Main

The protein product of the p21WAF1 gene is a negative regulatory element of the cell cycle whose function is mediated by the inhibition of the G1 cyclin-cyclin-dependent kinases complexes, thereby inhibiting the cell cycle progression and cell growth(1, 2).

The expression of this negative regulator is induced by the presence of wild type, but not mutant, p53, involved in apoptosis and cell cycle arrest in response to DNA damage(3). Alterations that abolish the normal function of genes up and down stream of the regulatory pathway of the p53 gene would be expected to have the same or similar effects than the mutations in this gene, that are known to be the most frequent alterations in sporadic human tumors(4). Therefore, mutations abolishing the normal functions of p21WAF1 or p53 would be complementary mechanisms involved in cell transformation due to deregulation of the G1 checkpoint.

The p21WAF1 protein is composed by 164 amino acids, and is encoded by three exons(3). We chose to analyze exon number 2 because it covers over 90% of the whole coding regions and easily yields a melting map susceptible to be analyzed by means of DGGE.

Losses of genetic material in the p21WAF1 locus (6p21.2) have been observed in several tumors(5, 6), although it is not the case in bone tumors. On the contrary, mutations altering the coding region of p21WAF1 have not been found in most of the tumors screened; instead, frequent genetic polymorphisms have been reported, and the effects of such changes in the encoded protein are being exhaustively analyzed(7–9).

We undertook this study to search for mutations of the p21WAF1 gene in a cohort of pediatric bone tumors in which alterations of p16INK4 and p53 tumor suppressor genes and the ras family of oncogenes have already been analyzed(10, 11).

METHODS

Patients and samples. Fresh or paraffin-embedded samples from primary tumors or metastases and peripheral blood lymphocytes were obtained from patients treated at the Clínica Universitaria in Pamplona, Spain. Seventy-four samples from 55 patients were examined: 45 osteosarcomas and 10 Ewing sarcomas. Matched normal DNA was available for 37 of these patients.

Genomic DNA was obtained by isolation and purification with proteinase K and extraction using conventional phenolchloroform procedures. Serial cuts were obtained from paraffin-embedded samples, and they were deparaffinized following conventional procedures with slight modifications(12).

PCR amplification. Codons 2-148 (exon 2) of the p21WAF1 gene were amplified using the following primers: direct, 5′-AGAACCGGCTGGGGATGTC-3′; and reverse, 5′-GCCCGCCGGCCCGACCCCCGCGCGTCCGGCGCCGCGCCCCGCTGGTCTGCCGCCGTTTTC-3′(7). In brief, 200 ng of genomic DNA were amplified in a 50-μL PCR reaction containing 15 pmol of each primer, 200 μM dNTPs, 20 mM Tris-HCl (pH 8.5), 16 mM SO4(NH4)2, 2.5 mM MgCl2, 150 μg/mL BSA, 5% DMSO, and 2 units of Thermus aquaticus DNA polymerase (BioTaq™, Bioprobe Systems). The amplification consisted in 30 cycles of denaturation at 94 °C/1 min, annealing at 57 °C/1 min, and extension at 72 °C/1 min(7).

DGGE analysis and sequencing. DGGE analysis was performed on 50-100% denaturing, 6.5% polyacrylamide gels (100% denaturant being 7 M urea and 40% formamide) at 60 °C for 3 h, 30 min. Before electrophoresis, heteroduplex formation was forced by incubation at 94 °C for 10 min followed by reannealing at 56 °C for 1 h. After electrophoresis, bands were visualized by staining with an 1 × Tris borate EDTA solution containing 0.5 μg/mL ethidium bromide and photographed under UV light.

DGGE conditions such as melting temperatures, gradients, and running times were obtained by analysis with program Melt87, kindly provided by Dr. L. Lerman(13).

Bands with altered electrophoretic mobilities were cycle-sequenced using the Thermo Sequenase™ cycle sequencing kit (Amersham Life Science).

RESULTS

In the analysis of the p21WAF1 gene, nine of the samples screened showed an altered DGGE pattern compared with normal controls (Fig. 1). Seven of them had the same band pattern, therefore suggesting the presence of the same alteration in all of them; the other two samples had different DGGE profiles. When DNA from peripheral blood lymphocytes of the same individuals was available (six out of the nine cases), we got identical profiles in every case in normal and matched tumoral samples, indicating that the alterations present in our samples corresponded to polymorphic variants. None of these samples carried an altered p53 gene.

Figure 1
figure 1

DGGE analysis showing altered band patterns in lanes 1, 3, 4, and 6, and normal ones in lanes 2, 5, and 7.

Full size image

We got DNA of enough quality to be sequenced in all cases but one, a paraffin-embedded biopsy sample whose DNA turned out to be impossible to sequence either due to the low quality and quantity of the DNA present in the sample or to the presence of traces of inhibitory elements for the cycle-sequencing reaction. Nevertheless, in the DGGE analysis the sample showed a clearly altered band pattern, but we cannot be sure of the identity of the alteration without confirmatory sequencing.

In the sequence analysis, seven of the eight samples available (6.4% of all chromosomes) were shown to carry the extensively reported polymorphism S31R, in which a C to A transversion at codon 31 changes the serine in that position into an arginine (Fig. 2). In the remaining sample an A64T alteration was present (0.9%), in which a G to A transversion changes the alanine into a threonine. Both these polymorphisms have already been described in the literature in tumors of different origins(8, 9) and also in the normal population.

Figure 2
figure 2

Sequence analysis of a patient's sample carrying the genetic polymorphism S31R (Ser31 → Arg).

Full size image

DISCUSSION

Human p21WAF1 is a transcriptional target and, therefore, a mediator of the cell cycle arrest induced by p53. Alterations in this pathway through mutations of either p53 or p21WAF1 may contribute to transformation of human cells.

In a previous analysis of p53 in our series we detected mutant patterns in 18.6% of the patients by means of DGGE analysis and confirmatory cycle sequencing(11).

Absence of inactivating mutations at the p21WAF1 locus is open to several interpretations. First, given that p21WAF1 induction blocks G1-S transition by inhibiting the activity of cyclin-cyclin-dependent kinases that phosphorylate pRB, p53-induced cell cycle arrest via p21WAF1 requires wild type pRB. Bone tumors frequently carry inactivated or altered pRB(14), so that this precise cell cycle arrest pathway may be already altered in the pediatric tumors, therefore alleviating the need for p21WAF1 inactivation(1, 2).

Other possible explanations for the low frequency of inactivation of p21WAF1 in human cancers may rely on the protective role that this gene seems to have over p53-induced apoptosis in conjunction with other trans-acting factors(15) and, as some authors have suggested, that complete abrogation of the normal functions of the gene may be incompatible with cell survival; therefore, p21WAF1 mutations will be scarcely found in human tumors(9).

Although the polymorphisms detected are not silent changes, the fact that they are found as well in the normal population and the presence of the same variants in the matched normal DNAs supports the hypothesis that they do not actually contribute to the development of pediatric bone tumors. Nevertheless, more analyses have to be made to determine the exact change in function that these alterations promote in the encoded protein.

It has to be taken into account that as only approximately 87% of the coding regions of the gene are being screened, we cannot exclude the presence of alterations in other regions of the gene or sequences involved in its regulation. To our knowledge, mutations of the p21WAF1 gene that impair the function of the protein have been described only in an invasive ductal breast carcinoma carrying a R94W alteration(16) and in a case of prostate cancer(17). Our results suggest that alterations affecting the p21WAF1 tumor suppressor gene are not frequent events in the development of pediatric bone tumors.