Med Microbiol Immunol (1992) 181:215-226

9 Springer-Verlag 1992

Determination of the minimal length of preS1 epitope recognized by a monoclonal antibody which inhibits attachment of Hepatitis B virus to hepatocytes I. Sominskaya 1, p. Pushko 2, D. Dreilina 2, T. Koziovskaya 2, and P. Pumpen 2 i Hepatological Center, Latvian Medical Academy,Riga, Latvia 2 Institute of Molecular Biology, Latvian Academy of Sciencesand Universityof Latvia, Krustpils Street 53, 226065 Riga, Latvia ReceivedApril 27, 1992

Abstract. The minimal amino acid sequence sufficient to be recognized efficiently by virus-attachment inhibiting murine monoclonal anti-preS 1 antibody MA18/7 has been determined. We have constructed a recombinant gene library using the cloned coat protein gene of Escherichia coli RNA bacteriophage fr as a carrier. Different fragments of preS 1 region from cloned hepatitis B virus (HBV) genomes, subtype ayw and adw, were inserted at position 2 of the 129 amino acid-long fr coat protein gene in the appropriate E. coli expression vectors. Fine mapping of preS 1 epitope recognized by MA18/7 was accomplished by bidirectional shortening of the preS1 within original recombinant preS-fr coat protein genes with Bal31 exonuclease. Immunoblot analysis of the obtained recombinant protein library revealed that the tetrapeptide Asp-Pro-Ala-Phe (DPAF), located at the position preS(31-34) and conserved in all known HBV genomes, is sufficient to bind MA18/7 antibody. Recognition of the preS1 region by MA18/7 occurred irrespective of the amino acid context surrounding this DPAF tetrapeptide. Further shortening of this minimal epitope from the left or from the right side completely prevented antibody binding in immunoblots.

Introduction The precise mapping of sequential epitopes recognized by monoclonal antibodies (mAb) or polyclonal antibodies of different origin is necessary to (1) understand the molecular mechanisms of immune response and/or host-virus, (2) study antigen-antibody binding on the level of epitope-paratope recognition, and (3) create vaccines and immunodiagnostic tools on the basis of artificial chimeric proteins carrying desired minimal epitopes on their surface. Such mapping requires the determination of the shortest stretches within protein molecule that are still capable of binding the appropriate antibodies. The Correspondence to: P. Pumpen

216 most prevalent mapping techniques presently used are based on one of two peptide libraries. The first consists of short overlapping synthetic peptides which cover the whole protein sequence in question (Geysen et al. 1984). Secondly, a library is used that consist of fused proteins constructed by cloning of overlapping DNA fragments of appropriate genes in suitable expression vectors, with subsequent shortening of putative epitope regions by exonucleases Bal31 or Exo III (for review, see Lenstra et al. 1990). Since examination of epitopes located within relatively long fused protein sequences can be more easily made independent of the epitope length and their surrounding than in the case of short synthetic peptides, the fusedgene approach seems to us more convenient to determine exactly minimal epitope sequences. Although this approach was used successfully in mapping real epitopes of different origins, including those from the preS region of hepatitis B virus (HBV) (Kuroki et al. 1990), the utmost possible definition of epitope length has not yet been achieved. For example, minimal epitope lengths that have been determined to be present are as a rule not less than seven to nine amino acid residues long (Kuroki et al. 1990; Lenstra et al. 1990). In the present study there were two main reasons for choosing the preS region of HBV as one of the models for the fine determination of epitope length. Firstly, the preS region possesses sequential, as opposed to conformational, epitopes (Milich et al. 1985, 1986, 1987; Neurath et al. 1984, 1985). Secondly, the preS sequences are practically very important from the point of view of vaccine design and creation of potential HBV diagnostic methods (for review, see Neurath and Kent 1988; Heermann and Gerlich 1992). The preS domains represent the outside part of the HBV envelope, which consists of three cocarboxy-terminal proteins designated as small (S, 226 amino acid residues), middle [M, i.e. S+55 amino acid residues ( p r e S 2 ) at the N terminus] and large [L, i.e. M + 108 (subtype ayw) or + 119 (other subtypes) amino acid residues (-preS1) at the N terminus] (Heermann et al. 1984; Stibbe and Gerlich 1982, 1983; Neurath et al. 1985). The preS 1 domain is responsible for some important properties of HBV such as binding of the virus particle to a hepatocyte receptor (Neurath et al. 1985, 1986; Petit et al. 1991; Pontisso et al. 1989a, 1989b) and the induction of virusneutralizing antibodies (Neurath et al. 1986a, 1988). As to the fine structure of the preS1 domain, it has been shown directly that the sequence preS (21-47) is responsible for the attachment of HBV to hepatocytes (Neurath et al. 1986b). More recent studies localized the binding site to shorter sequences preS (32-47) or (2749) (Neurath and Kent 1988; Pontisso et al. 1989b). Such binding of virus particles to cell membranes can be prevented by murine mAb which recognize the same region within preS1, e.g. by the first-established anti-preS1 mAb MA18/7 (Pontisso et al. 1989a, 1989b). This particular region also belongs to the immunodominant areas of preS. The preS1 sequence reveals two large immunodominant regions, preS(12-53) and preS(94-117), which contain major epitopes responsible for preSl-specific B cell immunoreactivity in humans, rabbits and mice (for review, see Neurath and Kent 1988). These regions also bear preSl-encoded HBV-protective epitopes localized originally within residues 12-47 and 94-117 (Thornton et al. 1989). The recent work on precise mapping of a new panel of murine anti-preS 1 mAb (Kuroki et al. 1990; Mimms et al. 1990) confirmed these previous conclusions and narrowed epitope boundaries to residues preS(27-35) and (72-78). The mAb from the first

217 group, which recognized preS(27-35), also competed with MA18/7 (Mimms et al. 1990). Moreover, epitope mapping based on synthetic peptides have shortened the recognition site of MA18/7 to six residues spanning positions 31-36 (Evstigneeva et al. 1990). Antibodies which compete with MA18/7 for its epitope occur in one third of HBsAg carriers who are negative for HBeAg. They also occur in one third of hepatitis B convalescent cases irrespective of the presence of anti-HBs (Deepen et al. 1990). Taking into account of both the adhesive and the immunodominant natures of the preS1 epitope recognized by antibodies of the MA18/7 type, we have chosen the latter for the determination of the exact minimal length of the sequential epitope. As a carrier for construction of a library of fused proteins we have used the cloned coat-protein gene of RNA bacteriophage fr (Kozlovskaya et al. 1986) which belongs, together with such well-known phages as f2, MS2, R17 etc., to the first serological group of small icosahedric Eseherichia coli RNA phages (for review, see Gren 1974; Zinder 1975). In this report we describe fine mapping of epitope recognized by MA18/7 antibodies to four-amino acid residues which appears to be the minimum possible epitope length which still allows effective antigen-antibody binding, irrespective of surrounding amino acid sequences.

Materials and methods

Monoclonal and polyclonal antibodies The preS 1-specificmAb MA 18/7 generated by immunization of mice with HBV particles detects the denaturation-resistant, subtype-independent epitope of the preS1 protein (Heermann et al. 1984). mAb produced from ascites fluid and purified by sodium sulfate precipitation was kindly supplied to us by W. H. Gerlich (Giessen, Germany). The anti-fr polyclonal antibodies obtained by immunization of rabbits with fr coat protein, prepared according to cold acetic acid method of Sugiyama et al. (1967), were a gift from V. Bauman (Riga). These antibodies recognize fr phage and recombinant fr empty capsids (Kozlovskayaet al. 1986)as well as the fr coat protein denatured by heating at 100~ for 5 min in SDS-gel electrophoresis sample buffer (Laemmli 1970) containing 2% SDS and 2% 2mercaptoethanol.

Construction of the fr CP expression vectors Expression vectors were constructed by a combination of standard techniques on the basis of the pBR322trp expression plasmid harboring a fragment of the E. coli tryptophan operon with the trp promoter adjacent to the Eco RI site (Ovchinnikov et al. 1982). To construct fr CP expression plasmid pFRA8 (Kozlovskaya et al. 1986), a 1243-bp long EcoRV-BglI fragment of cloned RNA phage fr genome (Berzin et al. 1986) containing 596 C-terminal nucleotides of fr A protein (AP) gene, 393 bp o f f r CP gene and 35 N-terminal nucleotides o f f r replicase (RP) gene, as well as AP-CP and CP-RP intercistronic regions, 25 and 33 bp long, respectively, was inserted downstream of the trp promoter into the Eco RI site ofpBR322trp, and the 1259-bp Eco RV-Bal I fragment from the original pBR322 sequence was deleted (Fig. 1).

218

A

pFR A 8 5264 bp

/

/

As Av Ba Bs C D El EV H K M N Ps Pv Sa Sc Sm Sp St Xb Xm

B i

C

pFRd8

i

[~3C~-C~C... M A1

82 N3 EVC

Pv

As

GClq" [CGATATCGATCGCI-F]CG,AAC...

pFR771 [ ~ M

A1

S

I

S

I

A

$2

C

EV C

AS pFRtQ

I

~

[~

A1

Ill

S

K

R

S Y

I

As El

Sc

IL

I

R

Y

R

S

F

E

K,XmSm Ba

L

Xb

Sa H

Ps

Sp

I

II

I

I

III

GCTFICGAA"I-TCGAGCTCGGTACCCGGGGATCCTCTAGAGTCGACCTGCAGGCATGCAAGCT I CGAAC._ M

pFR42

As

GCTr [CGAAGCGATCATATCGATATCGATCCTT1 CGAACTIq'GA M

pFR3G

N3

Asull Aval BamHI BstEII Clal Dral EcoRI EcoRV Hindll Kpnl Mfel Nrul Pstl Pvul Sail Sacl Smal Sphl Sty1 Xbal Xmal

A1

S

N

S

S

S

V

P

G

D

P

Sp

Ps Sa H

Xb

I

lil

I

L

E

Ba

s

T

Xm Sm

I II

C

R

H

A

S

$2

K

Sc El

As

I

Ill

N3

[-ATG] GC'FI" [CGAGCTFGCATGCCTGCAGGTCGACTCTAGAGGATCCCCGGGTACCGAGCTCGAATr] CGAAC... M A1 S S L H A C R s T L E D P R V P S S N $2 N3

Fig. 1A-C. Schematic diagram of the expression vectors for epitope mapping on the basis of cloned fr CP gene. A The parental pFRA8 plasmid was constructed from pB322trp as described under Materials and methods, while all other plasmids were derived from pFRA8 by insertion of polylinkers into theAsuII site located at the N terminus of the fr CP gene. The thin line represents pBR322 and the thick line represents trp operon sequences. The cloned fr CP gene and fragments of A protein (AP) and replicase (RP) genes are shown as open boxes, intercistronic regions of fr genome arefilled. Bla gene ofpBR322 and trp promoter are shown as open arrows. The direction of transcription is indicated by the thin arrows. B The fr CP gene with marked positions of restriction endonuclease sites which are suitable for further constructs with the aim of creating fused gene libraries is shown as an open box. Start Met codon is boxed, but numbering starts at the Ala codon following the initiating Met which is not present in mature protein. C N-terminal nucleotide and amino acid sequence of parental fr CP gene and recombinant fr CP vectors with polylinkers inserted at AsuII site are depicted. Symbols (explained in the figure) indicate the restriction sites. The single-letter amino acid code is used

219 To expand the in-frame insertion capacity at the N terminus o f the fr CP gene, synthetic polylinkers were inserted at the AsuII site. Vectors pFR771 and p F R 1 0 were constructed by cloning the oligonucleotide 5'-CGATATCGATCGCTT-3' 3'- T A T A G C T A G C G A A G C - 5 ' carrying restriction sites EcoRV, ClaI and PvuI into the AsulI site located at the N terminus o f f r C P gene. In-frame as well as out-of-frame recombinants o f the fr CP gene were selected: pFR771 represented one o f the in-frame recombinants with simple insertion o f a single copy o f the oligonucleotide, and p FR10 belonged to the out-of-frame recombinants in which the reading frame can be restored by insertion o f a foreign sequence. Such out-of-frame vectors had some advantages for direct immunoscreening o f fused proteins because o f the immunological inactivity o f parental clones. Vectors p F R 3 6 and p F R 4 2 were created by insertion o f the p U C 19 polylinker in two orientations after b o t h polylinker and AsuII sticky ends were filled-in by the Klenow fragment o f D N A polymerase. All fr C P expression vectors directed synthesis of fr CP or its N-terminal mutants at the level o f up to 20% o f total cellular protein with the exception o f p F R 4 2 and its derivatives which showed m u c h less synthesis.

Construction o f preS-fr CP expression gene library The entire preS1 region of HBV, subtype ayw, together with the 46 codon-long N-terminal fragment of preS2, as a 494- bp BstEII-XhoI fragment of cloned HBV ayw DNA (Bichko et al. 1985), and the preS1 region of HBV, subtype adw, as a 398-bp BstEII-EcoRI fragment of cloned HBV adw DNA (Meisel et al., in preparation) were inserted into the BarnHI restriction site of the polylinker region of the pFR42 vector plasmid. Correct reading frames in both cases were restored by means of restriction enzyme cleavage and fill-in reaction within the polylinker sequence (AsuII and Xba I) for pFR9618 and pFR 1018, respectively). For subfragmentation of preS l/ayw inserted at the N-terminal part of the fr CP gene, the 150bp BamHI-MstlI fragment of HBV ayw DNA spanning amino acid residues 31-80, and 60-bp Bam HI-Mva I fragment (amino acids 32-51) were cloned into the filled-in Asu II restriction site of pFR10. The 114-bp MstlI-MstlI fragment (amino acids 80-117) was cloned into EcoRV site of pFR771. For construction of the preS1/ayw deletion plasmid library, a pFR144 plasmid bearing the BamHI-MstlI fragment of HBV ayw DNA was cleaved with (a) BamHI (left-handed part of library, plasmid series pFR14), (b) with StuI or (c) simultaneously with StuI and Bali (righthanded part, plasmid series pFR13 and pFR94/104, respectively) and then exposed to Bal31 (Enzymas, Vilnius or Boehringer-Mannheim) exonuclease (0.3 U Bal31/ggDNA) for graded intervals of 10 s from 20 to 90 s at 20 ~C. After EGTA addition (11.7 mM final concentration) and ethanol precipitation, the DNA aliquots were recircularized with T4 DNA ligase and transformed into E. coli RRI. For insertion of a synthetic epitope, an oligonucleotide copy of the appropriate amino acid sequence (Asp-Pro-Ala-Phe) provided with the necessary sticky ends was synthesized and cloned in the Sail restriction site within fr CP gene.

Colony-immunoscreening analysis In-flame mutants were selected by immunoscreening of colonies with anti-fr antibodies according, in general, to a known technique (Wang and Esen 1985). Colonies transferred onto nitrocellulose filters (Bio-Rad, 0.2 ~t) were lysed with chloroform, incubated overnight at room

220 temperature with anti-fr (dilution of initial rabbit serum 1:100), then 1 h with Staphylococcusaureus A protein-horseradish peroxidase conjugate, and developed with 3,3'-diaminobenzidine peroxidase substrate. Deletions in selected in-frame mutants were identified by dideoxy sequencing of plasmid miniprep DNA, using synthetic 18-base long oligonucleotides as primers.

Polyacrylamide gel electrophoresis and immunoblot analysis E. colistrain K802 cells harboring the appropriate plasmids and demonstrating higher production level of fused proteins in comparison with E. coliRR1 strain were grown to saturation overnight in M9 synthetic medium supplemented with 2%g/1 Casamino acids (Difco). Bacteria were pelleted, suspended in SDS-gel electrophoresis sample buffer containing 2% SDS and 2% 2mercaptoethanol and lysed by heating at 100~ for 5 rain. Proteins were separated by Laemmli's polyacrylamide gel electrophoresis (PAGE) (Laemmli 1970) in a slab gel (150 • 150 • 0.75 mm) apparatus with gradient 12%-18 % running gel and 4% stacking gel. Western blotting was conducted in general as described by Towbin et al. (1979). Two parallel nitrocellulose sheets (HAWP, Millipore or Bio-Rad, 0.2 la) were incubated with (a) anti-fr (dilution 1: 100) and then A protein-peroxidase conjugate; on (b) MA18/7 (dilution 1:200) and then rabbit anti-mouse IgG and peroxidase-anti-peroxidase, mouse, complex and were developed with 3,Y-diaminobenzidine peroxidase substrate.

Results

Construction of a library of preSl-fr CP fusion proteins Native fr CP gene possesses some unique restriction sites which permit very easy insertion of foreign gene fragments at different parts within a 129-amino acid long fr CP molecule (Fig. 1). These sites are the AsuII spanning positions 1-3 of fr CP, SalI (positions 10-11), SacI (50-52), StyI (71-73), NruI (95-97), DraI (112-113), BstEII (115-117), MfeI (117-119) and AvaI (125-127). Insertion of various synthetic polylinkers into these natural restriction sites enabled us to broaden real resources of fr CP expression vectors from the point of view of diversity of reading frames in fragments to be inserted (Borisova et al. 1987; Kozlovskaya et al. 1988a). The preliminary experiments with preS fragments and also with other epitopes (from HBV, H I V and BLV proteins) showed that the restriction site AsuII located at the very N terminus of fr CP was the best choice in terms of stability and production level of expected fusion proteins. Furthermore, N-terminal recombinants of fr CP can be selected very simply by direct immunoscreening of in-frame fusions using polyclonal antibodies or anti-fr m A b (not shown). To maintain the correct reading frame of fragments prepared for insertion into the N termius of fr CP, we have constructed two variants of fr CP gene within plasmids pFR36 and pFR42 containing a polylinker sequence from pUC19 inserted at the AsuII site in both orientations (Fig. 1). The choice of vector was also made with regard to the reading frame of the appropriate fragments to be inserted. Initially, we constructed fr CP fused genes that contained the entire preS1 region from two cloned HBV genomes that represent two widely distributed and most distant HBV subtypes: ayw (Bichko et al. 1985) and adw (Meisel et al., in preparation), of Latvian and G e r m a n origin, respectively (Fig. 2). PreS1/ayw D N A was subcloned within the fr CP gene in nearly equally long fragments, and after preliminary tests on MA 18/7 recognition (not shown) preS 1reactive fr CP fusion pFR144, which contained preS1 amino acids 31-80, was

221 1 10 20 30 40 50 60 70 80 90 100 110 ........... MGQNLSTSNPLGFFPOHQLOPAFRANTANFDWDFNPNKDTMEIAN~G LGFTPPHGGLL~ILETLPANPPPASTNRQSC~~ k,~KPRKG T VP G ~N I H A Q V PL I T VSTI I BstEII

i=1 ,

BamHl r

Imlll

Mval ~

I

II

Stul

Mstll

I

I

119

MSltl

Xhol

EcoRI

BstEII

122

pFR10181m'w

,.11185

12 ~ p P ~ 1 2 8 3 1 ~ 1 pFR81 3~ i pFR14,P~ 3 3 r ~ pFR1418 38, _p,'~'R1412 _ 48 i pFR1310 31 pFR1312 31 pt=R13~831 pFR133731 pFR943 31 I I I l ~ l l l ~ i 4 0 pFR1041131 ~ 3 4 pFR945 31 E:~33

81 I i0 ~51

~80 ',~8(~ II 56

47

I117

78 ~ l l m 80 78 Imm180 7e m m l E I 78 m 80 74 i 8 0

75 c:~:::~80

Fig. 2. Schematic representation of the preS 1-fr CP expression library. The upper bar symbolizes the preS region of hepatitis B virus (HBV). Restriction endonuclease recognition sites that were used for creation of fused preSl-fr CP genes harboring fragments of HBV preS1, subtype ayw (top) or adw (bottom), are indicated. Symbols are explained in the figure. The amino acid sequences of preS1, subtype ayw (Bichko et al. 1985) and adw (Meisel et al., in preparation) are depicted. Boxes show the positions of preS 1 sequences cloned within chimeric fr-preS 1 genes as designated on the left. Numbers represent starting and ending amino acid positions of preS 1 fragments. Fragments that bind MA18/7 are shown asfilled areas, inactive fragments are shown as open areas. All segments are drawn to scale of upper restriction map. The single-letter amino acid code is used

subjected to shortening with Bal31 exonuclease. The following restriction sites were used: B a m H I for left-handed and Ball and Eco 1471 (StuI) for right-handed shortening of preS1 sequence. The structure of fused genes was determined by Sanger sequencing, and appropriate preSl-fr CP fusion proteins were applied directly to mapping of the MA18/7 epitope.

Determination of the minimal length of MA18/7 epitope In accordance with experimental data published earlier (Evstigneeva et al. 1990; Kuroki et al. 1990; Pontisso et al. 1989b), MA18/7 recognized the preS1 sequence of H BV subtype ayw as well as of adw. Rough subfragmentation of ayw preS 1 DNA within fr CP gene pointed at preS(31-80) as one bearing the MA18/7 epitope (Fig. 3). Left-handed shortening of the preS1 sequence within recombinant plasmid pFR144 showed at once that deletion of the first amino acid residue of preS1 Asp31 within plasmid pFR81 - already eliminated preS 1 reactivity of such fused proteins (Fig. 3, lanes 2 and 3). Further deletion of the second preS 1 amino acid residue Pro32 (pFR1425) did not restore the MA18/7 epitope (Fig. 3, lane4). Fused proteins lacking amino acid residues 31-37 (pFR1418), 31-42 (pFR1412) were also inactive in immunoblots (data not shown). These observations allowed us to identify preS1 amino acid residue Asp31 as a left boundary of MA18/7 epitope. To delimit the right border of the MA 18/7 epitope, the right-handed part of the p FR 144 deletion library was constructed. In this case we used two restriction sites first Eco 1471 and then Bali which is more proximal to the epitope. Figure 3 (lane 12) shows that elimination of a long stretch of amino acid residues including

222

Fig.3A-B. lmmunoblot analysis of preSl-fr CP hybrid protein library. Whole cell lysates of E. coli strain K802 harboring appropriate plasmids were prepared from overnight culture by boiling in Laemmli's sample buffer, separated by SDS-PAGE and electroblotted onto nitrocellulose membranes. For experimental details, see Materials and methods. Two parallel nitrocellulose sheets were specifically stained using polyclonal anti-fr antibody (A) or monoclonal anti-preS 1 antibody MA18/7 (B). Lanes: 1, pFR9618 coding for fr CP-preS(12-165) gene; 2, pFR144 - fr CPpreS(31-80); 3, pFR81 - fr CP-preS(32-51 ); 4, pFR 1425 - fr CP-preS(33-80); 5, pFR 1310 - fr CPpreS(31-69); 6, pFR1312 - fr CP-preS(31-65); 7, pFR1335 - fr CP-preS(31-55); 8, pFR1337 - fr CP-preS(31-47); 9, pFR943 - fr CP-preS(31-40); 10; pFR 10411 fr CP-preS(31-34); 11, pFR9413 fr CP-preS(31-34); 12, pFR945 - fr CP-preS(31-33); 13, pFRA8 (fr CP); 14, pFR1018 - fr CPpreS(1-122, adw); 15, pFRSal4 - fr CP-preS(31-34); 16, pFRSall3 - fr CP-preS(31-34). Schematic description of these constructs are given in Fig. 2 or in Table 1. The arrow points to the initial fr CP (lane 13). Lane 13 is a positive control for anti-fr antibody and a negative control for MAI8/7. The additional smaller protein band present in lane2 is probably a fusion-protein degradation product that lacks a discrete portion from the C terminus, since this band is visible with the MA 18/7 antibody. The positions of molecular size markers are indicated in the kDa scale

Phe34 (pFR945) was necessary to destroy M A 1 8 / 7 epitope. Both pFR10411 and pFR9413 which contained the entire region Asp31-Pro32-Ala33-Phe34 were recognized well by M A 1 8 / 7 (Fig. 3, lanes 10 and 11). This allowed us to proclaim Phe34 as a right b o u n d a r y of epitope.

Recognition o f the minimal M A 18/7 epitope within different surrounding amino acids After the tetrapeptide A s p - P r o - A l a - P h e was identified as the minimal recognition site for M A 1 8 / 7 , the question arose as to the possible role of the surrounding amino acids in epitope reactivity. The original H B V sequences of ayw and adw subtypes already presented two variants of the right-sided n e i g h b o r h o o d , i.e. Arg and Gly, and our m a p p i n g results offered two new variants, Leu and Glu (Table 1). To alter substantially the context in which the M A 1 8 / 7 epitope is situated, we inserted a synthetic D N A copy of the epitope into the SalI site spanning positions 10-11 within the fr CP gene. Plasmid pSal4 bore one copy and plasmid p S a l l 3 , the tandem repeat of the M A 1 8 / 7 epitope now with Val as left and right neighbors (Table 1). Both fused proteins were recognized equally well by M A 1 8 / 7 (Fig. 3).

223 Table 1. Antibody-binding properties of the minimal length MA18/7 epitope in dependence of surrounding amino acid residues Source of epitope Clone a

Epitope surrounding b

Immunoblot activity c

QL QL

DPAF DPAF

RA GA

+ +

AS AS AS AS LV LV

DPAF DPAF 9 PAF DPA. DPAF DPAF

EE LL RA LG VD V ~

+ +

Description of preS 1 fragment

Full-length preS 1: pFR9618 Subtype ayw pFR1018 Subtype adw Minimalized preS 1/ayw: pFR10411 pFR9413 pFR81 pFR945 pFRSal4 pFRSal 13

AsulI insertion AsulI insertion Asull insertion Asull insertion SalI insertion SalI insertion

VD

+ +

a Schematic structure of preSl-fr CP hybrid proteins (pFR9618, pFR1018, pFR10411, pFR81, pFR945) is depicted in Fig. 2 b Minimal length epitope recognized by MA18/7 is boxed. Two neighbouring amino acid residues are shown, amino acids adjacent to the minimal epitope are marked as bold-face letters9 Dots in amino acids sequence represent deletions c Experimental data are given in Fig. 3

These recombinants offer Val as the fifth possible variant for the right neighbour, and Val as the third for the left neighbour; Leu is the left neighbour in the native preS1 context and Ser in preSl-fr CP fused proteins.

Discussion Libraries of fused genes offer an easy method for fine mapping of sequential epitopes recognized by polyclonal antibodies or mAb. A disadvantage of short synthetic peptides is the sevese influence of length on their antigenic properties. Moreover, it m a y be very confusing to prove the antigenicity of short structures such as tetrapeptides if the latter are regarded as candidates for the role of minimal epitope (Khaitov and Andreev 1990). At first we needed to choose the expression system for the construction of the fused gene library. In our opinion, one of the most advantageous systems is based on the expression of the cloned coat-protein gene ofE. coli R N A phage fr (Borisova et al. 1987; Kozlovskaya et al. 1986, 1988a). The expression of the cloned fr CP gene under the control of the strong E. coli trp operon p r o m o t e r results in the highlevel synthesis of monomeric CP subunits that associate themselves in bacterial cells into stable capsid-like structures, indistinguishable morphologically and immunologically from mature fr phage particles (Kozlovskaya et al. 1986). Such fr recombinant capsids are used as a carrier for the exposure of inserted foreign sequences on their surface (Kozlovskaya et al. 1988b). The N terminus of the fr CP

224 gene is one of the most promising site for the insertion of such foreign sequences. Cloning of relatively short sequences, up to 20 amino acid residues, within the AsulI site at the extreme end of the terminus of the fr CP gene does not destroy, as a rule, the self assembly of chimeric monomers, but insertion of larger stretches of sequence exceeding 20 amino acids usually prevents the process of encapsidation and leads to formation of noncovalent dimers of fused proteins in bacterial cells (Pushko et al., in preparation). Chimeras bearing more than 100 amino acid insertions within the N terminus of the fr CP are usually insoluble. Production levels of fused protein depends heavily on the nature of inserted sequences, but in all cases studied previously, it was high enough to satisfy the immunoblot requirements. In addition, the expression system used has the advantage of a very easy and universal selection of in-frame chimeras, which can be used for the very sensitive search of all possible N-terminal fusions of fr CP by direct and/or Western blot immunoscreening with anti-fr antibodies. Insertion of various polylinkers into AsulI site of fr CP gene (1) broadens the sphere of simple in-frame insertions, (2) offers a method for preparation of non-symmetrically split vectors specialized for the insertion of synthetic DNA copies of immunological epitopes, and (3) allows the use of unidirectional shortening with ExolII, instead of bidirectional shortening with Bal31 (Pushko et al., in preparation). We have tried to solve the question of the minimal length of an epitope using as a model the well-known and extensively studied anti-preS1 antibody MA18/7 (Heermann et al. 1984; Pontisso et al. 1989a, 1989b). The MA18/7 epitope was narrowed in earlier works to seven, preS(30-36), and then to six, (31-36), amino acid residues (Evstigneeva et al. 1990). The epitope recognized by antibodies which were cross-reactive with MA18/7 was mapped at preS(27-35) (Kuroki et al. 1990). Our main idea was that the real epitope consisted, in fact, of a shorter amino acid stretch. In spite of the widespread adoption of the fused-gene technique for epitope mapping (Lenstra et al. 1990), there are not yet many data on such comprehensive minimizations of epitopes or on the real length of the typical sequential epitope. The data obtained by Kuroki et al. (1990) who have narrowed the immunodominant preS1 epitopes (27-35) and (72-78) to nine and seven amino acid residues, respectively, and by Salfeld et al. (1989) who located the HBV core antigen epitope within ten amino acid residues are among the most remarkable in this sense. As far as we know, therefore, the MA18/7 epitope minimized by our group, is the shortest sequential epitope localized using the fused-gene libraries. One of the most valuable minimizations was made, however, with synthetic peptides for antiHBe/b mAb (Sallberg et al. 1991). According to this report, a linear epitope sequence not shorter than four amino acid residues is needed for antibody recognition but only two of them are absolutely unchangeable. It seems very likely that also in our case the significance of epitope tetrapeptide residues is not equivalent. Unfortunately, we do not have data concerning the role of each amino acid residue within the DPAF peptide. Only the consecutive substitution of each residue against the whole set of 20 amino acids, work that is now in progress, can answer this question. In conclusion, the universal mapping system described here offers new possibilities for the further investigation of minimal epitope length and the possible role of surrounding amino acids. Minimization of immunodominant preS2 epitopes is now under way.

225

Acknowledgements. We thank Prof. Dr. W. H. Gerlich for a generous gift of MAI8/7 and for his permanent interest in our work, J. Ozols for skillful technical assistance, Dr. V. Bauman for preparation of polyclonal anti-fr antibodies, and Dr. E. Stankevica and her group for providing us with synthetic oligonucleotides. We also thank Prof. Dr. E. Grens for his encouragement throughout this work and for critically reading the manuscript. References Berzin VM, Avots A J, Jansone IV, Tsimanis AJ (1986) The primary structure of a fragment of the phage fr cDNA. Bioorg Khim 12:149-152 (in Russian) Bichko V, Pushko P, Dreilina D, Pumpen P, Gren E (1985) Subtype ayw variant of hepatitis B virus. DNA primary structure analysis. FEBS Lett 185: 208-212 Borisova G, Bundule M, Grinstein E, Dreilina D, Dreimane A, Kalis J, Kozlovskaya T, Loseva V, Ose V, Pumpen P, Pushko P, Snikere D, Stankevica E, Tsibinogin V, Gren EJ (1987) Recombinant capsid structures for exposure of protein antigenic epitopes. Mol Gen (Life Sci Adv) 6:169-174 Deepen R, Heermann K-H, Uy A, Thomssen R, Gerlich WH (1990) Assay ofpreS epitopes and preS 1 antibody in hepatitis B virus carriers and immune persons. Med Microbiol Immunol 179:49-60 Evstigneeva RP, Prokuronova EI, Zehltukhina GA, Smirnov VD, Semiletov IuA, Khudiakov Iu, Kalinina TI, Gerlich WH (1990) Synthesis and immunologic properties of the peptide fragment 30-36 from the preS-protein of the hepatitis B envelope. Dokl Akad Nauk SSSR 313:1135-1138 (in Russian) Geysen HM, Meloen RH, Barteling SJ (1984) Use of peptide synthesis to probe viral antigens for epitopes to a resolution of a single amino acid. Proc Natl Acad Sci USA 81 : 3998-4002 Gren EJ (1974) Regulatory mechanisms of RNA bacteriophage replication. Zinatne Publishers, Riga (in Russian) Heermann KH, Gerlich WH (1992) Surface proteins of hepatitis B viruses. In: MacLachlan A (ed) Molecular biology of hepatitis B viruses. CRC Press Boca Raton (in press) Heermann KH, Goldmann U, Schwartz U, Seyffarth T, Baumgarten H, Gerlich WH (1984) Large surface proteins of hepatitis B virus containing the pre-S sequence. J Virol 52: 396-402 Khaitov RM, Andreev SM (1990) Prospects for the use of synthetic antigens in immunodiagnosis. Biomed Sci 1:553-564 Kozlovskaya TM, Pumpen PP, Dreilina DE, Tsimanis AYu, Ose VP, Tsibinogin VV, Gren EJ (1986) Formation of capsid-like structures as a result of expression of cloned coat protein gene of RNA bacteriophage fr. Dokl Akad Nauk SSSR 287:452-455 (in Russian) Kozlovskaya TM, Pushko PM, Stankevich EI, Dreimane A J, Snikere D J, Grinstein EE, Dreilina DE, Vejina AE, Ose VP, Pumpen PP, Gren EJ (1988a) Genetically engineered mutants of the coat protein of RNA-containing bacteriophage fr. Mol Biol (Mosk) 22:731-740 (in Russian) Kozlovskaya TM, Sominskaya IV, Sergeyeva SM, Tsibinogin VV, Pumpen PP, Gren EJ (1988b) Genetically engineering's approaches to direct investigation of Dane particles and antibodies against them for diagnostic purpose. In: Bluger AF (ed) Novels in hepatology: methods, facts, conceptions. RMI Press, Riga, pp 40-49 (in Russian) Kuroki K, Floreani M, Mimms M, Ganem D (1990) Epitope mapping of the preS 1 domain of the hepatitis B virus large surface protein. Virology 176:620-624 Laemmli UK (1970) Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227:680-685 Lenstra JA, Kusters JG, van der Zeijst BAM (1990) Mapping of viral epitopes with prokaryotic expression products. Arch Virol 110:1-24 Milich DR, Thornton GB, Neurath AR, Kent SB, Michel M-L, Tiollais P, Chisari FV (1985) Enhanced immunogenicity of the pre-S region of hepatitis B surface antigen. Science 228:1195-1199 Milich DR, McLachlan A, Chisari FV, Kent SBH, Thornton GB (1986) Immune response to the pre-S(1) region of the hepatitis B surface antigen (HBsAg): a pre-S(1) specific T cell response can bypass nonresponsiveness to the pre-S(2) and S regions of HBsAg. J Immunol 137:315322

226 Milich DR, McLachlan A, Moriarty A, Thornton GB (1987) A single 10-residue pre-S(1) peptide can prime T cell help for antibody production to multiple epitopes within the pre-S(1), preS(2), and S regions of HBsAg. J Immunol 138:4457-4465 Mimms LT, Floreani M, Tyner J, Whitters E, Rosenlof R, Wray L, Goetze A, Sarin V, Eble K (1990) Discriminationof hepatitis B virus (HBV) subtypes using monoclonal antibodies to the preS1 and preS2 domains of the viral envelope. Virology 176:604-619 Neurath AR, Kent SBH (1988) The pre-S region of hepadnavirus envelope proteins. Adv Virus Res 34:65-142 Neurath AR, Kent SBH, Strick N (1984) Location and chemical synthesis of a pre-S gene coded immunodominant epitope of hepatitis B virus. Science 224:392-395 Neurath AR, Kent SBH, Strick N, Taylor P, Stevens CE (1985) Hepatitis B virus contains pre-S gene-encoded domains. Nature 315 : 154-156 Neurath AR, Kent SBH, Parker K, Prince AM, Strick N, Brotman B, Sproul P (1986a) Antibodies to a synthetic peptide from the preS120-145 region of the hepatitis B virus envelope are virus neutralizing. Vaccine 4:35-37 Neurath AR, Kent SBH, Strick N, Parker K (1986b) Identification and chemical synthesis of a host cell receptor binding site on hepatitis B virus. Cell 46:429-436 Neurath AR, Strick N, Kent SBH, Parker K, Seto B, Girard M (1988) Design of synthetic peptides mimicking the immunologic and biologic functions of the preS 1 sequence of the hepatitis B virus envelope protein. In: Ginsberg H, Brown F, Lerner RA, Chanak RM (eds) "Vaccines 88" Cold Spring Harbor Laboratory, Cold Spring Harbor, pp 229-234 Ovchinnikov YuA, Sverdlov ED, Tsarev SA, Khodkova EM, Monastyrskaya GS, Salomatina ES, Efimov VA, Chakhmakhchieva OG, Solov'ev VD, Kuznetsov VP, Zhdanov VM, Novokhatskii AS, Aspetov RD (1982) Direct expression of human leukocytic interferon F gene in Escherichia coli cells. Dokl Akad Nauk SSSR 265:238-242 (in Russian) Petit M-A, Dubanchet S, Capel F, Voet P, Dauguet C, Hauser P (1991) HepG2 cell binding activities of different hepatitis B virus isolates: inhibitory effect of anti-HBs and antipreSl(21-47). Virology 180:483-491 Pontisso P, Petit M-A, Bankowski MJ, Peeples ME (1989a) Human liver plasma membranes contain receptors for the hepatitis B virus pre-S1 region and, via polymerized human serum albumin, for the pre-S2 region. J Virol 63:1981-1988 Pontisso P, Ruvoletto MG, Gerlich WH, Heermann K-H, Bardini R, Alberti A (1989b) Identification of an attachment site for human liver plasma membranes on hepatitis B virus particles. Virology 173:522-530 Salfeld J, Pfaff E, Noah M, Schaller H (1989) Antigenic determinants and functional domains in core antigen and e antigen from hepatitis B virus. J Virol 63:798-808 Sallberg M, Ruden U, Wahren B, Noah M, Magnius LO (1991) Human and murine B-cells recognize the HBeAg/beta (or HBe2) epitope as a linear determinant. Mol Immunol 28:719-726 Stibbe W, Gerlich WH (1982) Variable protein composition of hepatitis B surface antigen from different donors. Virology 123:436-442 Stibbe W, Gerlich WH (1983) Structural relationships between minor and major proteins of hepatitis B surface antigen. J Virol 46:626-628 Sugiyama T, Hebert RR, Hartman KA (1967) Ribonucleoprotein complexes formed between bacteriophage MS2 RNA and MS2 protein in vitro. J Mol Biol 25:455-464 Thornton GB, Moriarty AM, Milich DR, Eichberg JW, Purcell RH, Gerin JL (1989) Protection of chimpanzees from hepatitis-B virus infection after immunization with synthetic peptides. Identificationof protective epitopes in the pre-S region. In: Brown F, Channock T, Ginsberg H, Lerner R (eds) Vaccines 89. Cold Spring Harbor Laboratory, Cold Spring Harbor, pp 467-471 Towbin H, Staehelin T, Gordon J (1979) Electrophoretic transfer of proteins from polyacrylamide gels to nitrocellulose sheets: procedure and some applications. Proc Natl Acad Sci USA 76:4350-4354 Wang SZ, Esen A (1985) Screening expression libraries with non-radioactive immunological probes. Gene 37:267-269 Zinder ND (ed) (1975) RNA phages. Cold Spring Harbor Laboratory, Cold Spring Harbor

Determination of the minimal length of preS1 epitope recognized by a monoclonal antibody which inhibits attachment of hepatitis B virus to hepatocytes.

The minimal amino acid sequence sufficient to be recognized efficiently by virus-attachment inhibiting murine monoclonal anti-preS1 antibody MA18/7 ha...
875KB Sizes 0 Downloads 0 Views