cleic Acids Res. 16:265-277.6.Sambrook, J., E.F. Fritsch and T. Maniatis.

1989. Molecular Cloning: A Laboratory Man-ual, 2nd ed. CSH Laboratory Press, ColdSpring Harbor, NY.

7.Schaefer, B.C. 1995. Revolutions in rapid am-plification of cDNA ends: new strategies forpolymerase chain reaction cloning of full-length cDNA ends. Anal. Biochem. 227:255-273.

8.Tillett, D. and B.A. Neilan. 1999. Simple n-butanol purification of dye terminator sequenc-ing reactions. BioTechniques 26:606-610.

9.Tillett, D. and B.A. Neilan. 1999. Xan-thogenate nucleic acid isolation from culturedand environmental cyanobacteria. J. Phycol.(In press.)

10.Troutt, A.B., M.G. McHeyzer-Williams, B.Pulendran and G.J.V. Nossal. 1992. Liga-tion-anchored PCR: A simple amplificationtechnique with single-sided specificity. Proc.Natl. Acad. Sci. USA 89:9823-9825.

This work was supported by grants fromthe National Health and Medical ResearchCouncil and the Australian Research Coun-cil. We are grateful to M. Cairns for the kindgift of the cordecypin modified oligonu-cleotide. Address correspondence to Dr.Brett Neilan, School of Microbiology andImmunology, University of New SouthWales, Sydney 2052, Australia. Internet:b.neilan@unsw.edu.au

Received 3 May 1999; accepted 25October 1999.

Daniel Tillett, Brendan P.Burns and Brett A. NeilanUniversity of New South WalesSydney, Australia

Monitoring and Purifica-tion of Proteins UsingBovine Papillomavirus E2 Epitope TagsBioTechniques 28:456-462 (March 2000)

ABSTRACT

We describe here the use of two newlymapped bovine papillomavirus type 1(BPV-1) E2 protein epitopes as tags. We

constructed several vector plasmids foroverexpression as well as for moderateexpression of single- or double-tagged pro-teins in either Escherichia coli or eukaryot-ic cells. The new tags were fused to severalproteins, and the activity of the tagged pro-teins was tested in different assays. The tagswere shown not to interfere with the func-tion of these proteins in vivo and in vitro. In-teraction of the monoclonal antibodies3F12 and 1E2 with their respective epitopeswas specific and had high affinity in a vari-ety of conditions. We have demonstratedthat the 3F12 antibody-epitope interactiontolerates high salt concentrations up to 2 M.This permits immunoprecipitation andimmunopurification of the tagged proteinsin high-salt buffers and reduction of thenonspecific binding of the contaminatingproteins. We also provide a protocol forDNA binding and DNase I footprinting as-says using the tagged, resin-bound DNA-binding proteins. The BPV-1 E2-derivedtags can be recommended as useful tools fordetection and purification of proteins.

INTRODUCTION

Epitope tagging is a recombinantDNA technique by which a protein ismade immunoreactive to a preexistingantibody. This technique simplifies de-tection, characterization, purificationand in vivo localization of proteins andhas become a standard method of mole-cular genetics (3). However, some tagsare not useful in certain applicationsdue to high background binding. Insome cases, affinity purification andimmunoprecipitation of a tagged pro-tein is problematic due to co-precipita-tion of contaminating proteins.

Here, we describe the use of two re-cently mapped bovine papillomavirustype 1 (BPV-1) E2 protein epitopes astags. We constructed several vector plas-mids for over-expression as well as formoderate-level expression of either sin-gle- or double-tagged proteins inEscherichia coli and eukaryotic cells.The new tags were fused to functionallydifferent proteins: a bacterial transcrip-tional activator, XylS, that agregates andbecomes nonfunctional at high levels ofexpression, several mutants of the tumorsuppressor protein p53 for overexpres-sion in E. coli and rat glyceraldehyde-3-phosphate dehydrogenase (GAPDH) for

expression in eukaryotic cells. Hereby,we report the use of BPV-1 E2-derivedtags and respective monoclonal antibod-ies for the identification of proteins onWestern blots, in the immunofluores-cence staining of cells and DNA band-shift assay. In addition, we describe anapplication of the resin-bound taggedprotein in DNA binding and DNase Ifootprinting assays.

MATERIALS AND METHODS

Plasmid Constructions

For the construction of pBR-3F12and pBR-1E2, we inserted the codingsequence of XylS with the N-terminal-ly fused influenza virus hemagglutinin(HA) epitope (2) from the pET11c-based parent plasmid pETSN117 (4)between the NheI and BamHI sites ofpBR322. The subcloned fragment con-tained in addition to a ribosome bind-ing site, an extra NdeI site and a startcodon preceding the tag sequence. Theresultant plasmid pBRSN 117 containsXbaI and BamHI sites for cloning of arecombinant coding sequence (4).Then, the coding sequence of the HAepitope between NdeI and XbaI siteswas replaced with double-stranded syn-thetic oligonucleotides encoding pep-tides GVSSTSSDFRDR and TTGH-YSVRD, recognized by anti-BPV-1 E2monoclonal antibodies 3F12 and 1E2,respectively (7).

For the construction of pBR-NC, weamplified xylS sequence by PCR usingthe 3-end primer containing a codingsequence for the peptide TSSDFRDR.The peptide recognized by 3F12 Mabwas fused in frame to the C-terminus ofXylS and flanked by KpnI and BamHIsites. The resultant PCR fragment wascloned into pBR-1E2 generating theexpression plasmid with cloning sitesXbaI and KpnI.

pET-3F12 was generated by cloningthe BamHI/NdeI fragment from pBR-3F12-xylS into the corresponding sites inpET-11c. Then, the XylS gene was re-moved by cleavage with XbaI andBamHI and replaced with coding se-quences for different mutant p53 proteinsbearing XbaI and BglII sites at the ends.

For the construction of pCG-3F12,double-stranded synthetic oligonu-

Short Technical Reports

456 BioTechniques Vol. 28, No. 3 (2000)

cleotide encoding 3F12 epitope and in-corporating multicloning sites was in-serted between the XbaI and BglII sitesof pCG (9). Plasmid pCG-3F12-GAPDH contains rat glyceraldehyde-3-phosphate dehydrogenase (GAPDH)inserted between BamHI and BglII sitesof pCG-3F12. Plasmid pCG-3F12-GAPDH-NLS contains nuclear local-ization signal of human p53 (aminoacids 305327) fused to the C-terminusof the GAPDH protein.

Immunoaffinity Bindingof 3F12-XylS

pBR-3F12-xylS bearing E. coliDH5a cells were grown in LB medium

supplemented with 100 mg/mL ampi-cillin at 20C to an A600 of approximate-ly 1.0. Cells were harvested, washedwith TBS and resuspended in 1/10 vol-ume of high-salt lysis buffer containing100 mM Tris-HCl, pH7.5, 1.5 M NaCland 5 mM EDTA, 20% (wt/vol) glyc-erol. Cell suspension was frozen in liq-uid nitrogen and stored at -70C.

For the preparation of crude extracts,cells were thawed and dithiothreitol(DTT; 10 mM), PMSF (100 mg/mL),aprotinin (1 mg/mL), and CHAPS(0.2%) were added. Cells were incubat-ed with lysozyme (0.5 mg/mL) on icefor 20 min and disrupted by sonication.The lysate was clarified by centrifuga-tion at 40 000 g at 4C for 30 min.

Affinity beads were prepared bycoupling 3F12 anti-BPV E2 monoclon-al antibody to divinylsulfon-activatedToyopearl HW65 TSK-gel (TOSOH,Tokyo, Japan). The 3F12-TSK affinitybeads were incubated with crude lysateat 4C for 1 h with end-over-end agita-tion. Beads were washed extensivelywith high-salt lysis buffer on glass fil-ter. These 3F12-XylS beads werestored at -20C in approximately 10 gelvolumes of storage buffer containing10 mM Tris-HCl, pH 7.5, 100 mMKCl, 10 mM MgCl2, 80 mM EDTA, 80mM DTT, 50% (vol/vol) glycerol, 100mg/mL PMSF and 1 mg/mL aprotinin.These beads were used for DNA bind-ing and footprinting assays.

Vol. 28, No. 3 (2000) BioTechniques 457

Table 1. Main Characteristics of Expression Vectors

DNA Precipitation Assay

End-labeled restriction fragments ofpUPM190 HpaII/HinfI digest and 50mL of the suspension of 3F12-XylSbeads were mixed with 0.3 mL of DNAbinding buffer containing 50 mM Tris-HCl, pH 7.5, 100 mM KCl, 2 mMEDTA, 10% (wt/vol) sucrose, 0.1%CHAPS, 700 mg/mL BSA, 1 mg/mLaprotinin and 1 mM meta-toluate. Themixture was incubated for 30 min atroom temperature with end-over-endagitation. Beads were washed 3 with 1mL of DNA binding buffer and incubat-ed with Proteinase K (50 mg/mL) at37C for 30 min in stop solution con-taining 200 mM NaCl, 5 mM EDTAand 1% SDS. The retained DNA was re-leased by phenol-extraction, ethanol-precipitated, and identified in 5% PAGEunder non-denaturing conditions.

DNase I Footprinting

The Om-containing EcoRI/XhoIfragment of pUPM190 was end-labeledat XhoI site in the lower strand withKlenow fragment. The labeled probewas incubated with 50 mL of the sus-pension of 3F12-XylS beads in DNAbinding buffer as for the DNA precipi-tation. The beads were washed 3 andsuspended in 100 mL of DNase I buffercontaining 10 mM Tris-HCl, pH 7.5,100 mM KCl, 5 mM MgCl2, 1 mMCaCl2, 100 mM DTT, 100 mM EDTA,100 mg/mL BSA and 2 mg/mL sonicat-ed salmon-sperm DNA. DNase I wasadded to concentrations of 0.63.0mg/mL, and the reaction mixture wasincubated for 1 min at 30C. The reac-tion was stopped by 200 mL of stop so-lution containing 600 mM NaCl, 120mM EDTA, 3% SDS and 150 mg/mLdextran, and the beads were treatedwith Proteinase K (50 mg/mL) at 37Cfor 30 min. DNA was extracted fromthe beads with phenol/chloroform, pre-cipitated with ethanol and analyzed onsequencing gel. DNase I cleavage ofthe unbound template was carried outin the same buffer at DNase I concen-trations of 30300 ng/mL.

Gel-Shift Assay

Gel-shift assay was performed aspreviously described (1). We used the

crude lysate preparations of p53 mu-tants. The double-stranded p53 con oli-go (5-GAT CCG GAC ATG CCCGGG CAT GTC CGG ATC-3) wasused as a probe. Protein-DNA com-plexes were separated from the un-bound DNA on 5% PAGE (55:1).

Immunodetection of Proteins

For Western blot analysis, cells ex-pressing the tagged proteins were lyzedin SDS sample buffer, total protein ex-tracts were separated on 12% SDS-PAGE and electroblotted onto nitrocel-lulose membrane filters. Blots wereblocked with 1% non-fat dry milk for 1h, incubated with anti-BPV E2 1E2,3F12 or anti-p53 pAb 240 monoclonalantibodies, followed by incubation withanti-mouse IgG alkaline phosphatase-conjugated secondary antibody.

For immunofluorescence staining,Saos-2 cells were transfected with plas-mids expressing 3F12-GAPDH-NLS or3F12-GAPDH proteins. Cells weregrown on microscopy cover glasses andfixed 24 h after transfection withmethanol at -20C. Proteins were detect-ed with 3F12 and anti-mouse Ig anti-body conjugated with FITC as a prima-ry and secondary antibody, respectively.

RESULTS AND DISCUSSION

Moderate Level Bacterial Expressionof Epitope-Tagged Proteins

The XylS protein, a transcriptionalactivator from the TOL plasmid pWWOof the soil bacterium Pseudomonas puti-da, like some other AraC/XylS familytranscription factors, does not toleratehigh level of over-expression and isprone to aggregation both inside the cellas well as in the solution, in the courseof purification (4,5). We expressed XylSand several truncated variants of theprotein at a near to native level in E. coliand tested their physiological activitiesin vivo. For that, we constructed vectorsfor the moderate level of expression ofepitope-tagged fusion proteins. The vec-tors were based on pBR322, and the tetpromoter of this plasmid was used forexpression of recombinant proteins. Thevectors pBR-3F12 and pBR-1E2 wereconstructed for the expression of pro-

teins with N-terminally fused peptidesGVSSTSSDFRDR and TTGHYSVRD,recognized by anti-BPV E2 monoclonalantibodies 3F12 and 1E2, respectively(7). Both vectors contain XbaI andBamHI sites for cloning of the recombi-nant sequence. We also constructed thevector pBR-NC for the expression ofproteins with different epitope tags inboth N- and C-termini. The N-terminaltag was TTGHYSVRD as in pBR-1E2and the C-terminal tag was TSSDFR-DR, a shorter version of the epitope rec-ognized by 3F12 Mab. The cloning sitesfor a coding sequence in pBR-NC areXbaI and KpnI (Table 1).

We transformed E. coli DH5a withplasmids expressing the tagged ver-sions of XylS and analyzed the expres-sion of the tagged XylS proteins byWestern blotting. The 3F12 antibodyrecognized both 3F12-XylS and NC-XylS proteins, while the 1E2 antibodyrecognized 1E2-XylS and NC-XylSproteins as single bands on the Westernblot (Figure 1A, lanes 16). No cross-reaction with cellular proteins was ob-served. However, when 3F12 Mab wasused for the detection of the double-tagged NC-XylS protein, a much weak-er signal was detected when comparedwith the 1E2 Mab signal. (Figure 1A,lanes 4 and 6). This could be explainedby the use of the shorter version of the3F12-specific epitope in the double-tagged protein. We studied the effect ofthese tags on the activity of XylS pro-tein in E. coli strain CC118Pm-lacZ,which carries a chromosomal copy ofthe XylS responsive Pm promoter fusedto the lacZ gene (6). The tags had no ef-fect on the transcriptional activation byXylS (data not shown).

Study of Specific DNA Bindingby the Matrix-AttachedEpitope-Tagged Protein

XylS is a DNA-binding protein,which specifically binds to the Om op-erator sequence and activates the re-sponsive Pm promoter. However, suit-able conditions to study the solubleXylS in vitro have not been found, asthe protein tends to aggregate. Todemonstrate the site-specific DNAbinding of epitope-tagged XylS in vit-ro, we used immunobound 3F12-XylSprotein that was attached to the TSK

Short Technical Reports

458 BioTechniques Vol. 28, No. 3 (2000)

beads through the N-terminal tag and3F12 antibody. Western blot analysisshowed that high salt concentrations,up to 2 M NaCl, do not hinder the inter-action of 3F12 MAb with the specificepitope (data not shown). Therefore, toavoid aggregation of XylS and coim-munoprecipitation of contaminatingproteins, the crude lysate was preparedin a high-salt lysis buffer, containing1.5 M NaCl, and 3F12-XylS was boundto the affinity resin carrying 3F12 Mabby a single-step, batchwise procedure.

The DNA binding assay was carriedout by mixing the 3F12-XylS beadswith a mixture of end-labeled restric-tion fragments of the Om-containingplasmid pUPM190 (4). After severalwashes, the bound DNA was releasedand identified by gel electrophoresis,using the input mixture of fragments asa marker. Figure 2A shows that only a

single Om-containing fragment of thepUPM190 digest was retained on thebeads, and binding of any other frag-ment could not be detected.

Further, we analyzed the interactionof the immobilized 3F12-XylS proteinwith Om by DNase I footprinting. Thespecific complexes of the Om-contain-ing DNA fragment and TSK-bound3F12-XylS were prepared identicallyas for DNA precipitation and weretreated with DNaseI. After the cleav-age, DNA was extracted from the beadsand analyzed on a sequencing gel. As acontrol, the unbound template wascleaved at lower concentrations ofDNaseI to obtain the equal rate ofcleavage. Figure 2B shows that 3F12-XylS protects a 44 bp area on the lowerstrand and four DNase I hypersensitivesites occur within the protected region.We obtained an identical DNaseI foot-

Vol. 28, No. 3 (2000) BioTechniques 459

Figure 1. Detection of the tagged proteins. (A) Western blot analysis of the tagged proteins. Total proteinextracts, separated by 12% SDS-PAGE and electroblotted, were analyzed using 1E2 (lanes 13), 3F12(lanes 49 and 1315) or anti-p53 pAb 240 (lanes 1012) primary antibodies and alkaline phosphatase-conjugated secondary antibody. Lanes 16: E. coli DH5a cells producing tagged XylS protein, bearingpBR-3F12 (lanes 1 and 4), pBR-1E2 (lanes 2 and 5) and pBR-NC (lanes 3 and 6) derived expression plas-mids. Lanes 712: E. coli BL21 (DE3) cells bearing pET-3F12 derived expression plasmids, producing3F12-tagged p53 variants DN39DC362 (lanes 7 and 10), DN39DC362trp248 (lanes 8 and 11) andDN61DC362 (lanes 9 and 12). Lanes 13 and 14: Saos-2 cells transfected with pCG-3F12 derived expressionplasmids producing 3F12-tagged GAPDH (lane 13) and GAPDH-NLS (lane 14). Saos-2 cells expressinguntagged p53 were used for a negative control (lane 15). (B and C) Subcellular localization of 3F12-GAPDH and 3F12-GAPDH-NLS proteins. Saos-2 cells were transfected with pCG-3F12-GAPDH (B) orpCG-3F12-GAPDH-NLS (C) and tagged proteins were detected with immmunofluorescence analysis.

print earlier with HA-epitope taggedXylS, using similar technical approach(4). DNA precipitation and DNase Ifootpriting with a resin-bound epitope-tagged protein (10) are the methods ofchoice for proteins that are prone to ag-gregation and are difficult to purify.

High Level Bacterial Expressionof Epitope-Tagged Proteins

For high-level bacterial expression ofN-terminally tagged proteins, we con-structed the vector pET-3F12, a deriva-tive of pET-11c that contains the tag-en-coding sequence and the cloning sitesidentical to pBR-3F12 (Table 1). Thecoding sequences for mutant p53 pro-teins were cloned into the vector. Mutantproteins DN39DC362 and DN61DC362had their transactivation and regulatoryparts deleted, but maintained the abilityto bind DNA, while the mutant proteinDN39DC362trp248 had lost its DNA-

binding ability due to the point mutationin its DNA-binding domain (8). Plas-mids, generated for the T7 promoter-di-rected expression of 3F12-tagged p53fusion proteins were transformed into E.coli strain BL21 (DE3).

Expression of the tagged p53 pro-teins was monitored by Western blotanalysis. The tag-specific 3F12 anti-bodies and p53-specific pAb240 anti-bodies were used for the detection ofthe proteins. Using this vector systemresulted in enormous overexpression ofthe protein detected with 3F12 Mab aswell as with pAb240 antibodies (Figure1A, lanes 712).

The DNA-binding activity of thetagged p53 proteins was studied in aband-shift assay (Figure 2C). The p53mutant DN39DC362trp248, carrying apoint mutation in its DNA-binding do-main, was used as a negative control.The mutants DN39DC362 and DN61-DC362 are functional in DNA binding

and produced a shifted band. To showthat the produced complex really con-tains p53, the protein-DNA complexeswere supershifted with the tag-specificmonoclonal antibody 3F12. This way,epitope tags can be used to verify speci-ficity of the shifted complex in a DNAband-shift assay without purification ofthe protein.

Expression of Epitope-TaggedProteins in Eukaryotic Cells

For eukaryotic expression, two cod-ing sequences of rat glyceraldehyde-3-phosphate dehydrogenase (GAPDH)gene were cloned into pCG-3F12 plas-mid (Table 1). The first had 3F12 epi-tope fused in frame to GAPDH aminoacids 2333 (pCG-3F12-GAPDH) andthe second also contained a nuclear lo-calization signal of p53 (amino acids305327) fused to the C-terminus ofthe GAPDH protein (pCG-3F12-GAP-

Short Technical Reports

DH-NLS).Saos-2 cells were transfected with 1

mg of expression plasmids and analysed24 h after transfection. Expression andlocalization of the proteins were deter-mined by Western blotting and im-munofluorescence analysis, respective-ly. The 3F12 antibody recognized bothproteins as single bands on the Westernblot (Figure 1A, lanes 13 and 14) andno cross-reaction with cellular proteinswas observed (Figure 1A, lane 15). Im-munofluorescence staining of transfect-ed cells with 3F12 antibody indicatedthat both proteins were localized in theappropriate compartment of the cell:The 3F12-GAPDH in the cytoplasmand 3F12-GAPDH-NLS in the nucleus(Figure 1, B and C). These results indi-cate that 3F12 epitope-tag can be usedfor the detection and determination ofthe localization of proteins expressed ineukaryotic cells.

Advantages of the BPV E2-DerivedEpitope Tags

We analyzed the expression of pro-teins tagged with the BPV E2-derivedepitopes in E. coli and eukaryotic cells.Detection of the tagged proteins bothon immunoblots and by immunofluo-rescence staining of cells indicates lowbackground activity, sensitivity andgood signal-to noise ratio of the usedepitope-antibody combinations. We didnot observe any cross-reaction with cel-lular proteins. Because of the highspecificity of the epitope-antibody in-teraction, our tagging system is espe-cially useful for the studies of proteinlocalization in the cells. In addition, wehave shown that interaction of 3F12MAb with the specific epitope is nothindered by high salt concentrations.That allows us to immunoprecipitateand immunopurify the tagged proteinsin high-salt conditions and to avoidcoimmunoprecipitation of contaminat-ing proteins as well as to avoid aggre-gation of the protein of interest in thecourse of purification.

REFERENCES

1.Abroi, A., R. Kurg and M. Ustav. 1996.Transcriptional and replicational activationfunctions in the bovine papillomavirus type 1E2 protein are encoded by different structural

determinants. J. Virol. 70:6169-6179.2.Field, J., J. Nikawa, D. Broek, B. MacDon-

ald, L. Rodgers, I.A. Wilson, R.A. Lernerand M. Wigler. 1988. Purification of a RAS-responsive adenylyl cyclase complex fromSaccharomyces cerevisiae by use of an epi-tope addition method. Mol. Cell. Biol. 8:2159-2165.

3.Jarvik, J. W. and C. A. Telmer. 1998. Epi-tope tagging. Annu. Rev. Genet. 32:601-618.

4.Kaldalu, N., T. Mandel and M. Ustav. 1996.TOL plasmid transcription factor XylS bindsspecifically to the Pm operator sequence. Mol.Microbiol. 20:569-579.

5.Kessler, B., M. Herrero, K.N. Timmis andV. de Lorenzo. 1994a. Genetic evidence thatthe XylS regulator of the Pseudomonas TOLmeta operon controls the Pm promoterthrough weak DNA-protein interactions. J.Bacteriol. 176:3171-3176.

6.Kessler, B., K.N. Timmis and V. de Lorenzo.1994b. The organization of the Pm promoterof the TOL plasmid reflects the structure of itscognate activator protein XylS. Mol. Gen.Genet. 244:596-605.

7.Kurg, R., J. Parik, E. Juronen, T. Sedman,A. Abroi, I. Liiv, U. Langel and M. Ustav.1999. Effect of bovine papillomavirus E2 pro-tein-specific monoclonal antibodies on papil-lomavirus DNA replication. J. Virol. 73:4670-4677.

8.Lepik, D., I. Ilves, A. Kristjuhan, T.Maimets and M. Ustav. 1998. p53 protein isa suppressor of papillomavirus DNA amplifi-cational replication. J. Virol. 72:6822-6831.

9.Tanaka, M. and W. Herr. 1990. Differentialtranscriptional activation by Oct-1 and Oct-2:interdependent activation domains induceOct-2 phosphorylation. Cell 60:375-386.

10.Ustav M., E. Ustav, P. Szymanski and A.Stenlund. 1991. Identification of the origin ofreplication of bovine papillomavirus and char-acterization of the viral origin recognition fac-tor E1. EMBO J. 10:4321-4329.

The study was supported by the Eston-ian Science Foundation Grant Nos. 2496,2497, 2315 and 2316. Address correspon-dence to Dr. Mart Ustav, Department of Mi-crobiology and Virology, Institute of Molec-ular and Cell Biology, Tartu University, 23Riia Street, Tartu, 51010, Estonia. Internet:ustav@ebc.ee

Received 14 July 1999; accepted 11November 1999.

Niilo Kaldalu, Dina Lepik,Arnold Kristjuhan andMart UstavTartu UniversityTartu, Estonia

Short Technical Reports

462 BioTechniques Vol. 28, No. 3 (2000)

Figure 2. Use of the tagged proteins in DNAbinding assays. (A) DNA precipitation by 3F12-tagged XylS. Radiolabeled HpaII/HinfI digest ofthe Om-containing plasmid pUPM190 was incu-bated with 3F12-beads containing 3F12-XylS.Unbound DNA was removed by washing. BoundDNA was extracted from the beads and analysedon non-denaturing TBE/PAGE (5%), using thefragment mixture as a marker (lane 1). Only asingle fragment, which contained the XylS bind-ing site Om, was retained on the beads (lane 2).(B) DNase I footprinting by resin-bound 3F12-tagged XylS. XhoI/EcoRI Om containing frag-ment from pUPM190 was end-labeled in the low-er strand at 3 terminus and incubated with3F12-beads containing 3F12-XylS. UnboundDNA was removed by washing. Both free andprotein-bound templates were subjected toDNase I cleavage. Lane 1, G-specific DNA se-quence marker; lane 2, DNase I digest of the un-bound DNA fragment; lane 3, DNase I digest ofthe fragment bound to 3F12-XylS. Brackets indi-cate the region protected from the DNase I cleav-age by the DNA-bound XylS. (C) Band-shift as-say of the 3F12-tagged p53 proteins. TheDNA-binding activity of the p53 proteins wasstudied by separating the protein-DNA complex-es from the unbound DNA on 5% PAGE (55:1).Crude E. coli lysates containing p53 mutantsDNDC362 and DNDC362 produce a shifted band(lanes 1 and 4). Supershift with tag-specific mon-oclonal antibody 3F12 was used to show that theprotein-DNA complexes contain p53 (lanes 2 and5). The DN39DC362trp248 p53 protein carryinga point mutation in its DNA-binding domain wasused as a negative control (lane 3).