CLIN. CHEM.35/9, 1838-1842 (1989)

1838 CLINICALCHEMISTRY,Vol. 35, No.9, 1989

Antigens Produced by Recombinant DNA TechnologyJ. Lawrence Fox and MIchael KIa8s

Some of the greatest beneficiariesof the revolutionaryad-vances in biotechnology over the past 15 years have beenproducers of diagnostic reagents, especially for the cloningand expressionof antigens,primarilyof viral origin.Recom-binant DNA technology provides methods for producingantigensfor diagnosticassays that are more highlypurified,morespecific,and safer to producethanviralcultureandthatare significantlyless expensive to manufacture.Antigenssoproducedcan be used for productionof antibodiesor anti-sera for competition assays, as reagents for mappingepitopes, as affinity-chromatographyligandsfor purificationof antibodies or protein, and as research reagents. Theirinitial use in some hepatitis B assays may be primarilyacost-reduction application, but in otherapplications(e.g., HIVdiagnostic tests) they present the first opportunity to com-mercially produce an otherwise very expensive antigen.Recombinant-DNA-producedantigensare alsobeingusedtodevelop safer vaccines, but not, however, without someconsideration of the structural nature of immunodominantepitopesand the adequacy of the immune response.

The emergence of recombinant DNA technology has hadfar-reaching consequences for diagnostic products, not onlyaffecting production methodology but, perhaps more impor-tantly, also opening many new doors for research efforts.Robert Gallo (personal communication, 1988) has statedthat without recombinant DNA technology it would havetaken at least two decades to understand the humanimmunodeficiency virus (HIV), with consequences muchmore disastrous than they in fact are.’

Molecular BIology Background

Central to an understanding of recombinant antigens isan understanding of recombinant DNA technology (thereader may want to consult a more extensive presentationof this technology; Weatherahl’s The New Genetics andClinical Practice is recommended). The flow of informationin biological systems is generally DNA -, RNA - protein.DNA serves as the genetic repository for all living orga-mama, but must be transcribed into a related molecule,RNA, to finally find its expression through translation intoprotein sequences.

The linear sequence of the nucleic acid bases in DNAmolecules codes for each of the 20 amino acids. When theDNA sequence is known, the amino acid it encodes isunambiguously identified. If the amino acid is known, onecan work backwards to the series of three nucleic acid basesin the DNA (a codon) that encodes it; however, this “re-verse” translation process is not unambiguous.

The nature of the process of expressing information

Corporate Molecular Biology, Abbott Laboratories, Abbott Park,IL 60064.

‘Nonstandard abbreviations: HIV, human immunodeficiencyvirus; mRNA, messenger RNA; cDNA, complementary DNA;MAb, monoclonal antibody; and CEA, carcinoembryonic antigen.

Received April 13, 1989; accepted June 16, 1989.

contained in DNA into a protein sequence of amino acids isdifferent for bacteria, which are prokaryotes (cells withoutnuclei), and organisms with eukaryotic (nucleated) cells,e.g., humans. In bacteria, the process of RNA transcriptionis directly coupled to protein translation. In higher plantsand animals, the process is complicated by the presence ofintrons, noncoding regions interspersed within the amino-acid-coding regions (exons) of the gene (see Figure 1). Thesynthesis of RNA is contained within the nuclei of eukary-otic cells and produces an RNA species that is much largerthan the finally processed messenger RNA (mRNA), typi-cally by an order of magnitude, because it is a transcript ofboth introns and exons. This primary RNA transcript mustundergo maturation processing, during which the intronsare spliced out. Only after this step is a mature RNAmolecule formed, mRNA, that can direct the synthesis of aprotein.

Recombinant DNA Background

Recombinant DNA technology is based on the manipu-lation of DNA molecules. Because working with DNA,which contains introns, increases the complexity of thetask by an order of magnitude, an alternative approach wasadopted. If one isolates mature mRNA and uses a viralenzyme that can use RNA as a template on which toreverse-transcribe DNA, then a complementary DNA(cDNA) can be made. This cDNA is a direct copy of themRNA and does not contain introns. This simplifies theproblem of DNA manipulation.

To utilize the DNA isolated from an organism or thecDNA generated from an organism’s mRNA, it must beinserted into another piece of DNA. This recipient DNA,called a vector, is typically a circular DNA double helix(Figure 1). To open the circle to insert the new DNA, thevector must be cut. A series of enzymes, restriction endo-nucleases, have been isolated that recognize and cut atspecific sequences of bases in double-stranded DNA. Endo-nucleases are ordinarily found in bacteria (their names arederived from the strains of bacteria from which they areisolated) and serve to restrict the kinds of DNA that can betaken up and used by a bacterial cell. To the bacterial cell,they are a defense mechanism against foreign DNA, but tothe molecular biologist, they are the tools required toengineer DNA.

Once the vector DNA is cut, any desired DNA can beinserted in the vector. The result of combining vector DNAwith a DNA insert creates a recombinant DNA. Bacterialcells can now be made to take up exogenous DNA, usuallyby treating them with calcium ions. The efficiency of thisbacterial transformation process (see Figure 2), the givingof new genetic information to a recipient or host cell, isabout 1 in iO.

Thus the major remaining problem is how to identifythose bacterial cells that contain the DNA insert of interestfrom a large number of cells that contain either nothing or,perhaps, other inserts. This screening process may beaccomplished by DNA probe screening or antibody expres-sion screening. The remaining task is to insert the DNA

Eukaryote

Secreted Proleln

cDNA Piasmid

Fig. 1. Expressionand insertionof geneticinformationin recombinantDNA techniquesThe genetic information is maintained in eukaryotic cells in the chromosomes contained in the cell’s nucleus.The processofRNA transcription initially creates a“very large”heteronuclear(hn)RNAmolecule that must undergo processing to evolve into the messenger ANA (mRNA), which ultimately codes for proteinsynthesis. The processincludesjoining theexons(expressedregions,shown in bold vertical line)bysplicingouttheintrons(interveningregions,theloops).Proteintranalationoccurs outside of the nucleusin the cytoplasm of the cell. For recombinant DNA manipulationsof eukaryotic proteins, it is most useful to isolate themRNA,whichnolongerhasthe lntronspresent,and use it as a templatewiththeviralenzyme“reversetranscriptase”to prepare complementary DNA(cONA).ThecDNAcan then be inserted into a bacterialplasmidand used totransforma bacterium(theinserts are highlightedby cross-hatching). Once insidea bacterium,a plasmid is typically replicatedinto50-100 copies

Replicator Repllcator

Eco RI

Foreign DNA

oOC’c

TranslormedE. coil

oo

Translormallon

Chromoeome

Expression

CLINICALCHEMISTRY,Vol.35, No. 9, 1989 1839

sequence of interest into a suitable expression system andproduce large quantities of protein. For a review see Dar-nell et al. (1).

Producing Antigens by Recombinant DNA Technology

General Advantages

By definition, an antigen is a foreign protein, i.e., virtu-ally any nonhuman protein recognized as foreign by ourimmune system (2, 3). Practically, however, antigens tendto be surface proteins of viral, bacterial, or protozoanorigin. In nature, they exist in relatively minute quanti-ties, which, in the past, made them intractable to experi-mentation or investigation and precluded any practical, letalone industrial, use. Recombinant DNA technology has ofcourse changed this entirely. Recombinant-DNA-producedantigens have become a powerful tool in both diagnostics

1

Bacterium 0

and vaccines.Such antigens offer several advantages:#{149}More abundanticonsistent supply#{149}Reduced cost of production#{149}Safety of manufacture#{149}Potential for genetic manipulationThe general advantages of recombinant antigens are

many. Today, even average-size (200 L) bacterial fermentervessels can produce gram quantities of an antigen (protein)free from nonhost proteins. For example, p41, the envelopeprotein of HIV-1, can be expressed in Escherichia coli sothat hundreds of milligrams of purified p41 can be producedfree from any other HIV-1 or non-E. coli proteins. Further-more, the yield of recombinant antigen is consistent; i.e.,repetition of the same protocol will yield identical product,thus reducing the variability observed when protein is

Fig. 2. Production of a recombinant DNA molecule by cutting a vectorwithan endonucleaseand insertingthe foreignDNACombiningthe two DNAsand covalentlyligatingthemtogetherproducesa recombinantDNAmolecule,whichcan be used totransforma bacterialcell. The vector(here it is a plasmid) can multiply many times in a bacterial cell, typically achievinga 50- to 100-foldamplification

1840 CLINICALCHEMISTRY,Vol. 35, No. 9, 1989

isolated from poorly controllable biological sources, e.g.,tissue, sera, etc.

Not only can consistent, pure, abundant quantities beobtained, but also the cost of this is significantly less thanthat for product isolated from natural sources, e.g., tissueculture, human tissues, or live animals. Although eachantigen is different, a general rule of thumb is that theproduction in E. coli and yeast (or equivalent prokaryote) is1/100 to 1/1000 the cost of production in tissue culture. Thisis especially true for some particularly virulent pathogens,e.g., HIV antigens, which previously had to be isolatedfrom intact virus grown in tissue culture in special contain-ment facilities by specially trained technicians-all ofwhich significantly increased the cost.

Safety is another general advantage of recombinantantigens. Because one is working with only one, or a few, ofthe genes of the intact infectious agent, the danger ofinfection, which requires a complete gene complement, iseliminated. Contrast this with the isolation of live, intact,infectious H1V virus from tissue culture in strictly con-trolled biological and physical containment facilities. Al-though it has been proposed that working with recombi-nant-DNA-produced antigens might lead to seroconver-sion, no incidents have been observed to date.

Finally, cloning the gene for the desired antigen empow-ers the investigator with all the tools of modern molecularbiology for making any desired modifications such as inser-tions, fusions, and (or) deletions to the recombinant anti-gen. Such modifications may actually improve the antige-nicity of the recombinant antigens. For example, removalof a cleavage site in the HW enu protein appears to increaseits antigenicity (4). Furthermore, specific deletions allow usto obtain antibodies to whatever region may be deemedmost important for a diagnostic assay, e.g., the mainimmunogenic region for H1V-1 (5). In addition, if a recom-binant-DNA-produced antigen is less immunogenic thandesired, it can be genetically fused with a protein of highimmunogenicity (6). Further, this technique allows theproduction of polyvalent antigens to induce immunity tomultiple infectious agents simultaneously (7).

Applications

Recombinant-DNA-produced antigens have many impor-tant uses in medical research:

#{149}in diagnostic assays#{149}in antibody induction#{149}as affinity chromatography ligands for antibody purifi-

cation#{149}in competition assays#{149}as reagents to map epitopes#{149}as basic research toolsRecombinant-DNA-produced antigens are beginning to

be used in an ever-increasing number of diagnostic assays.For example, recombinant-DNA-based diagnostic assaysfor hepatitis B (8, 9), HIV (10, 11), carcinoembryonicantigen (CEA) (12), and Haemophilus influenzae type b P2protein (13) are in various stages of use or development. Asearch of patent applications worldwide revealed that inthe past decade more than 87 applications have been filedfor diagnostic tests based on the use of recombinant-DNA-produced antigens.

As diagnostic reagents, these antigens are extremelyvaluable for the induction of both polyclonal and monoclo-nal antibodies. Once the gene encoding the desired antigenis cloned, virtually any region of the protein can be ex-

pressed and used to obtain region-specific antibodies (14).In addition, any part of the antigen may be expressed as afusion protein, similar to the practice of conjugating ahapten to a larger, highly immunogenic protein (6). Suchrecombinant fusion proteins can be important for stimulat-ing an immune response. Fusions between different pro-teins have also been used to construct polyvalent antigens.For example, a Vaccinia virus that expresses the surfaceproteins of influenza A virus hemagglutinin and herpessimplex type 1 (HSV-1) glycoprotein D induced antibodyresponse to both proteins, producing immunity in mice (7).

Another use for recombinant-DNA-produced antigens isin the purificationlisolation of specific antibodies. Antigencan be bound to a chromatography matrix to produce anaffinity column (15), through which antibody preparationsare passed. Depending on the part or domain of the antigenused, this affinity chromatography can yield preparationsof mono-epitopic antibodies. Such techniques have beenused to obtain both high- and low-affinity antibodies frompolyclonal sera, depending on the elution salts used on thecolumn.

In some diagnostic assays, the antibody used often cross-reacts with similar antigens that are usually present. Forexample, in diagnostic assays for CEA, a marker for coloncancer, many anti-CEA antibody preparations cross-reactwith related members of the CEA family (13). Such cross-reactivity increases the background response and may leadto false-positive results. One key to the solution of thisproblem lies in detecting regions of amino acid sequence ofthe desired antigen that are unique and not found in theordinarily cross-reactive antigens. The DNA encoding thisunique region is then cloned and expressed separately toyield a mono-epitopic antigen capable of inducing non-cross-reacting antibodies.

If this antigen is used to affinity-purify polyclonal sera,then only antibodies that bind this specific region of theantigen will be isolated. Similarly, such subdomains ofrecombinant-produced antigens can by themselves be usedto obtain monoclonal antibodies (MAbs). The polyclonalapproach has the advantage of consistently giving high-affinity antibodies, whereas most MAbs display low aflin-ities.

Recombinant antigens are also used in diagnostic assayconfigurations. They can provide superior quantities ofreagent to serve as positive controls, and they are usedextensively in antibody capture assays. For example, Ab-bott’s Second Generation HIV diagnostic assays are config-ured with recombinant antigens (HIV p41 and p24) boundto polystyrene beads. The sample is added and any anti-HIV antibodies top41 or p24 are bound to the recombinantantigens on the beads. These human antibodies are thendetected with an anti-human IgG conjugated to horserad-ish peroxidase (EC 1.11.1.7). A strong color reaction from asubstrate cleaved by horseradish peroxidase indicates thepresence of anti-H1V antibodies in the patient’s serum.

The HIV diagnostic assay EnvaCor (Abbott Labs.) isconfigured with human IgG anti-fflV bound to polystyrenebeads. The sample to be tested is mixed with a limitedamount of recombinant antigen (p41 envelope or p24 core).If there is no anti-HIV antibody in the patient’s serum, allthe recombinant antigen will bind to the anti-HIV antibodyon the bead. This bound antigen will then be detected bythe addition of horseradish peroxidase-conjugated MAb tolilY (p41 or p24). A strong color reaction caused by theperoxidase is indicative of a sample negative for anti-HIV

CLINICAL CHEMISTRY, Vol. 35, No. 9, 1989 1841

antibodies. If the patient’s sample does contain anti-HJVantibodies, they will compete for the recombinant antigen;this is indicated by a loss of color. Results with recombinantantigens have been comparable with (and are generallysuperior to) those with native antigens from viral lysates(10, 11).

Another advantage to using recombinant antigens indiagnostic assays is the ability to independently assaydifferent antigens, each of which may be diagnostic for adifferent disease state. For example, Decker and Dawson(10) found that anti-core antibody titers decreased dramat-

ically when persons infected with HIV entered the symp-tomatic state of AIDS. Thus, the ratio of anti-core to anti-enucould probably be used to follow disease progression (16-18).

CaveatsDespite the many advantages recombinant antigens

have as diagnostic reagents, it is important to recognizethat there are a number of caveats:

#{149}Post-translational modifications#{149}Conformational epitope requirements#{149}Oversimplification of epitopesRecombinant antigens expressed in prokaryotic orga-

nisms lack most, if not all, of the post-translational modi-fications typically found in eukaryotic proteins. Modifica-tions such as glycosylation, phosphorylation, etc., if theyconstitute immunologically important epitopes, would notoccur, so epitopes important for diagnostic assays may bemissing.

Furthermore, recombinant antigens expressed in pro-karyotic hosts will often not possess the same conforma-tional epitopes as the native antigen. If conformational ordiscontinuous epitopes constitute the major immunodomi-nant region of an infectious agent, then use of recombinantantigens from prokaryotic hosts may not detect antibodiesto the infectious agent. For example, normally protectiveMAbs to the Bordetella pertussis Si toxin will only bind tothe recombinant antigen expressed in E. coli after refoldingof the protein to allow the formation of a discontinuousepitope (19). Conformational or discontinuous epitopes arealso a major concern for the infectious agents causinghepatitis A (20-22), polio (23), and foot and mouth disease(24), to name but a few cases.

Recombinant antigens produced in prokaryotes do notreconstitute these epitopes. In such cases, special efforts arerequired to achieve the identical conformation and post-translational modifications found in eukaryotic proteins.These efforts include cloning and expression in appropriatetissue culture systems capable of the same post-transla-tional modifications. In the case of hepatitis A virus this isdone by growing the virus in tissue culture (25, 26).Unfortunately, as mentioned above, expression in tissueculture is more expensive than in bacteria, and workingwith live infectious agents is not desirable.

Finally, in configuring diagnostic assays that seek todetermine in the patient the presence of antibody directedto an infectious agent, one must always determine howmany epitopes are sufficient to adequately detect all in-fected individuals. For example, in an HN diagnostic testis it sufficient to use only p41 as the recombinant antigen orare p24 or other HIV proteins required? This questionunfortunately must be answered by the thorough screeningof thousands of confinned samples.

Recombinant Antigens for Vaccines

Recombinant antigens are becoming increasingly impor-

tant in the development of vaccines. Recombinant DNAtechnology can provide abundant quantities of antigenrequired for the development of vaccines that were previ-ously intractable owing to the scarcity of reagents. Not onlycan sufficient immune response be induced, but this can beaccomplished without the risk associated with the use oflive infectious agents. As such, recombinant antigens pro-vide a relatively inexpensive, abundant source of consis-tently pure material for vaccination. Vaccines for manyinfectious agents, e.g., cholera toxin B (27), hepatitis B(28-30), lilY (31), influenza A (32), malaria (33), etc.,prepared by using the respectively cloned genes and theresulting recombinant antigens, are in various stages ofdevelopment. Furthermore, because of the flexibility ofrecombinant antigens, polyvalent antigens can be obtainedand used to develop immunity to multiple infectious agentsat the same time (8).

Although recombinant antigens possess important ad-vantages for the development of vaccines, they are notnecessarily an automatic success. For example, HW vac-cine based on env expressed from recombinant Vacciniavirus in whole animals has elicited low titers of lilY-reactive antibodies (4).

Obviously, more factors must also be considered. First,the recombinant antigen must possess a sufficient numberof epitopes to induce adequate immunity. If such immunityrequires the presence of post-translational modifications orconformational epitopes, then vaccination with recombi-nant antigens produced in bacteria will most likely notsucceed. In addition, infectious agents with rapidly chang-ing or highly variable epitopes-e.g., common cold (Rhino-virus) (34), malaria (Plasmodium falciparum) (35, 36),sleeping sickness (Trypanosoma brucei) (37, 38), group Astreptococci (39), and, most likely, lilY-i (31, 40)-willprobably remain intractable to immunization by a singlerecombinant antigen. Newer, more clever and compleximmunization schemes mimicking the natural variation ofthese infectious agents or blocking their cell receptor sites,as in the case of CD4 for AIDS, will have to be developed.

Current efforts involving vaccination with live recombi-nant Vaccinia virus provide hope for circumventing thisproblem. Recombinant Vaccinia expressing the desiredantigens in the actual host facilitates the natural post-translational modifications and conformational folding (7,41). This is, perhaps, the most promising expression systemfor recombinant vaccines we possess today.

References

1. Darnell J, Lodish H, Baltimore D. Molecular cell biology. NewYork: WH Freeman & Co., 1986.2. Atassi MZ. Antigenic structures of proteins. Eur J Biochem1984;145: 1-20.3. Hood LE, Weissman JL, Wood WB. Immunology. Menlo Park,CA: Benjamin/Cummings Publishing Co., 1978.4. Kieney MP, Lathe R, Riviere Y, et al. Improved antigenicity ofthe HW e,w protein by cleavage site removal. Protein Eng1988;2:219-25.5. Chang TW, Kate I, McKinney S. et al. Detection of antibodies tohuman T-cell lymphotropic virus-ifi (HTLV-ffl) with an immu-noassay employing a recombinant Escherichia coli-derived viralantigenic peptide. Bio/Technology 1985;3:905-9.6. Sternberger LA. Immunocytochemistry, 2nd ed. New York:John Wiley & Sons, 1979.7. Flexner C, Murphy BR, RooneyJF, et al. Successful vaccinationwith a polyvalent live vector despite existing immunity to an

1842 CLINICAL CHEMISTRY, Vol.35, No. 9, 1989

expressed antigen. Nature (London) 1988;335:259-62.8. Mimms L, Staller J, Mushawhar 1K, et al. Production, purifi-cation, and immunological characterization of a recombinantDNA-derived hepatitis B e antigen. In: Viral hepatitis and liverdisease. New York: Alan R Liss, Inc., 1988:248-51.9. Decker RH, Kuhns MC, Brawner TA, et al. Future advanceddiagnostic techniques for hepatitis B. Ibid.:231-6.10. Decker RH, Dawson GJ. Recombinant antigens as diagnosticand screening reagents for HIV infections: experience with acommercial test. In: Luciw PA, Steimer KS, eds. HIV detection bygenetic engineering methods. New York: Marcel Dekker, Inc.,1989:77-97.11. Shoeman RL, Young D, Pottathil R, et al. Comparison ofrecombinant human immunodeficiency virus gag precursor andgag/env fusion proteins and a synthetic env peptide as diagnosticreagents. Anal Biochem 1987;161:370-9.12. Oikawa S, Nakazato H, Kosaki G. Primary structure ofhuman carcinoembryonic antigen (CEA) deduced from cDNAsequence. Biochem Biophys Res Commun 1987;142:511-8.13. Hansen EJ, Gonzales FR, Chamberlain NR, et al. Cloning ofthe gene encoding the major outer membrane protein of Haemo-philus influenzae type b. Infect Immunol 1988;56:2709-16.14. Vermersch PS, Klass MR, Bennett GN. Use of bacterialDHFR-II fusion proteins to elicit specific antibodies. Gene1986;41:289-97.15. Oroszlan S, Bova D, Gilden R. Isolation of mammalian type CRNA virus cross-reactive antigen and antibody by immuno-afflnitychromatography. Immunochemistry 1975;12:61-6.16. Kalyanaraman VS, Cabradillo CD, Getchell JP, et al. Anti-bodies to the core protein of lymphadenopathy-associated virus(LAV) in patients with AIDS. Science 1984;225:321-3.17. Barin F, McLane MF, Allan JS, et al. Virus envelope proteinof HTLV ifi represents the major target antigen for antibodies inAIDS patients. Science 1985;228:1094-.6.18. Lange JMA, Paul DA, Huismann JG, et al. Distinct IgGrecognition patterns during progression of subclinical and clinicalinfection with lymphadenopathy associated virus/human T lym-photropic virus. Br Med J 1986;292:228-30.19. Bartoloni A, Pizza M, Bigio M, et at. Mapping of a protectiveepitope of pertussis toxin by in vitro refolding of recombinantfragments-using monoclonal antibodies and site-directed mute-genesis; potential pertussis vaccine component. Bio/Technology1988;6:709-12.20. Ping L-H, Jansen RW, Stapleton JT, et al. Identification of animinunodominant antigenic site involving the capsid protein VP3of hepatitis A virus. Proc Natl Aced Sci USA 1988;85:8281-5.21. Johnston JM, Harmon SA, Binn LN, et al. Antigenic immu-nogenic properties of a hepatitis A virus capaid protein expressedin Escherichia coli. J Infect Dis 1988;157:1203-11.22. Gauss-Mueller V, Deinhardt F. Immunoreactivity of humanand rabbit antisera to hepatitis A virus. J Med Virol 1988;24:219-28.23. Ferguson M, Minor PD, Magrath DI, et al. Antigenic charac-terization of polio virus type 3 using monoclonal antibodies-

hybridoma generation and monoclonal antibody production. RevInfect Dis 1984;6(Suppl 2):5510-3.24. Pfaff E, Thiel HJ, Beck E, et al. Analysis of neutralizingepitopes on foot-and-mouth-disease virus-using monoclonal anti-body produced by hybridoma. J Virol 1988;62:2033-40.25. Robertson BH, Khanna B, Brown VK, et al. Large scaleproduction of hepatitis A virus in cell culture: effect of type ofinfection on virus yield and cell integrity. J Gen Virol1988;69:2129-.34.26. Gregersen J-P, Mehdi 5, Mauler R. Adaptation of hepatitis Avirus to high titre growth in diploid and permanent cell cultures.Med Microbiol Immunol 1988;177;91-100.27. Sanchez J, Holmgren J. Recombinant system for overexpres-sion of cholera toxin B subunit in Vibrio cholerae as a basis forvaccine development. Proc Natl Aced Sci USA 1989;86:481-5.28. McAleer WJ, Buynak EB, Maigetter RZ, et al. Human hepa-titis B vaccine from recombinant yeast. Nature (London)1984;307:178-80.29 Dandolos E, Roumeliotou-Karayannis A, Richardson SC, et al.Safety and immunogenicity of a recombinant hepatitis B vaccine.J Med Virol 1985;17:57-62.30. Murray K. Application of recombinant DNA techniques in thedevelopment of viral vaccines against hepatitis B virus; a review.Vaccine 1988;6:164-74.31. Koff WC, Hoth DF. Development and testing of AIDS vac-cines-recombinant vaccine preparation, monoclonal antibodyconstruction, etc.; a review. Science 1988;241:426-32.32. Tite JP, Russell SM, Dougan G, et al. Antiviral immunityinduced by recombinant nucleoprotein of influenza A virus. I.Characteristics and cross-reactivity of T cell responses. J Immunol1988;141:3980-7.33. Barr PJ, Gibson HL. Genetic engineering approaches toward amalaria vaccine-a review. Biotech USA 1987:222-9.34. Macnaughton MR. The structure and replication of rhinovi-ruses. Curr Top Microbiol Immunol 1982;97:1-26.35. Favaloro JM, Marshall VM, Crewther PE, et a!. cDNA se-quence predicting an octapeptide-repeat antigen of Plasmodiumfalciparum. Mol Biochem Parasitol 1989;32:297-300.36. Ravetch JV, Young J, Poste G. Molecular genetic strategiesfor the development of anti-malaria vaccines-review of currentresearch on malarial vaccine development. Bio/Technology1985:3:729-40.37. Borta P, Cross GAM. Molecular basis for trypanosome anti-gemc variation. Cell 1982;29:291-303.38. Mowatt MR, Wisdom GS, Clayton CE. Variation of tandemrepeats in the developmentally regulated procyclic acidic repeti-tive proteins of Trypanosonia brucei. Mol Cell Biol 1989;9:1332-5.39. Hruby DE, Hodges WM, Wilson EM, et al. Expression ofstreptococcal M protein in mammalian cells. Proc Natl Aced SciUSA 1988;85:5714-7.40. Godin N. A.ms: is a safe vaccine readily available? MedHypotheses 1989;28:3-5.41. Mackett M, Smith GL. Vaccinia virus expression vectors. JGen Virol 1986;67:2067-82.