Year: 2005
Type: letter
Abstract: Yamanoshita et al. recently published a TP53 mutation pattern obtained by analysing oesophageal cancerous tissues from 207 Chinese patients.1 With great interest, we read their article, which was mainly dedicated to the comparison of 2 mutation screening approaches: single-strand conformation polymorphism (SSCP) and denaturing high-performance liquid chromatography (DHPLC). Reliable TP53 mutations screening and identification methods are needed in numerous fields of cancer research. In molecular epidemiology and environmental toxicology, the establishment of mutational patterns provides interesting clues in the involvement of cancer risk factors.2 Two of the most famous examples are tandem CC to TT substitutions related to UV radiations in skin cancers and G to T transversions at codon 249 in aflatoxin B1-related liver cancers. Another application is the use of TP53 mutations as a tool to improve diagnosis, prognosis or prediction of response to therapy. In these areas, the precise determinations of mutation type and localisation inside the gene seem to bring more relevant information than a simple immunohistochemical determination of the p53 protein overexpression.3 Finally, the identification of TP53 mutations could be of great value if new therapies aiming to restore the protein p53 functionality confirm their promising preliminary results in the future.4 For all these reasons, many approaches are optimised to improve the detection and the following identification of TP53 mutations in human tissues. These last few years, the most popular of them have been SSCP, functional assay for the separation of alleles in yeast (FASAY), DNA microarrays dedicated to TP53 analysis, restriction methods and techniques based on double-stranded DNA denaturation such as DHPLC, denaturing gradient gel electrophoresis (DGGE) or temporal temperature gradient gel electrophoresis (TTGE). DHPLC has already been used to detect TP53 alterations in different cancerous tissues including oesophagus,5 stomach,6 ovary,7 blood cells,8 bladder9 or conjunctiva.10 As mentionned by Yamanoshita et al., this technique is sensitive, easy to use (in comparison to electrophoretic methods) and offers an interesting throughput. However, the authors reported a disappointing specificity with 50% of false-positives and considered this performance as “not yet understood.” As this issue is essential to obtain reliable results, we would like to provide here some clues to explain at least one part of this lack of specificity and some practical solutions for clinicians and researchers beginning to screen TP53 mutations. Indeed, we will demonstrate that some false-positives attributed to a lack of DHPLC specificity could in fact constitute genuine mutated DNA fragments. The inclusion of these additional positive results in tumor series could improve the sensitivity of the whole procedure (screening + sequencing) and increase mutation rates found in studies similar to that of Yamanoshita et al. These comments are based on our own experience5, 11, 12 and on technical information collected in other laboratories and published studies. DHPLC is indeed likely to give false-positives. In these cases, additional peaks or irregularities of the baseline can result from degradation products, aspecific DNA amplification or from mutations generated during the polymerase chain reaction (PCR). Provisions against these potential artefacts must be taken: (i) systematic use of well-characterized positive and negative DNA controls for each fragment; (ii) replication of any doubtful analysis (using a high-fidelity polymerase if it was not the case during the first analysis); and (iii) sequencing of all positive samples. This last point has been performed by Yamanoshita et al. However, in spite of the undeniable usefulness of direct sequencing to identify the mutations (e.g., determination of mutation type and localisation inside the gene), we would like to point out that this approach is far from a “gold-standard” approach. It has indeed often been demonstrated that direct sequencing has a low sensitivity compared to screening methods such as DGGE13 or DHPLC.14 As a consequence, DHPLC “false-positives” are in fact direct sequencing “false-negatives.” This issue is particularly problematic for mutations acquired during tumor development (as opposed to inherited mutations). Indeed, in these cases, mutated cancerous cells might sometimes be “diluted” amongst other cells. In oesophageal samples, the mutated cancerous cells might, for example, have been mixed with nonoesophageal cells (fibroblasts, endothelial cells, etc.), noncancerous oesophageal cells or cancerous cells without TP53 mutations. Indeed, oesophageal squamous cell carcinomas constitute a classic example of field carcinogenesis, and multiple independent foci of neoplastic or preneoplastic cells could coexist.15 Various approaches have been used to enrich the proportion of mutated cells or mutated DNA to be analysed, hence improving the sensitivity of the sequencing stage. The analysis can be restrained to cancerous cells by selecting them with laser capture microdissection (LCM).15, 16 By doing so, the sensitivity of the screening method and that of the sequencing are likely to be enhanced. A second solution is to limit the sequencing to mutated single-strand DNA after SSCP or to mutated double strands after TTGE, DGGE or DHPLC. This approach is commonly used by extracting and sequencing the DNA from excised bands containing mutated DNA after TTGE17 and DGGE5 or by collecting the DNA from an eluted mutated peak during DHPLC (by using a fraction collector).8 As an example, Figure 1 shows a TP53 DNA fragment (from an oesophageal squamous cell carcinoma biopsy) presenting positive signals with DGGE and DHPLC without demonstrating any mutation after direct sequencing. Excision of mutated bands from the DGGE gel, DNA extraction from this gel, PCR and a new sequencing stage allowed the identification of the mutation (exon 6, codon 203). DHPLC and DGGE results were therefore not false-positives. Denaturing high-performance liquid chromatography (DHPLC), denaturing gradient gel electrophoresis (DGGE) and sequencing analysis of a DNA sample extracted from an oesophageal squamous cell carcinoma (TP53 gene, exon 6). The mutated sample (198) presented a positive signal with DHPLC and DGGE (left), but direct sequencing did not exhibit any difference with the wild-type sequence (right). The mutation was finally identified after excision and amplification of heteroduplex bands from the DGGE gel (codon 203). The homoduplex mutated band (HoM) did not provide enough DNA to achieve a reliable amplification. Our letter is mainly dedicated to the specificity of the whole mutation analysis procedure. However, we would like to underline that specificity is, of course, not the only criterion necessary to achieve a reliable screening method. We could, for instance, add the sensitivity. As a short but useful example concerning DHPLC or DGGE, it has been demonstrated that some DNA sequences are particularly prone to false-negative results. This is the case of DNA sequences of high melting temperature embedded in lower melting temperature fragments. A careful choice of primer sets is necessary to avoid this problem.5 We hope that these short comments and advice will benefit researchers beginning to screen human tissues for TP53 mutations. We thank Dr. E.C. Smith (University of Leeds) for critical reading of the manuscript. Yours sincerely, Jean Breton, Mathilde Lechevrel, François Sichel.
Source: International Journal of Cancer
DOI: 10.1002/ijc.21223
Language: English
Subfield: Oncology
Field: Medicine
Domain: Health Sciences
Open Access status: bronze
OpenAlex ID: W2062467154
