Journal of Virological Merhods, 17 (1987) 211-217 Elsevier

211

JVM 00613

Purification of adenovirus hexon by high performance liquid chromatography

Scott A. Siegel, James E. Hutchins and Donald J. Witt

Immunodiagnostics Department. Becton Dickinson and Company, Corporate Research Center, Research Triangle Park. North Carolina, U.S.A.

(Accepted 15 April 1987)

Summary

Hexon is the major structural protein of adenovirus, and has significance in studies of virus structure and function, vaccine development, and immunodi- agnosis. We describe a simple, single-step, anion-exchange high performance liq- uid chromatography (HPLC) method for the high yield purification of hexon. Pu- rity of the isolated hexon was assessed by SDS-PAGE and HPLC methods. The isolated hexon was immunologically reactive with anti-hexon monoclonal antibody in a dot-blot assay. It also retained immunogenicity, as polyclonal antisera from rabbits immunized with hexon showed the desired antigen specificity. The en- hanced speed of this purification method allows for the efficient isolation of hexon from various serotypes, and thus may facilitate comparative studies of hexon im- munobiology.

Adenovirus; Hexon; Purification method; HPLC

Introduction

Adenoviruses are non-enveloped DNA viruses which cause a variety of respi- ratory and enteric diseases in man (Pettersson, 1984; Horowitz, 1985; Philipson, 1984). The adenovirus virion contains at least nine unique polypeptides, with hexon being the largest and most abundant capsid protein (Pettersson, 1984; Philipson, 1984). The molecular mass of hexon is approximately 324 kDa consisting of three

Correspondence to: Scott A. Siegel. Immunodiagnostics Department, Becton Dickinson and Com- pany. Corporate Research Center. Research Triangle Park. NC 27709. U.S.A.

0166-0934/87/$03.50 0 1987 Elsevier Science Publishers B.V. (Biomedical Division)

212

identical subunits of approximately 120 kDa each (Horowitz et al., 1973; Cornick et al., 1973; Grutter and Franklin, 1974). Hexon has been shown to be important in the immunology of adenovirus, with both type-common and type-specific epi- topes existing on hexons from the various serotypes (Norrby and Wadell, 1969; Pereira and Laver, 1970; Pettersson, 1971; Nasz et al., 1972; Stinski and Ginsberg, 1975; Kinloch et al., 1984; Adam et al., 1985a,b; Pettersson and Wadell, 1985). Multiple copies of identical or closely related epitopes may also exist on the in- dividual hexon molecules (Adam et al., 1986). A variety of techniques for the pu- rification of hexon have been published (Pettersson et al., 1967; Dowdle et al., 1971; Boulanger and Pouvion, 1973; Nasz et al., 1972; Jornvall et al., 1974), that require multiple separations, dialysis steps, or other tedious manipulations taking several days to complete.

Since current high-performance liquid chromatography (HPLC) techniques may offer unique advantages for protein purification. we undertook a study to deter- mine if the application of these techniques would produce a simpler and more rapid purification scheme for adenovirus hexon. We chose to adapt the classical ion-ex- change approach for hexon purification to the HPLC system. The technique re- ported herein requires only a single HPLC separation step, and the entire purifi- cation of hexon from culture takes only several hours to complete. The increased ease and speed of this method can allow for the efficient isolation of hexons from a variety of serotypes, and hence facilitates comparative studies of adenovirus im- munobiology.

Materials and Methods

Cells and virus

HeLa S3 and KB cells were obtained from the American Type Culture Collec- tion (Rockville, MD). HeLa S3 cells were grown as suspension cultures at 37°C in spinner modified minimal essential medium (MEM) (Gibco, Grand Island, NY) supplemented with 1% glutamine and 10% fetal bovine serum (fbs) (Hyclone, Lo- gan, UT). Cells were routinely tested for mycoplasma contamination.

For virus production, exponentially growing HeLa S3 cells (3-6 X 10s/ml) were concentrated to 1120 of the original volume. The cells were then infected with pur- ified adenovirus serotype 5 (obtained from Dr. Michael Katze, Sloan Kettering In- stitute Cancer Center, NY) at a multiplicity of infection (moi) equal to 20 (Maize1 et al., 1968). Following a 1 h adsorption period at 37”C, maintenance medium con- sisting of MEM, 2% glutamine, and 2% fbs was added to one-half the original vol- ume. Infected cells were incubated with spinning at 37°C for 48 h and harvested by centrifugation at 230 x g for 20 min at 5°C.

For SDS-polyacrylamide gel electrophoresis (SDS-PAGE) and immunoblot analyses, adenovirus-infected KB cell monolayers (moi=20) were washed with Earle’s balanced salt solution and harvested at 72 h post-infection, as described above.

213

Virus extraction and chromatography

Infected-cell pellets were resuspended in 10 mM sodium phosphate, pH 6.8 (buffer A) at about 3 ml per liter of infected cells. After 3 freeze-thaw cycles, the suspension was sonicated on ice for 1 min. Hexon was separated by anion-ex- change chromatography on a Waters system (Milford, MA), consisting of a model 840 data and chromatography control station, two model 510 pumps, a U6K man- ual injector, and a model 441 absorbance detector. A sample of the extract was applied to a Waters Protein Pak DEAE 5PW column (7.5 mm X 7.5 cm) equili- brated in buffer A at a flow rate of 1.0 mlimin. After a 5 min isocratic period, a 40 min linear gradient to 10 mM sodium phosphate, 0.5 M NaCl, pH 6.8 (buffer B) was applied to the column. The column was maintained in buffer B for 11 min followed by a 5 min ramp back to buffer A. A minimum of 20 min re-equilibration in buffer A was allowed between injections. Columns were maintained at ambient temperature, and ultraviolet absorbance at 280 nm was monitored. Peaks of hexon protein, eluting at about 38 min, were manually collected and precipitated by the addition of ammonium sulfate to 95% saturation. After allowing the precipitate to form for 1 h on ice, the hexon was collected by centrifugation at 4°C in a Sorvall RC2-B centrifuge (Newton, CT) using an SS-34 rotor at 10000 rpm for 30 min. The pellet was then resuspended in a small volume of phosphate-buffered saline (PBS), pH 7.2, and desalted on a Pharmacia (Uppsala, Sweden) PD-10 (Sephadex G2.S M) column. Pooled protein fractions were stored at -70°C until use.

Biochemical and immunochemical analysis

Protein concentrations were determined by the method of Bradford (1976). The Laemmli (1970) system was used for SDS-PAGE analyses. and gels were stained by the Kodavue (Kodak, Rochester, NY) method. Western blots were performed in the Hoefer Transphor Model TE.50 (San Francisco, CA) by the method of Tow- bin et al. (1979). Both Western and dot immunoblots used 0.45 km nitrocellulose (Schleicher and Schuell, Keene, NH) and were visualized by the method of Ei- senberg et al. (1985) using ‘2s1 protein A (New England Nuclear, Boston, MA) and subsequent autoradiography.

Po/yclonal antiserum production

Polyclonal antiserum was raised in female white New Zealand rabbits by in- tradermal injection of an equal mixture of the HPLC purified hexon and complete Freunds adjuvant (Boehring Diagnostics, LaJolla, CA). Booster inoculations were given at various intervals, and bleeds were collected by ear vein puncture at least 60 days past the initial inoculation. In all cases, pre-immune sera were collected and found to be non-reactive with adenovirus antigens by indirect immunoflu- roescence assay (IFA). Serum was prepared from test bleeds and stored at -20°C

until use.

214

w, pg 651 a-

:

P a 99

MINUTES

Fig. 1, HPLC chromatogram from adenovirus hexon purification. Adenovirus S-infected HeLa S3 ex- tracts were prepared as described in Materials and Methods. 1000 mV = 1.0 O.D. (280 nm).

Results and Discussion

Using the conditions described above, adenovirus hexon was successfully iso- lated from infected HeLa cells. A typical chromatographic profile for hexon pu- rification is shown in Fig. 1. Hexon was retained on the anion-exchange column, and eluted at about 36-38 min in our buffer-gradient system. The yield of purified hexon was found to be approximately 1 mgi4 x 10” adenovirus 5infected HeLa cells. The hexon peak was easily distinguishable from other proteins by virtue of its re- tention time, peak height, and the reproducible nature of chromatograms in the area just prior to its elution. SDS-PAGE analysis, under reducing conditions, re- vealed a single band of approximately 116K MW in the isolated hexon fraction (Fig. 2), which is well within the reported range for hexon monomer. Re-injection of isolated hexon onto the HPLC system revealed the extent of purity achieved (Fig. 3). Minor contaminants may be contained in small peaks eluting in the void vol- ume and at the tail end of the hexon peak. Materials eluting between 18 and 28 represent HPLC buffer contaminants, since they were also present in blank gra-

Hexon

A B PM

205K 116K 97.4K 66K

45K

29K

Fig. 2. SDS-PAGE analysis of purified hexon fraction. 9% gel stained by Kodavue method. A = in- fected KB cell extract, B = purified hexon, PM = protein marker.

215

MINUTES

Fig. 3. Be-injection of purified hexon on HPLC system. 100 mV = 0.1 O.D. (280 nm)

dients. Electron microscopic analysis of the purified hexon revealed that a major- ity of the protein was in the trimer form, a result obtained by previously described purification methods (Pettersson et al., 1967; Nasz et al., 1972; Boulanger and Pouvion, 1973).

The identity of the presumptive hexon was confirmed using immunological tech- niques. The isolated hexon reacted positively with a known hexon-specific mon- oclonal antibody (Chemicon, San Diego, CA) in an immunoblot (Fig. 4). This an- tibody also reacted with virions and infected-cell lysates, but not with uninfected cell controls (Fig. 4). These results also showed that the immunological reactivity of the hexon was preserved during I-IPLC purification. This was confirmed by ex- periments utilizing the hexon as capture antigen in a solid-phase ELISA, and by the ability of Chloramine T radioiodinated hexon to function in a liquid-phase ra- dioimmunoassay (LPRIA) with various monoclonal and polyclonal antibodies (data not shown). The ability of the HPLC purified hexon to elicit a specific immune response was assesssed by immunizing rabbits. Sera obtained from hexon-immu- nized rabbits were found to be adenovirus-specific by IFA, ELISA, and LPRIA analyses (not shown). The major reactivity of the rabbit antisera was to the hexon protein, as shown by Western blotting (Fig. 5). This result is generally indicative of antisera raised to a highly purified immunogen.

150 1:lOO 1:200

a

lib

HeLa, uninfected

KB, infected

KB, uninfected

Ad6 virions

Hexon (0.6ug)

Fig. 4. Dot-blot analysis of purified hexon. Purified hexon. positive controls (KB, infected, Ad5 vi-

rions) and negative controls (HeLa, uninfected, KB, uninfected) were spotted onto nitrocellulose and

reacted with Chemicon (San Diego, CA) anti-hexon monoclonal antibody at three dilutions. Reactions were visualized with “‘I protein A and subsequent auto~diograph~.

216

442

443

444

445

Fig. 5. Western blot analysis of rabbit anti-hexon antiserum. Rabbit antiserum raised against purilled

hexon was reacted with adenovirus 5 virions run on 9%~ SDS-PAGE gels as described in Materials and Methods. 442-445 = identification numbers for the various experimental animals (I: 100 dilution of each

serum).

It was found that the soluble antigen fraction from the top of the CsCl gradients used in purification of adenovirus virions (Maize1 et al., 1968) was adequate start- ing material for hexon purification, if dialyzed into buffer A. Thus a single culture could serve as starting material for both virion and hexon purification.

The HPLC method for adenovirus hexon purification yielded a highly purified and immunologically active protein. This technique offers the advantage of fewer steps and decreased time to obtain final product when compared to previously re- ported chromatographic methods. Application of this technique for the purifica- tion of hexons from a variety of adenovirus serotypes can facilitate comparative studies of adenovirus immunobiology.

Acknowledgements

The authors wish to acknowledge Dr. R. Compans, University of Alabama, Bir- mingham, for his invaluable assistance in performing the electron microscopic studies, and Karen Baldwin for her assistance in manuscript preparation.

References

Adam, E., Nasz. I.. Lengyel, A., Erdei, J. and Fachet. J. (198Sa) Mol. Immunol. 22, 967-971. Adam. E.. Erdei. J.. Lengyel, A., Berencsi. G.. Fachet. J. and Nasz. I. (198Sb) Intervirology 23,

222-227.

Adam. E.. Nasz, I.. Lengyel, A, Erdei. J. and Fachet. J. (1986) Mol. Immunol. 23. 755-759.

217

Boulanger, H.A. and Puvion. F. (1973) Am. .I. Biochem. 39. 37-42.

Bradford, M.M. (1976) Anal. Biochem. 72. 248-254.

Cornick, G., Sigler, P.B. and Ginsberg, H.S. (1973) J. Mol. Biol. 73. 533-537. Dowdle, W.R.. Lambriex, M. and Hierholzer, J.C. (1971) Appl. Microbial. 21, 718-722.

Eisenberg. R.J., Long. D.. Ponce De Leon. M.. Matthews. J.T., Spear, P.G., Gibson, M.G., Lasky, L.. Berman, P.. Golub, E. and Cohen. G.H. (1985) J. Viral. 53, 634-644.

Grutter. M. and Franklin, R.M. (1974) J. Mol. Biol. 89, 167-178.

Horowitz, M.S., Maizel, J.V. and Scharff, M.D. (1973) J. Viral. 6. 569-576.

Horowitz, M.S. (1985) In: Virology (Fields, B.N., ed.). pp. 477-495. Raven Press, New York. NY.

Jornvall. H.. Pettersson. U. and Philipson, L. (1974) Eur. J. Biochem. 48, 179-192.

Jornvall, H. and van Bahr-Lindstrom, H. (1981) J. Biol. Chem. 256. 6187-6198. Kinloch, R.. Mackay, N. and Mautner, V. (1984) J. Biol. Chem. 259, 6431-6436.

Laemmli. U.K. (1970) Nature (London) 277, pp. 68&685.

Lengyel, A., Adam, E.. Nasz, I.. Erdei, J. and Fachet, J. (1985) Acta Viral. 29. 362-372.

Maizel. J.V., Jr.. White, D.O. and Scharff. M.D. (196X) Virology 36, 115-125.

Nasz, I.. Lengyel. A. and Cserba, I. (1972) Arch. Gesamte Virusforsch. 36, X&92.

Norrby, E. and Wadell, G. (1969) J. Viral. 4, 663-670.

Pereira, H.G. and Laver, W.G. (1970) J. Gen. Viral. 9. 163-167.

Pettersson, U., Philipson, L. and Hoglund. S. (1967) Virology 33, 575-590.

Pettersson, U. (1971) Virology 43, 123-136.

Pettersson, U. (1984) In: The Adenoviruses (Ginsberg. H.S.. ed.). pp. 205-270. Plenum Press. New

York, NY. Pettersson. U. and Wadell, G. (1985) In: Immunochemistry of viruses. The basis for serodiagnosis and

vaccines (Van Regenmortel, M.H.V. and Neurath, A.R.. eds.). pp. 295-323. Elsevier Science Pub-

lishers B.V. Philipson, L. (1984) Curr. Top. Microbial. Immunol. 109, l-52.

Stinski, M.F. and Ginsberg. H.S. (1975) J. Virnl. 15, 89%905.

Towbin. H.. Staehelin. T. and Gordon. J. (1979) Proc. Natl. Acad. Sci. USA. 76, 435&4354.