Ataxia telangiectasia-mutated dependent DNA damage checkpoint functions regulate gene expression in human fibroblasts

Abstract

The relationships between profiles of global gene expression and DNA damage checkpoint functions were studied in cells from patients with ataxia telangiectasia (AT). Three telomerase-expressing AT fibroblast lines displayed the expected hypersensitivity to ionizing radiation (IR) and defects in DNA damage checkpoints. Profiles of global gene expression in AT cells were determined at 2, 6, and 24 h after treatment with 1.5-Gy IR or sham treatment and were compared with those previously recognized in normal human fibroblasts. Under basal conditions, 160 genes or expressed sequence tags were differentially expressed in AT and normal fibroblasts, and these were associated by gene ontology with insulin-like growth factor binding and regulation of cell growth. On DNA damage, 1,091 gene mRNAs were changed in at least two of the three AT cell lines. When compared with the 1,811 genes changed in normal human fibroblasts after the same treatment, 715 were found in both AT and normal fibroblasts, including most genes categorized by gene ontology into cell cycle, cell growth, and DNA damage response pathways. However, the IR-induced changes in these 715 genes in AT cells usually were delayed or attenuated in comparison with normal cells. The reduced change in DNA damage response genes and the attenuated repression of cell cycle-regulated genes may account for the defects in cell cycle checkpoint function in AT cells.

Figure 1

A. Protein extracts were prepared from AT fibroblast cell lines (AT1, AT2 and AT3), and normal human fibroblast cell lines (F1, F3 and F10), and 100 µg of total protein was analyzed by western immunoblot analysis with anti-ATM antibody. B. Inactivation of clonogenic survival by IR. Three normal fibroblast lines and three AT fibroblast lines were irradiated with IR and colonies were counted after a 14-day incubation. Results show the mean relative colony formation in irradiated cultures (mean ± SD, n=3). C. IR-induced G1 and G2 checkpoint functions in normal and AT fibroblasts. Checkpoint functions were determined by measuring BrdU incorporation 6–8 h after 1.5 Gy IR or sham treatment (G1) and phospho-histone H3 expression 2 h after 1.5 Gy IR or sham treatments (G2) in normal and AT cells. The values depicted for each fibroblast line are the mean percentages of inhibition of IR-treated cells relative to sham-treated cells (mean ± SD, n=3).

Figure 2

The time-course of cellular DNA synthesis and mitosis in normal and AT cells after IR-induced DNA damage. A. The fractions of cells in S were determined by measuring BrdU incorporation with flow cytometry as described in Materials and Methods. B. Mitotic cells were quantified using anti-phospho-histone H3 antibody with flow cytometry. Although representative flow profiles are shown, the numerical values depicted for each cell line are the mean percentages of sham- or IR-treated cells in each cycle phase compartment (n=3). MI = mitotic index.

Figure 2

The time-course of cellular DNA synthesis and mitosis in normal and AT cells after IR-induced DNA damage. A. The fractions of cells in S were determined by measuring BrdU incorporation with flow cytometry as described in Materials and Methods. B. Mitotic cells were quantified using anti-phospho-histone H3 antibody with flow cytometry. Although representative flow profiles are shown, the numerical values depicted for each cell line are the mean percentages of sham- or IR-treated cells in each cycle phase compartment (n=3). MI = mitotic index.

Figure 3

Stereotypic gene expression patterns in AT cell lines in response to IR-induced DNA damage. Four DNA-damage-responsive patterns of gene expression were extracted using EPIG, each pattern represented a group of genes that responded to IR in the same way. Each pattern is presented as the average of the expression levels of the 6 most highly correlated genes, with the level of expression of each gene being the average of the two dye-flip replicates. For each pattern and for each cell line, gene expression levels are shown as the log2 ratios of sample RNA against reference RNA. The log2 ratio of gene expression in the sham-treated control was adjusted to zero and IR-treated samples were adjusted by the same proportion. Sample treatments from left to right were sham-treated control (square), IR 2h (circle), IR 6h (triangle) and IR 24h (diamond). In each panel, the numbers of genes in each pattern are shown in parentheses.

Figure 4

Idiosyncratic expression of p53 targets or cell cycle-regulated genes in AT lines (AT1, AT2 and AT3) and stereotypic expression of these genes in normal lines (F1, F3 and F10). Details for each panel are as in Figure 3. Panels A and C show AT lines, panels B and D show normal lines. The names of the genes that were included in this analysis, including p53 target genes (A and B) and G2/M transition-regulated genes (C and D), are given on the right.

Figure 5

ATM-dependent gene expression patterns in response to IR-induced DNA damage. Gene expression profiles for three AT cell lines and three normal cell lines were analyzed using EPIG. Six expression patterns were extracted showing different responses to IR between normal and AT cell lines. Each pattern is presented as the average of the expression levels of the 6 most highly correlated genes with the level of expression of each gene being the average of the two dye-flip replicates. In each panel, cell lines from left to right are F1, F3, F10, AT1, AT2 and AT3 as depicted at the bottom of the figure; the numbers of genes in each pattern are shown in parentheses. For each pattern and for each cell line, gene expression levels are as described in Figure 3.

Figure 6

Potential pathways of checkpoint function and cell cycle gene regulation in AT fibroblasts after IR-induced DNA damage. Upon IR-induced DNA damage, the ATM-dependent signaling pathway is absent in AT cells, but the ATR-dependent pathway, although delayed, is activated and initiates G1 and G2 arrests. p53 plays a key role in ATR-dependent cell cycle checkpoint function by inducing important cell cycle kinase inhibitors and by repressing transcription of cell cycle-regulated genes. Not all AT cell lines followed the same rules. The AT1 cell line was defective in induction of some p53 target genes, including p21^Waf1(a), and AT2 was defective in repression of G2/M transition-regulated genes upon DNA damage, leading to escape from delayed G2 arrest (b).