Figure 1.
Comparison between copy number changes and gene expression by chromosomal arms.
On the left side, the number of SNPs located in each chromosomal arm is indicated, which were explored by the 100 K microarray. On the right side, the number of genes located in each arm is indicated, which were explored for changes in gene expression by the ST1.0 expression microarray. Each bar represents the percentage of recurrent altered SNPs (left) or deregulated genes (right) common to the 4 cell lines. The chromosomal arms are indicated in the middle column. Arms labeled with asterisks had a mean number of CN-altered SNPs higher and statistically significant (p<0.05, chi-square test) compared with the whole genome means. Arms with a statistically significant deregulated gene enrichment were labeled with “a” (identified with both chi-square test and PAGE), “b” (identified only with chi-square test, p<0.05) or “c” (identified only with PAGE).
Figure 2.
Genome amplification and deregulation of gene expression in Chr 5p.
Panel A shows the copy number log2 ratio of SNPs investigated in Chr 5 by the 100 K SNP microarray in HeLa, CaSki, SiHa, and CaLo. Panels B to D show the fold change of gene expression of genes evaluated by the ST1.0 expression microarray located in MRR 5-1 (n = 64), MRR 5-4 (n = 44), and MRR 5-5 (n = 37) at 5p. The bars represent upregulated genes, downregulated genes, and genes without change in gene expression. The genes are ordered according to position in the genome. The SAM method was used for the analysis, using cut-off values of fold change of ≥1.5 or ≤0.66 for up- or downregulated genes and fold discovery rate (FDR) of 0%. Genes previously reported associated with cervical cancer are labeled with asterisks (IPA system) or circles (PubMed).
Table 1.
Influence of copy number alterations in gene deregulation in cervical cancer cell lines.
Figure 3.
Comparison of gene expression of PARP1, MCM2, ECT2, NAALADL2, NLGN1, TNSF10, RFC4, TRIO and CLPTM1L between cell lines and controls.
Panel A shows the experiments of microarrays and panel B the qRT-PCR experiments. Panels show the mean ± standard error of expression intensity of 9 CN-altered genes located at 1q (PARP1), 3q (MCM2, ECT2, NAALADL2, NLGN1, TNSF10 and RFC4) and 5p (TRIO and CLPTM1L). For both methods intensities are expressed in relative units (see Materials and Methods).
Figure 4.
Trend of deregulated genes as CN-altered SNPs increased by gene or region.
The figure shows the percentage trends of deregulated genes as the number of CN-altered SNPs per region (panel A) or gene (panel B) increased. MRR includes the genes harbored by the minimal recurrent regions common to the 4 cell lines. The linear association between the variables in all but one plot (CNAs, panel A) was statistically significant, p<0.01, Mantel–Haenszel linear-by-linear association chi-squared test.
Figure 5.
Trend of down- and upregulated genes as CN-altered SNPs/gene increased in deleted and amplified MRRs.
In panel A, the percentage trend of deregulated genes is compared among the genes located in MRRs having 1–100, 101–500, and >500 SNPs. The trends of up- and downregulated genes are shown in panels B (47 deleted MRRs, <500 SNPs), C (51 amplified MRRs, <500 SNPs), and D (2 amplified MRRs, >500 SNPs). The total number of genes studied for expression and included in the analysis of panels B, C, and D was 390, 267, and 108, respectively. The numbers above the bars indicate the number of deregulated genes.
Figure 6.
Genome amplification and deregulation of gene expression in 3q.
Panel A shows the copy number log2 ratio of SNPs investigated in Chr 3 by the 100 K SNP microarray in HeLa, CaSki, SiHa, and CaLo. Panels B to D show the fold change of gene expression of genes evaluated by the ST1.0 expression microarray located at 3q26 (n = 73), 3q27 (n = 63), and 3q28–29 (n = 66). The genes are ordered according to the position in the genome. See the legend of Figure 2 for further information.
Figure 7.
Genome amplification and deregulation of gene expression in Chr 1q.
Panel A shows the copy number log2 ratio of SNPs investigated in Chr 1 by the 100 K SNP microarray in HeLa, CaSki, SiHa, and CaLo. Panels B to D show the fold change of gene expression of genes evaluated by the ST1.0 expression microarray located at MRRs 1-8, 1-9, 1-14 and 1-15. The genes are ordered according to position in the genome. For panel F, the mean ± S.D. of the log2 ratio signal of 110 SNPs located in the MRR-15 was plotted. In panel G is shown the copy number of PARP1 gene calculated by qPCR in triplicate experiments. See the legend of Figure 2 for further information.
Figure 8.
Copy number analysis of cytobands 5p15 and 3q26 with fluorescence in situ hybridization (FISH).
Representative experiments of FISH analysis of cytobands 5p15.33, 5p15.2, and 3q26 in 3 cell lines (CaLo, CaSki, and HeLa) are shown. Two sets of probes were used for the analysis of 5p15 and one for the analysis of 3q26 (see Materials and Methods). The sets for 5p15 included a target probe (green signals) and a control probe (red signals) located at 5q. The set for 3q26 included a target probe (green signals) and a control probe located at the centromere. Nuclei in interphase (first, third, and fourth rows) and Chr in metaphase (second row) were counterstained with DAPI.
Figure 9.
Canonical pathways where deregulated genes are involved.
Top 25 canonical pathways identified in the set of 3,122 genes deregulated in the four cell lines (A) and in the subset of 147 deregulated and recurrent CN-altered genes (B). The canonical pathways were identified with the Ingenuity Pathway Analysis (IPA) system. The −log (p-value), gray bars, and the ratio, black dots, were calculated comparing the number of genes of the pathways present in the datasets versus the human database. The p-value was calculated with the chi square or Fisher exact tests as appropriate and the values of −log (p-value)>1.3 (dash line) correspond to p<0.05.