A new whole genome amplification method for studying clonal evolution patterns in malignant colorectal polyps

Abstract

To identify the genetic drivers of colorectal tumorigenesis, we applied array comparative genomic hybridization (aCGH) to 13 formalin-fixed paraffin-embedded (FFPE) samples of early, localized human colon adenocarcinomas arising in high-grade adenomas (so-called "malignant polyps"). These lesions are small and hence the amount of DNA is limited. Additionally, the quality of DNA is compromised due to the fragmentation as a consequence of formalin fixation. To overcome these problems, we optimized a newly developed isothermal whole genome amplification system (NuGEN Ovation® WGA FFPE System). Starting with 100 ng of FFPE DNA, the amplification system produced 4.01 ± 0.29 μg (mean ± standard deviation) of DNA. The excellent quality of amplified DNA was further indicated by a high signal-to-noise ratio and a low derivative log(2) ratio spread. Both, the amount of amplified DNA and aCGH performance were independent of the age of the FFPE blocks and the associated degradation of the extracted DNA. We observed losses of chromosome arms 5q and 18q in the adenoma components of the malignant polyp samples, while the embedded early carcinomas revealed losses of 8p, 17p, and 18, and gains of 7, 13, and 20. Aberrations detected in the adenoma components were invariably maintained in the embedded carcinomas. This approach demonstrates that using isothermally whole genome amplified FFPE DNA is technically suitable for aCGH. In addition to demonstrating the clonal origin of the adenoma and carcinoma part within a malignant polyp, the gain of chromosome arm 20q was an indicator for progression from adenoma to carcinoma.

Figure 1

Example of an early, moderately differentiated adenocarcinoma arising in a tubular adenoma and invading below the muscularis mucosae (“malignant polyp”). Enclosed and marked with (1): invasive adenocarcinoma part, enclosed and marked with (2): adenoma part. HE, 20× magnification.

Figure 2

Size distribution of starting material and amplified products. A) Depending on the age of the FFPE block, isolated DNA showed a different degree of degradation, which increased with age (highest in samples >10 years old). B) Regardless of the initial degree of degradation, the Agilent DNA 1000 BioAnalyzer profiles of amplified DNA looked similar.

Figure 3

Parameters of whole genome amplification (WGA) and array comparative genomic hybridization (aCGH) as a function of the age of the FFPE blocks (0-5 years: n=8, 5-10 years: n=12, >10 years: n=6). A) Amount of amplified DNA using an input of 100 ng. B) Signal-To-Noise Ratio from aCGH. C) Derivative Log₂ Ratio Spread (DLRSpread) from aCGH. Horizontal lines and error bars show mean ± standard deviation.

Figure 4

Ideograms of chromosomal gains and losses for amplified and unamplified identical FFPE DNA. A) Genome view showing the same aberrations for both amplified (orange) and unamplified (blue) DNA. B,C) Chromosome views of chromosome 17 demonstrate that even small chromosomal aberrations within complex aberration patterns were amplified correctly (B: amplified; C: unamplified).

Figure 5

Frequency plot of copy number gains (green) and losses (red) identified by aCGH in adenoma and carcinoma components of malignant polyps.

Figure 6

Genome views of the adenoma (blue) and carcinoma (red) component of the same malignant polyp. The adenoma component shows losses of 8p, 17p and 18 and gains of 8q and 20. In addition to the aberrations present in the adenoma component, the carcinoma component shows a loss of 16q13-24.3 and gains of 13, 16p and 16q12.1-13.