Renal cell carcinoma primary cultures maintain genomic and phenotypic profile of parental tumor tissues

Abstract

Background: Clear cell renal cell carcinoma (ccRCC) is characterized by recurrent copy number alterations (CNAs) and loss of heterozygosity (LOH), which may have potential diagnostic and prognostic applications. Here, we explored whether ccRCC primary cultures, established from surgical tumor specimens, maintain the DNA profile of parental tumor tissues allowing a more confident CNAs and LOH discrimination with respect to the original tissues.

Methods: We established a collection of 9 phenotypically well-characterized ccRCC primary cell cultures. Using the Affymetrix SNP array technology, we performed the genome-wide copy number (CN) profiling of both cultures and corresponding tumor tissues. Global concordance for each culture/tissue pair was assayed evaluating the correlations between whole-genome CN profiles and SNP allelic calls. CN analysis was performed using the two CNAG v3.0 and Partek software, and comparing results returned by two different algorithms (Hidden Markov Model and Genomic Segmentation).

Results: A very good overlap between the CNAs of each culture and corresponding tissue was observed. The finding, reinforced by high whole-genome CN correlations and SNP call concordances, provided evidence that each culture was derived from its corresponding tissue and maintained the genomic alterations of parental tumor. In addition, primary culture DNA profile remained stable for at least 3 weeks, till to third passage. These cultures showed a greater cell homogeneity and enrichment in tumor component than original tissues, thus enabling a better discrimination of CNAs and LOH. Especially for hemizygous deletions, primary cultures presented more evident CN losses, typically accompanied by LOH; differently, in original tissues the intensity of these deletions was weaken by normal cell contamination and LOH calls were missed.

Conclusions: ccRCC primary cultures are a reliable in vitro model, well-reproducing original tumor genetics and phenotype, potentially useful for future functional approaches aimed to study genes or pathways involved in ccRCC etiopathogenesis and to identify novel clinical markers or therapeutic targets. Moreover, SNP array technology proved to be a powerful tool to better define the cell composition and homogeneity of RCC primary cultures.

Figure 1

Phenotypic characterization of ccRCC primary cultures. (a) Representative cellular morphology during in vitro growth. 100× magnification. (b) Representative micrographs of immunofluorescence staining (top) and FACS analysis (bottom) of pan-cytokeratin, vimentin, CD13 and CA9. DAPI counterstains nuclei in blue. 400× magnification. The positivity percentages for the different markers are reported in the FACS analysis as mean value (± SD) of the nine cultures. (c) Western Blot analysis of CA9 in all ccRCC primary cultures. B-actin was used as internal control.

Figure 2

Copy number alterations and LOH events in ccRCC primary cultures and parental tissues, as calculated by CNAG v3.0 software. On each chromosomal arm (p, short arm; q, long arm), amplifications (↑) and deletions (↓) and LOH events are reported for all samples. Color labels distinguish CN alterations (CNAs) detected by CNAG and signed in the color-coded "HMM-CN state" track (red for amplifications and dark green for deletions), and CNAs resulting below threshold to be visualized in the HMM-CN track (light green for deletions). Only LOH events reaching significant likelihood to be signed by CNAG in the HMM-LOH track are reported.

Figure 3

Whole-genome view of copy number profile in 81BPG primary culture at first (p1), second (p2) and third (p3) confluences, and in corresponding tumor tissue, using CNAG v3.0 software. Analysis was performed using CNAG v3.0 software, comparing primary culture at each passage and parental tumor tissue to the autologous blood sample. Chromosomes are represented horizontally, from 1 to 22 in different colors, separated by vertical bars. For each sample, the three tracks represent (on log scale): a) "copy number plot": copy number log ratio values of single SNPs; b) "copy number average": copy number log ratio values locally averaged on 10 contiguous SNPs; c) "allele-based analysis": copy number log ratio values for each allele (red and green lines).

Figure 4

Visualization of chr 3 in 66SML primary culture (upper panel) and parental tissue (lower panel) using CNAG v3.0 software. Chromosome 3 is shown from p to q end (from left to right). The upper two graphs represent single SNP copy number data on log2 scale (red dots) and copy number values locally averaged on 10 contiguous SNPs (blue line), whereas copy number values for each allele (red and green lines) are shown below. Green bars in the middle represent heterozygous SNP calls detected by the software comparing each sample to autologous blood. The three bars at the bottom represent the color-coded visualization of HMM-CN state (yellow, diploidy; pink, amplification; light blue, deletion) and of HMM-LOH state (blue, significant LOH; yellow, no LOH), with LOH likelihood indicated by the thickness of the third blue bar. Boxes on the left report mean CN log2ratio values and mean LOH likelihoods calculated for the whole deleted region in primary culture and tissue, respectively.