normal cytogenetics contain genomic copy number alterations that may effect treatment and outcome. Patient # FAB subtype Genomic location Gain or loss Size (Megabase) Dye-Swap confirmed 1 M0 7(q31.31q34) loss 2.0 Not done 1 11(q23.3qter) gain 16.5 Not done 2 M1 11(p14) loss 7.4 Yes 3 M1 11(q13.2q14.1) gain 15.8 Yes 3 19(p) gain 64.0 Yes 4 M2 17(p12pter) gain 8.6 Not Done 5 M2 19(p13.1pter) loss 14.8 Yes 5 12(q13) loss 5.0 Yes

Download Full-text

Novel DNA Copy Number Changes in Hematological Malignancies: A cDNA-Based CGH Microarray Screening of CML, AML and CLL Cases without Chromosomal Imbalances in G-Banding.

Blood ◽

10.1182/blood.v104.11.4418.4418 ◽

2004 ◽

Vol 104 (11) ◽

pp. 4418-4418

Author(s):

Tuija Lundan ◽

Anne Oikarainen ◽

Lena Hafren ◽

Maija Wolf ◽

Erkki Elonen ◽

...

Keyword(s):

Copy Number ◽

Myeloid Leukemia ◽

Array Cgh ◽

Reference Sample ◽

Copy Number Alterations ◽

Dna Copy Number ◽

Chromosomal Imbalances ◽

Single Allele ◽

Dna Copy Number Alterations ◽

Copy Number Changes

Abstract Copy number changes, such as small single allele losses and gains, have important roles in the mechanisms of cancer development. These alterations also have often prognostic significance. Genome wide screening of DNA copy number losses previously conducted by extensive LOH analyses can now be performed with array-based comparative genomic hybridization (array CGH). We assessed the utility of array CGH in the detection of single allele deletions and gains in a cohort of seven patients with chronic myeloid leukemia (CML), seven patients with chronic lymphatic leukemia (CLL) and three patients with acute myeloid leukemia (AML). All the CLL and AML patients had a normal karyotype as assessed by standard G-banding. In CML patients the only clonal abnormality detected by cytogenetics was the reciprocal Philadelphia translocation, t(9;22)(q34;q11). The derivative chromosome 9 [der(9)] deletion status of the CML patients was determined using fluorescence in situ hybridization (FISH) analysis. Four patients did not have the deletion, two had a der(9) deletion spanning both 5′ABL and 3′BCR regions and one patient had a deletion of the 5′ABL region alone. The array CGH experiments were performed using Agilent Technologies Human 1 cDNA microarray slides consisting of 13,000 clones. A total of 6 ug of fluorescently labeled DNA extracted from bone marrow samples was hybridized on cDNA array. Normal male or female DNA was used as the reference sample in the hybridization. The slides were scanned with the Agilent fluorescent scanner and intensity ratio data between the tumor and reference sample was processed using Feature Extraction software. The data was filtered and analyzed using SPSS (version 11) and Origin 7.0 softwares. The processed, untransformed red-to-green fluorescence signal ratio was used for evaluating gene dosage. Ratios greater than 1.1 were considered to indicate DNA copy number gains and ratios below 0.9 DNA copy number losses in tumor samples. In two CML patients who had deletions covering both the 5′ABL and 3′BCR regions in the translocation breakpoint of der(9), the deletion was detectable with the array CGH. In four patients with no deletion the red-to-green ratio profile for der(9) was 1. However, in one patient with an isolated 5′ABL deletion, the deletion was not visible in array CGH. No other obvious DNA copy number alterations were seen in CML patients. Array CGH detected deletions in three of the seven CLL patients. Deletions were found in 13q14, 2q32-33 and 14q24. One of the three AML patients studied showed an amplification in chromosome 9p. No alterations were seen in the other two AML patients. The FISH and array studies are being done on larger set of patient samples to confirm the results. We conclude that array CGH provides new information in patients without chromosomal imbalances in standard cytogenetics and enables the detection of novel small submicroscopic copy number alterations. Furthermore, a cDNA-based array platform can be used both for studies of DNA copy number alterations and gene expression analyses.

Download Full-text