| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
From the Departments of Pathology,* Neuropathology, Head and Neck Surgery, Biometrics, and Immunology,¶ Otto-von-Guericke University, Magdeburg, Germany
 |
Abstract |
|---|
 Top Abstract Materials and MethodsResultsDiscussionReferences |
|---|
The role of membrane death receptor-mediated apoptosis is well-established both in vitro and in vivo9,10 . When cell-surface Fas engages with Fas ligand (FasL), several proteins are recruited to the intracellular "death domain" of Fas to form a death-inducing signaling complex. Disturbances in apoptosis mediated via the Fas/FasL system play a key role in the pathogenesis of cancer.11 A p53-responsive element was found in both human and mouse Fas death receptor genes.12,13 Mutant p53 overexpression can also trigger continuous expression of Bax and Bcl2, proteins that regulate apoptosis via the mitochondrial signal cascade.14,15 Recently, it was reported that the allele constitution at codon 72 of the p53 gene (CGC-arginine or CCC-proline) contributes to the apoptosis induction of p53 mutant cells by influencing the interaction of p53 with p73.6,16 Furthermore, Arg72 seems to be preferentially mutated, whereas Pro72 is preferentially lost in various squamous cell cancers.6,7 The relationship between codon 72 status and apoptotic regulation in human SCCHN is still unclear. Since codon 72 is localized in a proline-rich region of the p53 gene that is essential for p53-mediated apoptosis, we investigated the association between p53-LOH, p53 mutations and p53 protein expression, and the expression of three apoptosis-related proteins (Fas, FasL, and Bcl2) in a group of 54 SCCHN.
|
Materials and Methods |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
We analyzed 54 patients suffering from advanced head and neck squamous cell cancers, treated in the Department for Head and Neck Surgery of the Otto-von-Guericke University of Magdeburg. Twelve (22.2%) tumors were located in the oral cavity, 22 (40.7%) in the oropharynx and 13 (24.1%) in the hypopharynx. Five (9.3%) tumors were supraglottic, and two (3.7%) were glottic cancers. There were 47 male and seven female patients; the average age was 57 years. Twenty-seven patients had stage pT4-cancer, 13 had a pT3-tumor, and 14 had a pT2-tumor, but none of them was affected by a small stage pT1 tumor. In 21 patients, cervical lymph node metastases did not occur (pN0). Ten patients were diagnosed at stage pN1, and 22 patients at stage pN2. One patient had cervical metastasis with a size of more than 6 cm (pN3). Only two of the carcinomas were well differentiated (GI), 40 carcinomas were moderately (GII), and 12 were poorly differentiated (GIII). In the analyzed group were 45 patients with a history of alcohol and nicotine abuse. Seven patients had neither alcohol nor nicotine abuse in their history, and two were smokers only. All patients were treated by radical surgery, and 45 patients underwent postoperative radiotherapy.
DNA Extraction
Tumor DNA and blood DNA were extracted using a standard proteinase K-phenol-chloroform extraction protocol.
Exon 4-LOH (Codon 72 of the p53 Gene)
The restriction fragment length polymorphism (RFLP) in codon 72 of exon 4 of the p53 gene was determined using the LightCycler (Roche Diagnostics, Mannheim, Germany). Primers and oligonucleotide hybridization probes were chosen using the TIB MOLBIOL computer program (Berlin, Germany: www.TIBMOLBIOL.de/oligo_ag.html) to ensure their total gene specificity. Primersequences (GenBank Accession No. X54156) were sense 5'-GATGCTGTCCCCGGACGA-3' (position 12053–12070) and antisense 5'-AGGGGCCGCCGGTGTAG-3' (position 12181–12163) amplifying a 128 bp fragment. The sequence of the 5'-fluorescine-labeled probe was 5'-GATGAAGCTCCCAGAATGCCAGAGGCT-3', position 12105–12131. The LCRed 640 probe was designed to detect the C-variant (5'-TCCCCCCGTIICCCCTGCACCA-3', position 12134–12155). If G is present, the resulting mismatch destabilizes the sample probe-hybrid, and this effect is further enhanced by inserting two additional inositol bases into the probe, yielding a melting temperature that is lower in mismatch (Tm = 62.5°C) than in normal-match cases (Tm = 69°C). The different melting curve profiles (Figure 1)
allow for the exact genotyping without post-PCR handling. Fluorescence melting peaks were obtained by plotting the negative derivative of fluorescence over temperature (-dF/dT) versus temperature (T) showing melting temperatures (Tm).
|
For PCR of exon 4 of the p53-gene, we used the LightCycler-DNA Master Hybridization Probes Kit (Roche Diagnostics). The 20 µl PCR reaction contained the exon 4 primer set (10pmol each), two hybridization probes (4pmol of the fluorescine-labeled probe and 2pmol of the LCRed 640 probe), 3 mmol/L MgCl2, 1x LC-DNA Master Hybridization Probe Mix, and 100 ng DNA.
PCR conditions were as follows: a first denaturation step (95°C, 30 seconds) followed by 45 cycles of denaturation (95°C, 10 seconds), annealing (60°C, ramping 0.1, for 10 seconds), and extension (72°C, 5 seconds), one melting cycle (45°C to 80°C, 20 seconds), and a final cooling step (40°C, 30 seconds).
p53 Intron-1 Polymorphism
The intron-1 polymorphic region was amplified by using the following touch-down PCR: 95°C for 5 minutes, 20 cycles of 95°C for 10 seconds, 66°C for 10 seconds and 72°C for 20 seconds, with a stepwise decrease of 2°C in each fifth cycle, and finally 30 cycles at 95°C for 45 seconds, 58°C for 1 minute, and 72°C for 1 minute. An aliquot of the PCR product was electrophoresed on native 8% polyacrylamide gels, cross-linked with piperazine diacrylamide, and visualized by silver staining. An allelic loss was considered in cases when, in comparison with non-tumor DNA, the signal of a tumor band disappeared or signal intensity was reduced by more than 50%. Evaluation was done twice, visually, or by densitometry (VDS, Pharmacia Biotech, Uppsala, Sweden) in unambiguous cases.
Immunohistochemistry
For Fas, FasL and Bcl-2 immunohistochemistry, we used monoclonal mouse antibodies and antigen retrieval by microwave heating of Fas (clone DX3, 1:10 dilution, EDTA (pH 8.0), Dako, Hamburg, Germany), FasL (clone 5D1, 1:50 dilution, Glycabuffer (pH 3.0), Novocastra, Hamburg, Germany) and Bcl-2 (clone 124, 1:10 dilution, EDTA (pH 8.0), Dako) after inhibition of endogeneous peroxidase activity. The primary antibodies were incubated for 1 hour at 37°C. Slides were subsequently incubated with a 1:10 dilution of normal swine serum (Vector, Burlingame, CA). After washing in PBS (pH 7.4), the samples were incubated with a 1:200 dilution of biotinylated anti-mouse secondary antibody (Vector) for 30 minutes at room temperature. The detection of bound antibody was accomplished using the avidin-biotin complex method (Dianova UniTect A.B.C. System XHC1, Hamburg, Germany). A 0.1% solution of 3,3'-diaminobenzidine (5 minutes) was used as a chromogen. Specificity for immunostaining was checked by omitting single steps in the immunohistochemical protocol and by replacing primary antibody with non-immune serum. Tissue from a xenotransplant of a dedifferentiated thyroid carcinoma was used as a positive control. Estimating 10 high-power fields, we found a very low staining intensity and an inhomogeneous staining pattern if less than 20% of cells showed positivity. Therefore we considered a section immunohistochemically positive if cytoplasm was strongly positively stained (with pronounced membrane staining for Fas and Fas-L) in more than 20% of tumor cells.
Apoptosis Detection by the Terminal Deoxy-Transferase-Mediated dUTP-Biotin Nick End Labeling Method
Apoptosis was detected using the terminal deoxy-transferase-mediated dUTP nick end labeling (TUNEL) method in situ apoptosis detection kit (Roche Diagnostics) according to the manufacturer’s instructions. Briefly, five paraffin sections of each Arg, Arg/Pro, and Pro status were mounted on glass slides and treated with xylene for 5 minutes and then in 100%, 95%, or 75% ethanol. The deparaffinized tissue samples were incubated with proteinase K (2 mg/ml) at room temperature for 15 minutes. After phosphate-buffered saline (PBS) washing, endogenous peroxidase was blocked by the addition of 3% H2O2. Tissues were then treated with terminal deoxynucleotidyl transferase and biotinylated dUTP. After stopping the reaction with TB buffer (30 mmol/L sodium chloride, 30 mmol/L sodium citrate), the samples were investigated by fluorescence microscopy (Axiovert 200, Zeiss, Intermedic, Germany).
Statistics
Statistical analyses were carried out using the 
2-test or Fisher's exact test in cross tables, and one-way analysis of variance (for comparison of group means). P values less than 0.05 were considered statistically significant, and P < 0.10 was regarded as a statistical trend. All statistical tests were two-sided. Calculations were carried out by SPSS-9.0 software (SPSS, Chicago, IL).
|
Results |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
P53 mutations were identified in 16 of 54 (29.7%) tumors (Table 1)
. Representative PCR and SSCP gels are shown in Figure 1
. We found seven missense (43.8%), five deletion/insertions (frame-shifts) (33.3%), three splicing mutations (18.7%), and one stop-mutation (6.2%). Seven mutations corresponded to the transition type (five G:A at CpG islands, one C:T and one G:A outside of CpG islands) and only three were transversions (A:C, A:T, and C:A). Five mutations occurred at known hot-spot codons (Arg175, Arg248, and Arg273). To date, four exon mutations have not been described in the Hainaut p53 database for head and neck cancer (tumors 29, 30, 8, and 41).17
|
p53-LOH was detected in 19 of 42 (45.2%) advanced head and neck carcinomas (Tables 2, 3, 4, and 5)
. The intron 1-VNTR-marker showed higher informativity (42 of 42 = 100%) versus exon 4-RFLP marker (23 of 42 = 54.7%). Exon 4-LOH was found in 9 of 23 informative cases (39.1%), (Figure 2)
. Intron 1-LOH was detected in 19 of 42 informative cases (45.2%). There were six cases with partial deletions; in five cases, the deletion involved both markers; in eight cases only the intron 1-marker showed allelic loss.
|
|
Nuclear accumulation of p53 protein was seen in 21 of 54 (38.9%) tumors (5 cases 
30% and 16 cases >30% immunopositive nuclei, respectively). All tumors with missense mutations at hot-spot codons (tumors 1, 2, 4, 7, 15, 24, and 41) showed nuclear p53 protein accumulation, whereas the five tumors with frame-shift mutations (tumors 8, 10, 29, 30, and 49), the single tumor with a stop-mutation (tumor 38), and one of three tumors with splice mutations (tumor 16) showed clear null-staining, confirming that such mutations result in a truncated p53 protein molecule (Tables 1 to 5)
. In addition, 8 of 28 (28.6%) wild-type p53 tumors identified by PCR-SSCP-sequencing analysis also expressed p53 protein. To reevaluate these discordant results, separate PCR reactions were performed and the PCR products were reanalyzed. In none of these p53 immunopositive tumors was a p53 mutation identified.
Arg/Pro Status and p53 Alterations (Mutations + LOH)
The Arg/Pro status of patients blood was Arg/Arg [17 of 42 (40.5%)], Arg/Pro [23 of 42 (54.8%)], and Pro/Pro [2 of 42 (4.8%)]. Codon 72 status of these 42 cases was Arg/Arg [25 of 42 (59.5%)], Arg/Pro [14 of 42 (33.3%)], and Pro/Pro [3 of 42 (7.1%)] (Tables 2 and 6)
. Nine of 23 informative cases (39.1%) showed allelic loss for the exon 4-marker (Figure 2)
. There was a preferential loss of the proline allele [8 of 9 cases (88.9%), Table 2
]. Intron 1-LOH was mostly found in tumors bearing the Arg status [17 of 19 (89.5%)], (Tables 2 to 5)
.
P53 mutations tended to occur more frequently in the Arg allele; allelic losses (exon 4 and exon 1) were found significantly more often in tumors bearing at least one Arg allele (P = 0.01). Interestingly, tumors having both alterations were only of the Arg type (P = 0.021), (Table 6)
.
|
Representative immunostains are given in Figure 3
. Fas protein expression was detected in 17 of 54 (31.5%) tumors, FasL in 20 of 54 (37%) tumors, whereas Fas/FasL protein expression simultaneously occurred in only 5 of 54 (9.2%) tumors, (Table 7)
. Bcl2 protein expression was found in 33 of 54 (61.1%) tumors (Table 7)
. Fas/FasL and Bcl2 expression were inversely correlated (P = 0.024), Fas/FasL did not correlate with p53 protein expression (P = 0.89). Sixteen of 21 (76.2%) p53 immunopositive tumors expressed Bcl2. In contrast, no expression of the anti-apoptotic protein Bcl2 was detected in five p53 immunopositive tumors. Sixteen of 33 (48.5%) p53 immunonegative tumors were Bcl2-positive (P = 0.09).
|
|
Pro status of the tumor significantly correlated with high Fas/FasL expression (P < 0.01). Simultaneous Fas/FasL expression never occurred in heterozygous Arg/Pro and in Arg tumors (Table 7
, Figure 3
). The apoptosis-inducing feature of tumors carrying the Pro allele was underlined by the frequent demonstration of DNA strand breaks in the TUNEL assay (Figure 4)
. Expression of the anti-apoptotic protein Bcl2 was never observed in tumors having the Pro status; this protein was expressed in only 1 of 16 (6.3%) heterozygous Arg/Pro tumors. p53 immunopositive tumors tended to show the Arg status.
|
There was a tendency for p53 alterations to occur more frequently in the proximal tumors (P = 0.07). We did not find a significant correlation between, sex, age, tumor grade, pT-status, nodal status, and Arg/Pro status of the tumor, p53 mutation status, or overexpression of any of the apoptosis-related immunohistochemical markers examined. Interestingly, non-alcoholics and non-smokers never showed p53-LOH (P = 0.055). p53 alterations were more frequently observed in tumors of the oral cavity, oropharynx, and hypopharynx, whereas they were rare in larynx carcinomas (P = 0.07).
|
Discussion |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
SCCHN develop in a multi-step process, acquiring different molecular alterations during carcinogenesis. So far, mutations of the p53 tumor suppressor gene have been demonstrated to play a major role in the development of SCCHN.18 Regarding the frequency of hot-spot mutations, type of base changes (transition versus transversion) and the distribution of mutations, our spectrum of p53 mutations was essentially identical to that of the IARC database.17 When excluding chain-terminating mutations (premature stop codon, splicing, and frame-shift mutations), we found transdominant mutations preferentially at Arg72. There was one recessive mutation (codon 168: CAC to CCC) at the Pro72 allele. Codon 72 in the mutant alleles was Arg in 11 of 33 (33.3%) tumors and Pro in 1 of 5 (20%) cases. Mutant Arg72 allele thus tended to be over-represented in the tumors. This is consistent with the observation of Marin et al6 who found that mutant alleles containing Arg72 are preferentially selected during tumorigenesis.
There is an expanding body of literature suggesting that host factors, including genetic polymorphisms, may explain some of the individual differences in cancer occurrence.19 These studies correlate allelic status at codon 72 in the blood of patients with manifestation of specific tumor types. In a recent work, Bergamaschi et al8 reported that patients with p53-mutant head and neck cancer carrying codon 72 Arg allele have a worse response to chemotherapy than whose having the Pro allele. However, in contrast to our study these authors focused on p73-dependent apoptosis, but our results add evidence that the codon 72 allele status is important for the induction of apoptosis in SCCHN.
The allelic status at codon 72 of the p53 gene has been already suggested to influence the induction of apoptosis in human cancer.6 As codon 72 is located in a proline-rich region of the p53 gene that is essential for p53-mediated apoptosis, the frequent occurrence of p53 mutations in SCCHN may affect the regulation of apoptotic tumor cell death, thus influencing the biological properties and prognosis of these tumors. That p53 is involved in the regulation of apoptosis in SCCHN has been demonstrated by the observation that the introduction of wild-type p53 decreases telomerase activity level, suppresses cell growth in SCCHN cell lines,20 and induces apoptosis.3,20 Furthermore, Marin et al have shown that the p53 mutation status of tumor cells influences oxygenic stress-induced apoptosis.6 In the latter study, all p53 mutants bound to p73 contained the Arg allele at codon 72, suggesting that the presence of Arg allele in the p53 mutants prevents tumor cells from apoptotic cell death.6 We found that allelic loss at the p53 gene in SCCHN occurs generally at the Pro allele while the Arg allele is retained. This imbalance resulted in markedly reduced apoptosis in tumors bearing the Arg allele, which, in turn, could influence the clinical outcome of these patients. In many cancers, such as bladder and gastric carcinomas, special codon 72 variants may serve as risk factors for neoplasia and may play a role in modulating environmental risk factors (smoking, alcohol, chemical toxins) for cancer development.21,22
The Fas/FasL system is involved in the induction of apoptosis in a variety of fetal and adult tissues.23 FasL (Fas ligand) is a key molecule in normal immune development, homeostasis, modulation, and cell function, that induces apoptosis by binding to its receptor Fas. There is evidence that the Fas receptor can be regulated at the transcriptional level by the tumor suppressor p53.24 Drug-induced up-regulation of the Fas receptor is dependent on the p53 status of tumor cell lines.12 The up-regulation of Fas receptor has been found to be reconstituted in p53 null cells by exogeneous wild-type p53 transfection.12 Our findings are in agreement with the literature where we show that Fas receptor protein expression is absent in most tumors of high immunopositivity for 53. In contrast, the regulation of FasL clearly involves p53-independent mechanisms.25 Indeed, FasL was overexpressed in approximately two-thirds of tumors showing high p53 immunopositivity.
Despite the multiple levels at which p53 seems to control Fas-induced apoptosis, few studies have addressed this relationship in vivo. In the esophagus, Fas expression can be used as a marker for differentiating between normal gastric mucosa and gastric metaplasia.26 Interestingly, simultaneous expression of Fas and FasL does not necessarily lead to apoptotic cell death. Various anti-apoptotic mechanisms may exist in Fas/FasL co-expressing cells that protect these cells against apoptosis.27 In our study, we found that Fas or Fas/L was expressed only in about one third of the tumors, and only a minority of tumors expressed Fas and Fas/L simultaneously. Expression of Fas and FasL in SCCHN in situ was heterogeneous; there were FasL-positive and FasL-negative regions of the tumor. Similar to Bennett et al28 and Gastman et al,29 we observed that the numbers of infiltrating lymphocytes were higher in SCCHN regions with higher FasL expression, but lower in tumor areas with low FasL expression. It is known that in contrast to its receptor, Fas-L is expressed on cytotoxic T-lymphocytes where it contributes to their cytotoxic function and mediates the elimination of peripheral T-cells following immune response.30
Tumors having the Arg72 constitution showed lack of co-expression of Fas and FasL and high expression of Bcl2 proteins, which was associated with impairment or lack of apoptosis. The expression of these proteins was reversed in tumors having the Pro72 constitution and this was associated with improvement of apoptosis induction. In heterozygous Arg/Pro tumors, we found neither Fas/FasL co-expression nor lack of Bcl2 protein expression except for only one tumor. The Arg allele seems to prevent apoptosis. Thus, there was a significant correlation between codon 72 genotype and apoptosis induction.
We conclude that apoptosis is correlated with the codon 72 allelic status in SCCHN. Homozygous Pro72 appears to be an important regulator of apoptosis via the Fas/FasL pathway in SCCHN.
|
|
|
|
Acknowledgements |
|---|
|
Footnotes |
|---|
Supported in part by the "Forschungsaentrum für Immunologie" Germany, BMBF01ZZ0110. A.G. is a fellow of the "Graduiertenkolleg" of the Otto-von-Guericke University.
Accepted for publication December 11, 2003.
|
References |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  N. A. Ellis, D. Huo, O. Yildiz, L. J. Worrillow, M. Banerjee, M. M. Le Beau, R. A. Larson, J. M. Allan, and K. Onel MDM2 SNP309 and TP53 Arg72Pro interact to alter therapy-related acute myeloid leukemia susceptibility Blood, August 1, 2008; 112(3): 741 - 749. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. Chen, Z. Hu, L.-E Wang, E. M. Sturgis, A. K. El-Naggar, W. Zhang, and Q. Wei Single-nucleotide polymorphisms at the TP53-binding or responsive promoter regions of BAX and BCL2 genes and risk of squamous cell carcinoma of the head and neck Carcinogenesis, September 1, 2007; 28(9): 2008 - 2012. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S Mitra, S Banerjee, C Misra, R K Singh, A Roy, A Sengupta, C K Panda, and S Roychoudhury Interplay between human papilloma virus infection and p53 gene alterations in head and neck squamous cell carcinoma of an Indian patient population J. Clin. Pathol., September 1, 2007; 60(9): 1040 - 1047. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Z.-Z. Zhu, A.-Z. Wang, H.-R. Jia, X.-X. Jin, X.-L. He, L.-F. Hou, and G. Zhu Association of the TP53 Codon 72 Polymorphism with Colorectal Cancer in a Chinese Population Jpn. J. Clin. Oncol., May 1, 2007; 37(5): 385 - 390. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H. H. Nelson, M. Wilkojmen, C. J. Marsit, and K. T. Kelsey TP53 mutation, allelism and survival in non-small cell lung cancer Carcinogenesis, October 1, 2005; 26(10): 1770 - 1773. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Tommiska, H. Eerola, M. Heinonen, L. Salonen, M. Kaare, J. Tallila, A. Ristimaki, K. von Smitten, K. Aittomaki, P. Heikkila, et al. Breast Cancer Patients with p53 Pro72 Homozygous Genotype Have a Poorer Survival Clin. Cancer Res., July 15, 2005; 11(14): 5098 - 5103. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
). Tu, tumor; NT, non-tumorous tissue
