- Email: [email protected]
High incidence of mutations of the p53 gene detected in ovarian tumours by the use of chemical mismatch cleavage
Cancer Letters, 68 (1993) 83 - 89 Elsevier Scientific Publishers Ireland
83 Ltd
High incidence of mutations of the p53 gene detected in ovarian tumours by the use of chemical mismatch cleavage E. Sheridan, Department
B.W. Hancock
and M.H. Goyns
of Clinical Oncology, Royal Hallamshire Hospital, Sheffield, S10 2JF /UK)
(Received 23 October 1992) (Accepted 3 November 1992)
SUOBeUWy
IntrodnctIon
investigated a series of ovarian evidence of mutations in the ~53 tumour suppressor gene. In this study we have made use of the chemical mismatch cleavage technique which, from analyses of other genes, has been shown to consistently identify all point mutations present within a region of DNA. This approach revealed mutations of p53 in 11/20 tumours studied, mainly in exons 5 or 7. After sequencing the relevant
Ovarian cancer is the fourth most common cause of cancer death in women in England and Wales. In 1985 it was estimated that there were 4400 new cases and 3800 deaths from the disease in this country [ 11. The majority of these cases occurred in the over-45 age group and, because symptoms were initially vague, usually resulted in late presentation. It was reported in one British series that 67% of patients had stage 3 or 4 disease at diagnosis [Z]. Due to its relatively high frequency and to the complexities of the treatment regimes, the management of ovarian cancer was the subject of the First Report of the Standing Subcommittee on Cancer by the UK Department of Health. Highlighted in that document was the conclusion, ‘Perhaps the greatest benefit that can be hoped for, is one that will not show in the mortality statistics - a reduction in the morbidity caused by inappropriate treatment’ [3]. To achieve even this limited goal, it is essential to identify prognostic variables of greater reliability than those that are currently available. The p53 gene is currently regarded as the most important gene in the evolution of human malignant disease. Not only is it associated with a wide spectrum of neoplasias, it is also found to be frequently mutated or
We have turnouts for
regions of the gene it was shown that ten of these mutations would have resulted in an amino acid subsfitution in the protein and only one represented a polymorphism. The observed incidence of p53 missense mutations in our series (50%) was the highest recorded in ovarian tumours and demonstrated the potential of the mismatch cleavage technique as a reliable method for the detection of p53 mutations in human turnours.
Keywords: ~53; ovarian cancer; chemical mismatch cleavage; mutation
Correspondence to: M.H. Goyns, Department cology, Royal Hallamshire Hospital, Sheffield,
03043835/92/$05.00 Printed and Published
of Clinical OnSlO 2JF, UK.
0 1992 Elsevier Scientific Publishers in Ireland
Ireland Ltd
84
deleted within any one tumour type [4,5]. Its protein product is a nuclear phosphoprotein [6,7], which appears to be constitutively expressed at low levels in all mammalian cells. By contrast the p53 protein in transformed cells appears to be present at elevated levels, due primarily to its longer half-life in these cells [8,9]. The most common mechanism which results in the presence of a stable form of p53 protein is the occurrence of a missense mutation. Mutations of p53 are common events, especially in aggressive tumours such as lung cancer [lo] and it is possible that they may be useful as markers of clinical outcome [ll]. A number of studies of ovarian cancer have demonstrated allelic loss of chromosome 17 [12] and positivity of p53 immunohistochemical staining [13], which have indicated that p53 is also mutated in these tumours. To evaluate the usefulness of p53 mutations as markers of response to therapy in ovarian cancer, we therefore had to adopt a method which would allow us to efficiently screen large numbers of samples for the presence of these mutations. To this end we have carried out an initial study on a series of 20 fresh ovarian tumours, which were analysed for p53 mutations by using the chemical mismatch cleavage technique [14]. The results presented here demonstrate that this is an efficient method for quickly identifying mutations and indicate that it could readily be applied to much larger studies of archival material. Materials
and Methods
Clinical samples Tumour samples were collected at the time of primary surgery, snap frozen in liquid nitrogen and stored at - 70°C until required for DNA extraction. All samples were diagnosed as epithelial ovarian cancer, details of which are summarised in Table I. DNA extraction High molecular weight DNA was extracted by grinding the frozen tissue to a fine powder under liquid nitrogen and incubating with
Table
1. Clinical details vestigated in this study.
of ovarian
tumours
Patient
Histology
Grade
Stage
OCOOl ocoo2 ocoo3 ocoO4 ocoo5 OCO06 ocoo7 OCOO8 ocoo9 OCOlO OCOll oco12 oco13 oco14 oco15 OC016 oco17 OC018 oco19 oco20
Serous N/A Serous N/A Serous N/A Anaplastic Anaplastic Endometrioid Uncertain Undifferentiated Serous Clear cell Anaplastic Mutinous Serous Anaplastic Endometrioid Serous
Poor N/A N/A Poor N/A Moderate N/A Poor Poor Moderate Poor Poor Moderate Well Poor Moderate Moderate Poor Moderate Poor
3c 3c N/A 3c N/A N/A N/A 3c 4 lc N/A 2c 3a N/A N/A 3b N/A N/A N/A N/A
in-
N/A, information not available.
RNAse and proteinase-K in a SDS buffer at 50°C, overnight. The sample was then exthree tracted times with (25/24/ 1) phenol/chloroform/iso-amyl alcohol. The DNA was collected by ethanol precipitation and spooling onto a glass rod. It was finally washed with 70% ethanol, air dried and then resuspended in TE buffer at pH 7.5 PCR amplification
Amplification of DNA was carried out by using the polymerase chain reaction (PCR) method of Saiki et al. [15]. A 1-,~g quantity of genomic DNA was amplified using 500-ng primers per reaction in a total reaction volume of 50 ~1, containing 1.5 mM MgC12, 200 PM dNTPs and 1 unit Amplitaq (Cetus).The amplification cycle was 95OC for 7 min, followed by thirty cycles of 94OC for 1 min, 55OC for 1 min and 70°C for 3 min. The
85
oligonucleotides used as primers were all 20 mers and their sequences were as follows: Exons Sense Antisense Exons Sense Antisense Exons Sense Antisense
2- 4 ‘5-TGGATCCTCTTGCAGCAGCC ‘5-GCAACTGACCGTGCAAGTCA 5-6 ‘5-TTCCTCTTCCTGCAGTACTC ‘5-AGTTGCAAACCAGACCTCAG 7- 9 ‘5-GTGTTGTCTCCTAGGTTGGC ‘5-AGACTTAGTACCTGAAGGGT
The PCR products were purified on a 0.8% (w/v) agarose gel (SeaKem FMC BioProducts) . In each case the amplification generated a single product corresponding to a band of the expected fragment size on the gel. This DNA fragment was cut out from the gel, extracted by electroelution and resuspended in 20 ~1 TE buffer. Approximately 2 pg of purified product was obtained per reaction. Chemical mismatch cleavage Mutations were detected by the use of chemical cleavage as described by Montandon et al. [14] and modified by Prosser et al. [16]. A 40-ng quantity of 32P-labelled wild-type DNA was allowed to anneal with 400 ng target DNA in 0.3 M NaCI, 0.1 M (Tris pH 8), in a total reaction volume of 10 ~1. This was overlayered with paraffin and incubated at was 65OC overnight. The heteroduplex precipitated with 1.5 pg mussel glycogen and then taken up into 13 ~1 TE buffer. The hydroxylamine cleavage was carried out as follows. Twenty microlitres of freshly prepared 4 M hydroxylamine solution (Aldrich) titrated to pH 6 with diethylamine (Aldrich), were added to 7 ~1 heteroduplex solution and incubated at 37OC for 2 h. The reaction was stopped by the addition of 200 ~1 stop buffer (0.3 M sodium acetate, 0.1 mM EDTA), the heteroduplex precipitated with 15 pg of mussel glycogen and 750 ~1 of 100%
(v/v) ice-cold ethanol. Fifty microlitres of fresh 1 M piperidine (Aldrich) were added to the lightly dried pellet, vortexed and incubated at 90°C for 30 min. The reaction was then stopped and precipitated as before. The pellet was resuspended in 2 j.d loading buffer and 5 ~1TE buffer, and run on a 6% (w/v) denaturing polyacrylamide gel. The osmium tetroxide cleavage was carried out as follows. Fifteen microlitres of fresh 0.8% (v/v) osmium tetroxide (Johnson & Matthey) were added to 6 ~1 heteroduplex solution with 2.5 ~1 10 x buffer (100 mM Tris, 10 mM EDTA, 15% (v/v) pyridine) and incubated at 37OC for 10 min. The cleavage reaction was carried out in the same way as that with hydroxylamine. Sequencing
the PCR product
The PCR products were purified as above and sequenced using the method of Winship [17]. A 200-ng quantity of purified PCR product and 280 ng primer were used. Extension and 35S-labelling were performed using the Sequenase Version 2 enzyme (USB) and the reaction products run on an 8% (w/v) denaturing gel. Results and Discussion
We have investigated mutations in the p53 tumour suppressor gene in a series of 20 fresh ovarian tumour samples. In an attempt to identify all mutations present, we have used the chemical mismatch cleavage technique to analyse PCR-amplified regions of the gene. We amplified those regions which contained exons 2-4, exons 5-6 and exons 7 -9, as these were known to contain all of the missense mutations that had been identified in other tumour types. The use of hydroxylamine cleavage allowed 83% of all mutations to be identified, the remaining mutations being identified by the osmium tetroxide cleavage [18]. An example of a mutation of p53 detected in this way is shown in Fig. 1. In this figure, six of the samples showed no mutations, but one sample (Fig. 1, lane 1) exhibited a typical
Fig. 1. Detection of mutations in the p53 tumour suppressor gene in human ovarian tumours. Exons 5 - 6 of the P53 gene were amplified by PCR and then subjected to (a) osmium tetroxide cleavage or (b) hydroxylamine cleavage. The cleavage product in lane 1 is indicated by an arrow. The samples used in each of the lanes were: (l), OC016; (Z), 0C017; (3), OCO18; (4), OCO19; (5), OC013; (6), OC014; (7)) 0C015. Lanes 8 and 9 are positive and negative mlutation controls (i.e. mutated and wild type samples of the Factor IX gene).
87
ACGTACG
T
b Fig. 2. Nucleotide sequence of the region of p53 identified as havil lg a mutation by the chemical mismatch clea lvage procedure shown in Fig. 1. The sequence of the mutated gene from sample 0C016 (a) is shown for comparison with that from a normal p53 sample (b). The C to T mutation is indicate d by an arrow.
cleavage product. This sample was further analysed by direct sequencing of the relevant PCR product, which revealed that there was a C to T base change present (Fig. 2). In total we observed point mutations in 11 of the 20 samples analysed. Only one of these (Patient OCOlS) represented a polymorphism, whereas the other 10 all would have resulted in an amino acid substitution in the protein product (Table II). These missense mutations appeared to be clustered in exons 5 - 7, which are known to be hot spots for p53 mutation in a wide spectrum of tumour types [19]. The mutations we identified occurred in dif-
ferent histological subtypes of ovarian cancer and although they were all associated with stage 3 or stage 4 tumours (Table I), this might simply be a reflection of the fact that many of the patients presented with advanced stages of the disease. The most obvious way to detect mutations in genes in tumour cells would be to sequence the respective gene in all of the samples under study. Such an approach is, however, impracticable in most situations, especially when large numbers of tumours are intended to be screened. As a result, several techniques have been developed to try and detect mutations reliably
88 Table II. tumours.
Mutations
in the
~53
gene
in ovarian
Patient
Exon
Base change
Amino acid substitution
oco2 oc17 OCOl oco4 oco5 oc20 OC16 0C06 oc13 oc15 OC18
4 4 5 5 5 5 6 7 7 7 7
C to G
Pro Pro Ala Ala Arg Glu Leu Ser Met Met
G C C C A C C T T G
to to to to to to to to to to
T G G A T T T G G A
to to to to to to to to to to
Arg Ser Gly Gly Ser Val Phe Phe Arg Arg
Arg
and rapidly before selecting samples for sequencing. These methods, of which there are two main types, ail involve the amplification of one region of the gene by PCR. The first group of techniques rely on the physico-chemical interactions between the DNA fragments and the gels. These include single strand conformationai polymorphism analysis [ZO] and density gradient gel eiectrophoresis [Zl]. The second group of techniques rely upon the specific reactivity of mismatched bases and include RNAse protection assays [22] and chemical mismatch cleavage [ 141. The chemical mismatch cleavage method has particular advantages, not least of which is that it enables ail possible base changes to be recognised. It also does not require experimental conditions to be altered when anaiysing different DNA fragments or different mutations, which is often the case when using other techniques. One of the main advantages of mismatch cleavage is that it is very sensitive and can detect mutations in only a small number of DNA fragments in the population. This is particularly important, because tumour material is often very heterogeneous; both malignant and normal supporting stromai ceils
are usually present in an ovarian tumour sampie. The use of this technique, which does not produce false negatives or false positives, therefore ensures that the presence of a mutation can be determined with certainty. This means that one need obtain sequence data only to define the nature of the mutation. The results presented here demonstrated that mutations in the p53 gene could be readily detected using the mismatch cleavage technique and that the detected missense mutation frequency of 50% was much higher than previously reported frequencies of 29% [23] and 36% [24] that were identified by single strand conformational polymorphism analysis. This is interesting with respect to a recently reported immunohistochemicai study of ~53 protein in ovarian tumours, which also reported an incidence of 50% [25]. There has been some concern that immunohistochemicai positive staining for p53 protein may not always reflect the presence of missense mutations in the p53 gene 1261, but in ovarian cancer these two events appear to be associated. It should be possible therefore to use the mismatch cleavage method to rapidly and efficiently anaiyse a large series of archival material. The latter would enable us to evaluate the incidence of p53 missense mutations as a marker of prognosis, but in particular would also allow us to assess whether these mutations could indicate which ovarian cancer patients were likely to respond favourabiy or unfavourabiy to therapy.
Acknowledgement
This work was supported by the Yorkshire Cancer Research Campaign. References 1 2
Cancer Statistics: Registrations. England and Wales 1985 (1990) HMSO, London. Shepperd J.H. (1991) Clinical Gynaecological Oncology,
89
3
4 5
6
7
8
9
10
11 12
13
14
15
2nd Edition, pp. 187-207. Editors: J.H. Shepherd and J.M. Monaghan. Blackwell Scientific Publications, Oxford. Report of a working group of the Standing Subgroup on Cancer of the Standing Medical Advisory Committee, Chair: Scott J.S. (1991) Management of Ovarian Cancer Current Clinical Practices. HMSO, London. Levine, A.J., Momand, J. and Finlay, C.A. (1991) The p53 tumor suppressor gene. Nature, 351, 453 -456. Hollstein, M., Sidransky, D., Vogelstein, B. and Harris, C.C. (1991) p53 mutations in human cancer. Science, 253, 49-53. De Leo, A.B., Jay, G., Appella, E., Dubois, G.C., Law, L.W. and Old, L.J. (1979) Detection of a transformation related antigen in chemically induced sarcomas and other transformed cells of the mouse. Proc. Natl Acad. Sci. U.S.A., 76, 2420-2424. Lane, D.P. and Crawford, L.V. (1979) T antigen is bound to a host protein in SV40transformed cells. Nature, 278, 261- 263. Linzer, D.I.H. and Levine, A.J. (1979) Charactexisation of a 54 K dalton cellular SV40 tumor antigen present in SV40-transformed cells and uninfected embryonal carcinoma cells. Cell, 17, 43-52. Crawford, L.V., Pim, D.C., Gurney, E.G., Goodfellow, P. and Taylor-Papadimitiou, J. (1981) Detection of a common feature in several human tumour cell lines. PIOC. Natl Acad. Sci. U.S.A., 78, 41-45. Iggo, R., Gatter, K., Bartek, L., Lane, D. and Harris, A.L. (1990) Increased expression of mutant forms of ~53 oncogene in primary lung cancer. Lancet, 335, 675 - 679. Vogelstein, B. and Kinzler, K.W. (1992) ~53 function and dysfunction. Cell, 70, 523 - 526. Russell, S.E.H., Hickey, G.I., Lowry, W.S., White, P. and Atkinson, R.J. (1990) Allele loss from chromosome 17 in ovarian cancer. Oncogene, 5, 1581- 1583. Eccles, D.M., Brett, L., Lessells, A., Gruber, L., Lane, D., Steel, C.M. and Leonard, R.C.F. (1992) Overexpression of the p53 protein and allele loss at 17~13 in ovarian carcinoma. Br. J. Cancer, 65, 40-44. Montandon, A.J., Green, P.M., Gianelli, F. and Bentley, D.R. (1989) Direct detection of point mutations by mismatch analysis. Nucleic Acid Res., 17, 3347 -3358. Saiki, R.K., Gelfand, S., Stoffel, S.J., Scharf, R., Higuchi,
16
17
18
19 20
21
22
23
24
25
26
R., Horn, G.T., Mullis, K.B. and Erlich, H.A. (1988) Primer directed amplification of DNA with a thermostable DNA polymerase. Science, 239, 487-491. Prosser, J., Elder, P.A., Condie, A., MacFadyen I, Steel, CM. and Evans, H.J. (1991) Mutations in ~53 do not acount for heritable breast cancer: a study in five affected families. Br. J. Cancer, 63, 181- 184. Winship, P.R., (1989) An improved method for directly sequencing PCR amplified material using dimethyl sulphoxide. Nucleic Acid Res., 17, 1266. Prosser, J., Thompson, A.M., Cranston, G. and Evans, H.J. (1990) Evidence that ~53 behaves as a tumour suppressor gene in sporadic breast cancer. Oncogene, 5, 1573 - 1579. Vogelstein, B. (1990) A deadly inheritance. Nature, 348, 681- 682. Ortta, M., Youichi, S., Takao, S. and Mayashi, K. (1989) Rapid and sensitive detection of point mutations and DNA polymorphisms using the polymerase chain reaction. Genomics, 5, 74 - 79. Myers, R.M., Lumelsky, N., Lereman, L. and Maniatis, T. (1985) Detection of single base substitutions in total genomic DNA. Nature, 313, 495-498. Myers, R.M., Larin, Z. and Maniatis, T. (1985) Detection of single base substitutions by rtbonuclease cleavage at mismatches in RNA/DNA heteroduplexes. Science, 230, 1242 - 1246. Okamoto, A., Sameshima, Y., Yokoyama, S., Terashima, Y., Sugimura, T., Terada, M. and Yokota, J. (1991) Frequent allelic losses and mutations of the p53 gene in human ovarian cancer. Cancer Res., 51, 5171-5176. Maws, R., Pujol, P., Maudelonde, T., Jeanteur, P. and Theillet, C. (1991) p53 mutations in ovarian cancer: a late event? Oncogene, 6, 1685- 1690. Marks, J.R., Davidoff, A.M., Kerns, B.J., Humphrey, P.A., Pence, J.C., Dodge, R.K., Clarke-Pearson, D.L., Iglehart, J.D., Bast, R.C. and Berchuck, A. (1991) Overexpression and mutation of p53 in epithelial ovarian cancer. Cancer Res., 51, 2979- 2984. Wynford-Thomas, D. (1992) P53 in tumour pathology: can we trust immunocytochemistry? J. Pathol., 166, 329 - 330.
83 Ltd
High incidence of mutations of the p53 gene detected in ovarian tumours by the use of chemical mismatch cleavage E. Sheridan, Department
B.W. Hancock
and M.H. Goyns
of Clinical Oncology, Royal Hallamshire Hospital, Sheffield, S10 2JF /UK)
(Received 23 October 1992) (Accepted 3 November 1992)
SUOBeUWy
IntrodnctIon
investigated a series of ovarian evidence of mutations in the ~53 tumour suppressor gene. In this study we have made use of the chemical mismatch cleavage technique which, from analyses of other genes, has been shown to consistently identify all point mutations present within a region of DNA. This approach revealed mutations of p53 in 11/20 tumours studied, mainly in exons 5 or 7. After sequencing the relevant
Ovarian cancer is the fourth most common cause of cancer death in women in England and Wales. In 1985 it was estimated that there were 4400 new cases and 3800 deaths from the disease in this country [ 11. The majority of these cases occurred in the over-45 age group and, because symptoms were initially vague, usually resulted in late presentation. It was reported in one British series that 67% of patients had stage 3 or 4 disease at diagnosis [Z]. Due to its relatively high frequency and to the complexities of the treatment regimes, the management of ovarian cancer was the subject of the First Report of the Standing Subcommittee on Cancer by the UK Department of Health. Highlighted in that document was the conclusion, ‘Perhaps the greatest benefit that can be hoped for, is one that will not show in the mortality statistics - a reduction in the morbidity caused by inappropriate treatment’ [3]. To achieve even this limited goal, it is essential to identify prognostic variables of greater reliability than those that are currently available. The p53 gene is currently regarded as the most important gene in the evolution of human malignant disease. Not only is it associated with a wide spectrum of neoplasias, it is also found to be frequently mutated or
We have turnouts for
regions of the gene it was shown that ten of these mutations would have resulted in an amino acid subsfitution in the protein and only one represented a polymorphism. The observed incidence of p53 missense mutations in our series (50%) was the highest recorded in ovarian tumours and demonstrated the potential of the mismatch cleavage technique as a reliable method for the detection of p53 mutations in human turnours.
Keywords: ~53; ovarian cancer; chemical mismatch cleavage; mutation
Correspondence to: M.H. Goyns, Department cology, Royal Hallamshire Hospital, Sheffield,
03043835/92/$05.00 Printed and Published
of Clinical OnSlO 2JF, UK.
0 1992 Elsevier Scientific Publishers in Ireland
Ireland Ltd
84
deleted within any one tumour type [4,5]. Its protein product is a nuclear phosphoprotein [6,7], which appears to be constitutively expressed at low levels in all mammalian cells. By contrast the p53 protein in transformed cells appears to be present at elevated levels, due primarily to its longer half-life in these cells [8,9]. The most common mechanism which results in the presence of a stable form of p53 protein is the occurrence of a missense mutation. Mutations of p53 are common events, especially in aggressive tumours such as lung cancer [lo] and it is possible that they may be useful as markers of clinical outcome [ll]. A number of studies of ovarian cancer have demonstrated allelic loss of chromosome 17 [12] and positivity of p53 immunohistochemical staining [13], which have indicated that p53 is also mutated in these tumours. To evaluate the usefulness of p53 mutations as markers of response to therapy in ovarian cancer, we therefore had to adopt a method which would allow us to efficiently screen large numbers of samples for the presence of these mutations. To this end we have carried out an initial study on a series of 20 fresh ovarian tumours, which were analysed for p53 mutations by using the chemical mismatch cleavage technique [14]. The results presented here demonstrate that this is an efficient method for quickly identifying mutations and indicate that it could readily be applied to much larger studies of archival material. Materials
and Methods
Clinical samples Tumour samples were collected at the time of primary surgery, snap frozen in liquid nitrogen and stored at - 70°C until required for DNA extraction. All samples were diagnosed as epithelial ovarian cancer, details of which are summarised in Table I. DNA extraction High molecular weight DNA was extracted by grinding the frozen tissue to a fine powder under liquid nitrogen and incubating with
Table
1. Clinical details vestigated in this study.
of ovarian
tumours
Patient
Histology
Grade
Stage
OCOOl ocoo2 ocoo3 ocoO4 ocoo5 OCO06 ocoo7 OCOO8 ocoo9 OCOlO OCOll oco12 oco13 oco14 oco15 OC016 oco17 OC018 oco19 oco20
Serous N/A Serous N/A Serous N/A Anaplastic Anaplastic Endometrioid Uncertain Undifferentiated Serous Clear cell Anaplastic Mutinous Serous Anaplastic Endometrioid Serous
Poor N/A N/A Poor N/A Moderate N/A Poor Poor Moderate Poor Poor Moderate Well Poor Moderate Moderate Poor Moderate Poor
3c 3c N/A 3c N/A N/A N/A 3c 4 lc N/A 2c 3a N/A N/A 3b N/A N/A N/A N/A
in-
N/A, information not available.
RNAse and proteinase-K in a SDS buffer at 50°C, overnight. The sample was then exthree tracted times with (25/24/ 1) phenol/chloroform/iso-amyl alcohol. The DNA was collected by ethanol precipitation and spooling onto a glass rod. It was finally washed with 70% ethanol, air dried and then resuspended in TE buffer at pH 7.5 PCR amplification
Amplification of DNA was carried out by using the polymerase chain reaction (PCR) method of Saiki et al. [15]. A 1-,~g quantity of genomic DNA was amplified using 500-ng primers per reaction in a total reaction volume of 50 ~1, containing 1.5 mM MgC12, 200 PM dNTPs and 1 unit Amplitaq (Cetus).The amplification cycle was 95OC for 7 min, followed by thirty cycles of 94OC for 1 min, 55OC for 1 min and 70°C for 3 min. The
85
oligonucleotides used as primers were all 20 mers and their sequences were as follows: Exons Sense Antisense Exons Sense Antisense Exons Sense Antisense
2- 4 ‘5-TGGATCCTCTTGCAGCAGCC ‘5-GCAACTGACCGTGCAAGTCA 5-6 ‘5-TTCCTCTTCCTGCAGTACTC ‘5-AGTTGCAAACCAGACCTCAG 7- 9 ‘5-GTGTTGTCTCCTAGGTTGGC ‘5-AGACTTAGTACCTGAAGGGT
The PCR products were purified on a 0.8% (w/v) agarose gel (SeaKem FMC BioProducts) . In each case the amplification generated a single product corresponding to a band of the expected fragment size on the gel. This DNA fragment was cut out from the gel, extracted by electroelution and resuspended in 20 ~1 TE buffer. Approximately 2 pg of purified product was obtained per reaction. Chemical mismatch cleavage Mutations were detected by the use of chemical cleavage as described by Montandon et al. [14] and modified by Prosser et al. [16]. A 40-ng quantity of 32P-labelled wild-type DNA was allowed to anneal with 400 ng target DNA in 0.3 M NaCI, 0.1 M (Tris pH 8), in a total reaction volume of 10 ~1. This was overlayered with paraffin and incubated at was 65OC overnight. The heteroduplex precipitated with 1.5 pg mussel glycogen and then taken up into 13 ~1 TE buffer. The hydroxylamine cleavage was carried out as follows. Twenty microlitres of freshly prepared 4 M hydroxylamine solution (Aldrich) titrated to pH 6 with diethylamine (Aldrich), were added to 7 ~1 heteroduplex solution and incubated at 37OC for 2 h. The reaction was stopped by the addition of 200 ~1 stop buffer (0.3 M sodium acetate, 0.1 mM EDTA), the heteroduplex precipitated with 15 pg of mussel glycogen and 750 ~1 of 100%
(v/v) ice-cold ethanol. Fifty microlitres of fresh 1 M piperidine (Aldrich) were added to the lightly dried pellet, vortexed and incubated at 90°C for 30 min. The reaction was then stopped and precipitated as before. The pellet was resuspended in 2 j.d loading buffer and 5 ~1TE buffer, and run on a 6% (w/v) denaturing polyacrylamide gel. The osmium tetroxide cleavage was carried out as follows. Fifteen microlitres of fresh 0.8% (v/v) osmium tetroxide (Johnson & Matthey) were added to 6 ~1 heteroduplex solution with 2.5 ~1 10 x buffer (100 mM Tris, 10 mM EDTA, 15% (v/v) pyridine) and incubated at 37OC for 10 min. The cleavage reaction was carried out in the same way as that with hydroxylamine. Sequencing
the PCR product
The PCR products were purified as above and sequenced using the method of Winship [17]. A 200-ng quantity of purified PCR product and 280 ng primer were used. Extension and 35S-labelling were performed using the Sequenase Version 2 enzyme (USB) and the reaction products run on an 8% (w/v) denaturing gel. Results and Discussion
We have investigated mutations in the p53 tumour suppressor gene in a series of 20 fresh ovarian tumour samples. In an attempt to identify all mutations present, we have used the chemical mismatch cleavage technique to analyse PCR-amplified regions of the gene. We amplified those regions which contained exons 2-4, exons 5-6 and exons 7 -9, as these were known to contain all of the missense mutations that had been identified in other tumour types. The use of hydroxylamine cleavage allowed 83% of all mutations to be identified, the remaining mutations being identified by the osmium tetroxide cleavage [18]. An example of a mutation of p53 detected in this way is shown in Fig. 1. In this figure, six of the samples showed no mutations, but one sample (Fig. 1, lane 1) exhibited a typical
Fig. 1. Detection of mutations in the p53 tumour suppressor gene in human ovarian tumours. Exons 5 - 6 of the P53 gene were amplified by PCR and then subjected to (a) osmium tetroxide cleavage or (b) hydroxylamine cleavage. The cleavage product in lane 1 is indicated by an arrow. The samples used in each of the lanes were: (l), OC016; (Z), 0C017; (3), OCO18; (4), OCO19; (5), OC013; (6), OC014; (7)) 0C015. Lanes 8 and 9 are positive and negative mlutation controls (i.e. mutated and wild type samples of the Factor IX gene).
87
ACGTACG
T
b Fig. 2. Nucleotide sequence of the region of p53 identified as havil lg a mutation by the chemical mismatch clea lvage procedure shown in Fig. 1. The sequence of the mutated gene from sample 0C016 (a) is shown for comparison with that from a normal p53 sample (b). The C to T mutation is indicate d by an arrow.
cleavage product. This sample was further analysed by direct sequencing of the relevant PCR product, which revealed that there was a C to T base change present (Fig. 2). In total we observed point mutations in 11 of the 20 samples analysed. Only one of these (Patient OCOlS) represented a polymorphism, whereas the other 10 all would have resulted in an amino acid substitution in the protein product (Table II). These missense mutations appeared to be clustered in exons 5 - 7, which are known to be hot spots for p53 mutation in a wide spectrum of tumour types [19]. The mutations we identified occurred in dif-
ferent histological subtypes of ovarian cancer and although they were all associated with stage 3 or stage 4 tumours (Table I), this might simply be a reflection of the fact that many of the patients presented with advanced stages of the disease. The most obvious way to detect mutations in genes in tumour cells would be to sequence the respective gene in all of the samples under study. Such an approach is, however, impracticable in most situations, especially when large numbers of tumours are intended to be screened. As a result, several techniques have been developed to try and detect mutations reliably
88 Table II. tumours.
Mutations
in the
~53
gene
in ovarian
Patient
Exon
Base change
Amino acid substitution
oco2 oc17 OCOl oco4 oco5 oc20 OC16 0C06 oc13 oc15 OC18
4 4 5 5 5 5 6 7 7 7 7
C to G
Pro Pro Ala Ala Arg Glu Leu Ser Met Met
G C C C A C C T T G
to to to to to to to to to to
T G G A T T T G G A
to to to to to to to to to to
Arg Ser Gly Gly Ser Val Phe Phe Arg Arg
Arg
and rapidly before selecting samples for sequencing. These methods, of which there are two main types, ail involve the amplification of one region of the gene by PCR. The first group of techniques rely on the physico-chemical interactions between the DNA fragments and the gels. These include single strand conformationai polymorphism analysis [ZO] and density gradient gel eiectrophoresis [Zl]. The second group of techniques rely upon the specific reactivity of mismatched bases and include RNAse protection assays [22] and chemical mismatch cleavage [ 141. The chemical mismatch cleavage method has particular advantages, not least of which is that it enables ail possible base changes to be recognised. It also does not require experimental conditions to be altered when anaiysing different DNA fragments or different mutations, which is often the case when using other techniques. One of the main advantages of mismatch cleavage is that it is very sensitive and can detect mutations in only a small number of DNA fragments in the population. This is particularly important, because tumour material is often very heterogeneous; both malignant and normal supporting stromai ceils
are usually present in an ovarian tumour sampie. The use of this technique, which does not produce false negatives or false positives, therefore ensures that the presence of a mutation can be determined with certainty. This means that one need obtain sequence data only to define the nature of the mutation. The results presented here demonstrated that mutations in the p53 gene could be readily detected using the mismatch cleavage technique and that the detected missense mutation frequency of 50% was much higher than previously reported frequencies of 29% [23] and 36% [24] that were identified by single strand conformational polymorphism analysis. This is interesting with respect to a recently reported immunohistochemicai study of ~53 protein in ovarian tumours, which also reported an incidence of 50% [25]. There has been some concern that immunohistochemicai positive staining for p53 protein may not always reflect the presence of missense mutations in the p53 gene 1261, but in ovarian cancer these two events appear to be associated. It should be possible therefore to use the mismatch cleavage method to rapidly and efficiently anaiyse a large series of archival material. The latter would enable us to evaluate the incidence of p53 missense mutations as a marker of prognosis, but in particular would also allow us to assess whether these mutations could indicate which ovarian cancer patients were likely to respond favourabiy or unfavourabiy to therapy.
Acknowledgement
This work was supported by the Yorkshire Cancer Research Campaign. References 1 2
Cancer Statistics: Registrations. England and Wales 1985 (1990) HMSO, London. Shepperd J.H. (1991) Clinical Gynaecological Oncology,
89
3
4 5
6
7
8
9
10
11 12
13
14
15
2nd Edition, pp. 187-207. Editors: J.H. Shepherd and J.M. Monaghan. Blackwell Scientific Publications, Oxford. Report of a working group of the Standing Subgroup on Cancer of the Standing Medical Advisory Committee, Chair: Scott J.S. (1991) Management of Ovarian Cancer Current Clinical Practices. HMSO, London. Levine, A.J., Momand, J. and Finlay, C.A. (1991) The p53 tumor suppressor gene. Nature, 351, 453 -456. Hollstein, M., Sidransky, D., Vogelstein, B. and Harris, C.C. (1991) p53 mutations in human cancer. Science, 253, 49-53. De Leo, A.B., Jay, G., Appella, E., Dubois, G.C., Law, L.W. and Old, L.J. (1979) Detection of a transformation related antigen in chemically induced sarcomas and other transformed cells of the mouse. Proc. Natl Acad. Sci. U.S.A., 76, 2420-2424. Lane, D.P. and Crawford, L.V. (1979) T antigen is bound to a host protein in SV40transformed cells. Nature, 278, 261- 263. Linzer, D.I.H. and Levine, A.J. (1979) Charactexisation of a 54 K dalton cellular SV40 tumor antigen present in SV40-transformed cells and uninfected embryonal carcinoma cells. Cell, 17, 43-52. Crawford, L.V., Pim, D.C., Gurney, E.G., Goodfellow, P. and Taylor-Papadimitiou, J. (1981) Detection of a common feature in several human tumour cell lines. PIOC. Natl Acad. Sci. U.S.A., 78, 41-45. Iggo, R., Gatter, K., Bartek, L., Lane, D. and Harris, A.L. (1990) Increased expression of mutant forms of ~53 oncogene in primary lung cancer. Lancet, 335, 675 - 679. Vogelstein, B. and Kinzler, K.W. (1992) ~53 function and dysfunction. Cell, 70, 523 - 526. Russell, S.E.H., Hickey, G.I., Lowry, W.S., White, P. and Atkinson, R.J. (1990) Allele loss from chromosome 17 in ovarian cancer. Oncogene, 5, 1581- 1583. Eccles, D.M., Brett, L., Lessells, A., Gruber, L., Lane, D., Steel, C.M. and Leonard, R.C.F. (1992) Overexpression of the p53 protein and allele loss at 17~13 in ovarian carcinoma. Br. J. Cancer, 65, 40-44. Montandon, A.J., Green, P.M., Gianelli, F. and Bentley, D.R. (1989) Direct detection of point mutations by mismatch analysis. Nucleic Acid Res., 17, 3347 -3358. Saiki, R.K., Gelfand, S., Stoffel, S.J., Scharf, R., Higuchi,
16
17
18
19 20
21
22
23
24
25
26
R., Horn, G.T., Mullis, K.B. and Erlich, H.A. (1988) Primer directed amplification of DNA with a thermostable DNA polymerase. Science, 239, 487-491. Prosser, J., Elder, P.A., Condie, A., MacFadyen I, Steel, CM. and Evans, H.J. (1991) Mutations in ~53 do not acount for heritable breast cancer: a study in five affected families. Br. J. Cancer, 63, 181- 184. Winship, P.R., (1989) An improved method for directly sequencing PCR amplified material using dimethyl sulphoxide. Nucleic Acid Res., 17, 1266. Prosser, J., Thompson, A.M., Cranston, G. and Evans, H.J. (1990) Evidence that ~53 behaves as a tumour suppressor gene in sporadic breast cancer. Oncogene, 5, 1573 - 1579. Vogelstein, B. (1990) A deadly inheritance. Nature, 348, 681- 682. Ortta, M., Youichi, S., Takao, S. and Mayashi, K. (1989) Rapid and sensitive detection of point mutations and DNA polymorphisms using the polymerase chain reaction. Genomics, 5, 74 - 79. Myers, R.M., Lumelsky, N., Lereman, L. and Maniatis, T. (1985) Detection of single base substitutions in total genomic DNA. Nature, 313, 495-498. Myers, R.M., Larin, Z. and Maniatis, T. (1985) Detection of single base substitutions by rtbonuclease cleavage at mismatches in RNA/DNA heteroduplexes. Science, 230, 1242 - 1246. Okamoto, A., Sameshima, Y., Yokoyama, S., Terashima, Y., Sugimura, T., Terada, M. and Yokota, J. (1991) Frequent allelic losses and mutations of the p53 gene in human ovarian cancer. Cancer Res., 51, 5171-5176. Maws, R., Pujol, P., Maudelonde, T., Jeanteur, P. and Theillet, C. (1991) p53 mutations in ovarian cancer: a late event? Oncogene, 6, 1685- 1690. Marks, J.R., Davidoff, A.M., Kerns, B.J., Humphrey, P.A., Pence, J.C., Dodge, R.K., Clarke-Pearson, D.L., Iglehart, J.D., Bast, R.C. and Berchuck, A. (1991) Overexpression and mutation of p53 in epithelial ovarian cancer. Cancer Res., 51, 2979- 2984. Wynford-Thomas, D. (1992) P53 in tumour pathology: can we trust immunocytochemistry? J. Pathol., 166, 329 - 330.