Abstract
Mutation of the tumor suppressor gene p53 is the most common genetic alteration observed in human tumors. However, the relationship between the mutation point of p53 and the transcriptional specificity is not so obvious. We prepared Saos-2 cells with various mutations of p53 that are found in human tumors, and examined the resulting transcriptional alterations in the cells. Loss of function and gain of function were observed in all p53 mutants. Hot-spot mutations of p53 are frequently found in tumor cells. We compared hot-spot mutations and other mutations of p53 and found that a more than 2-fold transcription of CADPS2, PIWIL4 and TRIM9 was induced by hot spot mutations, but not by other mutations. As PIWIL4 suppresses the p16INK4A and ARF pathway, restraining cell growth and genomic instability, induction of PIWIL4 expression may be one reason why hot-spot mutations are frequently found in tumor cells.
The tumor suppressor gene p53 plays a central antiproliferative role through induction of apoptosis, senescence and cell-cycle arrest in response to various types of stress, including DNA damage (1-3). The p53 status of a tumor has deep implications for the tumor growth potential. Mutation in p53 remains the most frequent genetic change identified in human cancer (4). Most somatic mutations in p53 are single-nucleotide substitutions, and most missense mutations are located in the DNA-binding domain that encompasses p53 exons 5-8; six of these (Arg175, Gly245, Arg248, Arg249, Arg273 and Arg282) have been labeled hot-spot codons because of their high mutation frequency (5). Mutant forms of p53 differ in their properties according to the location of the mutations. Using a large series of p53 mutants, Crook et al. found that not all transcriptionally-active mutants retained their ability to suppress transformation, and that some tumor-derived point mutations conferred both transforming and transactivating activity (6). Some mutant forms of the p53 gene do not merely induce the functional equivalent of p53 loss (7). The resulting p53-mutant proteins can contribute to both a loss of function and a gain of function. However, the relationship between the mutation point of p53 and transcriptional specificity has not been studied extensively. Therefore, we prepared 14 types of Saos-2 cells harboring mutant forms of p53 (130V, 143A, 157F, 168R, 175H, 195T, 242F, 244C, 245S, 273H, 277F, 280T, 282W and 286K) that are found in human tumors; the 175H, 245S, 273H and 282W mutations are hot-spot mutations. We then examined the transcriptional expression in each mutant cell and compared that for the hot-spot mutants with that for the other mutants.
Materials and Methods
Cell culture. Human osteosarcoma Saos-2 cell line (RCB0428: Riken Cell Bank, Tsukuba, Japan), which is null for p53, in Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% fetal calf serum (FCS). Transformants were then selected in a medium containing 100 μg/ml hygromicin or 400 μg/ml G418.
Plasmids. We used the LacSwitch inducible mammalian expression system (Stratagene, La Jolla, CA, USA) (8). The expression plasmid was pOPI3CAT, and the CAT gene was replaced by mutant or wild-type p53 genes at the Not I site. Saos-2 cells were transfected with 2 μg of expression plasmid using the calcium phosphate precipitation method. The p3’SS plasmid, expressing the lac gene encoding the lac repressor protein, was used as a regulation plasmid. We then co-transfected the expression and regulation plasmids into the cells. The expression of each mutant p53 gene was almost suppressed by the lac repressor in this LacSwitch system.
In vitro mutagenesis. We performed mutagenesis with a QuikChange site-directed mutagenesis kit (Stratagene), in accordance with the manufacturer's protocol (8). We introduced wild-type p53 cDNA (pPro Sp53: CD 104; Japanese Cancer Research Resources Bank, Tokyo, Japan) into the M13 vector. Oligonucleotides for the mutations were ordered from Hokkaido System Science, Sapporo, Japan. We checked the mutant p53 DNA sequences with a DNA sequencing kit and a model 373A DNA sequencer (Applied Biosystems, CA, USA).
Radiation treatment. Cells were irradiated with a PS-3100S Cs-137 γ-ray machine (Pony Industry Co. Ltd., Osaka, Japan) 3 or 6 Gy at a dose of 0.993 Gy/min at room temperature, and examined the cells 24 h after irradiation for the induction of gene expression by irradiation (9).
Microarray expression analysis. We extracted mRNA from each cell type and synthesized the cDNA. After we had synthesized, in turn, cRNA from the cDNA, we labeled the former with biotin and hybridized it with a DNA microarray (Gene Chip, Human Genome U133 Plus 2.0 Array; Affymetrix, Santa Clara, CA, USA) containing over 54,000 probe sets (10). After staining and washing, fluorescence was read using a scanner. The expression value (signal) of each gene was calculated and normalized to adjust for minor differences between the experiments using GeneSpring (Agilent Technologies, Santa Clara, CA, USA). In order to obtain the mean basal expression level of each gene, the signal values for parental Saos-2 cells were used as the standard for the analysis. The change in value (signal log ratio) for each gene was calculated using comparison analysis in the software. Microarray data analysis was complied with the Minimum Information About Microarray Experiments (MIAMI) standard (11).
Results and Discussion
Expression of p53 downstream genes. We examined the expression of genes in cells harboring p53 mutations. As mutation of p53 may not merely lead to full functional loss of p53, we selected genes downstream of p53 showing an increase in gene expression of more than 2-fold in wild-type cells. Table I summarizes the genes induced by wild-type p53: the cells harboring wild-type p53 (WT) are located on the far left, those harboring hot-spot mutant p53 (175H, 245S, 273H and 282W) on the right, and the cells harboring other mutant forms of p53 (130V, 143A, 157F, 168R, 195T, 242F, 244C, 277F, 280T and 286K) are in the middle. Cells harboring mutant p53 lacked expression of many genes that were induced in cells harboring wild-type p53, and the 175H mutant cell line did not show any increase in the expression of these genes. The expression of the gene GDF15, also known as PDF, MICI, PL4B, MIC-1, NAG-1 and PTGFβ (12), was increased by more than 2-fold in most mutant cells, as well as in cells harboring wild-type p53, but not in the 157F and 175H mutant cells. Expression of some genes was increased by more than 2-fold in some mutant cells, whereas the expression of other genes was not increased more than 2-fold in all mutant cells. We were unable to find any well-defined differences between hot-spot mutant cells and other mutant cells, except for the 175H mutant.
Induction of gene expression by irradiation. Expression of p53-dependent genes is induced by stress, including DNA damage due to ionizing radiation. Expression of the gene CDKN1A, also known as p21, WAF1, CIP1 and SDI1, concerned with growth arrest (13), was increased by more than 2-fold in only 244C and 282W mutant cells, but not in other mutants (Table I). The BAX gene, which induces apoptosis (14), was not increased by more than 2-fold in any mutant cells (Table I). As expression of these two genes, CDKN1A and BAX, was induced by irradiation of wild-type cells, we examined their expression in mutant cells after exposure to irradiation. Figure 1 shows the expression of CDKN1A and BAX 24 h after irradiation. In cells with wild-type p53, but not in cells with mutant p53, the expression of these genes increased with the irradiation dose. The expression of CDKN1A was reduced by irradiation in the 224C mutant cells, and was almost unchanged in the 282W mutant cells, as was the case in other mutant cells. The expression of BAX was not increased by more than 2-fold in any of the p53-mutant cells. The mechanism of induction of p53 target genes by irradiation depends on the accumulation of p53 protein (15). These results indicate that mutation of p53 affects the ability of cells not only to regulate the transcription of target genes, but also to accumulate p53 protein as a result of irradiation (8). Indeed, the lifespan of p53 protein is prolonged by mutation, and mutant p53 proteins often accumulate in the nucleus of tumor cells (16). Therefore, the system for p53 accumulation in response to stress may be disregulated in cells harboring mutant p53.
Hot-spot p53 mutant-specific gene expression. Among the losses of function in p53 mutants, we were unable to find any precise differences between hot-spot p53 mutations and other forms of p53 mutation. Because hot-spot p53 mutations are found more frequently in human tumor cells than other forms of p53 mutation, the transcriptional changes in hot-spot p53 mutants might make them specifically prone to carcinogenesis. One of the cancer-related processes in cells with mutant p53 is gain of function, in which mutant p53 is expressed in cells previously lacking wild-type p53, thus enhancing their tumorigenic properties (17). However, it remains unclear whether gain of function is a universal property of all hot-spot p53 mutants (18).
We examined the gains of function in p53 mutants, especially of hot-spot p53 mutants. Table II shows that some genes were expressed by more than 2-fold in all hot-spot mutants, but less than 2-fold in wild-type cells. Thirty-four genes were expressed in hot-spot p53 mutants, but their properties varied. If specific transcriptional expression of genes conferring tumor-like properties occurs in hot-spot p53 mutant cells, this would be evident in all hot-spot p53 mutant cells, but not in cells harboring other forms of p53 mutation. CADPS2, PIWIL4 and TRIM9 were simultaneously expressed by more than 2-fold in all hot-spot p53 cells, but less than 2-fold in all other p53-mutant cells. CADPS2 was originally cloned as a Ca+-dependent activator of the secretion protein family (19), and deletion of this gene contributes to autism spectrum disorders and persistent hyperplastic primary vitreous (20). TRIM9 is the gene for brain-specific E3 ubiquitin ligase, and its expression is repressed in the brain of patients with Parkinson's disease and dementia with Lewy bodies (21). It is of considerable interest that both genes are related to central nervous system function, but there is very little information about the roles of these genes in tumor cells. For example, the expression of TRIM9 is reportedly increased in the breast epithelium of patients with breast cancer (22). PIWIL4 is a member of the Argonaute protein family, which plays an important role in stem cell self-renewal, interference (RNAi) and translational regulation (23). Argonaute proteins are highly conserved between species and can be divided into the Ago and Piwi sub-families. In humans, there are four members of the PIWI-like family, PIWIL1, PIWIL2, PIWIL3 and PIWIL4, and they are primarily expressed in testis and embryonic tissues.
PIWIL4 is highly expressed in tumorous tissues and may promote tumor invasion (24). PIWI4 induces methylation of H3K9 at the p16INK4A locus, and an elevated level of histone methylation results in a down-regulation of the p16INK4A gene (25). The p16INK4A protein is a potent cell-cycle inhibitor that regulates the RB pathway. p16INK4A is encoded within the human CDKN2A locus, from which ARF is an alternative product. Su et al. have reported that PIWIL4 regulates the growth of cervical cancer cell lines and is involved in down-regulation of the expression of ARF and p53 (26). ARF signaling is complex, and involves a p53-dependent or - independent pathway aimed mainly at restraining abnormal cell growth and maintaining genomic stability. PIWIL4 is highly expressed in hot-spot p53-mutant cells (Table II), and is thought to prevent the p16INK4A and ARF pathway from conferring tumorigenicity on cells. These finding suggest that PIWIL4 may be a candidate gene related to the high frequency of hot-spot mutations in tumor cells.
In this study we examined only four hot-spot mutations and 10 other mutations of p53. More p53-mutants will need to be examined to clarify the specificity of p53 hot-spot mutations. It will also be necessary to examine microRNAs, because p53 regulates the expression of mRNAs and microRNAs, which can change in tumor cells (27).
Acknowledgements
This work was partly supported by the Global Center of Excellence (GCOE) of the Ministry of Education, Culture, Sports, Science and Technology of Japan.
- Received January 14, 2013.
- Revision received February 15, 2013.
- Accepted February 15, 2013.
- Copyright© 2013 International Institute of Anticancer Research (Dr. John G. Delinassios), All rights reserved