J Clin Pathol 1997;50:143-147 143

A novel quantitative immunoassay system for p53 using antibodies selected for optimum designation J Clin Pathol: first published as 10.1136/jcp.50.2.143 on 1 February 1997. Downloaded from of p53 status

M D Thomas, G G McIntosh, J J Anderson, D M McKenna, A H Parr, R Johnstone, T W J Lennard, C H W Horne, B Angus

Abstract The prediction of likely outcome in cancer Aim-To develop a highly sensitive and would permit allocation of patients to appro- specific enzyme linked immunosorbent priate aggressive therapeutic regimens in in- assay (ELISA) system for analysis of p53 stances where prognosis is poor; and con- protein in cancer lysates. versely, patients with good prognosis could be Methods-The anti-p53 monoclonal anti- spared dangerous and debilitating treatment. bodies D07, 1801, BP53.12, and 421, and Mutations of the tumour suppressor gene p53 anti-p53 polyclonal antiserum CM1 were can result in dysfunction, with accumulation of assessed by immunohistochemistry and the inactive/mutated protein which increases genomic instability leading to the expression western blot analysis to identify those and progression of the metastatic phenotype.' most suitable for determining p53 status Accumulation ofp53 has been shown in several of cancer celis. Antibodies with desired different cancers.2" In breast cancer, we and characteristics were used to develop a others have shown that accumulation of p53 is non-competitive sandwich type ELISA associated with poor clinical prognosis.57 system for analysis of p53 expression in Although p53 overexpression can be detected cancer cytosols. Using the ELISA, p53 immunohistochemically, only a qualitative esti- protein concentrations were measured in mation can be made. It has, however, been a small series of breast cancers, and the proposed that quantitative analysis of p53 pro- quantitative values compared with p53 tein concentrations in cancer will provide addi- immunohistochemical data of the same tional objective prognostic information.8 Sev- cancers. eral groups have described non-competitive Results-DO7 and 1801 gave the most sandwich type immunological based assay sys- specific and reliable results on immuno- tems, using monoclonal/polyclonal antibody http://jcp.bmj.com/ histochemisry and western blot analysis. combinations, for quantitative analysis of p53 Using these two antibodies, a non- in tumour cytosol.9-' Using immunohisto- competitive sandwich type ELISA system chemistry, we have recently reported that was developed to p53 assessment of p53 status in breast cancer tissue analyse quantita- sections is critically dependent upon the tively. Analysis of the breast cancer series antibodies used for screening, and influences showed a good correlation between immu- the significance of the prognostic data on September 23, 2021 by guest. Protected copyright. nohistochemistry and the ELISA- generated.5 Similarly, the choice of antibodies tumours were generally positive using to be applied in quantitative p53 assays may both techniques. Discrepancies were exert a notable effect on the final estimate of noted however: some cancers were immu- tumour p53 concentrations. Selection of the Department of nohistochemically negative but ELISA Pathology, Royal most appropriate antibodies will assure the Victoria Infirmary, positive. One explanation for this may be specificity and reproducability of the data on University of that the ELISA is more sensitive than which subsequent assessments of patient prog- Newcastle upon Tyne, immunohistochemistry. nosis may be based. Newcastle upon Tyne Conclusion-The p53 ELISA system is a NEl 4LP double monoclonal M D Thomas non-competitive anti- Methods sandwich G G McIntosh body method, using D07 and TUMOUR TISSUE J J Anderson 1801 which have been shown to be highly Samples of mammary carcinoma tissue were D M McKenna specific for p53 protein by immunohisto- collected at surgery. The tumour was dissected A H Parr and western blot The R Johnstone chemistry analysis. free of superfluous tissue and representative C H W Horne lower threshold of the assay is 0.1 ng/nd samples processed for the production of frozen B Angus analyte in an enriched recombinant p53 sections and cytosols. preparation. As p53 is now regarded as a Department of Clinical protein associated with prognosis in Surgery CYTOSOL PREPARATION T W J Lennard breast and other cancers, the assay may p53 was quantified in cytosol extracts prepared have clinical applications. for routine estimation of oestrogen/ Correspondence to: (7 Clin Pathol 1997;50:143-147) progesterone receptor analysis. Frozen tumour Dr B Angus. tissue (approximately 5 mm2) was powdered Accepted for publication Keywords: p53; p53 antibodies; immunohistochemis- using a micro-dismembranator and resus- 5 November 1996 try; western blot; quantitative ELISA. pended in 1 ml HEPES/EDTA buffer (20 mM 144 Thomas, McIntosh, Anderson, McKenna, Parr, J7ohnstone, et al

HEPES, 1.5 mM EDTA, pH 7.4) containing WESTERN BLOT ANALYSIS

protease inhibitors: aprotinin (2 jg/ml), leu- Specificity of the p53 antibodies was assessed J Clin Pathol: first published as 10.1136/jcp.50.2.143 on 1 February 1997. Downloaded from peptin (2 jg/ml), phenylmethylsulphonyl fluo- using western blot analysis against a cell lysate ride (PMSF) (50 jig/ml), and pepstatin (1 jg/ of the breast adenocarcinoma cell line MDA- ml). Further disruption was achieved using MB-231. Cells mechanically harvested from sonication. Owing to the heat lability of the confluent 250 ml tissue culture flasks were proteins to be assayed, the samples were main- washed three times with PBS and lysed in tained on ice throughout the process. The 1.5 ml ice cold Laemmli buffer (0.125 M Tris, tissue/cell lysate was clarified by ultracentrifu- 4% SDS, 40% glycerol,1 % bromophenol blue, gation, 50 000 x g at 4°C for 40 minutes. and 5% mercaptoethanol, pH6.8). The lysate Cytosol fractions were collected, and stored at was held on ice, sonnicated using a single five -70°C prior to analysis for p53. Samples were second burst, and stored at -70°C until analysed within two weeks of preparation. required. Samples of lysate were fractionated by discontinuous SDS-polyacrylamide gel electrophoresis, using a 10% T, 2.6% C, poly- ESTIMATION OF TOTAL PROTEIN CONCENTRATION acrylamide gel. Proteins were transferred onto Total protein concentration of the cytosol nitrocellulose membrane using a semi-dry preparation was determined in relation to electroblotting system. Blots were probed with bovine serum albumin (BSA) using bicin- p53 antibodies D07, 1801, BP53.12, 421, choninic acid (BCA) protein assay kits (Pierce each at 0.5 jig/ml and with CM1 at dilutions of and Warriner Ltd). 1 in 1000 and 1 in 2000. Reacted p53 antibod- ies were detected using alkaline phosphatase conjugated goat anti-mouse immunoglobulin p53 ANTIBODIES or goat anti-rabbit immunoglobulin polyclonal Anti-p53 monoclonal antibodies D07, 1801 anti-sera. Blots were developed with 0.1 mg/ml and BP53.12, and the rabbit anti-p53 polyclo- nitroblue tetrazolium/0.15 mg/ml BCIP (5- nal antiserum CM1 were all obtained from bromo-4-chloro-3-indolyl phosphate), in de- Novocastra Laboratories Ltd, Newcastle upon velopment buffer (0.1 M Tris buffer, 0.1 M Tyne, UK. Anti-p53 monoclonal antibody 421 NaCl and 50 mM MgCl2, pH 9). was obtained from Oncogene Science Ltd (Sera Laboratories Ltd, Crawley Down, UK). SELECTION OF p53 ANTIBODIES FOR ELISA The concentration of monoclonal antibodies DEVELOPMENT 1801, D07, and BP53-12 was determined p53 antibodies for the development of the using IgG radial immunodiffusion. ELISA were selected on the basis of their per- formance in immunohistochemistry, for opti- IMMUNOHISTOCHEMISTRY mum assignment of p53 status, and their

To assess the immunohistochemical perform- specificity for p53 in western blot analysis. http://jcp.bmj.com/ ance of antibodies directed against p53 for the Using these criteria, monoclonal antibodies assignment of p53 status, we stained serial 1801 and D07 were selected for use in assay cryostat sections of breast cancers with test development. monoclonal antibodies D07, 1801, BP53.1 and 421, and the rabbit polyclonal antiserum PREPARATION OF BIOTINYLATED Do7 CM1. Sections, 3 jim thick, of cryopreserved Monoclonal antibody D07 was purified from

hybridoma culture supernatant by absorption on September 23, 2021 by guest. Protected copyright. tumour tissue were cut and air dried onto with protein A (Sigma). Antibody was eluted at microscope slides. Endogenous tissue peroxi- pH 3 and dialysed against 0.1 M borate buffer, dase activity was quenched using 0.6% hydro- pH 8.8. Purified antibody was then reacted gen peroxide in methanol, for five minutes. with ester of biotinamidocaproate N-hydroxy- Subsequently, non-specific immunostaining succinimide, in dimethyl sulphoxide (DMSO), was blocked by incubating the sections with (100 jg ester/mg monoclonal antibody) for 5% lamb serum/phosphate buffered saline four hours at room temperature. The conjuga- (PBS) for 30 minutes. Test p53 monoclonal tion reaction was terminated by the addition of antibodies at 0.5 jg/ml, or rabbit anti-serum 1 M ammonium chloride (259 l/mg mono- CM1 at a dilution of 1 in 1000 were applied clonal antibody), and the products extensively and incubated with sections for one hour at dialysed against 0.1 M Tris/HCl, pH 7.4. room temperature. Binding of p53 antibody was detected using biotinylated goat anti- ELISA PROCEDURE mouse or goat anti-rabbit polyclonal anti- A sandwich type ELISA system was developed serum and streptavidin/biotinylayed horseradish for the quantitative analysis of p53. Solid phase peroxidase complex (Dako, High Wycombe, monoclonal antibody 1801 was used as the p53 UK). Sections were developed with 3,3- immobiliser, and biotinylated D07 as the p53 diaminobenzidine tetrahydrochloride (Sigma, detector. In brief, Immulon-1 96-well ELISA Poole, Dorset, UK) in Tris buffered saline trays (Dynatech Laboratoies, Billinghurst, UK) (TBS) containing 0.3% hydrogen peroxide, were coated overnight at 4°C with 50 jl/well counterstained with Meyer's haematoxylin, de- 10 jg/ml 1801 in PBS. Unbound antibody was hydrated stepwise in ethanol, cleared in xylene, removed by three washes with PBS (this buffer and mounted. Tumours were scored according was used for all subsequent washing steps), and to the proportion of cells showing nuclear reac- the wells incubated for one hour at 37°C with tivity: -, no reactivity; +, <25% reactivity; ++, 200 jl 5% BSA in PBS, to block non-specific 25-50% reactivity; and +++, >50% reactivity. immunoreactivity. Wells were then washed Novel quantitative immunoassay system for p53 145

once and incubated for three hours at 370C Table 1 Percentage of breast cancers immunohistochemically scored as p53 positive using with 50 tld standard recombinant p53 protein, antibodies D07, 1801, 421, BP53. 12, and CM1 tumour cytosol at 2 mg/ml, 1 mg/ml, 0.5 mg/ J Clin Pathol: first published as 10.1136/jcp.50.2.143 on 1 February 1997. Downloaded from ml, and 2.5 mg/ml, or patient serum diluted 1 Antibody Type/species Per cent positive cases in 10 and 1 in 100. In each instance, 2% BSA D07 Monoclonal/mouse 46 in PBS was used as the diluent. After washing 1801 Monoclonal/mouse 45 three times, the wells were next incubated for 421 Monoclonal/mouse 42 two hours at 37°C with 50 gl biotinylated D07 CM1 Polyclonal/rabbit 17 diluted 1 in 10 in 2% BSA in PBS (this buffer BP53. 12 Monoclonal/mouse 16 was used for all subsequent antibody dilu- tions). The wells were washed again three times and then incubated for one hour at 37°C with 50 gl sheep anti-biotin polyclonal antiserum (The Binding Site, Birmingham, UK) diluted 1 in 1000. Specificity of this reagent was assured by extensive pre-absorption with mouse IgG- agarose. After washing three times, the wells were incubated for one hour at 37°C with 50 gl horseradish peroxidase conjugated donkey 69 K- anti-sheep immunoglobulin anti-serum (Stra- tech Scientific Ltd, Luton, UK) diluted 1 in 46 K- 4000. The wells were washed a final four times and developed with 50 gl 3,3' 5,5'- tetramethylbenzidine substrate. The reaction was stopped with 0.2 M sulphuric acid and results recorded at 250 nm using a Titertek multi-scan plate reader. The assay was stand- ardised using an enriched preparation of recombinant p53 protein (kindly supplied by Professor D Lane). Test absorbance values were corrected against control values derived from wells processed in the absence of 1 2 3 4 5 biotinylated D07. p53 concentrations were Figure 1 Western blot analysis ofp53 antibodies against a calculated with reference to the linear range of cell lysate ofthe breast carcinoma cell line MDA-MB-231. Lane 1, D07; lane 2, 1801; lane 3, BP53. 12; lane 4, CMJ standard recombinant p53 titration curves. (diluted 1 in 1000);lane 5, CMJ (diluted 1 in 2000). Cytosol p53 values were expressed as p53 ng/ mg total cytosol protein, and serum values as p53 ng/ml serum. cers. All samples scored positive using http://jcp.bmj.com/ BP53.12, 421 or CM1 were also positive on staining with D07 and 1801. Results IMMUNOHISTOCHEMISTRY WESTERN BLOT ANALYSIS Using a series of breast carcinomas, we Specificity of the p53 antibodies for p53 assessed the immunohistochemical perform- protein was assessed by western blot analysis ance of p53 antibodies for the assignment of against lysate ofthe breast adenocarcinoma cell on September 23, 2021 by guest. Protected copyright. tumour p53 status. Each antibody was assessed line MDA-MB-23 1. Figure 1 shows that on consecutive sections cut from each tumour monoclonal antibodies D07,1801 and BP53.1 sample. For immunohistochemistry, cryopre- exhibit specificity for a single protein, the served tissue was used in preference to forma- molecular weight of which is consistent with lin fixed, paraffin wax embedded tissue as the that of p53 (monoclonal antibody 421 at 1 jg/ results generated would best reflect the per- ml failed to detect p53). In comparison, CM1, formance of the antibodies in ELISA for the at different concentrations, reacted with several analysis of p53 in cancer cytosol. Positive cell proteins in addition to a weak reaction with staining with any of the monoclonal antibodies p53. We suggest that this polyreactivity is tested was restricted to the nucleus, and was analogous to CM1 cytoplasmic staining of consistent with the nuclear localisation of p53. breast cancers which are p53 negative-that is, In comparison, however, CM1 commonly no nuclear reactivity with CM1, D07, 1801, exhibited reactivity within the cell cytoplasm and 421. which was independent of nuclear staining, suggesting reactivity with cellular antigens in ELISA addition to p53. On this basis, cancers assessed Results generated from immunohistochemical for p53 using CM1 were scored p53 positive and western blot analysis of the p53 antibodies only if nuclear staining was apparent. suggest that ELISA systems incorporating Table 1 shows the percentage of cancers combinations of monoclonal antibodies 1801 scored as p53 positive using each of the and D07 would provide a highly specific antibodies tested. A higher percentage of system for the analysis of p53 in most cancers. cancers were scored p53 positive with mono- Therefore, an ELISA system was developed clonal antibodies D07 and 1801 than any of using solid phase 1801 as the p53 immobiliser the other antibodies tested, indicating that and biotinylated D07 as the p53 detector. D07 and 1801 identify most p53 positive can- Subsequent detection and amplification of 146 Thomas, McIntosh, Anderson, McKenna, Parr,Johnstone, et al J Clin Pathol: first published as 10.1136/jcp.50.2.143 on 1 February 1997. Downloaded from Horseradish peroxidase Ue conjugated donkey anti- 0 0 0 sheep immunoglobulin 40in E Sheep anti-biotin C._ 0E ++ F- 0 0 0 nylated p53 antibody D07 C

E EQ + - O 0 0m 0 Solid phase E E Figure 2 Schematic representation ofthe p53 quantitative r- ELISA. GD )CXD (D 0.5 0 5.0 10.0 15.0 20.0 Cystosol p53 concentration (ng/ml) Figure 4 Comparison between immunohistochemical p53 staining (using 1801 and D07) ofbreast cancer tissue and corresponding cytosolp53 values (assessed using the 0.4 quantitative ELISA).

chemically scored as p53 positive, using DO7/ 1801, and corresponding positive cytosol values was shown. In no case were any cancers .3 0 scored positive by immunohistochemistry (N which did not have a cytosol p53 value statisti- C.)0) cally distinguishable from the background value. Several cancers immunohistochemically .0 o 0.2 scored p53 negative were, however, positive for p53 in ELISA. p53 was not detected in any of the serum samples tested. http://jcp.bmj.com/ Discussion 0.1 _ Mutation ofthe tumour suppressor gene p53 is a common event in carcinogenesis, and can lead to accumulation of the mutated protein.'" Accumulation of the p53 protein has been 0.0 assessed qualitatively using immunohisto-

0.1 1 10 chemistry and recently quantitatively using on September 23, 2021 by guest. Protected copyright. p53 concentration (ng/ml) immunoassay based systems. Both qualitative and have shown that Figure 3 Typical titration curve for detection of quantitative analyses p53 recombinant p53 using the quantitative ELISA. overexpression in breast cancers is commonly associated with poor prognosis.5 7 If assessment D07 binding was achieved using sheep anti- of p53 overexpression is to have any important biotin and horseradish peroxidase conjugated clinical implications in cancer prognosis, we donkey anti-sheep immunoglobulin polyclonal believe it is necessary to have screening systems anti-sera (fig 2). A typical calibration curve for which are highly sensitive and specific for the the p53 assay using an enriched preparation of analyte. For this reason, we have developed a standard recombinant p53 is shown in fig 3. highly specific and sensitive immunoassay sys- The detection limit, the concentration of p53 tem which can be used to analyse quantitatively which could be distinguished from the zero p53 concentrations in most cancers. The assay value with 95% confidence, was 0.1 ng/ml of system incorporates two monoclonal antibod- the enriched p53 protein. ies directed against p53 which exhibit high Concentrations of p53 were measured using specificity for the antigen, and detect immuno- the ELISA system in a small series of breast histochemically a high proportion of p53 posi- cancer cytosols and in serum samples from tive cancers. patients with breast cancer. Cytosol values In a recent paper, we have shown that the ranged from zero (values which could not be immunohistochemical assessment of p53 status distinguished from the control with statistical in formalin fixed, paraffin wax embedded cancer significance) to 14.9 ng/mg cytosol protein. tissue is critically dependent on the p53 Comparisons of the quantitative values with antibodies used for screening.5 In the present immunohistochemical staining of frozen sec- study, using cryopreserved, non-fixed breast tions of the same cancer tissue, using antibod- cancer tissue, immunohistochemical assessment ies 1801 and D07, are shown in fig 4. Good of p53 has shown that a higher percentage of correlation between cancer immunohisto- cancers are scored positive using monoclonal Novel quantitative immunoassay system for p53 147

antibodies 1801 and D07, in comparison with general, high immunohistochemically p53 421 and BP53.12, and that polyclonal antibody positive cancers had corresponding high quan- CM1 performs poorly. These findings are simi- titative p53 values. About 38% of the series J Clin Pathol: first published as 10.1136/jcp.50.2.143 on 1 February 1997. Downloaded from lar to those reported in routinely processed scored immunohistochemically negative had tissue.5 In addition to the immunohistochemical positive values in ELISA, with values ranging assessment of p53 antibodies, we have also ana- from 0.3 to 2 ng/mg. Owing to the high specifi- lysed their specificity for p53 by western city of the antibodies used in the ELISA, we blotting. CM1 was again found to perform suggest that these cancers are indeed p53 posi- poorly, exhibiting only weak reactivity with p53 tive, but the levels expressed are below the and stronger reactivity with other cellular threshold value for immunohistochemical de- antigens. In comparison, monoclonal antibodies tection. Alternatively, the variation between 1801, D07 and BP53.12, exhibit specificity for cytosol and immunohistochemical results may a single protein with molecular weight indicative simply reflect heterogeneity within a tumour, of p53. From analysis of these data, we suggest suggesting that more relevant cancer p53 data that quantitative immunoassay systems for p53 will be obtained from assessment of samples using CM1 may not produce consistently taken from several different sites of the primary specific and reproducible results. As monoclonal cancer. In line with previous findings, p53 was antibodies D07 and 1801 were preferentially not detected in the serum samples." selected for both their detection of p53 in In conclusion, we suggest that our method is immunohistochemistry and their specificity for a precise and sensitive immunologically based p53 in western blot analysis, we suggest that the system for the quantitative analysis of p53 pro- quantitative immunoassay system we describe tein. Whether the precise estimation ofp53 will has advantages over other systems for analysis of provide more significant prognostic data will be tumour p53 concentrations. assessed by the analysis of a larger series of D07 and 1801 bind epitope regions located cancer cytosols. at the N-terminus of the p53 molecule: 1801 recognises an epitope mapping between amino 1 Lane DL. P53 guardian of the genome. Nature acid residues 32-7912 and D07 within the 1992;358:15-16. 2 Hollstein M, Sidransky D, Vogelstein B, Harris CC. P53 region defined by amino acids 20-25.'3 Both mutations in human cancers. Science 1991 ;253:49-53. regions are distant from the central portion of 3 Nigro JM, Baker SJ, Preisinger AC, Jessop JM, Hostetter R, Cleary K, et al. Mutations in the p53 gene occur in diverse the p53 molecule which is most commonly human tumour types. Nature 1989;342:705-8. affected by point mutations.'4 We, therefore, 4 Bartek J, Bartkova J, Voijesek B, Staskova Z, Lukas J, Rejthar A, et al Aberrant expression of the p53 onco-protein is a suggest that these epitope regions will be struc- common feature of a wide spectrum of human malignan- turally conserved in a high proportion of cies. Oncogenel991;6:1699-703. 5 Horne GM, Anderson JJ, Tiniakos DG, McIntosh GG, cancers, and, thus, it is expected that in most Thomas MD, Angus B, et al P53 protein as a prognostic cancers p53 protein concentrations can be indicator in breast carcinoma: a comparison of four antibodies for immunohistochemistry. Br J Cancer 1986; measured using this antibody combination. No 73:29-35. competition of p53 binding was observed 6 Elledge RM, Fuqua SAW, Clark GM, Pujol P, Allred G, http://jcp.bmj.com/ McGuire WL. Prognostic significance of p53 gene between the two antibodies when using solid alterations in node negative breast cancer. Breast Cancer Res phase 1801 and D07 as the p53 detector. To Treat 1993;26:225-35. 7 Iwaya K, Tsuda H, Hiraide H, Tamaki K, Takamura S, accommodate the reduction in amplification Fukutomi T, et al. Nuclear p53 immunoreaction associated accompanying the use of a monoclonal anti- with poor prognosis ofbreast cancer. J Cancer Res 1991;82: 835-40. body as p53 detector in comparison with a 8 Thomas MD, Parr AH, McIntosh GG, Anderson JJ, Ellis I, polyclonal anti-serum-for example, CM 1, we Nicholson R, et al. Measurement of p53 protein in human

breast cancer: comparison between immunohistochemistry on September 23, 2021 by guest. Protected copyright. labelled D07 with biotin and used sheep anti- and semi-quantitative ELISA values. Pathol 1995; biotin anti-sera and horseradish peroxidase 176(Suppl): 172. 9 Vojtesek B, Fisher CJ, Barnes DM, Lane DP. Comparison conjugated donkey anti-sheep immunoglobu- between p53 staining in tissue sections and p53 protein lin anti-sera. This approach permitted the levels measured by ELISA technique. BrJ' Cancer 1993;67: 1254-8. detection of 0.1 ng/ml of p53 in an enriched 10 Levesque MA, Diamandis EP, Yu H, Sutherland JA. Quan- recombinant p53 preparation. This is less than titative analysis of mutant p53 protein in breast tumour 10 cytosols and study of its association with other biochemical that described by other assays using CM1.9 prognostic indicators in breast cancer. Breast Cancer Res We are, however, confident in suggesting that Treat 1994;30:179-95. 11 Hassapoglidou S, Diamandis EP, Sutherland DJA. Quanti- the analyte measured in our system is p53. fication of p53 protein in tumour cell lines, breast tissue The ELISA system was used to assess quan- extracts and serum with time-resolved immunofluorom- etry. Oncogene 1992;8:1501-9. titative p53 values in a small series of breast 12 Banks L, Matlashewski G, Crawford L. Isolation of human cancers, and the results compared with immu- p53 specific monoclonal antibodies and their use in the

study of human p53 expression. Eur _ Biochem 1986;159: nohistochemical staining of frozen sections 529-34. from representative samples of the same cancer 13 Stephen CW, Heiminen P, Lane DP. Characterisation of epitopes on human p53 using phage displayed peptide tissue, using antibodies 1801 and D07. All libraries: Insight into antibody peptide interactions. 3 Mol cancers with positive immunohistochemical Biol 1995;248:58-78. 14 Stephen CW, Lane DP. Mutant conformation of p53. status were similarly positive in ELISA, with Precise epitope mapping using filamentous phage epitope p53 values ranging from 1 to 14.9 ng/mg. In library.3 Mol Biol 1992;225:577-83.