Mutagenesis, Vol. 18, No. 3, 299-306,
May 2003
© 2003 UK Environmental Mutagen Society/Oxford University Press
Simultaneous use of DGGE and DHPLC to screen TP53 mutations in cancers of the esophagus and cardia from a European high incidence area (Lower Normandy, France)
Groupe Régional d’Etudes sur le CANcer, Université de Caen/Basse-Normandie, Laboratoire de Cancérologie Expérimentale and 1 Laboratoire d’Anatomie Pathologique, Centre Françiois Baclesse, Route de Lion/Mer, 14 076 Caen cedex 05, France and 2 Département d’Hépato-Gastroentérologie, CHU Côte de Nacre, 14 033 Caen, France
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and methodsResultsDiscussionReferences |
|---|
We investigated TP53 mutation patterns in cancers of the esophagus and cardia of patients coming from Lower Normandy, a region situated in the highest incidence area in Europe. To screen tumor samples, we first used denaturing gradient gel electrophoresis (DGGE), a well-characterized technique which constituted our reference method. Then the results were compared with those obtained by denaturing high performance liquid chromatography (DHPLC), a recent and automatic screening technology. Analysis of the TP53 mutations profile showed that the detected alterations were mainly point mutations. Ninety-seven percent (33/34) of esophageal squamous cell carcinoma samples presented at least one mutation or polymorphism. The proportion of somatic, non-silent and sequence-confirmed mutations was 76% (26/34). The most common substitutions were G
A transitions, which could be related to nitrosamines, acetaldehyde or factors prone to producing mucosal irritation, like hot beverages. GT transversions, which were also frequently detected, could originate from benzo[a]pyrene in tobacco smoke. AT transversions were not revealed in our series, which constitutes a discordance with mutational spectra already performed in north-western France. Concerning adenocarcinoma of the esophagus and cardia, the alteration frequency was 69% (11/16), with a majority of GA transitions at CpG dinucleotides. They are probably related to endogenous process mediated by inflammatory diseases like gastro-esophageal reflux and Barrett’s esophagus. The main advantage provided by DHPLC was its ease of application. However, the optimization steps turned out to be quite critical, especially for sequences with high melting temperatures embedded in lower melting temperature fragments. Considering only the common sequences analyzed by the two techniques, four of the 46 positive samples detected by DGGE were not revealed by DHPLC. This result stresses the limited sensitivity of DHPLC compared with DGGE under the conditions described in this study. |
Introduction |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
In 2000, esophageal cancer was the eighth most common cancer in the world by incidence (Parkin, 2001
). Its main epidemiological feature is its large geographical variability, attributed to various environmental factors. As regards Europe, the highest risk area is located in north-western France. In western countries, tobacco and alcohol are the leading etiological agents for squamous cell carcinomas (SCC). Barrett’s esophagus and gastro-esophagus reflux disease (GERD) patients are prone to the less frequent but increasing adenocarcinoma (ADC) of the esophagus. These cancers of the lower esophagus share clinical, epidemiological and molecular characteristics with adenocarcinoma of the gastric cardia. Some authors therefore make the assumption that these two adenocarcinomas may constitute the same disease (Gleeson et al., 1998; Dolan et al., 1999). Concerning genetic modifications, various kinds of damage have been found in esophageal tumors, including gene deletions (Rb), gene amplifications (cyclin D1, c-myc, EGFR, etc.) or allelic loss. However, one of the most common alterations is TP53 mutations, particularly in SCC, where they constitute early genetic lesions in the tumor development process (Montesano and Hainaut, 1998).
The tumor suppressor gene TP53 plays a key role in carcinogenesis. The activation of p53 protein by genotoxic or non-genotoxic stress is implicated in cell cycle control, apoptosis and DNA repair through a multiple transcription factor activity and protein–protein interactions. The relevance of research concerning TP53 is also linked to the more than 50% of human malignancies showing TP53 alterations (Vogelstein and Kinzler, 1992). Among these, somatic mutations are known to show a very diverse pattern, including various transitions, transversions, deletions or insertions, with a majority of the 393 codons being affected. Some of these alterations could reflect exposure to genotoxic compounds which are known to leave characteristic mutations in TP53, as has been noted in liver, lung and skin cancers (Semenza and Weasel, 1997). Thus, the dietary carcinogen aflatoxin B1 induces GT transversions at codon 249, found with a high prevalence in hepatocellular carcinoma. GT transversions at hot-spots, including codons 157, 248 and 273, have also often been reported in lung tumors. There is growing evidence that these mutations arise from polycyclic aromatic hydrocarbons contained in tobacco smoke. Finally, tandem CCTT mutations retrieved in SCC and basal cell carcinoma of the skin are considered the signature of UV radiation. One of the purposes of our team is to establish such relationships in esophageal cancers in order to characterize carcinogenic factors. The first step of this approach is to identify the most frequent modifications in the TP53 sequence in patients suffering from esophageal cancer or adenocarcinoma of the cardia.
The construction of a TP53 mutational spectrum from human tumor specimens requires a fast and reliable screening tool to identify mutated samples. The most common methods are single-strand conformation polymorphism (SSCP), denaturing gradient gel electrophoresis (DGGE) and direct sequencing. A fourth technique, denaturing high performance liquid chromatography (DHPLC), appeared in the mid 1990s (Oefner and Underhill, 1995). This automated device seems to offer improved performance over former methods regarding sensitivity, ease of use and sample throughput. Its principle is based on reversed phase and ion pair chromatography. During their migration in a heated column, DNA double strands are partially denatured, which lowers their retention. This principle enables the distinction of heteroduplexes from homoduplexes because of distinct denaturation behaviors linked to mismatches. Few studies have been published on the use of DHPLC to detect TP53 mutations. To our knowledge, this work is the first report comparing a well-established technique (DGGE) with DHPLC used to screen samples for TP53 mutations. This comparison has been carried out on 53 tumor samples (esophagus and cardia) from Lower Normandy (north-western France). Besides technical data, we will therefore be able to describe and comment on the mutation pattern obtained in the highest risk area in Europe.
|
Materials and methods |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
Patients
The 53 consenting patients were recruited between 1996 and 2000 within the context of a case–control study intended for the identification of various biomarkers suitable for molecular epidemiology of esophageal cancers. This project has been approved by the local ethical board, Comité Consultatif pour la Protection des Personnes en Recherche Biomédicale. Diagnosis was performed by the Hepato-gastroenterology Department of Caen University Hospital and by the Anatomo-pathology Department of the Françiois Baclesse Center. The study included 34 cases with SCC, eight with ADC of the esophagus, eight with ADC of the cardia, one with small cell carcinoma of the esophagus and two undifferentiated forms. Siewert criteria were adopted for the delicate distinction between ADC of esophagus and gastric cardia (Siewert and Stein, 1996
).
Biological samples
Tumor samples were collected at the Caen University Hospital and Françiois Baclesse Center from patients who underwent biopsy or esophagectomy. Carcinoma specimens were obtained before radio- or chemotherapy. Two patients (54 and 199) had undergone radiotherapy for primary head and neck cancers, respectively, 9 and 4 years previous to esophagus cancer diagnosis. Blood samples (20 ml) were collected and white blood cells were isolated using Unisep Maxi tubes (Novamed, Jerusalem, Israel). Biological samples were stored in liquid nitrogen until genomic DNA isolation. This step was performed by proteinase K digestion and phenol/chloroform extraction. DNA concentrations ranged between 50 and 1000 µg/ml.
Polymerase chain reactions (PCR)
PCR were carried out in a volume of 25 µl containing 1.5 µl of genomic DNA solution, 2 mM MgCl2, 2.5 µl of GeneAmp 10x PCR buffer II (Applied Biosystems, Courtaboeuf, France), 200 µM each deoxynucleotide triphosphate, 800 nM each primer (ESGS, Evry, France), which allowed examination of the exon–intron borders, and 0.5 U AmpliTaq Gold DNA polymerase (Applied Biosystems). PCR were performed in a PTC 100 thermocycler (MJ Research, Watertown, MA).
PCR preceding DGGE
Primer design and cycling conditions have been previously described (Hamelin et al., 1993). The conditions were carefully chosen to detect genetic changes whatever their location throughout exons 5–8. This was particularly delicate for exon 5, which contains a strong proportion of G and C nucleotides forming a high temperature melting domain. This difficulty has been resolved by analyzing the 5' (proximal) and 3' (distal) regions of this exon separately. The last cycle was followed by additional denaturation and annealing steps to enable heteroduplex formation.
PCR preceding DHPLC
Primers for exons 5/6, the distal part of exon 5, exon 7 forward and exons 8/9 forward were the same as those used for DGGE (without GC clamps). The exon 7 reverse and exons 8/9 reverse primers have already been reported by Gross et al. (2001) (Table I). Cycling conditions were the same as for DGGE without the final two steps. After amplification, half of the PCR product was mixed with normal amplified DNA, which had been confirmed to be free of mutation by sequencing. Mixed and unmixed amplified DNA samples were submitted to a heteroduplex formation step: heating at 95°C for 30 s followed by cooling down to 58°C with 30 s for each temperature.
|
DGGE
GC-clamped PCR products were mixed with loading buffer (1/1 v/v) and 15 µl of the resulting solution was loaded onto a 6% polyacrylamide gel containing a chemical denaturing gradient as described in Table II
. The gels were run at 60°C in TAE buffer (40 mM Tris, 20 mM acetate, 1 mM EDTA, pH 8). Electrophoresis conditions are indicated in Table II. After running, the gels were stained with ethidium bromide and visualized with a UV transilluminator. Positive and negative controls were included in each electrophoresis run.
|
DHPLC
The DHPLC we used was the Helix system (Varian, Les Ulis, France). Initial column temperatures (Ti) were determined using the Stanford DHPLC Melt program (www.insertions.stanford.edu/melt.htm). These temperatures were further optimized as recommended in the supplier’s instructions: the retention time (tr) of each fragment was determined while the temperature was increased from Ti -5°C to Ti +5°C. The chosen temperatures were the point at which tr corresponded to 75% of (trmax - trmin). This empirical determination did not allow the identification of all mutations previously found by DGGE. Other conditions were therefore tested for these particular cases. The selected column temperatures are described in Table I
. Buffer A (Varian) comprised 100 mM triethylammonium acetate (TEAA), 0.1 mM EDTA (pH 7) while the composition of buffer B (Varian) was 100 mM TEAA, 0.1 mM EDTA, 25% (v/v) acetonitrile (pH 7). Runs were performed at a flow rate of 0.45 ml/min with the buffer gradients described in Table I. DNA was detected by UV absorbance at 260 nm. Analysis lasted 10 min. Two runs at two temperatures were necessary to ensure sufficient sensitivity. The selectivity of the system was evaluated before each run series by injecting a pUC 18 HaeIII digest (Sigma-Aldrich, Saint-Quentin Fallavier, France) and a mutated DNA sample used as a positive control. A negative control from peripheral blood mononuclear cells of a healthy subject was also introduced to compare chromatograms with a known non-mutated profile (verified by sequencing). For each DNA fragment we injected ‘pure’ PCR products in order to duplicate the conditions for DGGE and PCR products mixed with amplified wild-type DNA samples. This approach allows the detection of homozygous polymorphisms and mutations in tumors without any contaminant non-mutated cells or with loss of heterozygosity (LOH) affecting the TP53 locus.
Sequencing
Mutated exons detected by DGGE and/or DHPLC were amplified. PCR products were loaded onto a 2% agar gel containing ethidium bromide. After electrophoresis in TAE buffer (15 min, 100 V), DNA fragments excised from the gel were isolated on Ultrafree DA columns (Millipore, Saint-Quentin Fallavier, France). Then, samples were purified using Microcon YM 30 centrifugal filter devices (Millipore). Sequencing was performed on both strands using the ABI Prism BigDye Terminator Ready Reaction Cycle Sequencing Kit (Applied Biosystems) on an ABI Prism 377 DNA Sequencer (Applied Biosystems). When direct sequencing did not allow us to characterize the mutation, mutated DNA fragments were excised from a new DGGE gel, amplified by PCR and sequenced.
Statistical analysis
Fisher’s exact test was used to test the difference in mutation rates between SCC and ADC.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
Performances of the screening methods
Determination of the TP53 mutational spectra was carried out by DGGE and DHPLC. Mutated samples detected by these screening methods were then sequenced. Examples of signals obtained during the experiments are shown in Figure 1
. DGGE detected 46 tumors containing at least one mutation or polymorphism. DHPLC was then evaluated with temperature conditions empirically determined by the Stanford DHPLC Melt program and time of retention-related criteria. The first attempts explored exons 5/6, exon 7 and exons 8/9. A comparison with the DGGE results showed a lower sensitivity of DHPLC with this first approach. The following optimization steps improved the method by increasing the denaturation temperatures and modifying some primer sets. This was the case, for example, for exons 5/6: 20 mutations were found with DHPLC whereas 28 were detected by DGGE. Among the eight undetected mutations, the six that had been sequenced were all in the GC-rich domain of exon 5. Changes in temperature were not conclusive in this case and it was therefore decided to screen the proximal and distal parts of exon 5 separately, as was already done by DGGE. Finally, the simultaneous study of exons 5/6 and 5d fragments permitted recovery of all DGGE positive samples.
|
The chromatogram profile varied according to the type of mutation. Deletions always appeared as two well-resolved peaks. For single substitutions, our protocol did not succeed in showing the four theoretically distincts peaks resulting from the hybridization of mutated and non-mutated alleles every time. DHPLC did not allow isolation of one mutation in exon 7 and three in exon 8. For this fragment, the choice of new primers did not improve the result (data not shown). DHPLC detected five genetic changes (three in intron 7, one in exon 9 and one in intron 9) undetermined by DGGE because their loci were not contained in the fragments studied by this technique. It is worth noting that the analysis of DNA with added non-mutated sequences did not reveal new mutations, suggesting that the number of non-mutated cells in tumor samples was sufficient to form heteroduplexes during PCR. A lot of mutations were at first undetected by sequencing using genomic DNA solutions despite a positive result with the screening methods. This was probably due to the fact that sequencing is less sensitive. To sequence these mutations, mutated fragments had been excised from DGGE gels. Despite this modification, eight mutations were not found, mainly due to poor quality DNA amplification. For these samples, DGGE and DHPLC were performed at least in triplicate, showing a positive result every time. This seemed to exclude false positives in DGGE and DHPLC. It should be pointed out that the sequencing of intron 7 systematically showed (even with normal DNA controls) divergence from the GenBank TP53 sequence (accession no. X54156): nucleotide 14168 is a T in our samples instead of a G.
TP53 mutations in tumors of esophagus and cardia
DGGE and DHPLC for exons 5–9 revealed 47 abnormal samples out of 53 patients (89%) (Table III). Ninety-seven percent of SCC (33/34) showed anomalous bands or chromatograms while the proportion was 69% for ADC (11/16), with 87.5 (7/8) and 50% (4/8) for esophagus and cardia, respectively. Sequencing enabled us to characterize 42 mutations and four polymorphisms. The tumors with somatic, non-silent and sequence-confirmed TP53 mutations were 76 (26/34) and 56% (9/16) for SCC and ADC, respectively. These levels were 89% (26/29) for SCC and 64% (9/14) for ADC when samples in which sequencing did not allow confirmation of mutations were not taken into account. Seven cases harbored two mutations or one mutation and one polymorphism. Table IV summarizes the types of changes detected in SCC and ADC. Substitutions were more common, with 86% (25/29) for SCC and 78% (7/9) for ADC. Regarding frameshift alterations, no insertions were found.
|
|
As for SCC, G
A transitions were more prevalent mutations (9/29, 31%), followed by GT transversions (8/29, 28%). AT transversions did not appear. Concerning ADC, there was a majority of GA transitions (6/9, 67%), all occurring at CpG dinucleotides. The small cell carcinoma tumor contained two nonsense point mutations in exon 5. Regarding locations, and taking into account all tumor types, 17 changes were in exon 5 (33%), nine in exon 6 (20%), six in exon 7 (13%), 11 in exon 8 (22%), one in exon 9 (2%) and five in introns (splice sites) (11%). For four samples, DGGE and DHPLC showed a positive result for exons 5/6 but sequencing did not permit localization of the mutations to a particular exon. Among SCC cases, single base substitutions were found in codons 175 (two cases), 179 (two cases), 220 and 273, which have already been characterized as hot-spots in European cases. Concerning ADC, three of the seven point mutations were at codon 248 with one each at codons 175, 273 and 282, all of which are frequently affected in this tumor type.
One patient (1.9%) was homozygous for a CGACGG polymorphism at codon 213 and three cases (5.7%) were heterozygous for concomitant CT and TG polymorphisms localized at nucleotides 14181 and 14201, respectively (intron 7). Sequencing of white blood cell DNA verified these polymorphisms.
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
Comparison of DGGE and DHPLC
Our objective was to evaluate DHPLC as a method for TP53 mutation screening. We examined 53 tumor samples by a well-characterized technique (DGGE) and then using the Helix DNA system (Varian). DGGE is based on the migration of double-stranded DNA in a polyacrylamide gel including a chemical denaturing gradient. The denaturation process increases the volume of the molecule, which is progressively immobilized in the gel. Mismatches in the duplex modify its stability and therefore create shifts with non-mutated fragments. When particular care is taken with various melting temperature domains in the analyzed sequence, DGGE shows sensitivities close to 100%. This constitutes an advantage over another commonly used technique, SSCP, for which the highest sensitivities reported in the literature using basic protocols rarely exceed 90–95% for TP53 analysis (Moyret et al., 1994
; Kourkine et al., 2002).
Validation studies evaluating DHPLC sensitivity and specificity with known mutated genes report performances close to 100% for both criteria (Xiao and Oefner, 2001). The main difficulty encountered with TP53 is the number of mutations likely to be detected in a tumor sample. Each of these mutations could need its own analysis conditions, according to its particular type or localization. Few studies related to DHPLC and TP53 mutations have been reported at this time. Gross et al.(2001) screened exons 2–11 in ovarian tumors of unknown genetic status. Keller et al.(2001) examined colorectal carcinomas (exons 5–8) and a third report used DHPLC to look for TP53 mutations (exons 5–8) in leukemia and myelodysplastic syndrome consecutive to ovarian cancer treatment (Leonard et al., 2002). None of these studies were performed with the Varian system but with the Wave DNA Fragment Analysis System (Transgenomic, San Jose, USA).
Our experience showed a need for known mutated DNA samples to finalize DHPLC conditions. Indeed, data obtained by DGGE allowed us to note the lack of sensitivity provided by the conditions initially determined with the melting temperature program. As already pointed out, the recommended temperatures are sometimes too low (Xiao and Oefner, 2001). The most critical domains to analyze are high melting temperature sequences inserted in regions melting at lower temperatures. This drawback can be illustrated by the exons 5/6 and 8/9 analyses (Figure 2). In exons 5/6, mutations not detected during our first DHPLC attempt were all localized in the highest melting temperature domain. The DGGE protocol circumvented this difficulty by using GC clamps and shortening the DNA fragment (exons 5p and 5d). Applied to DHPLC, this second approach proved to be efficient. That was not the case for exons 8/9, as the new primer sets did not permit detection of mutations at codons 281 and 282. This drawback is reinforced by the fact that codon 282 is a mutational hot-spot. We note here that among the 44 mutations detected in the three reports previously mentioned, none were localized at these positions. Finally, regarding common areas analyzed by the two techniques, DHPLC missed four abnormalities discovered by the preceding DGGE analysis (one in exon 7 and three in exon 8). Another weakness to be underlined is the progressive loss of stability of the stationary phase: in our laboratory, columns were not used for more than 1300 injections. To evaluate this phenomenon, a pUC 18 HaeIII digest and a known mutated DNA sample had to be injected before each analysis series.
|
These limitations should not mask the benefits provided by DHPLC. We should first note its ease of application. After the introduction of unpurified PCR product into the autosampler, the process is completely automatic. Secondly, expensive GC clamps do not have to be added to the PCR primers. Moreover, the technique shows a throughput compatible with molecular epidemiology studies and does not require handling of hazardous chemicals like acrylamide or ethidium bromide. Some studies also reported an attractive limit of detection provided by DHPLC. Using dilutions of a TP53 polymorphism in exon 6 and of a transition at codon 273, Keller et al.(2001)
demonstrated that a proportion of 10% of one allele is sufficient to generate a positive signal. Experiments described by Leonard et al.(2002) lowered this limit to 3.25% mutant allele diluted in wild-type sequences. These works show the limited usefulness of adding wild-type DNA before runs. Although this mixture of analyzed and normal DNA did not allow the detection of new mutations in our experiments, we consider that the need to detect homozygous polymorphisms or potential mutations in tumors, including LOH, makes it worthwhile to apply such a protocol.
It can be underlined that sequencing did not succeed in characterizing some mutations previously revealed by DGGE and DHPLC. This lack of sensitivity has already been demonstrated versus DGGE (Trulzsch et al., 1999) and DHPLC (Jones et al., 2001). This drawback is especially true when somatic mutations in low amounts have to be detected in tumor DNA samples.
Interpretation of mutation patterns
Concerning SCC, it must be remembered that two patients (nos 54 and 199) underwent radiotherapy for former head and neck cancer. As radiation is implicated in TP53 mutations, these cases have not been taken into consideration in linking mutation types to environmental factors. Twenty-six of the 28 tumor samples successfully sequenced (93%) from patients never subjected to radiotherapy revealed at least one mutation (silent mutations included). One sample contained no mutation and one had only a polymorphism. We could assume that the screening of all 11 TP53 exons and the application of additional screening techniques could lead to mutation rates close to 100%. This mutation frequency obtained by sequencing exons 1–11 has already been achieved in head and neck squamous cell cancers, which share common etiological factors with esophageal SCC (Kropveld et al., 1999).
The high rate found in our work confirms previous reports showing proportions exceeding 80% of SCC bearing TP53 mutations in north-western France, especially in Normandy and Brittany (Hollstein et al., 1991; Audrézet et al., 1993; Robert et al., 2000). The origin of the high SCC incidence rates observed in these two French regions remains unclear. However, epidemiological studies have disclosed that traditional hot spirit (hot Calvados) and high butter consumption seem to explain at least part of the increased incidence in Normandy (Launoy et al., 1997, 1998).
In France, the percentage of cancers with mutations correlates with the striking variation in incidence within the country. Indeed, in tumors collected in Lyon (a low risk region), TP53 mutations were detected in only 36% of cases (Tanière et al., 2001). In comparison with other high incidence areas throughout the world, the frequencies of mutations found in Lower Normandy are among the highest. Maximum rates recently measured in China did not exceed 77% (Hu et al., 2001), whereas the proportion of TP53 mutations is 35% in southern Brazil (Putz et al., 2002) and 65% in northern Iran (Biramijamal et al., 2001). If we overlook possible discrepancies between sensitivities of various screening methods, these variations can be attributed to environmental etiological factors varying between countries or even between distinct regions within a country. This notion is strengthened by studying the pattern of mutations (Figure 3). In our samples, GA transitions are among the more common point mutations (29%), as already recorded in the International Agency for Research on Cancer (IARC) database (25% of the point mutations for Western European populations; www.iarc.fr/p53/) (Olivier et al., 2002). These changes, especially at CpG dinucleotides, are linked to endogenous mutagenesis. Some authors suggest that the formation of endogenous nitrogen reactive species (nitric oxide) is enhanced by mucosal irritation caused by scalding beverages. This drinking habit could explain GA transitions at CpG found in SCC in southern Brazil (Putz et al., 2002), northern Iran (Biramijamal et al., 2001) and in our report. GA transitions could also be related to exogenous N-nitroso compounds. Nitrosamines or their precursors are contained in tobacco smoke and food (for example in pickled vegetables consumed in China). They are known to produce O6-methylguanine adducts which could initiate the GA transitions retrieved in rat esophageal tumors induced by N-nitrosomethylbenzylamine (Wang et al., 1996). Acetaldehyde has also been associated with GA changes in the HPRT reporter gene (Noori and Hou, 2001). This primary metabolite of ethanol could also be implicated in mutations occurring at A:T base pairs. Once again, N-nitrosamines in cigarette smoke could constitute another lead to explore for this type of mutation (Hollstein et al., 1991). Changes at A:T base pairs constituted only 21% of our spectrum, with a majority of AG transitions, whereas IARC data indicate 40% of mutations at A:T base pairs with many AT changes (17%). Such transversions have not been detected in our work, which is inconsistent with mutation patterns already elaborated in Normandy and Brittany (Hollstein et al., 1991; Audrézet et al., 1993; Robert et al., 2000). This discrepancy could be due to the inability of our protocol to detect these alterations. However, one has been retrieved in a histologically undifferentiated case. We can also assume that the patients recruited had not been exposed or were not susceptible to etiological factors producing AT transversions. A third type of change should be noted. GT transversions constitute 18% of IARC recorded mutations and 29% of the pattern in the present report. In lung cancers, these very common mutations often occur at codons known to constitute a target for bulky DNA adducts originating from polycyclic aromatic hydrocarbons (Smith et al., 2000). Among these, benzo[a]pyrene contained in tobacco smoke is most often blamed. It must be added that a strand bias appears when the spectrum of lung cancer patients exposed to tobacco smoke is examined: GT transversions with G on the non-transcribed strand are much more frequent than CA transversions. Despite the small number of cases in our study, the same distribution is observed, with only one CA mutation versus seven GT transversions (Figure 3). This imbalance has been attributed to slow repair of bulky adducts along the non-transcribed strand of DNA (Denissenko et al., 1998). However, if the link between benzo[a]pyrene and these TP53 mutations seems to be clearly established for lung tumors, this type of relationship remains to be proved concerning esophageal cancers. Another hypothesis is that the mutations could derive from oxidative stress and 8-hydroxy-2'-deoxyguanosine adducts, which are able to generate GT transversions (Marnett, 2000).
|
If we examine the localization of point mutations, we can note potential repercussions on p53 functionality. Our work mainly explored TP53 regions encoding the central DNA-binding domain, which constitutes the most mutated fragment. Two mutations affecting codons 175 and 179 were detected. They are located in a highly conserved region of the gene. Residue H179 is one of the four amino acids implicated in the connection with a zinc atom which stabilizes the structure of the protein. R175H is an oncogenic mutation also related to the stability of p53 by modifying the conformation of the L2 loop (Sigal and Rotter, 2000
). Another cluster of point mutations (six in our series) can be observed in a conserved region situated in exon 8 (codons 270–286). Among these, the R273H mutation affects a residue directly in contact with DNA.
Mutations were less frequent in ADC cases in comparison with SCC proportions in our high incidence area (64% versus 93%, P = 0.031, Fisher’s exact test). Fifty percent of cardia cancer cases showed TP53 modifications whereas the percentage reached 87.5% for esophageal tumors. This difference is difficult to interpret because of the limited number of cases (n = 8 for each cancer type). Data available in the literature indicate rates of 42 (n = 50) (Fléjou et al., 1999), 43 (n = 14) (Liang et al., 1995) and 63% (n = 41) (Gleeson et al., 1998) for cardia cancers. Concerning ADC of esophagus, the proportions are between 44 (n = 59) (Schneider et al., 2000) and 85% (n = 27) (Muzeau et al., 1996). The assumption that ADC of cardia and esophagus constitute the same cancer remains controversial. Disagreement exists concerning TP53 mutations: Montesano and Hainaut (1998) observed that these alterations were less common in cardia (25%) than in esophagus (71%) whereas Gleeson et al.(1998) and Ireland et al.(2000) found a similar pattern for the two localizations. Such discrepancies could be explained by the various criteria used to distinguish the two forms (anatomical location, clinical symptoms, histology) or the various protocols used to screen mutations. This research area is of particular interest because of the significant increase in these adenocarcinomas in some developed countries (Botterweck et al., 2000), which needs to be better understood from an epidemiological and molecular point of view.
Regarding types of mutations, GA substitutions at CpG sites were the most commonly detected (67%), as already found in studies of adenocarcinomas. They constitute 27 of the 63 mutations (42.8%) recorded in the IARC mutations database R6 for Western European populations. These CpG mutations are often associated with cancers poorly related to exogenous carcinogenic compounds, resulting in particular from spontaneous deamination of 5-methylcytosine. This endogenous mechanism could be favored by the production of nitrogen oxides during chronic inflammation processes (Ambs et al., 1997). This hypothesis is consistent with the implication of GERD and Barrett’s esophagus as risk factors for adenocarcinomas.
Of the seven mutations sequenced, one each were observed at codons 175, 273 and 282 and three at codon 248. These four localizations belong to the six hot-spots common for all cancers (175, 213, 245, 248, 273 and 282) and play a major role in p53 functionality. Residue 248 is particularly involved in direct contact with DNA.
In summary, DHPLC can be considered as a technique well suited to studies of a large number of samples because of its throughput and the minimal handling needed. However, its application to the detection of TP53 mutations can require delicate optimization steps due to the heterogeneity of alterations affecting this gene. The analysis of esophageal and cardia tumors confirmed the good level of sensitivity of our protocols and verified the diverse patterns and high frequency of TP53 mutations in these cancers. The next step to improve understanding of esophageal carcinogenesis should be to link specific mutations to etiological factors. This aim could be achieved by in vitro experiments or by molecular epidemiology studies associating mutation profiles with exposition data collected in large patient series.
|
Acknowledgments |
|---|
We thank S. Daon, the Laboratoire de Biochimie Clinique et Oncologique (Centre Françiois Baclesse) and the Département Génétique et Reproduction (Hôpital Clémenceau, Caen) for their technical assistance. We are grateful to surgeons of the Hepato-gastroenterology Department at the Caen University Hospital. We also thank Dr G.Launoy and M.Ingouf (Registre des Tumeurs Digestives du Calvados, Caen) for their participation in the recruitment of patients. The DHPLC system was acquired through a grant from the Association pour la Recherche sur le Cancer and with financial support received from the Contrat de Plan Etat-Région (Pôle de Recherche en Biologie Médicale et en Epidémiologie–Université de Caen). The work of the authors is supported by the Ligue Nationale Contre le Cancer (Comités de l’Orne et de la Manche). J.B. was the recipient of a fellowship from the Association C
ur et Cancer (Cherbourg-Octeville) and from the Conseil Régional de Basse-Normandie. |
Notes |
|---|
3 To whom correspondence should be addressed. Tel: +33 2 31 45 50 70; Fax: +33 2 31 45 51 72; Email: j.breton{at}baclesse.fr
|
References |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
-
Ambs,S., Hussain,S.P. and Harris,C.C. (1997) Interactive effects of nitric oxide and the p53 tumor suppressor gene in carcinogenesis and tumor progression. FASEB J., 11, 443–448.[Abstract]
Audrézet,M.P., Robaszkiewicz,M., Mercier,B., Nousbaum,J.B., Bail,J.P., Hardy,E., Volant,A., Lozac’h,P., Charles,J.F. and Gouérou,H. (1993) TP53 gene mutation profile in esophageal squamous cell carcinomas. Cancer Res., 53, 5745–5749.
Biramijamal,F., Allameh,A., Mirbod,P., Groene,H.J., Koomagi,R. and Hollstein,M. (2001) Unusual profile and high prevalence of p53 mutations in esophageal squamous cell carcinomas from northern Iran. Cancer Res., 61, 3119–3123.
Botterweck,A.A., Schouten,L.J., Volovics,A., Dorant,E. and van den Brandt,P.A. (2000) Trends in incidence of adenocarcinoma of the oesophagus and gastric cardia in ten European countries. Int. J. Epidemiol., 29, 645–654.
Denissenko,M.F., Pao,A., Pfeifer,G.P. and Tang,M. (1998) Slow repair of bulky DNA adducts along the non-transcribed strand of the human p53 gene may explain the strand bias of transversion mutations in cancers. Oncogene, 16, 1241–1247.[CrossRef][ISI][Medline]
Dolan,K., Sutton,R., Walker,S.J., Morris,A.I., Campbell,F. and Williams,E.M. (1999) New classification of oesophageal and gastric carcinomas derived from changing patterns in epidemiology. Br. J. Cancer, 80, 834–842.[CrossRef][ISI][Medline]
Fléjou,J.F., Gratio,V., Muzeau,F. and Hamelin,R. (1999) p53 abnormalities in adenocarcinoma of the gastric cardia and antrum. Mol. Pathol., 52, 263–268.[Abstract]
Gleeson,C.M., Sloan,J.M., McManus,D.T., Maxwell,P., Arthur,K., McGuigan,J.A., Ritchie,A.J. and Russell,S.E. (1998) Comparison of p53 and DNA content abnormalities in adenocarcinoma of the oesophagus and gastric cardia. Br. J. Cancer, 77, 277–286.[ISI][Medline]
Gross,E., Kiechle,M. and Arnold,N. (2001) Mutation analysis of p53 in ovarian tumors by DHPLC. J. Biochem. Biophys. Methods, 47, 73–81.[CrossRef][ISI][Medline]
Hamelin,R., Jego,N., Laurent-Puig,P., Vidaud,M. and Thomas,G. (1993) Efficient screening of p53 mutations by denaturing gradient gel electrophoresis in colorectal tumors. Oncogene, 8, 2213–2220.[ISI][Medline]
Hollstein,M.C., Peri,L., Mandard,A.M., Welsh,J.A., Montesano,R., Metcalf,R.A., Bak,M. and Harris,C.C. (1991) Genetic analysis of human esophageal tumors from two high incidence geographic areas: frequent p53 base substitutions and absence of ras mutations. Cancer Res., 51, 4102–4106.
Hu,N., Huang,J., Emmert-Buck,M.R. et al. (2001) Frequent inactivation of the TP53 gene in esophageal squamous cell carcinoma from a high-risk population in China. Clin. Cancer Res., 7, 883–891.
Ireland,A.P., Shibata,D.K., Chandrasoma,P., Lord,R.V., Peters,J.H. and DeMeester,T.R. (2000) Clinical significance of p53 mutations in adenocarcinoma of the esophagus and cardia. Ann. Surg., 231, 179–187.[CrossRef][ISI][Medline]
Jones,A.C., Sampson,J.R. and Cheadle,J.P. (2001) Low level mosaicism detectable by DHPLC but not by direct sequencing. Hum. Mutat., 17, 233–234.[CrossRef][ISI][Medline]
Keller,G., Hartmann,A., Mueller,J. and Hofler,H. (2001) Denaturing high pressure liquid chromatography (DHPLC) for the analysis of somatic p53 mutations. Lab. Invest., 81, 1735–1737.[ISI][Medline]
Kourkine,I.V., Hestekin,C.N., Buchholz,B.A. and Barron,A.E. (2002) High-throughput, high-sensitivity genetic mutation detection by tandem single-strand conformation polymorphism/heteroduplex analysis capillary array electrophoresis. Anal. Chem., 74, 2565–2572.[Medline]
Kropveld,A., Rozemuller,E.H., Leppers,F.G., Scheidel,K.C., de Weger,R.A., Koole,R., Hordijk,G.J., Slootweg,P.J. and Tilanus,M.G. (1999) Sequencing analysis of RNA and DNA of exons 1 through 11 shows p53 gene alterations to be present in almost 100% of head and neck squamous cell cancers. Lab. Invest., 79, 347–353.[ISI][Medline]
Launoy,G., Milan,C., Day,N.E., Faivre,J., Pienkowski,P. and Gignoux,M. (1997) Oesophageal cancer in France: potential importance of hot alcoholic drinks. Int. J. Cancer, 71, 917–923.[CrossRef][ISI][Medline]
Launoy,G., Milan,C., Day,N.E., Pienkowski,M.P., Gignoux,M. and Faivre,J. (1998) Diet and squamous-cell cancer of the oesophagus: a French multicentre case-control study. Int. J. Cancer, 76, 7–12.[CrossRef][ISI][Medline]
Leonard,D.G., Travis,L.B., Addya,K. et al. (2002) p53 mutations in leukemia and myelodysplastic syndrome after ovarian cancer. Clin. Cancer Res., 8, 973–985.
Liang,Y.Y., Esteve,A., Martel-Planche,G., Takahashi,S., Lu,S.H., Montesano,R. and Hollstein,M. (1995) p53 mutations in esophageal tumors from high-incidence areas of China. Int. J. Cancer, 61, 611–614.[ISI][Medline]
Marnett,L.J. (2000) Oxyradicals and DNA damage. Carcinogenesis, 21, 361–370.
Montesano,R. and Hainaut,P. (1998) Molecular precursor lesions in oesophageal cancer. Cancer Surv., 32, 53–68.[ISI][Medline]
Moyret,C., Theillet,C., Puig,P.L., Moles,J.P., Thomas,G. and Hamelin,R. (1994) Relative efficiency of denaturing gradient gel electrophoresis and single strand conformation polymorphism in the detection of mutations in exons 5 to 8 of the p53 gene. Oncogene, 9, 1739–1743.[ISI][Medline]
Muzeau,F., Fléjou,J.F., Potet,F., Belghiti,J., Thomas,G. and Hamelin,R. (1996) Profil des mutations du gène p53 et expression anormale de la protéine p53 dans les deux formes de cancer de l’oesophage. Gastroenterol. Clin. Biol., 20, 430–437.[ISI][Medline]
Noori,P. and Hou,S.M. (2001) Mutational spectrum induced by acetaldehyde in the HPRT gene of human T lymphocytes resembles that in the p53 gene of esophageal cancers. Carcinogenesis, 22, 1825–1830.
Oefner,P.J. and Underhill,P.A. (1995) Comparative DNA sequencing by denaturing high-performance liquid chromatography. Am. J. Hum. Genet., 57, A266.
Olivier,M., Eeles,R., Hollstein,M., Khan,M.A., Harris,C.C. and Hainaut,P. (2002) The IARC TP53 database: new online mutation analysis and recommendations to users. Hum. Mutat., 19, 607–614.[CrossRef][ISI][Medline]
Parkin,D.M. (2001) Global cancer statistics in the year 2000. Lancet Oncol., 2, 533–543.[CrossRef][ISI][Medline]
Putz,A., Hartmann,A.A., Fontes,P.R., Alexandre,C.O., Silveira,D.A., Klug,S.J. and Rabes,H.M. (2002) TP53 mutation pattern of esophageal squamous cell carcinomas in a high risk area (Southern Brazil): role of life style factors. Int. J. Cancer, 98, 99–105.[CrossRef][ISI][Medline]
Robert,V., Michel,P., Flaman,J.M., Chiron,A., Martin,C., Charbonnier,F., Paillot,B. and Frebourg,T. (2000) High frequency in esophageal cancers of p53 alterations inactivating the regulation of genes involved in cell cycle and apoptosis. Carcinogenesis, 21, 563–565.
Schneider,P.M., Stoeltzing,O., Roth,J.A., Hoelscher,A.H., Wegerer,S., Mizumoto,S., Becker,K., Dittler,H.J., Fink,U. and Siewert,J.R. (2000) P53 mutational status improves estimation of prognosis in patients with curatively resected adenocarcinoma in Barrett’s esophagus. Clin. Cancer Res., 6, 3153–3158.
Semenza,J.C. and Weasel,L.H. (1997) Molecular epidemiology in environmental health: the potential of tumor suppressor gene p53 as a biomarker. Environ. Health Perspect., 105 (suppl. 1), 155–163.[Medline]
Siewert,J.R. and Stein,H.J. (1996) Carcinoma of the cardia: carcinoma of the gastroesophageal junction—classification, pathology and extent of resection. Dis. Esophagus, 9, 173–182.
Sigal,A. and Rotter,V. (2000) Oncogenic mutations of the p53 tumor suppressor: the demons of the guardian of the genome. Cancer Res., 60, 6788–6793.
Smith,L.E., Denissenko,M.F., Bennett,W.P., Li,H., Amin,S., Tang,M. and Pfeifer,G.P. (2000) Targeting of lung cancer mutational hotspots by polycyclic aromatic hydrocarbons. J. Natl Cancer Inst., 92, 803–811.
Tanière,P., Martel-Planche,G., Saurin,J.C., Lombard-Bohas,C., Berger,F., Scoazec,J.Y. and Hainaut,P. (2001) TP53 mutations, amplification of P63 and expression of cell cycle proteins in squamous cell carcinoma of the oesophagus from a low incidence area in western Europe. Br. J. Cancer, 85, 721–726.[CrossRef][ISI][Medline]
Trulzsch,B., Krohn,K., Wonerow,P. and Paschke,R. (1999) DGGE is more sensitive for the detection of somatic point mutations than direct sequencing. Biotechniques, 27, 266–268.[ISI][Medline]
Vogelstein,B. and Kinzler,K.W. (1992) p53 function and dysfunction. Cell, 70, 523–526.[CrossRef][ISI][Medline]
Wang,D., Weghorst,C.M., Calvert,R.J. and Stoner,G.D. (1996) Mutation in the p53 tumor suppressor gene in rat esophageal papillomas induced by N-nitrosomethylbenzylamine. Carcinogenesis, 17, 625–630.
Xiao,W. and Oefner,P.J. (2001) Denaturing high-performance liquid chromatography: a review. Hum. Mutat., 17, 439–474.[CrossRef][ISI][Medline]
Received on October 31, 2002; revised on January 13, 2003; accepted on January 16, 2003.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  V. Loyant, A. Jaffre, J. Breton, I. Baldi, A. Vital, F. Chapon, S. Dutoit, Y. Lecluse, H. Loiseau, P. Lebailly, et al. Screening of TP53 mutations by DHPLC and sequencing in brain tumours from patients with an occupational exposure to pesticides or organic solvents Mutagenesis, September 1, 2005; 20(5): 365 - 373. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||