DK173597B1 - Process for the preparation of a membrane-free truncate from a membrane-bound polypeptide - Google Patents

Description

l DK 173597 B1

The invention relates to a method of producing a membrane-free truncate of a membrane-bound polypeptide, which truncate lacks membrane-binding domain, whereby the polypeptide truncate is free of this membrane, and which truncate has exposed antigenic determinants capable of inducing neutralizing antibodies. in vivo exposure to a pathogen, characterized in that DNA encoding the truncate is expressed in a stable eukaryotic cell line that is transiently infected with the DNA.

The analysis of the * immune response to a variety of infectious agents has been limited by the fact that it has often been found difficult to grow pathogens in sufficient quantities to enable the isolation of important cell surface antigens. The advent of molecular cloning has overcome some of these limitations by providing an agent by which gene products from pathogenic agents can be expressed in virtually unlimited amounts in a non-pathogenic form. Surface antigens 20 from such viruses as influenza (1), foot-and-mouth disease (2), hepatitis (3), vesicular stomatitis (4), rabies (5), and herpes simplex (6) virus have now been expressed in 'E. coli and S. cerevisiae, and promises to provide improved subunit vaccine in the future. "

However, it is clear that the expression of surface antigens in lower organisms is not entirely satisfactory because potentially important antigenic determinants could be lost due to incomplete processing (e.g., proteolysis, glycosylation) or by denaturation 5Q during purification. the cloned gene product.

Oette is especially the case for membrane proteins which, due to hydrophobic transmembrane domains, tend to aggregate and become insoluble when expressed in E. coli. Cloned genes encoding membrane proteins can be expressed in mammalian cells, where the host cell provides the factors necessary for proper processing, polypeptide folding and incorporation into the cell membrane (7, 8). In these studies, membrane proteins can be expressed on the surface of a recombinant host cell, and, e.g. (8) * ~ that a truncated membrane protein lacking the hydrophobic carboxy terminal domain can be slowly secreted from the host cell 5. rather than being bound to it, it is not clear that the thus-expressed membrane-bound protein or ... the thus-truncated truncated protein will in fact be capable of inducing antibodies which are 'effective against. from which the protein is derived.

10 Herpes simpléx (HSV) is an ONA virus that occurs in two related, but distinct, forms of human infections. At least four of the large number of virus-encoded proteins have been found to be glycosylated and present on the surface of both the virion and the infected cells 15 (9). These glycoproteins, designated gA / B, gC, gD and gE, are found in both HSV type 1 (HSV1) and HSV type 2 (HSV2), while in the case of HSV2 an additional glycoprotein (gF) has been reported. Although their functions remain somewhat of a mystery, these glycoproteins are undoubtedly involved in viral association with cells, cell fusion, and a host of different host immunological responses to the viral infection (11). Although HSV1 and HSV2 only show approx. 50% DNA sequence homology (12), glycoproteins for the majority appear to be type-common. Thus, gA / B, gD and gE display a large number of type-common antigenic determinants (13 - 16), while gC, previously believed to be completely type-specific (17, 18), has also been found to type-common determinants.

However, type-specific antigenic determinants can be detected using monoclonal antibodies for some of the glycoproteins (10, 19), showing that some amino acid changes have occurred since HSV1 and HSV2 were separated.

35 One of the most important glycoproteins in terms of virus neutralization is gD (11). Considerable material has been provided which strongly suggests that the respective gD proteins of HSV1 and HSV2 are related. Fweks. have recombination mapping · localized the respective genes * to the colinear regions of both viral genomes. Amino acid analysis showed great homologue! between the two proteins. The gD proteins induce neutralizing antibodies for both type 1 and type 1i virus in a type-common manner. (19, -21). In addition, most monoclonal antibodies to these glycoproteins are type-shared, also suggesting a high degree of structural kinship between the two types of glycoproteins (20). However, some monoclonal antibodies were found to respond type-specifically, suggesting significant differences between the proteins (19). Peptide maps of the proteins also confidently revealed such differences (22a). However, although these results suggest that these polypeptides are related, they are insufficient to indicate exactly how close the kinship is.

To investigate the nature of the type community of HSV1 and HSV2 gD proteins, the DNA sequences of the gO genes 20 from HSV1 and HSV2 were determined. The deduced amino acid sequences showed similarity. The resulting derived protein sequences were also analyzed for structural differences using a program designed to determine hydrophobic and hydrophilic regions of the protein. This analysis showed a high degree of conservation at a rough structure level ·

Although several amino acid substitutions existed between the two glycoproteins, the majority of these replacements were conservative, suggesting an important structural requirement in this glycoprotein for the virus.

30 Unlike HSV1, HSV2 appears to encode yet another glycoprotein, designated gF (22b, 10, 22c, 22d). Although HSV2-gF had an electrophoretic mobility that was much faster than HSV1-gC, recombinant virus mapping studies revealed that this protein was encoded in a region of the approximately colinear HSV2 genome. with the gene for HSV1-gC '(22d, 22d). Furthermore, it has recently been demonstrated that a monoclonal antibody to HSV2-gF will. cross-react "weakly with 5 HSV1-gC (22f) and that a polyclonal antiserum Prepare lightly for HSV1-verion hyster proteins. gF (22d), suggesting a possible structural homogeneity between the two glycoproteins. Thus occurred that a possible homologue to HSV1-gC was the HSV / 2-gF protein.

This kinship was investigated in accordance with the present invention.

BIO / TECHNOLOGY, April 1983, pages 146-147, describes initial methods for preparing vaccine candidates which will require further processing, testing and modification before they can become vaccines. The article on speeches works by Gething, Sveda and Rose. on page 146, first column, about Gething's method of facilitating the preparation of enzymes, receptors and viral antigens for vaccines.

20 About Sveda's protein is said on page 147, first column, that the secreted form was not structurally identical to the membrane-bound form, perhaps because some of the polypeptide needed for proper processing had been deleted.

25 Rose is stated further down the same column to have emphasized that Gething's merit lay in the fact that she had obtained extremely high yields of the protein; there is no mention of obtaining a secreted product which, in immunogenic terms, corresponds to the membrane bound form.

30 Furthermore, Gething's work is said a few lines further down that perhaps the simplest application of this technology is the production of viral antigens for subunit vaccines; and some lines are continued later that anchorless mutants of such proteins could be an excellent source of vaccine antigen. A technical problem with these subunits is that they are monoclonal and therefore less immunogenic than the aggregates constituted by isolated molecules containing the membrane anchor peptide. Dennis Kleid emphasizes that the ease of isolation of antigen (the trait highlighted by Gething and others) is of secondary importance to whether the protein is immunogenic. Dr. Gething is quoted as having replied that this would be accomplished by cross-linking them or perhaps 10 by other methods (which are not specified) and that polyvalent complexes are clearly needed to achieve immunogenicity but are not difficult to and that simple purification from the medium is more important than a potential disadvantage of making the proteins polyvalent *

All of the above points away from what is described in connection with the present invention, as one skilled in the art who reads the article alone will see that this could be a good way to facilitate the purification of normally membrane-bound proteins. . The resulting products might constitute source material for various types of useful end products. Among such possible end products are vaccines, but it would be unlikely that the truncated proteins would be immunogenic per se. Rather, they would require additional processing such as cross-linking to make them candidates for a vaccine.

Thus, as can be seen from the above, on the basis of the article from BIO / TECHNOLOGY, it was surprising that only in the method of the invention are membrane-free truncates capable of inducing neutralizing antibodies by in vivo exposure to a pathogen.

35 In examining the relationship between HSV1 and HSV2, in this specification it has been determined that a DNA sequence of a 2.29 kb region in the HSV2 genome is colinear with the HSV1-gC gene. Translation of a large open reading frame in this region shows that a protein that has substantial homology to HSV1-gC is encoded in this region.

It is believed that this region encodes the HSV2 gF gene and that the gF protein is the HSV2 homologous to HSV1 gC.

In light of this information on the structure of the gD protein, as described in more detail below, it was decided to express gD pro-protein DNA in mammalian cells to see if this was possible and, if possible, whether the expressed protein would bind to the host cell membrane, and whether a truncated form of the protein lacking the membrane binding domain would be secreted from the host cell and, in each of the latter cases, whether the expression product proteins could induce antibodies effective against HSV1 and / or HSV2 . As the results set forth below show, these objectives have been achieved. In particular, the invention makes it possible to use these proteins obtained by recombinant DNA processes as components of a vaccine effective against HSV type 1 and HSV type 2. Thus, a protective vaccine against the presence of 2Π is described. herpes infection and a reduction in the incidence and frequency of recurrent herpes infection in individuals already infected.

Also disclosed is another class 25 of glycoproteins obtained by recombinant DNA processes useful as components of a vaccine against HSV type 1 and / or HSV type 2. More specifically, this class of glycoproteins includes HSV1-gC (effective against HSV1), HSV2-gF (more correctly termed an HSV2-gC, effective against HSV2) or combinations of the two proteins (effective against both viruses). Other such glycoproteins include gA, g8 and gE. It is believed that a vaccine based on the combined glycoproteins, gC and gD, would be significantly more effective than a vaccine based on each glycoprotein alone.

7 DK 173597 B1

An embodiment of the invention relates to a method for producing a polypeptide with antigenic determinants capable of specifically eliciting complementary antibody to Herpes simplex type 1 and 5 Herpes simplex type 2 virus, but which is not functionally associated with the surface membrane. As explained in more detail below, such a polypeptide is a truncated *, membrane-free derivative of a membrane-bound polypeptide. The derivative is formed by omitting a membrane binding. domain, from the polypeptide, allowing it to be secreted from the recombinant host cell system # in which it has been produced. In another embodiment, the polypeptide forms a functional compound with. a surface membrane, 6g is then dissolved in the polypeptide, preferably in a nonionic surfactant, to release it from the membrane.

In this specification, the term "recombinant" refers to cells that have been transfected with vectors 20 constructed using recombinant DNA technology and thus transformed with the ability to produce the polypeptide of this invention. "Functional compound" means bound to the membrane, typically by protruding to both sides of the membrane, in such a way that antigenic determinants lie merely folded into a native conformation which can be recognized by antibody triggered against The native pathogen. "Membrane-bound" with reference to the polypeptides of the present invention refers to a class of polypeptides commonly produced in eukaryotic cells and characterized by having a signal sequence believed to aid in their secretion. through various cell membranes as well as a membrane-binding domain (usually of a hydrophobic nature and occurring at the C-terminal end), which are thought to be extinct.

35 close their complete secretion through the cell membrane. Thus, the polypeptide remains functionally associated with or bound to the membrane. This invention is particularly directed to the utilization of the membrane-bound polypeptides associated with pathogenic organisms, e.g. herpes virus.

In this specification, the terms "HSV2-gF", "HSV2-gCM and HgC-2" are optionally used to designate a glycoprotin moiety of HSV2 ', soro is highly homologous to HSV1-gC and capable of to. to induce sufficient anti-5 drugs to be useful as a vaccine.

When the antigenic determinants of the polypeptides prepared by the method of the invention are provided by functional connection to the surface membrane, the membrane can then be removed from the polypeptides without destroying the antigenic properties. Thus, the membrane-bound polypeptide, e.g. is removed from the membrane by solubilization with a suitable solution, preferably one containing a nonionic surfactant, to remove the polypeptide from the membrane. An advantage of doing this is that the polypeptide is isolated from foreign cell material, which increases the potential potency of its use in a vaccine. A technique for removing the membrane from the polypeptide is described below.

20 Membrane-free preparations can also be obtained by creating a secretory system. As described in more detail below, such a secreted polypeptide has at least some of the antigenic centers necessary for antibody stimulation.

2 5 'Brief explanation of the drawings

FIG. 1A and 1B show the DNA sequence and the deduced amino acid sequence of the HSV1 and HSV2 gD genes and surrounding * flanking regions.

FIG. Figure 2 shows a hydropathy analysis of the gD proteins of the 50 HSV1 and HSV2 proteins.

FIG. 3 is a diagram of the plasmid p'gO-dhfr, constructed to express a membrane-bound form of HSV1 glycoprotein D.

9 DK 173597 B1

FIG. 4 shows the result of labeling gDl2 cells with human antibodies to HSV, with (A) a phase contrast optics visualization (B) a fluorescence image of the same cells * 5. 5. shows radioimmunoprecipitations of cloned gQ from the present gD12 cell line and native gD from HSV1-infected human cells.

FIG. Figure 6 shows the binding of human anti-HSV antibodies to gD12 cells and the starting CHO cell line.

FIG. Figure 7 is a schematic representation of HSV1-gO protein and illustrates the locations of signal sequence and membrane binding domain.

FIG. 8 is a diagram of the expression plasmid pgDtrunc-dhfr for a secreted form of HSV1-gO protein.

FIG. 9 shows radioimmunoprecipitates from the gD10.2 cell line of the invention.

FIG. Figure 10 shows radio-immuno-precipitates from amplified and amplified gD10.2 cell lines.

FIG. Figure 11 shows the degree of enhancement achieved with the Mtx enhanced gD10.2 cell line.

FIG. 12 shows the fragments of pgL'2Sal2.9 which were subjected to DNA sequence analysis.

FIG. Figure 13 shows the DNA sequence derived from pgC2SA12.9 compared to the ONA sequence of the HSV1-gC region.

FIG. 14 illustrates Southern immersion analysis of HSV2 genomic DNA and pgC1, Sal2.9 DNA.

DK 173597 Pg 10

FIG. 15 illustrates the translation of the large open HSV2 reading frame and comparison with the HSV1-gC amino acid sequence.

FIG. 16 illustrates hydropathy analysis of the HSV1-gC protein and protein from the large open HSV2 reading frame.

The invention is further illustrated by the following examples.

EXAMPLE 1

This example deals with gD protein.

10 Virus growth and isolation of viral DNA.

HSV1 (strain Hzt) and HSV2 (strain G) were grown on Hep 2 cells at 37 ° C and 33 ° C, respectively. The viral DNA was isolated from infected cell cultures by digestion with proteinase K and forming bands with CsCl 15 (23).

Cloning of the gD genes from HSV1 and HSV2.

Prior mapping and cloning studies have located the HSV1 gD gene to a ca. 6.6 kb Bam HI fragment (6.24). HSV1 DNA was digested with BamHI and the 6-20 7 kb region was isolated by agarose gel electrophoresis.

This fragment was ligated into Bam HI-degraded pBR322 and the resulting mixture was used to transform E. coli strain 294 (ATCC No. 31446). The plasmids from the ampicillin-resistant tetracycline-sensitive clones were screened for the correct HSV1 fragment by restriction enzyme degradation. The correct gD-containing SstI fragment was subcloned into Sstl-degraded plasmid pFM3 (EP Publication No. 0060693).

• 11 DK 173597 B1

Although the H5V2 gD gene was previously mapped by recombination with HSV1, the "exact location of this gene was unknown. Therefore, an approximately 10 kb HindIII fragment from the small unique region of .HSV2- the genome.

5 (4) ligated into the HindHI center of the bacteriophage - ^ - cloning vector 590 (25). In vitro packed, phage were plated at low density and screened by the Benton-Davis procedure with a β-labeled subclone of the HSV1 gD gene (26). Positively.

hybridizing spots were grown, the DNA isolated, 10 and the gD gene localized by Southern blotting and hybridization with the P-labeled HSV1 gD gene (27). The positively hybridizing HSV2-gD-containing fragments were subcloned into plasmid pUC9 (28).

DNA sequencing and data analysis.

15 Various fragments from the HSV1 and HSV2 gD genes were subcloned into ml3 phage vector mp9 (29) and sequenced by the dideoxynucleotide method described by Sanger (30).

The nucleotide sequences were analyzed using the HOM program (31). The hydropathy of the derived protein sequence was analyzed using a width of 12 and a jump of 1 (31a);

Cloning of the gD regions from HSV1 and HSV2.

Other studies had located the HSV1 gD gene 25 to 6.6 kb Bam HI-J fragment according to the nomenclature given by Roizman (6, 12, 24). Isolation and sequencing of a portion of this fragment showed that this fragment contained the HSV1 gD gene. As one would expect that the DNA sequences of the HSV1 gD gene would be relatively homologous to the HSV2 gD gene, this fragment was used as a probe to isolate the gD gene from the HSV2 genome.

DK 173597 B1 1Å

Since most of the genes from the HSV1 and HSV2 genomes * · appear to be co-linear on the map (35), the region from the small unique region of the H5V / 2 genome corresponding to the HSV1 gD region ( HindIII-L fragments.t (12)), cloned into a ^ V * phage vector. Screening of the resulting spots with a P-labeled HSV1-gD gene subclone revealed positive hybridizing spots, suggesting that there really was nucleic acid sequence homology between the two viral genomes in this region. Isolation of the phage D'NA 10 and subsequent Southern blotting analysis revealed the region of this fragment corresponding to the gD gene.

This region was subcloned for DNA sequence analysis.

Coding regions.

FIG. 1 illustrates the two gD DNA sequences compared with the HOM program (31). Nucleotide # 1 is selected as the A in the ATG start codon encoding methionine. Openings have been introduced by the HOM computer program to maximize the sequence homologies (31). Nucleotide differences are shown by the symbol (*), while amino acid differences are shown by framing. Amino acid differences between the HSV1 gD sequence reported here, determined for the Hzt strain of H5V1, and that reported by Watson et al.

(6) for the Patton tribe, is indicated by the symbol (+).

The onset of the HSV1 gD gene transcription, shown by a 25 arrow, is from Watson et al. (32). Possible N-linked glycosylation centers are shown in grayscale. Two possible MTATA "sequences are shown 5 'for the start of the gO transcription, while a third possible" TATA "sequence is shown 5' for another open reading frame at the 3 'end of the HSV2 sequence.

Two areas of non-coding sequence homology should be noted 5 'for the gD genes and 5' for the second open reading frame from the HSV2 sequence.

13 DK 173597 B1

The hydropathy of gD proteins.

The hydropathy of each glycoprotein was analyzed using the program developed by Hopp. Et al. (51a). As shown in FIG. 2, a hydrophobic .5 transmembrane domain is obtained. 3 'end of. gene. Twelve amino acids, long stretches were analyzed and the average hydropathy calculated. Differences in residues between the two glycoproteins are shown with conserved changes marked (*) and non-conserved changes marked (+). A) HSV1-gD-protein hydropathy, 8) HSV2-gD-protein hydropathy.

DNA sequence analysis shows that the HSV1 and HSV2 gD proteins are 80% homologous. Most of the differences found between these two proteins were in the amino and carboxyl-terminal regions. The amino-terminal region of these proteins contains a highly hydrophobic region which contains an arginine residue close to the amino-terminal methionine. This hydrophobic domain is the signal sequence that is characteristic of secreted and membrane-bound proteins and which presumably has the function of directing at least a portion of the protein into the space of the endoplasmic reticulum (33). A comparison of the first 20 amino-terminal amino acids showed that there were a total of 12 differences between the type 1 and type 2 gene. Virtually all differences, however, are conservative as they encode other hydrophobic amino acids. The exceptions are the gly-arg replacement at residue # 3 and the arg-gly replacement at residue # 7. Although these replacements are not conservative, they do not, however, alter the network structure of the signaling network. Both genes retain a positively charged residue in the first 10 amino acids.

The hydropathy curve of FIG. 2 revealed a hydrophilic carboxyl terminal domain after a hydrophobic region. This structure is characteristic of membrane-bound glycoproteins and has previously been found in other viral surface antigens (5, 34). Its function is to anchor the protein in the cellular and viral membranes and as such it plays an important role in viral infection. 5 12 amino acid changes were found in this region of the gD proteins from residue 333 to residue 362, most of which are conservative.

This suggests that the only criterion for the amino acids in this region is that they are predominantly apolar to span the lipid bilayer. In addition, the region 10 shows the membrane domain (residues 363 - 375), which mainly serves to anchor the protein in the membrane (33), to 5 changes in its first 13 amino acid residues followed by a homologous stretch. This result suggests that the first 10 to 15 residues in the carboxyl-terminal hydrophilic domain have only one anchoring function and therefore need only be charged, while the subsequent 23 residues may have another function which is specifically important for gD protein.

Although many other amino acid changes exist through these two proteins, the vast majority of the changes are conservative. This is emphasized by the structure revealed by the one shown in FIG. 2 hydropathy program. As can be seen in this comparison, the two glyco proteins show very similar curves. The amino acid changes which are not conservative do not appear to alter the hydropathy of the protein. -

Expression of HSV1-gD.

To establish a permanent cell line producing membrane-bound gD, the gD-containing fragment was ligated 30 (Fig. 3) into a mammalian expression vector (36) containing the selectable marker, dihydrofolate reductase (dhfr). FIG. Figure 3 shows a diagram of the plasmid pgO-dhfr constructed to express HSV1 glycoprotein D.

15 DK 173597 B1

The expression plasmid consisted of the origin of replication and the β-lactamase gene (smpr) derived from the E. coli plasmid pBR322 (37), a cDNA insert encoding mouse dhfr (36, 38) Under the control of the early SV40 promoter 5 and a 4.6 kb HindIII-BamHI fragment containing the gD gene also under the control of the early SV40 promoter. The HindIII end of this fragment is 74 bp. 5 * side of the starting methionine codon and includes the mRNA cap center. The HindIII center is located 250 bp to the 3 'side of the Goldberg-10 Hogness box in the SV40 promoter. The coding region of the gD coding fragment is 1179 bp long and joins a large (1.9 kb) 3 'region containing at least a portion of the gE gene (24, 32), a translation stop codon and a polyadenylation center .

The plasmid pgD-dhfr was constructed as follows: The 4.6 kb HindIII-BamHI fragment containing the entire gO code sequence was isolated from the BamHI fragment cloned from the HSV1 genome (see above). The 2.8 kb HindilI-SalI fragment containing an SV40 origin / early promoter and the pBR322 ampicillin resistance gene and origin of the ENA replication was isolated from the plasmid pEHBal 14. 2.1 kb SalI-BamHI the fragment containing a mouse dihydro-folate reductase cDNA clone under the control of another SV40 origin / early promoter was isolated from plasmid 25 pE348HBV E400D22 (36). These 3 fragments were ligated together in triplicate ligation using T4 DNA ligase and the resulting mixture was used to transform E. coli strain 294. The resulting colonies were grown and plasmid DNA1 screened by The correct DNA construct, pgD-dhfr (Fig. 3), was used for further transfection studies.

The plasmid was introduced into Chinese hamster ovary (CHO) cells deficient in production of dhfr (39), using the calcium phosphate precipitation method (40). Colonies capable of growing in media lacking hypoxanthine, glycine and thyroidin were obtained and 9 dhfr + clones were assayed. Of these, gD could be detected in 5 colonies using anti-HSV1 antibodies in radioimmunoprecipitation and indirect immunofluorescence assays. One of the five lines (gD12) was designated for further study. To characterize the cloned gD gene product, gD12 cells were metabolically labeled with '55S-methbine or 3 H-glucosamine and analyzed by radioimmunoprecipitation. The procedure used was as follows: cells were grown in Ham's F12 medium (Gibco) supplemented with 7% commercially dialyzed fetal bovine serum (Gibco), penicillin (100 units / ml) and streptomycin (100 units / ml).

When the cultures were about 80% confluent, the medium was removed, the cells were washed twice with phosphate-buffered saline (PBS), and a labeling medium was added (Dulbecco's modified Eagle's medium containing either 1/10 of the normal concentration of methionine 2 or glucose ) to a final concentration of 0.064 ml / cm.

Either S-me thionine (SJ.204, Amersham Int.) (50 - 75 µ, µCi / ml) or 3 H-glucosamine (100 µCi / ml) was added and the cells were grown for an additional 18-20 hours.

After labeling, the medium was harvested and the cells were washed twice in PBS and removed from the culture dishes by treatment with PBS containing 0.02% EDTA. The cells were then solubilized in lysis buffer consisting of: PBS, 3% NP-40, 0.1% bovine serum albumin, 5 x 10 4 M phenyl-30-methylsulfonyl fluoride and 0.017 TIU / ml apoprotinin, and the resulting lysate was cleared by centrifugation at 12,000. x. For immunoprecipitation reactions, the cell lysates were diluted triple with PBS, and aliquots (typically 180 µl) were mixed with 2-5 µl of antisera 35 and incubated at 4 ° C for 30 minutes. Immune complexes were then adsorbed to fixed S. aureus cells by Kessler's method (40a) and re-precipitated by centrifugation at 12000 x G for 30 seconds. The S. aureus cells were then washed 3 times with wash buffer (PBS, 1% NP-40, 0.3 Sf sodium dodecyl sulfate), and the immune complexes were eluted with 20 µl polyacrylamide gel samples (62.5 mM). Tris "HCl buffer, pH 6.8 containing 10% glycerol, 5% 2-mercaptoethanol, 0.01% bromophenol blue) at 90 ° C for 3 minutes. After centrifugation for 30 s, the supernatants were applied to 10% polyacrylamide gel plates according to Laemmli's method (45).

FIG. 4a compares autoradiographs obtained with the g012 cell line and HSV1-infected cells: control immunoprecipitation from the gD12 cell lysate with normal rabbit serum (lane 1) immun immunoprecipitation of native gO grown in HEL cells (lane 15 2) and A549 cells (lane 3 ) with the monoclonal anti-gD antibody, 55-S (41); immunoprecipitation of cloned gO from the gD12 cell lysate with polyclonal rabbit antibodies (Dako *

Corp.) for HSV1 (lane 4) and the monoclonal antibody, 55-S (lane 5); immunoprecipitation of the cloned gD from the gD12 cells metabolically labeled with 1 H -glucosamine with polyclonal rabbit anti-HSV1 antibodies (lane 6).

It is seen (lanes 4 and 5) that a 59-60 kilodalton diffuse band was specifically precipitated from the gD12 cell line using either rabbit anti-HSV1 antibody or the 55-S monoclonal antibody, which is specific for the HSV1 protein (41). This molecular weight is in good agreement with that reported for gD isolated from HSV1-infected KB cells (42). It is seen that the same monoclonal antibody precipitated proteins of similar but different molecular weights from HSV1-infected human cell lines. The main product that was precipitated from the human lung carcinoma cell line A549 (lane 2) was 53 kilodaltons, and that precipitated from the human embryonic lung cell line (HEL) was 56 kilodaltons (lane 3). Previous studies (43) have shown that the molecular weight of HSV glycoproteins varies depending on the host cell and that this difference is due to differences in glycosylation. To determine if that.

5 in CHO cells produced gD. protein was actually glycosylated, the cells were metabolically labeled with 3 H-glucosamine.

When the bands of identical molecular weights (lanes 5 and 6) were decomposed after metabolic labeling with ³²S-methionine or ³H-glucosamine, we concluded that the gD protein produced in 10 CHO cells is glycosylated.

The human cell lines A549 (ATCC CCL 185) and HEL 299 (ATCC CCL 137) were grown to confluence in 3.5 cm life culture dishes and infected with HSV1 at a multiplicity of 10 pfu per ml. cell. Virus-infected cells were labeled 15 by a method similar to that described by Cohen et al. (44). 4 Hours after infection, the medium was feathered - in mesh and cells were washed once with fresh, medium * (Dulbecco's modified Eagle's medium) and once with phosphate buffered saline (PBS). Then, 20 fresh medium containing 1/10 of the normal concentration of methionine to the cells was added with 3 S-methionine (Amersham, International) to a final concentration of 75 µCi per ml. ml of medium. The cells were cultured for an additional 20 hours and then harvested by treating the 25 washed cells with PBS containing EDTA (0.02%). Viral proteins were solubilized in lysis buffer consisting of PBS, 3% NP-40, 1% bovine serum albumin, 5 x 10 8 M phenylmethylsulfonyl fluoride and 0.017 TIU / ml apoprotinin.

The resulting lysate was clarified by centrifugation at 12000 x G in a microcentrifuge. For immunoprecipitation reactions, the cells or virus lysates were diluted triple with phosphate buffered saline, mixed with 2-5 µl of the appropriate antiserum and incubated for 30 minutes at 4 ° C. Antibody-antigen complexes 35 were removed from the reaction medium by addition of 25 µl of a 10% solution of fixed S. aureus (Kessler (40a)) and precipitated by centrifugation at 12000 x G for 30 s. The aureus cells were then washed 3 times with wash buffer (PBS, 1% HP-40, 0.3% sodium dodecyl sulfate), and the cells were suspended in 20 ml of polyacrylamide gel sample buffer (10% glycerol, 5% 2). mercaptoethanoyl 0.0625 M with respect to pH 6.8 "Tris" buffer, 0.01% bromophenol blue) and incubated at 90 ° C for 3 minutes.

After centrifugation (12000 x G) for 30 s, the supernatants were applied to 10 S polyacrylamide gel plates (45).

To further explore the post-translational whole processing of the cloned gD, impulse-pursuit studies were performed. FIG. Figure 4B shows immunoprecipitation of cloned gD from gD12 cells with rabbit anti-HSV1 antibodies (Dako, 15 Corp.) at various times after pulse labeling with 35 S-methionine. FIG. Figure 4B shows an impulse labeling of g012 cells. In these studies, cells were grown to confluence in 10 cm culture dishes and labeled with 3 S methionine as described above except that the msrkning reaction was carried out for 15 minutes on ice, the cells washed 3 times with fresh medium and then returned. The reactions were terminated by washing the cells in cold phosphate buffered saline and solubilizing the cells as described above Proteins were immunoprecipitated at the following times after pulse labeling: lane 1, 5 min 2, 15 minutes; lane 3, 30 minutes; lane 4, 60 minutes; lane 5, 120 minutes Precursor form of gD having a molecular weight of 51 kd was specifically precipitated from the gD12 cell line 5 minutes after a pulse of methionine, and this precursor was followed into the higher molecular weight form (59 kd) after about 60 minutes. From these studies we estimate the half-life of this post-trans latio no given time to be about 45 minutes. The precursor link between the 51 kd band and the 59 kd band closely resembles that reported for virus-produced gO (14, 42, 46, 47), and the kinetics of this process is similar to that described. by Cohen et al.

5 (42). In virus-infected cells, the difference in molecular weight between the precursor and the product has been attributed to both N-linked and O-linked oligosaccharides (48).

To determine if gD was exported to the cell surface, indirect immunofluorescence studies were performed.

10 During these studies, rabbit, mouse, and human anti-HSU antibody was reacted with unfixed cells under conditions that do not permeabilize the cell membrane (49). g012 cells and the starting (CHO) cells (ratio 1: 1) were plated on glass cover plates (2.2 x 2.2 cm) and cultured until the cells were about 60% confluent.

Human serum, known to contain antibodies to HSV1 (50), was diluted 40 times with phosphate buffered saline (PBS) and 100 µl pipetted onto washed cells and the cells incubated for 30 min at room temperature in a humidified chamber. The cells were immersed 3 times in PBS to wash away unbound antibody and then incubated with 100 µl 20 times diluted tetramethylrhodamine isothiocyanate-labeled goat anti-human IgG (Cappel Laboratories) antibodies for an additional 30 minutes. The unbound labeled antibody was washed away with PBS and the cells were dehydrated in ice-cold 50% ethanol and 100% ethanol and rehydrated with glycerol on a microscope slide (49). The cells were then considered under Phase Contrast and Fluorescence optics in 30 fluorescence microscope (Zeiss). FIG. Figure 5 shows: A, g012 and CHO cells visualized with Phase control topics; B, fluorescence image of the same cells as in A. Comparison of the phase contrast images with the fluorescence images (Fig. 5) showed that the gD12 cells were strongly labeled, whereas the starting CHO cells only gained little or no labeled antibody. In control experiments with normal mouse sera, normal rabbit sera or human sera known to be negative for HSV antibodies, no specific labeling of the cells could be detected. These 5 studies indicated that the gD protein was exported right to the cell surface. Experiments with CHO and gD12 cells fixed prior to labeling with agents known to permeabilize the cell membrane (methanol or acetone) gave different labeling patterns. In these studies 10, strong perinuclear labeling of the gD12 cells was observed with anti-HSV1 antibodies and no specific labeling of the CHO cells.

To determine whether gD12 cells expressed antigenic determinants relevant to human HSV1 and HSV2 infections, the binding of antibodies from individuals known to have anti-HSV1 or anti-HSV2 antibodies was examined (50). Radioimmunoprecipitation of lysates from metabolically labeled gD12 cells gave results comparable to those obtained with rodent anti-HSV sera 20 (Fig. 4). Similarly, human anti-HSV1 sera gave specific labeling of gD! 2 cells by an indirect immunofluorescence assay (Fig. 5) and did not label the starting CHO cell line. Taken together, the results obtained with various rodent anti-HSV1 and HSV2 antisera, 25 monoclonal anti-gD antibodies and human anti-HSV anti-sera = evidence that gD expressed on the surface of gD12 cells have a number of antigenic determinants in common with the native virus and that the structure of these determinants does not depend on the interaction with other HSV1 proteins.

The fact that one of the monoclonal antibodies (1-S) tested is known to neutralize HSV1 in vitro (41) and shows that the gO produced in CHO cells has at least one of the neutralizing antigenic determinants in common with the native virus.

22 DK 173597 B1

To obtain a quantitative target for the binding of anti-HSV1 antibodies to g012 cells, an enzyme-linked immunosorption assay (ELISA) (52) was developed. In these studies, gD12 cells and CHO cells were plated and chemically fixed to alternate holes in 96 holes of microtiter tissue culture plates. Various antisera, which are known to contain antibodies to HSV, were then diluted and allowed to react with the fixed cells.

At the end of the test, the absorbance in every 10 holes was measured and normal bonding curves were constructed. *

The specific binding of antibodies to the gD12 cells was determined by subtracting the values obtained with the starting CHO cells from those obtained with the gD12 cells. Specific binding of high-titer sera could be detected in dilutions of 1: 10000.

Serum titers determined using the gD12 cell ELISA assay were compared to anti-HSV1 and anti-HSV2 titers determined by conventional methods. Human sera which were already titrated (50) to HSV by conventional tests, i.e. inhibition of hemagglutination (IHA) or complement fixation (CF), were diluted into holes on microtiter plates containing either gD12 cells or the starting CHO cell line, and the binding of anti-gD antibodies was checked by an ELISA assay. . The gD12 cells and the starting CHO cells were seeded in alternating holes on 96-hole microtiter tissue culture plates (Falcon Labware) and grown to confluence in F12 medium (GIBCO) containing 10% fetal bovine serum.

The cells were washed 3 times with phosphate buffered saline (PBS) and then chemically fixed with 0.0625% glutaraldehyde in PBS. The cells were again washed 3 times with PBS and stored until use at A ° C in PBS containing 1% bovine serum albumin, 100 mM glycine, 1 mM NaN 2. To measure anti-gD antibody titers, the cells were washed with PBS and series thin antisera were allowed to react with the fixed cells (50 µl final volume) for 1 hour at room temperature. Unbound antibody was washed away and cells were incubated with 50 µl 1: 2000 diluted goat anti-human IgG coupled to horseradish peroxidase (Tago »Inc.). The enzyme-linked antibody was allowed to react for 1 hour at room temperature and the cells were then washed 3 times with PBS. After incubation, the peroxidase substrate, o-phenylenediamine (200 µl) was added and the reaction allowed to proceed for 10 minutes.

The reaction was terminated by the addition of 2.5 M 2 SO 2 (50 µl) and the absorbance of the reaction medium from each hole was determined with an automated plate reading spectrophotometer (Titertek). In FIG. 6, the serum, represented by the open and closed circles, exhibited an HSV1-15 CF titer of 128 and HSV1 and HSV2 IHA titers of 4096.

The serum, represented by open and closed squares, showed an HSV1 CF titer of <8 and HSV1 and HSV2 IHA titers of <8. A, closed circle and closed square indicate binding to gD12 cells, open circle and open square 20 indicate binding to CHO cells. B, closed circle and closed square represent the specific binding to gD12 cells calculated by subtracting the values in A. In FIG. 6, it can be seen that a serum with a high anti-HSV titre determined by conventional tests yielded a high ELISA titre, while another serum with low anti-HSV titers gave no detectable binding by gD12-ELISA.

The studies described show that stable cell lines constitutively express on their surface a transfected gene product which associates with antibodies produced by herpes virus infection.

Immunization of mice with gD12 cells.

20 BALB / c female mice (5 weeks old) were obtained from Simonsen 24 DK 173597 B1

Laboratories (Gilroy, California). The mice were divided into two groups of 10 mice each: a "test" group and a "control" group. Each mouse in the test group was injected with gD12 cells known to express HSV1-5 glycoprotein D on their surface. mice in the control group were injected with the starting CHO cell line from which the gD12 cell line was derived. For immunization of mice, both types of cells were grown to confluence in 15 cm culture dishes. The CHO cells were grown in 10 Hams F12 medium (GIBCO) supplemented with 7% commercially dialyzed fetal bovine serum (GIBCO), penicillin (100 units / ml) and streptomycin (100 units / ml), the gD12 cells were grown in the same medium without glycine, hypoxanthine and thymidine. each dish 15 was washed twice with 15 ml of phosphate buffered saline (PBS) and then treated with 15 ml of PBS containing 0.02% EDTA. After 15-20 minutes, the cells were removed from the dish and pelleted by centrifugation for 5 minutes at full speed. a clinical centrifuge (IEC module 20 CL clinical centrifuge, rotor model 221). The supernatant was discarded and the cells were resuspended in PBS to a final concentration of 1 ml PBS per ml. every 15 cm bowl of cells. Each mouse was then injected with 0.5 ml of cell suspension (approximately 5 x 10 ° cells) distributed as follows: 0.25 ml injected intraperitoneally and 0.25 ml injected subcutaneously into the loose skin of the neck. The mice were then amplified twice with fresh cells (prepared as described above) on the 38th day and the 55th day after the original immunization. Mice 30 were bleed via the tail vein on day 68 to obtain sera for in vitro neutralization studies. The mice were exposed to H5V1 (MacIntyre strain) on day 70. The virus exposure occurred by an intraperitoneal injection of 2 x 10 7 pfu virus in each mouse. The mice were rated 35 daily for mortality and every other day for guard change and paralysis onset. All the mice in the control group died within 7 days of the virus exposure, while all the test mice were protected and showed no evidence of infection. These studies lead to the conclusion that immunization with the gD12 cells protects against a lethal HSV1 exposure.

Of course, a variety of transfection schemes is possible using a variety of selectable markers. Eg. For example, mouse L cells can be usefully transfected using an innate dhfr gene as a selectable marker. The gD gene was transfected into such cells via a vector containing such a marker. In principle, the strategy described could be applied to any situation where expression of a membrane protein is desired.

Expression of a truncated form of the gD gene.

The foregoing description relates to the preparation of membrane bound gD protein. As discussed above in connection with FIG. However, in analysis of the amino acid sequences of the gD protein in HSV1 and HSV2, in each 20 cases, a hydrophobic / hydrophilic carboxyl-terminal membrane-binding domain was identified (Figure 7).

A schematic diagram of HSV1 glycoprotein D (gD).

Hydrophobic (grayscale) and hydrophilic (labeled +) regions of the protein were determined from the hydropathy analysis of the 25 gD protein sequence derived from the gene sequence. Only those areas that are thought to be important for membrane localization and binding are shown. The functional domains are: a) the signal sequence (33), b) the hydrophobic transmembrane domain, and c) the charged membrane anchor. The 3 putative 30 N-linked glycosylation sites are shown with the letter G. The expression plasmid consisted of the bacterial pBR322-26 DK 173597 B1 replication origin and ampicillin resistance gene, a cDNA insert that codes for the mouse dihydrofolate reductase gene during promoter (53) and a HindIII-Hinfl fragment encoding 5 for the first 300 amino acids of gD under the transcriptional control of another early SV40 promoter. The HindIII center of this fragment lies 74 bp to the 5 'side of the gD gene's start methionine codon. The HindIII center in the early SV40 region of the vector (36) lies 250 bp to the 3 * -10 side of the SV40 promoter Goldberg-Hogness box. The gene center (blunted with Klenow DNA polymerase and 4 deoxynucleotide triphosphates) is ligated to the Hpal center in the 3 'untranslated region of the hepatitis B virus surface antigen gene (36). This method is also useful for producing a truncated HSV2 gene. It results in the sequence creating a stop codon (TAA) immediately after amino acid 300 in the gD gene. Transcription termination and polyadenylation centers for the truncated gD gene transcript are encoded by the 31-untranslated region of the 20 hepatitis B surface antigen gene (36).

The plasmid pgDtrunc.dhfr was constructed as follows: the 2.9 kb gD-containing SacI fragment was isolated from the BamHI fragment cloned from the HSV1 genome (see above) into plasmid pFM3 (see above) cut with SacI. A 1.6 kb HindIII BstNI fragment containing the entire gD gene was subcloned into HindiII-BstNI digested pFM42 (EP Application No. 68693). This plasmid was then cut with HinfI, blunted with Klenow DNA polymerase and 4 deoxynucleotide triphosphates and then cut with HindIII.

The 960 bp HindiII blunt HinfI fragment containing the truncated gD gene was isolated and ligated to HindIII-HpaI-digested pEHBall4. The resulting construct (pgDCos trunc) contained the truncated gD gene with the hepa-titis B surface antigen gene at its 3 'end. A 2.3 kb HindIII-BamHI fragment containing the truncated B1 gD gene was isolated from pgOCos trune. The 2.8 kb fragment containing the original and early SV40 promoter and the ampicillin resistance gene and bacterial replication origin of pBR322 was isolated from plasmid 5 pEHBal 14. The 2.1 kb fragment containing the mouse dihydrofolate reductase cDNA clone under transcriptional control of another SV40 early promoter was isolated from the plasmid pE348HBVE400D22 (36). These 3 fragments were ligated together with T4 DNA ligase and the resulting mixture was used to transform E. coli strain 294. Plasmid DNA from the resulting colonies was screened with Sac 2 and the proper construct pgDtrunc.dhfr (Fig.18 ') was used for further transfection studies.

Plasmid pEHBal 14 was constructed by cleavage of pE342AR1 (described below), an SV40 hepatitis chimera, with Xba I, which cleaves once in the coding region of the HBV surface antigen, and then sequentially removes sequences surrounding this XbaI. center using the nuclease Bal 31. The plasmid was ligated in the presence of the synthetic oligonucleohide 5 · -AGCTGAATTC, which links the HBV DNA to a HindIII restriction center.

Resulting plasmids were screened for an Eco RI-HindIII fragment of ca. 150 bp. pEHBal 14 was sequenced, confirming that a HindIII center had been located at a point just above where the HBsAg start codon is normally found. Thus, this construct places a unique HindIII center suitable for cloning at a position where a highly expressed protein (HBsAg) begins translation. All putative signals required for high expression of a protein should be present on this 5 'leader sequence.

The plasmid pE342, which expresses HBV surface antigen 28 DK 173597 B1 (also referred to as pHBs348-E), has been described by Levinson et al., EP Publication No. 0073656, which is considered to be incorporated in this specification by this reference. (Briefly, the origin of monkey virus 5 SV40 was isolated by digestion of SV40 DNA with HindIII and conversion of HindIII ends to EcoRI ends by addition of a converter (AGCTGAATTC). This DNA was cut with PvuII and RI linkers added. After digestion with EcoRl, the 348 bp fragment spanning the original was isolated by polyacrylamide gel electrophoresis and electroelution and cloned into pBR322. The expression plasmid pHBs348-E was constructed by cloning the 1986 bp fragment from EcoRl and BglII digestion of HBV (Animal Virus Genetics, (Ch. 5) Acad. Press, N.Y.

15 (1980)) (spanning the gene encoding HBsAg) into the p ismid pML (Lusky et al., Nature, 293: 79 (1981)) at the EcoRl and BamHI centers. (pML is a derivative of pBR322 which has a deletion that eliminates sequences that are inhibitory for plasmid replication in monkey cells).

The resulting plasmid (pRI-Bgl) was then linearized with EcoRI, and the 348 bp fragment representing the SV40 origin region was introduced into the EcoRI center of pRI-Bgl. The origin fragment can enter both orientation and orientation. Since this fragment encodes both the early and late S1 / 40 protnotor in addition to the origin of replication, HBV genes will be expressed under the control of one and the other promoter depending on this orientation (pHBs348-E represents HBs expressed under control of the early promoter). pE342 is modified by partial degradation with EcoR1 filling in the cleaved center using Klenouz DNA polymerase I and ligation of the plasmid back together, removing the EcoRI center prior to the SV40 origin in pE342.

The resulting plasmid is designated pE342ARI.

The resulting sequence creates a stop codon (TAA) immediately partitioned after amino acid 300 in the gO gene. The transcription termination and polyadenylation centers of the truncated gO gene transcript are encoded by the 3 'untranslated region of the hepatitis B surface antigen gene (36) ·.

The resulting vector was transfected (40) into a dhfr "CHO cell line (39), and a suitable clone gG10 * 2 was selected which produced the truncated gD protein and secreted it to the surrounding medium. The protein was extracted from the medium and the cells were tested for immunogenic activity. FIG. Figure 9 shows the results of immunoprecipitations of intra- and extracellular β-methionine-labeled extracts.

Radioimmunoprecipitation of cell-linked and secreted forms of gD.

Fifteen cells were grown in Ham's F12 medium (Gibco) supplemented with 7S commercially dialyzed fetal bovine serum (Gibco), penicillin (100 units / ml) and streptomycin (100 units / ml).

When the cultures were about 80% confluent, the medium was removed, the cells were washed twice with phosphate buffered saline (PBS), and a labeling medium (Dulbecco's modified = Eagle's medium containing 1/10 of the normal concentration of methionine) was added to 2 A final concentration of 0.05 ml / cm. S-methionine SJ.204, Amersham Int.) Was added to a final concentration 25 of 50-75 µCi / ml and the cells were grown for an additional 18-20 hours. After labeling, the medium was harvested and the cells washed twice in PBS and removed from the culture dishes by treatment with PBS containing 0.02% EDTA. The cells were then solubilized in 1 buffered in light buffer 30 consisting of: PBS? 3% NP-40, 0.1% bovine serum albumin, 5 x 10 8 M phenylmethylsulfonyl fluoride and 0.017 TlU / ml apoprotinin, and the resulting lysate was cleared by centrifugation at 12000 x G. For immunoprecipitation reactions, the cell lysates were diluted triple with PBS, and aliquots (typically 180 µl) were mixed with 2-5 µl antisera and incubated at 4 ° C for 30 minutes. To immunoprecipitate the secreted form of gD, 500 µl of 5 conditioned medium was incubated with 2 µl antisera for 30 minutes at 4 ° C. Immune complexes were then adsorbed to fixed S. aureus cells by Kessler's method (40a) and precipitated by centrifugation at 12000 x G for 30 s. The S. aureus cells were then washed 3 times 10 with wash buffer (PBS, 1% NP). 40, 0.3% sodium dodecyl sulfate), and the immune complexes were eluted with 20 µl! polyacrylamide gel sample buffer (62.5 mM "Tris" HCl buffer (pH 6.8) containing 10% glycerol, 5 S 2-mercapto-ethanol and 0.01% bromophenol blue at 90 ° C for 3 minutes.

After centrifugation for 30 s, the supernatants were applied to 10% polyacrylamide gel plates according to Laemmli's method (45). A, Immunoprecipitation of full-length membrane-bound gD from the gD12 cell line. B, Immunoprecipitation of the cell-linked form of the truncated gD from lysates of two 20 independently derived cell lines (1 and 2). C, Immunoprecipitation of the truncated gD from the culture supernatants of the two cell lines from B. (-) shows control rabbit antiserum (+) shows rabbit anti-HSV1 antiserum (Dako Corp.).

An intracellular form of 35000 daltons 25 and a secreted and apparently glycosylated extracellular gD protein are clearly seen.

Preparation of truncated gD used for immunization.

gD10.2 cells were grown to confluence in polystyrene wafer culture roller bottles (Corning 25140) in F12 medium supplemented with 7% commercially dialyzed fetal calf serum, 50 µg / ml streptomycin and 0.3 µg glutamine. After confluency was reached, the medium was removed and the cells washed 3 times in the same medium without fetal calf serum and supplemented with 2 mg / ml Hepes-puFfer (serum-free medium). The cells were then cultured for 3-4 days in serum-free medium, and the conditioned medium was then harvested and stored at -20 ° C. The medium was then thawed at 37 ° C and centrifuged at 5000 rpm.

for 20 minutes in a Sorvall GS-3 rotor. After centrifugation, the pill was discarded and the supernatant was concentrated in an ultrafiltration apparatus (Amicon) equipped with a YM-5 ultrafiltration membrane. The resulting preparation was concentrated about 150 times relative to the starting material and contained about 8 mg protein per day. liter. The preparation was then extensively dialyzed against phosphate buffered saline (PBS) and used for immunization without further purification.

15 Immunization of mice

Every 8 weeks old BALB / c mice were immunized with 36 µg protein contained in 200 µl of an emulsion consisting of 50% aqueous antigen and 50% "Complete Freund's adjuvant." Each mouse was immunized with multiple intradermal 20 and subcutaneous sites as follows: 25 µl in each hind foot cushion, 50 µl in the tail and 100 µl distributed among 3-5 intradermal sites along the spine. Four weeks after the primary immunization, the mice were gain-immunized with 36 µg of the protein as above except that the emulsion was prepared with "incomplete

Freund's adjuvant. "For enhancement immunization, each mouse received 200 µl of the antigen emulsion distributed as follows: 50 µl in the tail, 150 µl distributed between 5 intradermal sites along the spine. The sera obtained from this tapping were used for in vitro neutralization studies (see below). 37 Days after amplification, the mice were used for virus exposure studies. Control mice, 32 DK 173597 B1, which matched the experimental mice for age , gender and strain, were immunized with human serum albumin (15 µg per mouse) using the same regimen as with the experimental mice.

5 In vitro neutralization.

Sera from 11 mice immunized with concentrated gD10.2 carbon. trip supernatant was tested for the ability to neutralize HSV1 in vitro. Row-diluted mouse serum (double dilutions: 1: 8 to 1: 16384) was incubated with about 10 40 pfu HSV1 for 1 hour at 37 ° C in Dulbecco's modified

Eagle's medium (DMEM). After the serum incubation, each dilution was applied to about 40,000 Vero cells contained in each hole on a 96-well tissue culture plate. After 3-4 days, virus growth was determined by staining each 15 hole with 0.5% crystal violet. Holes in which there was virus growth showed no staining. Neutralization titers were calculated by determining the highest serum dilution that prevented virus-induced cell death. All tested sera (n = 10) from mice immunized with gD10.2 supernatant showed H.SV1 neutralization activity (range.

1:16 to 1: 512) and the HSV2 neutralization activity (range 1: 8 to 1:16). Control mouse sera (n = 8) did not provide any neutralization. Serum obtained from a mouse immunized with HSV1 gave a neutralizing titer of 1:32.

Virus exposure 11 Mice immunized with concentrated gD10.2 supernatant and 13 control mice immunized with human serum albumin were exposed to 10000000 pfu HSV1 (Maclntyre strain) by intraperitoneal injection. 14 Days after injection 30 of virus, none of the gD10.2 immunized mice showed any evidence of viral infection. In the control group, 7 of the 13 mice were dead by the 14th day, 3 showed severe loss and 33 paralysis, and 3 looked healthy. Statistical analysis (two tailed Fisher exact test) showed that the difference between the immunized group and the control group was significant at the p = 0.002 level (see Table 1).

34 DK 173597 B1 jj c © —I <1> <—I • h o cn

C -U li- H C

not c) to> · Η (-lo?

X) Ό CO 0) Η Ό O

(U C <-I fA .. E H {J C Cj 1 * - O II

tf) (D r — i h Η · Η D d> »0) ΙΛ O 'I 3 E

r — I> Q) 4J £ 10 4) T3 C-t O Tue E 33

Od) jj λ Ct © © «- (-« £ (-1 jj f — i 1 * - CO - 4 JJ Cj O .C O 'CM) a -H 0)

Jj Q) t-4 Cj Di <· - CD Ή H JJ 0) 0)

Bi © -H © E © Ό (0 W CO

tf) (DO)> o? D K \ C E Cj C JJ

TIO | . OM) N £ P <U (0 (j O -H C

O sj O C0 «D> lA O Ό» TO:> © U_ C rt

T31 Ό 3WW »B C Q. Ό Tue E

EX o <u i? Λ I) h 3 on jj © ce cj c eb x: rH OD Cj X) X) -U CD 0 4) 1) 4) i χ> © —i © to co (j ia c ro co © JJ fl) > - (OE © Ό to -HD Cj Ό

~ jj © X) —l £ 3 ·· "> <D T) C E O CO

<M © JJ C © © JJ © D-H e) ®Ci_c

—4 E -H Door CTjJ Cj> EB

> E O fA> 4) H C> (0 · Η O I) (4 E

tn CO -H Cj> · Η Cj C 1-tJJ —4 —I © x i — i jjotn co © ο. © qVjq-pj) ^ e. X (j u. »C» e- c 'CO © © Γ ~ · Η Ό OM) Φ (0 U- —4 © «4- O JJ © C - (

OIJC «Η C B * - (JJ ©> -H © CO

, -JC EB © (4 © Cj Cj O '> • HC (4 OG) © X σ »¢) 4) ® s · Η JJ Cj © I i — I © ro C (!) (Ι) Ό i 3 VO o. Ό OHH H YY (4 (4 -r (T <U »—I BQ and U O) O YY © © C On 8)

c ·· · Η σιΓ 0) O E 1 * -C7I O M 3 —I

t -I i-I C> E «CO © -HEC

»—I CMIO CJ JJ © 3» ©> - (© CEN ©

> CO Cj © O © JJ C D-HD

—I. in · »J) h OOH (I) o> O 'E t <Ό U) χ - | D © X) O £> © HU D E © Ή ©

CD © C C O .Ω YY —1 '"-v-t Cj> E

C b IXt O Di £) EB

I— I © T3 ΟΙΌ © IA> E Cj JJ

n Or-u) ro o · <-) © ω · α o © —t> O (0 (i Ό> CD O'— I Cj t> - Cj • to o> .X s –th cd α © JJ CJ X e li- <t J- * -H Jj CO · 3, -t X) © * co © —i © cj ro c ro -h cm

© lA CO © Cj ϋ Cj C CO jJ Tue 3 © CO C O

C ·· O E C * - © cn © JJ ΕΣΌΌ 3 O

i — i> jj © cro o de -4 I O I C O © »4- (4 O« · (-1 E o

> NO <—I O <* —I UJ X) Jj x X) CO ©> -H

tn r— (> -HjQD · ηχ © c jj © il x ·· tn jj T) · »tj ε ·“ · ** - · “· J3 r — ix ro © cn ©> jj © · c od o o .

C E C Ό O O CL -—- Jj W · -)

(J -Η · Η 10 H i CO © © S (4 © X) JJ

O U 4) ΌΗ O EB (4 E (4 <—I © C ©

CfA i * - O © C Q.> JJE © CL CO © (0 D

© © ro cj> ·. cj © to cro co o x co d ro Π C JJ 0> © JJ © C-4 D H 4) -r-ι d E © • rt D © CO J3 (-4 Jj © Cj CO c: <fSH>

Jjf4> O 'D O H jJ -.- 4 Cj I D Jj O -H

C JJ C © C Di <4- JJ -M> O © E C · · (4 C

<£ Oin (4 -1-4 C O Jj ti-u e CO (4 Jj Jj cn x α C -r! (4 ti OT ©> - -H> © © C X?

Di © Di CPX) 4) 4) C D X »Cj O. © 3 »Cj>> -H C © coco> '-> © © ->

© EB © X E * · "· Cj CO * -4 * r4 © X) Jj CO« X

—4 H 4) H CjX) Ό o 4) -4 B BH .tn JJ

CO O Cl) D · M 1) C 3 Σ JC X) C C X C

4) »Η ΙΛ Cj O (J CO Ή -—4 O) CO © D © '- CO

CD-1 ^ 4 ro O —- O © --4 c> «4- CO - CE CO Di <X t-, ti- <t» —I «—I Cj © —4C DX) COECC -> - 4 © ao r- —i ro o —i> -h ecjj-hmh «4- CO C IA tA D Cj ti- X) t / SIDD CO Cj E ·« - »

©AKE NO £ 4) X — I Cj © © CO © D C

Gi (D Ό <—I XJ DE CO> © Cj Xl C CO X) G »

© D C ·· © nO © O D ** - tn> U- D -H -H —4 -H

CO O Ω X CO —I> ON Z CO X <0 X XL 'CO C C ro tn

Cj ON CTN

O O O · · U_ lA vA —I CM ΙΑ <ί 35 DK 173597 B1

It was found that the truncated protein released to the medium from gD10.2 cells was effective in protecting mice against a lethal infection of HSV1.

Antigen preparation for HSV2 virus exposure.

5 Enriched gD10.2.2 cells, grown in the presence of 250 nM methotrexate, were seeded in roller culture flasks 2 (850 cm) and grown in Ham's F12 medium (GIBCO) supplemented with 7% fetal bovine serum. After the cells reached confluence (about 5 days), culture medium was removed, cells were washed 3 times in phosphate-buffered saline (PBS) to remove serum proteins, and new "serum-free" culture medium was added. The serum-free medium consisted of Ham's F12 medium containing 25 mM Hepes buffer. The cells were then cultured for 3 days, 15 and the resulting conditioned medium was harvested and used for antigen preparation. Then, fresh serum-free medium was added to the cells and the conditioned medium harvesting cycle at 3 day intervals was repeated an additional 1 or 2 times until cells 20 died or no longer adhered to the culture surface.

The gD10.2.2 conditioned serum-free medium was then filtered and centrifuged at low speed to remove cell debris, and the resulting material was concentrated 10-20 times with an ultrafiltration device (YM-10 membrane, Amicon). The concentrated medium was then dialyzed overnight against PBS (3 replacements of PBS, 1 liter per replacement). The resulting material was tested to determine the protein concentration and analyzed by polyacrylamide gel electrophoresis 30 to determine the protein composition and assess the purity of the preparation. The material prepared by this method was then used to immunize animals against HSV2 infection as described below.

36 DK 173597 B1

Immunization of mice against H5V2 infection.

40 BALB / c female mice were obtained from the Charles River Laboratories (Boston, MA) and immunized with the secreted gO protein (gDtrunc) or human serum albumin 5 (HSA) at the age of 12 weeks. For the primary immunization against the secreted gO protein, the antigen was adjusted to a concentration of about 70 µg / ml in phosphate-buffered saline and emulsified with an equal volume of complete Freund's adjuvant. Each mouse was then immunized with 200 µl of this emulsion distributed as follows: 50 µl subcutaneously at a site ca. 1 cm from the tail root, 25 µl subcutaneously in each pageant tree pad and 100 µl distributed among 3-5 intradermal sites along the spine. The mice were then gain-immunized with the same antigen 1 month after the primary immunization. For the enhancement immunization, the antigen s were prepared by the same procedure as the primary immunization except that "complete Freund's adjuvant" was replaced by "incomplete Freund's adjuvant". For enhancement immunization, 200 µl of antigen emulsion was injected into each mouse, distributed as follows: 50 µl in the tail, 25 µl subcutaneously in the loose skin over each thigh, and 100 µl distributed among 3-5 intradermal sites along the spine. The control group of mice was immunized according to the same specification as the experimental group of mice, except that human serum albumin replaced the secreted gO protein as the immunogen. Serum was collected from the mice 24 days after enhancement for use in in vitro neutralization studies.

30 HSV2 virus exposure.

Both the experimental group (injected with secreted gD) and the control group (injected with HSA) of mice were exposed by an intraperitoneal injection of HSV2 (MS strain) at 31 days post-amplification immunization. Each mouse received 2 x 10 6 pfu virus in 100 µl of Dulbecco's modified Eagle's medium (DMEM) containing 10% fetal bovine serum. LD50 experiments showed that this amount of virus 5 represented 100 - 500 times the amount of virus required to kill 50% of a population of normal (uninjected) BALB / c mice. The virus-injected mice were observed for a period of 3 weeks. All control mice (injected with HSA) died within 9 days of virus exposure-10. All mice vaccinated with the secreted gD protein survived every 3 weeks and appeared normal (ie, they did not show any loss or paralysis).

38 DK 173597 B1 »“ I «i c» —i ε © -u ii- io i g>

(D Ό tO AJ E C

• Ό 00 © W D C

0) C iA O h ·· EH >> h 0)

OT © * —I 01 Η -H 3 ti 0) O

ι-Ι> -P --- ^ O) (O CJ U. 0)

4-1 »—I dl OH-UK (DU

jj PP 4-1 »01 O · γΙ 0) Λ W (l cn σοι> * (i) 0) tdtjI π οι nro ή μ D Bl α ΙΛ Ό TO> S © X> [CN 3 (Λ MU £ i ) σ \ ex © c »—1-PECO)

© P TI © CO

ΡΌϋ.-0-ρ-σβ »·

0) 01H O 0) P C

I - * -> + 1UH £ α © CN © * »- t C © © P <0 © 01 * —i ε α>> * m <o ρ * -t; =» ε O -H 4J HC <4 - © ©

co © -p ρ> -h to p -P

Srt -X O LO 01H 0) 4-1 «l_ χ p c © -h ta © © © p e ό w

Cns * 0-H 0) E h T3

(0ΗΗ ID «D

r-H <J- -H ta © <—I -P 3> I

• cn p et; p o © tn cn 4-1 o © © I O © © 3 => d r-ι m · 'o o jo c to

© ·. -Η Ω O P>> X

c r-ι oi— (oi.c © © · © · <- (II (0 OHOO Ό CN CN CN P -P © J3> C ©> rt -P © O TO> -) cn mdp © □ © - p UJ X ·· © © "OO CO © © p CO —ICCCO TO CO © <IP © <t Ό O '

I— CN © ~ U ffl · r — I

03> UH4l <i P4J ra rH tn © © P Ό © 4- »• OX TO 3 * OPD Di © 4J CN E« 4- © W Ό D .. DO © · © 4J>> ©.-IO © P Qi.'p © TO> - (

Cl gt- uh 3 -Ή I -d · I> P · 4-1 1-ttNO · —I C O © · -H ©

> O> O ^ H 0) ©> h C

cn · —I to -H jQ "O P ©> n

X '. Χ-ΡΌ (0ffi © 4-> TO

-I ra © o * · —i + ιυ · π c ρ c ε c ex-π © 4-> ©

o · Η -Η P 4-1 P OT SI

(^ ϋ4ΐυ e »u σ cm o © C4-> o> ccm © © JJ S h λ-H c -H -H © σ c ©> © 4J.p-hxp ε • H 3> to_qpopp ©

4-1 p QCTi30P © 0ot P

C 4-1 <C P C SS U- SS MU --- C O

<J O LO O. Ή C · * -0 · Η i — 1 «* -

Dl I C -Η P Ε Ή E TO C

> SS © TO O P © © P >> TOPW4->>

—1 K © X P 4-> 3 O

• —1 jQ 4-1 ^ © 3 · '© ra OT J3. ^ 4 © O »cc 4-i ot mm S p cj - (3Ξ c © CD i — I CN DO ^ oD · Η © 4-1 <cepo _c · ό c? M © CO r- il C -H 2 £ wcmo vo ©> \ r — i ©

© © N> (J \ H -U i) oJ

Ol TO TJ —I Ό O P -P

O DC •• © C-HOC © ot oo 2ira ^ -i> ©> ti-OT tn ρ cn o \ o r- r- ~ * · u_ mm <—i cn 39 DK 173597 B1

Truncated glycoprotein 0 was purified from culture medium conditioned by the growth of the gD10.2 cell line previously described. The culture medium was concentrated by ultrafiltration and truncated gD was reset by immunoaffinity chromatography using a monoclonal anti-gD-1 antibody coupled to "Sephadex® 4B". Truncated HSV1 glycoprotein D (gD-Ι) was isolated from serum-free medium conditioned by the growth of gD10.2 cells as previously described. Cell culture medium 10 was concentrated by ultrafiltration using commercially available membranes (Amicon Corp.) and ammonium sulfate precipitation. gD-Ι was then purified to near homogeneity by immunoaffinity chromatography.

The immunoaffinity column was prepared by coupling a monoclonal antibody produced against HSV1 to cross-linked sepharose * (Pharmacia Fine Chemicals) and eluted by a method similar to that described by Axen et al. Nature 214, 1302-1304 (1967). The main product of the medium of unfractionated culture medium conditioned by the gD10.2 cell line is the mature truncated form of gD-Ι (about 43 - 46 kd) and a precursor form of gD (about 38 - 40 kd). On average, the gD protein represents 20 - 50% of the protein present in growth-conditioned medium. Fractionation of this material by immunoaffinity chromatography resulted in a significant enrichment of gD. It can be seen that the eluted material is free of all contaminating proteins detectable by silver staining. To determine whether purification of the protein by this specification denatured the protein or broke the antigenic structure of the molecule, antigenicity studies were performed with a variety of monoclonal antibodies. In these studies, it was found that all tested antibodies except those reactive with the carboxyl end reacted with the purified preparation. No difference in antibody binding behavior could be detected with the purified preparation prepared relative to the material found in unfractionated culture supernatants.

To determine if the purified gD-1 protein could be used effectively as a basis for a purity vaccine to protect against genital infection with HSV2, guinea pigs were vaccinated with gD-1 compounded in various addition solutions. In the first studies, purified gD was incorporated into complete Freund's adjuvant and injected into intramuscular and subcutaneous sites on Hartley female 10 guinea pigs. Hartley Hurim guinea pigs, 2 months old and weighing about 250 g, were purchased from Charles River Laboratories (Portage, MI). For studies using "Freund's adjuvant", the primary immunization of the injection consisted of 30 µg gD-Ι emulsified 15 in 50 λ "complete Freund's adjuvant" distributed as follows: 0.5 ml injected subcutaneously into the loose skin of the neck and 0 , 5 ml injected intramuscularly into the thigh. After 31 days, the animals were fortified immunized with the same amount of antigen incorporated into "incomplete Freund's adjuvant".

The control animals were injected according to the same regulation as the test animals, except that additive solution was injected alone. The experimental and control animals were exposed to intravaginal infection with HSV2 19 days after enhancement immunization. In studies using alum-added gD-30, 30 µg of gD-Ι was incorporated into either alum phosphate or alum hydroxide (0.15 ml) used for both primary immunization and enhancement immunization. Alum added protein was injected by intramuscular injection into the hind legs.

The animals were gain-immunized 51 days after primary immunization and exposed to live virus 27 days later. Each animal received one primary immunization containing 30 µg of purified protein incorporated into "complete Freund's adjuvant" and one enhancement immunization (31 35 days later) with the same amount of antigen incorporated 41 "incomplete Freund's adjuvant". All animals were exposed by intravaginal inoculation of HSV2 19 days after enhancement immunization. Table 3 lists the results obtained from these studies. It can be seen that the animals 5 vaccinated with gD produced high levels of antibodies capable of preventing both HSV1 and HSV2 virus infection by an in vitro virus neutralization assay. It was found that the sera from these animals neutralized HSV1 slightly more effectively than HSV2. This result 10 is reasonable considering that the immunogen was derived from HSV1 and that there are known to be type-specific anti-genetic determinants on gD-Ι (Éisenberg, RJ et al., J. Virol. 3: 428 (1980); Pereira, L. et al., Infect., And Immun. 2: 724 (1980); Showalter, J.D. et al., Infect.

And Immune. "34, at 684 (1981)). More strikingly, all animals vaccinated with gD-Ι were completely protected against the clinical manifestations of viral infection (ie, redness, vesicle formation, ulceration, loss of urine retention and fatal encephalitis).

20 13 Of the 14 animals injected with the additive part alone, severe primary infections developed.

There were numerous vesicles that typically coalesced to form acute ulcerations. In contrast, the animals vaccinated with gD-gav gave no indication of virus infection. These results clearly showed that gD-Ι incorporated in "complete Freund's adjuvant" can provide effective protection against genital HSV2 infection.

Since "complete Freund's adjuvant" is not acceptable for human use, it was then sought to determine if gD-Ι could provide protection against HSV2 infection when combined with an additive suitable for human use. To this end, studies with alum-precipitated protein complexes have been initiated (J.S. Garvey et al., Methods in Immunology (1977), p. 185 (17)).

In Table 3, results obtained using gD-1 incorporated in the additives alum hydroxide and alum phosphate are compared. In control studies, animals were vaccinated with additive alone. It can be seen that both of the alum-based preparations elicited high 5 levels of neutralizing antibodies to HSV1 and that the neutralizing titers against HSV1 were comparable to those triggered against gD-1 incorporated in complete Freund's. ". However, the titers of antibody capable of neutralizing HSV2 were significantly lower with gD-1 incorporated in one of the alum preparations than with gD-1 incorporated in "complete Freund's adjuvant". This result suggests that incorporation of gD-1 into alum results in the loss of one or more antigenic determinants common to HSV1 15 and HSV2, or that the recognition of cross-reactive anti genes is more effective when the protein is incorporated into Freund's adjuvant. These results also suggest that alum hydroxide is a more effective additive than alum phosphate, since the neutralizing titers above 20 for HSV1 and HSV2 are substantially higher with the former than with the latter.

Although the protection afforded by alum additives was less effective than that achieved with "Freund's adjuvant" preparations, it was nonetheless significant. Although a number of animals showed signs of viral infection, the severity of the infections was significantly less than that observed in control animals injected with additive alone, thus the mean lesion number was 0.9 in animals vaccinated with the alum-phosphate vaccine composition and 0.7 in animals vaccinated with the alum-hydroxide vaccine composition, compared with an average lesion number of 3 , 2 for the control animals injected with additive According to the ή + scale used to assess lesion rate, the reduction from 35 averaged 3.2 to 0.7 to a reduction 43 in clinical symptoms from several large vesicles (rating of 3) to less significant redness and swelling (rating of 0.5) Interestingly, the average in vitro neutralization titer for these 5 studies, HSV2 correlated with the severity of clinical disease.

The results described above show that the clinical manifestations of primary genital HSV2 infection can be significantly reduced by vaccination with recombinant gD-1. The results obtained show that a single HSV1. 10 derived glycoprotein may provide complete protection against genital HSV2 infection when. it is administered in conjunction with a powerful additive.

44 DK 173597 B1 in d -h

C · - 1 CO

¢ 0 <f N Η N to (9 t3 ^ C t · * w λ -v · Η E H t3 • h ico h i — i * —i * —ι o α>. Ε -O rH C · JZ o

E 01 O + 1 +1 + «+1 D D -DO D

O "O -H + O 4) O D>, 0) <U

0 £ Ό W · —I ΙΛ Ό N> L <T> O Li Li Ό 0] -H © CO ~ - 004) 0) 0)

Li s: · - «- * -» Li 4- 0) 4- U 4- O 'C O * O <3

Ό CO Λ H H O U

o> * o) o »h d E i i) -r | V <Φ 3 · Η

In JJ -Η Π (0 C

C oa) φ tn ho to.

o a-> jx t ~> ~ ω u ε · H CL 03 O Γ "O CD Lt C -> O 0) 4) d ε * h o <r vo. -t tu οοΗβ-υ c cl

<0> »4J r-t -H D -H CO .C 0) C CL

C (0 to * - l -U 4) 0 -h <ce ό σ ^ -l> »co σ '* - · o -, - rlao) + h σ» OD to 4- rH 0) 0) CO- * ^ C * ^^ h

> * g -H t »V. to O.

ία) o to co o: li -ri> Ό o ^ CO (O 3 · J) * 4) Γ to -ucn -rt k_j-u t ω

> CL-H o K \ vO ΓΛ O D; » -H 0) <r ε -H

O ε-L »0 \ 1Λ <t 0 \ f-ι X)> - 4) JO

C> x (0 S 3 X-H · - 4J 4

OC in ε O- 1 (3 L H 4 -H S

-η · η ε jj ati to o f-t * o

-L> 0) O 4) CO 0) ε Ή CO

JX D Li * C '-' CjQLi ^ 0)

0) 0 D Li -it NtA C C * C O * H D JZ

4- Li -H0) - - 0) ω ω CO C CO 0) 0)

-) CQ. 3D f-t N »H O '4) +) XI (D T) -H

UI * HO -ri CM 0) V) CO> D

CD, I O C D 3 »+ J ΓΛ + Ι + ΙίΛ> -H C O O

<C CM> W -Γ-tCQLO VV 0) D (!) 4- C 4) C7> ε »-> —ID) 3C O VO Γ- H m'OXOOh oo C /) o cn c -—- - - x) ec σ 'odd

mi ft-rCN CO 1Λ VO O -H D C D CL

• H »—I Li Li D w 4> -H CO DE

Ή> D 0) CT 0) CL> β> n to tn 10 O 00 NN CO ε O 4) - a> 10 -Dm c Ή rt - - · —I >> _ * Li o Li Ό • h cn ri —- o cm .ho) co co a) oa> C ε <0 <—I CT H to ·· Η 3 Ό 0) CU Li 3 »+1 K \ + l-flK \ O 0) D 4) 0 ) 4- CT CD C /) VV CO Lt (0 X) Ό DO) C 3 X Ι * "* N Ot D8EC + 0) 0 • O φ o) - - ω o ci o> i * '» cx:

O CD C σ \ σ \ O Ό D <0 D CT

ε.-t C DE D · - OD

3 O >> g tg (4 DO) C cn co o 4- ω Ό ti -HC tf \ <J Ov O 'OO' HEAD 3 4>> ι — I ri * —t C · Η 4) \ i £ > (0 · Η D (i - Ε Ή 3 Li D Lc 0) ε E <0 4) 0)

CO Q) D i-C 0) CO <D · —I CO

H D "O CO EH-Pto> JO JD

= CO -H 0) -ri Q 0) O O

4- D _C X C —i XJ> »·» CO «—I

and cqloo) co ω r- 'co ε ο »η

two to CO Li r-c D Ό C tO N to D CO

3> Ο T3 CO D O OCO COD

Μ T) O JO> - 3> * d (0 O '»C0 CO)

Ή C '-} CL C D Q) 0) D. ~ t r-i 4- OC

0) 3TDIIC0 C rH ^ Cl) ^ -HO

d <u to c c .c x) ω ε tu to -η

D Li 330. O 4- S Γ> · Η a CO

>> C U. S CO · —t · —I CO LiCCeOsO ·· ΉΜ

4) = D- CO CO O D -Η · Η Ό ΙΛ O Lt HH

w σ co jo -H> co4) Ctae) 0) Ο -H CO C -Η -H Q.> (0 3 0 ·· C Ό4)

CD C> 34) I Li Li I VO I - Ο Ό I

3 rH 3 0) C * -t · —IOC C0-H <r <jj-f -H -HC0 ε I Ό Li α>] I 3 '- I s:> + ii | Lu «—t OQ ή * * —i C3 CD - CO CO O '<0 * * + 45 DK 173597 B1

The advantage of using the truncated protein for diagnostic and vaccine applications is that, because it is secreted to the extracellular medium, it is contaminated with far fewer proteins than would be found in a whole-cell preparation.

It will be appreciated that, according to the present invention, a permanent cell line is used to produce the protein. Upon transfection, the vector is incorporated into the cell line's genome and can produce the protein without cell lysis. Thus, the cell line can be used for continuous production of the protein, especially in the truncated form secreted from the cell. Eg. For example, the cells expressing truncated protein can be continuously used in a perfusion system by constantly removing antigenic medium from the cells and replacing it with fresh medium.

The particular cell line used here was a CHO line deficient in dhfr production transfected with a vector containing a dhfr marker. By exposing the cell line to methotrexate (Mtx) under suitable conditions (54), dhfr production and thus the associated gD production can be enhanced. Three cell lines derived by transfection of the truncated gD gene into dhfr "CHO cells were plated in parallel, labeled with 5S methionine and immunoprecipitated as described in FIG. 2. Lanes 25 1 and 2 indicate the amount of secreted gD immunoprecipitated from 500 µl of culture medium conditioned by two independently isolated cell lines before selection with methotrexate.

Lane 3 indicates the amount of truncated gD immunoprecipitated from an equal volume of culture medium from a cell line (gD10.2.2) selected for growth in 250 nM methotrexate. Rabbit anti-HSV1 antibodies (Dako Corp.) were used for the immuno-precipitates shown in lanes 1-3. Lane 4 represents a control immunoprecipitation of 500 µl of medium conditioned by the gOlO.2.2 cell line with normal rabbit serum.

To quantify the relative amounts of truncated gD secreted into the culture medium of cell lines before and after selection in methotrexate, a competitive EL1SA assay gD12 cells expressing a membrane-bound form of gD were plated and fixed. with glutaraldehyde to the surface of 96-hole microtiter plates as previously described. Conditioned medium from various cell lines known to produce the truncated gD was diluted across the microtiter plate and incubated with a fixed amount (2 µl) of rabbit anti-HSV1 antibody (Dako Corp.) for 1 hour. hour at 20 ° C. Unbound antibody and soluble truncated gD antibody complexes were removed by washing each 15 hole 3 times with PBS. Horseradish peroxidase coupled to goat anti-rabbit IgG was then reacted with the fixed cells for 1 hour at 20 ° C and unbound antibody was removed by washing 3 times with PBS. The colorimetric substrate, OPD (o-phenylenediamine), was then added to every 20 holes and allowed to react with the bound horseradish peroxidase-antibody complexes for 15 minutes. The reaction was terminated by adding sulfuric acid to a final concentration of 0.25 N. The absorbance of the OPD in each hole was determined using an automated microtiter plate scanner (Titertek multiskan) and dilution curves were deposited. The binding of anti-HSV1 antibodies to the starting CHO cell line was used to measure the degree of non-specific binding at each dilution.

The amount of truncated gD in each culture supernatant was inversely proportional to the amount of absorbance in each hole. Monkey circle, binding of anti-HSV1 antibodies to gD12 cells in the presence of medium conditioned by cells that secreted truncated gD before amplification with methotrexate. Closed circle, binding of anti-HSV1-35 antibodies to gD12 cells in the presence of medium from 47 g 171297.2 cells selected for growth in 250 nM methotrexate. Open square, binding of anti-HSV1 antibodies to gD12 cells in the presence of 100-fold concentrated medium from unstained cells that secreted truncated gD.

This procedure was performed on the gD10.2 cell line to produce an amplified gD10.2.2 cell line capable of growing in 250 nM methotrexate and which secreted about 20 times more depressed gD into the culture medium than the starting gD10.2 cell line. cell line (see Figs. 10 and 11). The 10 dhfr marker / amplification system can be used with other cells capable of recording and stably incorporating foreign DNA.

The invention's success in demonstrating that a truncated form of a membrane-bound protein lacking that portion of the hydrophobic-hydrophilic carboxyl-terminal region responsible for binding it to the membrane may be immunogenic, nevertheless, similar results can be expected. with other immunogenic membrane-bound proteins, thus providing an improved source of 20 vaccine against viruses, parasites and other pathogenic organisms.

In order to remove the membrane binding character, the protein is secreted to the surrounding medium.

EXAMPLE 2

This example deals with an HSV2-gC protein (formerly referred to as a gF protein).

Cells, viruses and DNA isolation.

HSV2 (strain G) was grown on HEp 2 cells after infecting the cell culture at an input multiplicity of 0.1 for 3 days at 33 ° C in Dulbecco's modified Eagle's 10 medium containing 10% fatal bovine serum and antibiotics. HSV2 DNA was isolated by digestion with proteinase K followed by CsCl ultracentrifugation as described (23).

pNA manipulations.

15 Restriction enzymes, DNA polymerase Klenour fragment, T4 DNA ligase and T4 polynucleotide kinase were purchased from Bethesda Research Labs and used according to the supplier's guidelines.

Molecular cloning of HSV2 DNA restriction fragments.

The EcoRI "P" fragment corresponding to the approximate map position is approx. 0.650 of the HSV2 genome was isolated from EcoRI degraded HSV2 DNA on 5% acrylamide gels. The isolated fragment was cloned into Eco RI digested pUC9 (28). This plasmid was called pUD-RIP.

The 25 pUC-RIP subclone was then used to locate a SacI fragment of the HSV2 gene containing the EcoRI-PM fragment. Southern Dilution Experiment (27) showed that a 4.9 kb Sac fragment of HSV2 contained the EcoRI - Pn fragment. This fragment was isolated on 0.7 / O 3Q3TOS6 gels and cloned into a plasmid derived from pBR322 which contained a unique SacI center (55). Another plasmid 5 was called pBRSacI- "A. Further restriction enzyme analysis of peRSed® E" detected a 2.9 kb SalI fragment having sequences homologous to the EcoRI-TT fragment subcloned into SalI-degraded pUC9 soro described above. This plasmid was called pgC2Sal2.9.

10 DNA sequence analysis of cloned HSV2 DNA.

Most of the DNA sequences were determined using the dideoxynucleotide chain termination technique.

Various fragments were subcloned into the replicative form of the ml3 phage vectors mp7, mp8 and mp9, and the DNA sequence was determined as previously described (29).

32 In some cases, fragments were P-labeled at their 32 5 'ends with γ P-ATP and T4 polynucleotide kinase, and the DNA sequence of the fragment was determined using the chemical degradation method (56). Computer assisted analysis of DNA and protein sequence data was performed using the HOM program (57). The hydropathy of the deduced amino acid sequences was analyzed using a width of 12 amino acids and a jump of 1 (31a).

Southern immersion analysis of HSV2 DNA.

Restriction endonuclease digested HSV2 DNA and plasmid DNA were fractionated on 1.5 K agarose gels and immersed on nitrocellulose using standard procedures.

The single-stranded ends of the Sac2 fragment, marked 30 with an asterisk in FIG. 12, was filled in with the KLenow fragment of DNA polymerase I, and the resulting strop-ended fragment was ligated to SmaI digested replicative

Form of ml3mp7 (29) with TA-ONA ligase. The single-stranded goat DNA prepared from this ligation and transfection was used as a template for the synthesis of P.-rosified single-stranded probe DNA with high specific activity.

ISLAND

5 te.t (1 x 10 counts / min. Μg) using

Klenov / fragment of DNA polymerase I. Hybridizations were performed using standard procedures (27, 58).

Results 10 Molecular cloning of the gF coding region of the HSV2 genome.

The strategy adopted for isolating the gF gene from HSV2 was based on the assumption that this gene was colinescent with the HSV1 gC gene. This assumption was supported by the recent recognition that a 75,000 dalton glycoprotein, gF, with antigenic determination for HSV1 glycoprotein C is found in HSV2, and that the gene for this protein is approximately colinear with the HSV1-gC gene (22d, 59). . Furthermore, the isolation of a monoclonal antibody that binds to both HSV1-gC and HSV2-F further suggested that these two proteins may be homologous to each other (22f). Thus, it was decided that DNA sequence analysis of the HSV2 genomic region, which is colinear with the HSV1-gC gene, would result in the derivation of protein sequence information that would localize the HSV2-25 gF gene.

The 600 bp EcoRI - P1 fragment of the HSV2 genome has been shown to belong to position ca. 0.650 on the card (12).

This region is approximately colinear with the known coding region for the HSV1-gC gene, which is between ca.

30 0.630 and approx. 0.660 on the HSV1 genome map (59).

This fragment was isolated from an EcoRI digestion of HSV2 DNA cloned into plasmid pUC9 (28) and its DNA sequence determined (29, 56). Comparison of the resulting sequence with the HSV1-gC sequence (59) revealed a significant degree of sequence homology between the EgoRI "P1 * fragment and the 3 'end of the HSV1-gC coding region. The fragment then used as a probe to isolate a Sac1 restriction endonuclease fragment from HSV2 genomic DNA that overlapped the Eco RI - P1 fragment sufficient to include the remainder of the HSV2 gene homologous to HSV1-gC. FIG.

Fig. 12 illustrates the steps taken to isolate a 2.9 kb SalI fragment from the HSV2 genome containing the EcoR fragment and used for subsequent DNA sequence analysis.

DNA sequence analysis of the Eco RI -, P - region of the HSV2 genome.

The 4.3 kb βΒαΙ-Έ1 fragment, which was isolated from the HSV2 genome based on its sequence homology to the EcoRI "Ρ" fragment, was further digested to obtain a 2.9 kb SalI fragment, which was designated pgC2Sal2.9.

FIG. 12 illustrates the fragments of pgC2Sal2.9, which were subjected to DNA sequence analysis using either the * dideoxynucleotide sequencing procedure (29) or the chemical degradation procedure (569. In addition, this figure shows the position of the EcoRI-P1 fragment in pgC2Sal2. 9, as well as the position of a BglII cell, which corresponds to the right end of the BglII "N" fragment at position about 0.628 in the HSV2 genome (12).

Specifically, FIG. 12 cloning of pgC2Sal2.9, the HSV2 region which is colinear with HSV1-gC on the map.

The area of HSV2 genome mapping from ca. 0.61 to 50 approx. 0.66 was cloned as a SacI fragment (pBRSac-EM) using the 600 bp EcoRI "P" fragment as probe. A SalI subclone of pBRSac- "E.", pgC2Sal2.9, was used for ONA sequence analysis. Arrows refer to the sequenced regions and the location of a larger open reading frame of 479 amino acids derived from the sequence is elucidated. Various restriction centers have been elucidated, including the EcoRI centers that delineate the EcoRI "P" fragment and the BglII center located at the right end 5 of the BglII "N" fragment (short position about 0.628) (26) ..

The Sad I fragment marked with an asterisk (*) was used in Southern immersion experiments to investigate the deletion that appears in this region (see results).

Other centers were used for DNA sequencing experiments. Sm: Narrow, 5a: Sac2, Rs: Rsal, Bg: Bgl2, Pv:

Pvu2, RI: EcoRI.

FIG. Figure 13 illustrates the DNA sequence obtained from pgC2Sal2.9 compared to the DNA sequence of the HSV1-gC region (59). The HSV1-gC region (HSV-1) and the sequence obtained from 15 pgC2Sal2.9 (HSV-2) were compared using the HOM program (57). Since various deletions were used to maximize sequence overlap, all positions, including spaces, have been numbered for clarity. Asterisks are placed over non-conforming nucleotides. The underlined "A" residue at position 43 of the HSV1 sequence is the approximate transcription start center of the gC mRNA. (59). "TATA" 1 and "TATA" 2 are the likely transcriptional control regions for the HSV1-gC mRNA * and 730 base 25 mRNA, respectively (59, 60). The inserted T residue at position 1728 in the HSV1 sequence was detected by repeated sequencing of this region (M. Jackson, unpublished) and was found to introduce an in-phase stop codon at positions 1735 - 1737 which was homologous to the stop codon for the 30 larger open HSV2 reading frame. The position of the 730 base mRNA start codon in HSV1 is shown at position 2032 - 2034 as is the position of another HSV2 start codon at position 1975 - 1977. 1 fig. 13, the illuminated derived sequence of HSV2 was compared to the DNA sequence of the gC genome region of HSV1 (59), which showed that the overall sequence homology between these two fragments was about 68%. However, certain areas of the sequence showed either a much higher or a lower degree of sequence homology than others. Eg. the 5 sequences between positions 0 and 570 in the HSV1 and HSV2 sequences showed only 51% homology, whereas the region between positions 570 and 1740 showed a much higher degree of sequence homology (80%). An additional hi-homologous region (70%) was also found at the end of the two sequences 10 from position 1975 to position 2419. In addition to the nucleotide sequence changes, the two genomes showed different deletions or insertions when compared to each other. Most notable was an 81 bp region at positions 346 - 426 in the HSV1-gC sequence, 15 missing in the HSV2 genome. From this overall sequence comparison, it appears that there was a high degree of sequence homology between the HSV1-gC region and the HSV2 region determined here.

Frink et al. (59) have found that the 5 'end of the 2520 base mRNA encoding HSV1-gC extends to the below dashed A residue at position 43 of FIG. 13. In addition, they designated an AT rich "TATA" box sequence (60) about 22 bp 5 'for this residue. Comparison of the two in FIG.

13, the HSV1 and HSV2 sequences 25 both contained the identical sequence CGGGTATAAA, in this region. This sequence is identical to that reported previously by Whitton et al. (61) found to occur at the "TATA" box regions in many of the HSV1 and HSV2 sequences so far determined.

This conserved sequence is also followed by a G-rich region in both viral genomes. In addition to this putative transcription control region, another "TATA" box was found in both sequences at positions 1845 - 1849 in FIG. 13th

This second "TATA" box has been proposed to control the transcript ion of a 730 base mRNA in the HSV1 genome (59).

54 GB 173597 B1 Both HSV1 and HSV2 contain this sequence surrounded by GC-rich flanking sequences, including a CGGGCG sequence similar to the CGGG sequence preceding the first "ΤΑΤΑ" box. In addition, both genomes encode 5 for open reading frames 3 'for these other "TATA" boxes, as will be discussed below.

To determine whether the above-described 61 bp deletion was truly present in the HSV2 genome, or whether it was one art product of cloning or sequencing, Southern immersion analysis of the HSV2 genomic DNA and the cloned HSV2 DNA was performed. A β-tag was prepared from a Sac2 fragment (see the fragment in Fig. 12) spanning the region lacking the 81 nucleotides. If the HSV2 genomic DNA is missing 81 bp in area, then a SmaI-BglII fragment spanning that region will be 576 bp, a SmaI fragment will be 662 bp, and a Sac2 fragment will be 195 bp.

FIG. 14 illustrates Southern immersion analysis of HSV2-genomic DNA and pgC2Sal2.9 ~ DNA. The area spanning 20 over the 81 bp range, which is missing in the one shown in FIG. 13 HSV2 sequence (HSV2 positions 346 - 426), was analyzed using the Sac2 fragment labeled with an asterisk in FIG. 12, which overlaps the deleted area. Lanes 1 - 3 are restriction digests of 25 HSV2 genomic DNA and lanes 4 - 6 are restriction zyme digests of pgC2Sal2.9. The digested DNAs were electrophoresed on 1.5% agarose gels, denatured, immersed on nitrocellulose and probed with the P-labeled Sac2 fragment. (The arrows show the position of the 564 bp Hindi 11 fragment of phage λ-ΟΝΑ.) Lanes 1, 6: Saml + Bgl2; lanes 2, 5: Narrow; lanes 3, 4: Sac2.

The 14 shows that the predicted 55 restriction centers surrounded the region lacking the 81 base pairs in both the HSV2 genomic ONA and the cloned HSV2 DNA. In addition, the HSV2 genomic fragments and the carbonaceous fragments exactly migrated together, proving that deletion is not an art product of cloning or sequencing.

Analysis of the larger open reading frame of the 2.9 kb HSV2-SalI fragment

Analysis of the potential coding sequences in the 2.9 kb. Sal I DNA-10 fragment of H5V2 revealed an open reading frame of 479 amino acids, beginning with the methionine encoded at position 199 - 201 of the 13 showed HSV2 sequence and ended with the TAA termination codon at position 1735 - 1737 of the HSV2 sequence in this figure. As can be seen from FIG. 13, both the HSV1-gC protein and the open HSV2 reading frame start at about the same position in the two sequences relative to the "TATA" box homologies. In addition, while it initially appeared that the open HSV2 read frame found 20 in this region terminated 12 codons before the HSV1 gC gene, then repeated sequencing of the carboxyl terminal region of the gC gene sequence (M. Jackson, unpublished) of HSV1 strain F revealed that Frink et al (59) reported sequence lacking a thymidine nucleotide 25 at position 1727, and insertion of this residue resulted in a translated HSV1-gC protein ending at the same site as the open HSV2 reading frame (1735 - 1737 Thus, considering the various deletions and inserts as illustrated in Figure 13, the HSV1-gC gene and the open HSV2 reading frame show a very high degree of overlap.

FIG. 15 illustrates translation of the large open HSV2 reading frame and comparison with the HSV1-gC amino acid sequence.

Oer was the single-letter amino acid symbols used. HSV-1. gC refers to the HSV1 gC sequence, and HSV-2 gF refers to the open HSV2 reading frame sequence. Proteins 5 were compared using the HOM program, which maximized homologies by inserting openings where needed (57). Stars are located above non-homologous amino acids. Presumed N-linked glycylation centers (NXS or NXT) (62) are grayed out and 10 cysteine residues (C) are framed. Only amino acids and not spaces are the number. Figure 15B illustrates the translation of the second open HSV2 reading frame and comparison with the 730 base mRNA HVS1 protein. HVS-2,730 bp ORF is the incomplete amino acid sequence of the second open HSV2 reading frame from position 1975 to position 2606 of the one shown in FIG. 13 shows the HVS2 sequence. HVS-1 730 bp ORF is the amino acid sequence derived for the protein encoded by the 730 base mRNA of HSV1 (59). Conservative amino acid changes in charge are labeled (C) and non-conservative changes in charge are labeled (N) in both Figs. 4A and 4B.

FIG. 15 illustrates the high degree of sequence homology between the HSV1-gC gene and the open HSV2 reading frame of 479 amino acids. The first 19 amino acids contain about 25 80% sequence homology where the changes in the first 25 amino acids are all conservative in charge.

From residue 124 in HSV1-gC (residue 90 in HSV2 sequence) to the end of both proteins, there is about 74% sequence homology, with 75% of the amino acid changes retaining 30 in terms of charge. Five putative N-seeded glycosylation centers (NXS or NXT (62)) are conserved between the two proteins, and all seven cysteine residues are located at homologous positions relative to the C-terminal end. In addition to the overall conservation of sequences in the 35 carboxy terminals le 3/4 of the proteins, there are also large regions of contiguous amino acid sequence homology with a length of up to 20 residues (i.e., positions 385 - 405 in the HSV1 sequence and 352 - 372 in the HSV2 sequence). It can be concluded from this sequence comparison that the 5 open reading frame in this region of the HVS2 genome encodes a protein homologous to HSV1-gC.

While the HSV2 β protein encoded by these regions shows a significant degree of sequence homology with the HSV1-gC-se sequence, there are several notable differences between the two sequences. The most striking frog is a deletion of 27 amino acids in the HSV2 sequence found in the HSV1-gC sequence from residue 50 to residue 76 (Fig. 15), which corresponds to the 81 bp deletion described above, in addition to this large deletion. both sequences show less deletions on one or two amino acids. All of these divisions are found in the amino-terminal regions of the proteins. In addition to these deletions, there are a large number of amino acid changes in the amino-terminal region of the proteins, which are clumped between residues 29 and 123 of the HSV1-gC-20 sequence (residues 31 - 90 of the HSV2 sequence). Only 30% of the amino acids in this region are homologous, and much of the homology is due to conserved proline residues. 43% of the amino acid replacements in this region are non-conservative in charge. The only other regions that showed such a large number of changes are a carboxyl-terminal hydrophonic domain (residues 476 - 496 in the HVS1 sequence and 443 - 463 in the HSV2 sequence), where the proteins are 55% homologous, but all the alterations are conservative, uncharged, hydrophobic amino acids, and the carboxyl ends of the proteins where the sequences are only 25% homologous but where the overall amino acid composition is similar (residues 500-512 in the HSV1 sequence and 467-479 in the HSV2 sequence). While 5 of the designed N-linked glycosylation centers are conserved between the two proteins, the HSV1-gC sequence contains two more centers than the HSV2 sequence (nine to seven in total). The HSV1-gC sequence contains 2N-linked glycosylation digests in the 27 amino acids deleted from HSV2 sequence 5 and an overlapping pair of centers between residues 109 and 112 of FIG. 15. The HSV2 sequence contains 2N-linked glycosylation centers not found in the HSV1 sequence, one of which is proximal to the amino terminus.

To more fully investigate the possible structural homologies between HSV1 and HSV2 sequences, a hydropathy analysis (31a) was performed. FIG. Figure 6 illustrates the hydropatiana light of the HSV1-gC protein and the protein encoded by the larger open HSV2 aL reading frame. The hydropathy of each protein was determined using the program 15 reported by Hopp and Woods (31a). Hydrophobic areas are above the midline and hydriphilic areas are below the midline. Stretches of 12 amino acids were analyzed and the average hydropathy calculated. Presumed asparagine-linked glycosylation centers (62) are labeled (0).

20 gC-1: HSV1-gC protein hydropathy. gC-2 (gF): hydropathy of the protein encoded in the larger open HSV2 reading frame.

FIG. 16 shows that both proteins exhibited an extremely high degree of structural homology based on amino acid synthesis. 25 the hydrophilic and hydrophobic properties of the quince. Each shows an N-terminal hydrophobic domain followed by a stretch of hydrphilic amino acids containing either 6 of a total of 9 (HSV1) or 3 of a total of 7 (HSV2) putative N-linked glycosylation centers. The peaks and valleys that follow this hydrophilic region are very similar in both proteins, including the hdyrophilic domain containing the final N-linked glycosylation center. The carboxyl ends of both proteins show a highly hydrophobic region of 20 residues followed by a hydrophilic carboxyl end. The 27 consecutive amino acids found exclusively in the HSV1-gC protein show you to encode a relatively hydrophilic region between residues 50 and 76 (Fig. 16). In conclusion, this analysis reveals that the hydropathy features of the HSV1-gC and HVS2 proteins are very similar and that the least conserved amino-terminal regions of the proteins fade into hydrophilic regions which have the ability to be highly glycosylated.

10 Analysis of the second open HSV2 relief frame

Translation of the last 431 base pairs of one in fig. 13, HSV2 sequence (residues 1975 - 2406) revealed another open reading frame of 105 amino acids. Whether the sequence information reported here is insufficient to contain the entire second open HSV2 read frame revealed a comparison of this sequence with the open read frame encoded by the 730 base HSV1 base mRNA, as reported by Frink et al. (10), also a high degree of sequence homology. As can be seen in FIG. 15B, the two sequences showed 75% sequence homology in the overlapping regions where about 90% of the amino acid changes were conservative in charge. The main fuselage between the two sequences was a 19 amino acid N-terminal region found in the HSV2 sequence, but not in the HSV1 sequence. Thus, although the function of the protein encoded in this region is unknown, the proteins of HSV1 and HVS2 showed a clear degree of sequence homology.

Discussion 50 The above results show that the HVS2 genome encodes a colinearly mapped homolog to HVS glycoprotein C. The colinearity of the sequences found here is corroborated by finding a sequence 3 'for the larger open HSV2 reading frame, which apparently encodes a homologous 5 to 730 base mRNA for HSV1 (10). Previous mapping of the HSV2 gF gene (33) together with the properties described herein for the larger open reading frame in this region of the HSV2 genome including several potential N-linked glycosylation centers and an apparent 10 amino-terminal signal sequence (5) as well as a putative carboxyl terminal transmembrane domain (28) allows the conclusion that the HSV2 protein described herein is the glycoprotein gF. In addition, the size of a translated HSV2 protein (about 52,000 daltons) is similar to that reported with the endoglycosidae H treated native HSV2-gF (54,000 daltons) (22d). Finally, the high degree of amino acid sequence homology as well as the preservation of several potential N-linked glycosylation centers and all 7 cysteine residues indicates a structural homology between HSV1-gC and HSV / 2-gF. Thus, these results strongly suggest that the HSV1-gC protein and HSV2-gF protein are homologous to each other.

These results help explain previous results which showed that 'H5V2-gF and HSV1-gC were essentially 25 type-specific, but that they had type-common determinants (17, 22d, 22f, 43). Among other early studies (17, 18, 43) showed that these proteins induced predominantly type-specific antibodies, it is reasonable that the most antigenic regions of proteins are found in the more 30 divergent N-terminal sequences following the putative hydrophobic signal sequences. The hydrophilic nature of the divergent regions together with their high content of potential N-linked glycosylation centers (62) suggests that these regions will be located on the surface of the protein.

Exposure of these divergent sequences to the exterior of the protein may be responsible for the production of type-specific antibodies directed against these divergent epitopes. However, type-dropped antibodies can probably also be produced by the more highly conserved carboxyl-terminal 3/4 of the proteins, since hydrophilic regions conserved between gC and gF could be exposed on the outside of the proteins and, in one case, may be glycosylated (the residues). 363 - 366 in HSV1-gC and 330 - 332 in HSV2-gF).

Thus, HSV1-gC and HSV2-gF have both type-specific and type-fall determinants, but it seems that the type-specific determinants are more antigenic.

Although an explanation of the type-specific and type-case determinants in gC and gF is unknown, it is possible that the proteins have at least 2 functions, one of which is important for the viability of both viruses, the type-case domain, and the other is specific to each virus type, the type-specific domain. Although the functions of gC and gF are currently unknown, and although viable gC mutants of HSV1 have been isolated in vitro (65), it is not clear that neither gC nor gF can be avoided by the viruses during in vivo infection of the human and the establishment of latency. It is possible that at least some of the biological differences between HSV1 and HSV2, including predisposition to infection site and virulence, may be due to the significant structural differences between the amino-terminal regions of gC and gF.

It can be concluded, even without any functional knowledge of these proteins, that different selection pressures must act on the divergent and conserved domains of gC and gF; Earlier sequence comparisons of the gO genes in HSV1 and HSV2 (58) showed that the amino-terminal signal sequence (63) and the carboxyl-terminal transmembrane domain (64) were able to tolerate a large number of mutations as long as the introduced amino acids. was hydrophobic. gC and gF sequence comparison shows a similar result in the carboxyl terminal putative transmembrane domain (64) from residues 476 - 496 in gC and 443 - 463 in gF.

The large number of heterologous hydrophobic substitutions in this region suggests that, as in gD, any amino acid that is lipid soluble can be tolerated in this region.

However, unlike gD, the amino-terminal signal sequences 10 in gC and gF are highly homologous in the first 19 residues. Thus, this region either has an important conserved function in addition to directing the glycoproteins into the crude endoplasmic reticulum (3), or there may be an overlapping gene or other functional sequence 15 in this region of the genome which must be preserved (866).

Although insufficient HSV2 sequence has been presented for a complete comparison, the region -3 'of the onset of HSV1-gC-roRNA transcription shows an identical sequence, CGGGTATAA, in both the HSV1 and HSV2 genome. In addition, both sequences of a G-rich region are followed immediately prior to the onset of transcription. Thus, as previously found for the gD genes in HSV1 and HSV2, upstream sequence homologies exist between the two virus types, suggesting the possibility that these regions are involved in transcriptional regulation of these genes.

Interestingly, the second MTATA "box homology, soro found in both viral genomes, and which probably controls the transcription of the 730 base mRNA (59, 60), also shows a relatively high degree of sequence homology in HSV1 30 and HSV2. these "TATA" boxes are CG-rich sequences that are similar, but not identical, to those preceded by the first "TATA" regions shown in FIG.

13, and they are both followed by a 14 bp range showing approx. £ 80 sequence homology. The entire region of homology that surrounds this region is only 33 bp with a total sequence homology of about 75 X. If this region is involved in transcriptional regulation of the 730 base mRNA, it appears that a relatively short sequence may be sufficient for recognition of transcriptional regulatory elements.

5 · The results show that HSV1-gC and HSV2-gF are highly homologous and that they encode type-common and type-specific domains. Since the two proteins exhibit substantial sequence homology and since they appear to be colinear on the map, we support the suggestion made by 10 Zeulak and Spear (22d) to change the name of HSV2-gF to HSV-gC or gC-2. In addition, the sequence determination data reported here opens the way for a functional analysis of the gC-1 and gC-2 proteins by mutually exchanging different type-specific regions between the two 13 proteins in vitro and expression of chimeric sequences in mammalian cells (67) or reincorporation. of these areas in the virus (68).

It is believed that the cloned gC-2 glycoproteins can be expressed and prepared for a vaccine in a manner analogous to that. are given in Example 1.

Furthermore, it is believed that a vaccine which includes a mixture of such recombinant gC and gD glycoproteins will be significantly more effective as a vaccine for HSV1 and HSV2 than one based on one of the 25 glycoproteins alone.

The references which are included in the subsequent bibliography and which are referred to by parentheses in the preceding text are considered to be incorporated in this description by this reference.

64 DK 173597 B1

Library F in 1. Emtage et al., Hature 283, 171 (1980); Davis et al., Proc.

Natl. Acad. Sci. (USA) 78, 5376 (1981); Tøeiland et al., Mature 292, 851 (1981).

2. Kupper et al., Nature 289, 555 (1981); Kleid. et al., Science 214, 1125 (1981).

3. Charnay et al., Nucleic Acids Research 7. » 335 (1979);

Valenzuela et al., Nature 298, 347 (1982).

4. Rose et al., Proc. Natl. Acadj.Sci. (USA) 78, 6670 (1981).

5. Yelverton, et al., Science 219, 614 (1983).

6. Watson et al., Science 218, 381 (1982).

7. Gething et al., Nature 293, 620 (1981); Liu et al., DNA 2 »213 (1982); Goodenow et al., Science 215, 677 (1982); Goodenow et al., Nature 300, 231 (1982); Crowley et al., Kolec. and Cell. Biol. 3y 44 (1983).

8. Rose et al., Cell 30, 753 (1982).

9. Spear, P.G., (1930), Herpesviruses, p709-750, in H.A. Blough and J.M. Tiffaney (ed)., Cell Hembranes and Viral Envelopes, Vol.

2nd, Academic Press, Inc., New York.

10. Balachandran et al., J. Virol. £ 4, 344 (1982).

11. Norrild, Curr. Top. Microbiol Immunol. 90, 67 (1930).

12. Roizman, Cell l6, 481 (1979).

65 DK 173597 B1 13. Baucke et al., J. Virol. 32, 779 (1979).

14. Cohen et al., 0. Virol. 27, 172.

15. Eberle et al., J. Virol: 36, 665 (1980).

16. Norrild et al., J. Virol. 26, 712 (1978).

17. Powell et al., Mature 249, 360 (1974).

18. Eberle et al., Infect, ϊπκηυη. 31, 1062 (1981).

19. Pereira et al., Infect. Immune. £ 9, 724.

20. Sim, C., et al., 0. Gen. Virpl. 19, 217 (1973).

21. Showalter et al., Infect. Immune. 34, 684 (1981).

22a. Eisenberg et al., 0. Virol. 41, 1099.

22b. Balachandran et al., J. Virol. 39, 438 (198 ().

22c. Para et al., I.e., Virol. 41, 137 (1982).

22d. Zezulak. one al., 0. Virol. 47, 553 (1983).

22f. Zweig et al., I.e., Virol. 47, 185 (1983).

23. Anderson et al., 0. Virol. 30, 805 (1979).

24. Tea et al., I.e., Virol. 43, 41 (1982).

25. Murray et al., Mol. Genet. , 150 (53).

26. Benton et al., Science J96, 180 (1977).

66 DK 173597 B1 27. Southern, J. Hol. Biol. 98, 503 (1975).

28. Vieira et al., Gene 19, 259 (1982).

29. Messing et al., Nuc. Acid. Res. Jl, 309 (1981).

30. Sanger et al., Proc. Natl. Acad. Sci. (USA) 74, 5436 (1977).

31. Atlas of Protein Sequence and Structure V.5, Supplement 2, 1976, M.O. Dayhoff, ed., The Biochemical Research Foundation, Spring, Maryland, p 311.

31a. Jump et al., Proc. Natl. Acad. Sci. (USA) 78, 3324 (1981).

32. Watson et al., Nucl. Acid. Res. 11, 1507 (1983).

33. Biobel, Proc. Natl. Acad. Sci. (USA) 77, 1746 (1980).

34. Rose. et al., Proc. Natl. Acad. Sci. (USA) 77, 3884 (1980).

35. Ruyechan et al., 0. Virol. 29 ', 677 (1979); Roizman, Cel! 26, 481 (1979).

36. Simonsen et al., Proc. Natl. Acad. Sci. (USA) 80, 2495 (1983).

37. Lusky. et al., Nature 293, 79 (19S1).

38. Nunberg et al., Cell 19, 355 (1980).

39. Urlaiib et al., Proc. Natl. Acad. Sci, (USA) 2Ζ. » 4216 (1930).

40. Graham et al., Virol. 52, 456, (1973).

40a. Kessler, J. Immuno. 115, 1617 (1975).

67 DK 173597 B1 41. Showalter et al., Infect, and Immun. 34, 684 (1981); Monoclonal anti-gO antibodies, 1-S and S5rS were kindly provided by Or.

Martin Zweig of the Laboratory of Molecular Oncology, National Cancer Institute, Frederick, Maryland 21701.

42. Cohen, et al., 0. Virol. 36, 429 (1980).

43. Pereira et al., Proc. Hatl. Acad. Sci. (USA) 78, 5202 (1981). 44. Cohen et al., J. Virol. 27, 172 (1973).

45. Laemmli, Nature 227, 680 (1970).

46. Honess. et al., 0. Virol. 16, 1308 (1975).

47. Spear, J. Virol. 17, 991 (1976).

48. 'Campadel 1 i-Fiuaie et al., 0. Viro). 43, 1061 (1982); Johnson et al., Cell 32, 987 (1933); Cohen et al., 0. Virol. 46, 679 (1983).

49. Bloch, J. Cell. Biel. 82, 629. (1979).

50. Human herpetic serum titrated against HSV-1 and HSV-2 by hemaglutination inhibition and complement staining tests was kindly provided by Dr.

John A. Stewart For the Centers for Disease Control, Atlanta, Georgia.

51. Rector, et al., Infect, and Immun. 38, 168 (1982).

52. Kennett in Monoclonal Antibodies, K. Kerrett, T. McKearn, and B. Bechtel, eds. (Plenum Press, K.Y., 1980), pp. 376-377.

53. Fiers et al., Nature 273, 113 (1978); Gluzman, Cell 23, 275 (1981).

68 DK 173597 B1 54. Lee et al., Nature 294, 223 (1981); Kaufman et al., Mol. and Cell. Biol. 2, 1304 (1983); Kaufman et al., J. Mol. Biol. 159, 601 (1982).

55. Kleid et al., Science 214, 1125 (1981).

56. Maxam et al., Methods Enzymol. 65, 499 (1980).

57. Dayhoff, M., Ed. Atlas of Protein Sequence and Structure, Vol.

5, Supplement 2, National Biochemical Research Foundation, '

Silver Spring, Maryland, p. 311 (1976).

58. Lasky et al., ONA, in press (1984).

59. Frink et al., J. Virol. 45, 634 (1983).

60. McKnight et al., Science 217, 316 (1982).

61. Whitton et al., Nucl. Acids Res. 13, 6271 (1983).

62. Hubbard et al., Ann. Rev. Biochem. 50, 555 (1981).

63. Blobel, Proc. Natl. Acad. Sci. USA 77, 1491 (1980).

64. Sabatini et al., 0. Cell. Biol. 92, 1 (1982).

65. Cassai et al., Intervirology, 212 (1975).

66. Hall et al., 0. Virol. 43, 594 (1982).

Claims (9)

69 DK 173597 B1 A method for producing a membrane-free truncate of a membrane-bound polypeptide, which truncate lacks a membrane-binding domain, whereby the polypeptide truncate is free of this membrane, and which truncate has exposed antigenic determinants capable of inducing neutralizing antibodies by in vivo exposure to a pathogen, characterized by expressing DNA encoding the truncate in a stable eukaryotic cell line transfected with the DNA. The method of claim 1, further comprising recovering the truncated polypeptide as a secretion product. 15 The method of claim 1 or claim 2, comprising the preliminary steps of producing DNA encoding the membrane-bound polypeptide but lacking coding for membrane-binding domain, incorporating the DNA into an expression vector, and transfecting the eukaryotic host cell with vector. The method of any one of claims 1, 2 or 3, wherein the transfected host cell is a mammalian cell line. 25 The method of claim 4, wherein the cell line is deficient in the production of dhfr and the vector contains a selectable dhfr marker. The method of any one of claims 1 to 5, wherein the membrane-free truncate is a glycoprotein D truncate of a herpes simplex virus type 1 or type 2 and the pathogen is herpes simplex virus type 1 and / or type 2. 70 DK 173597 B1 The method of any one of claims 1 to 5, wherein the membrane-free truncate is a truncate of a glycoprotein C from a herpes simplex virus type 1 or type 2 and the pathogen is herpes simplex virus type 1 and / or type 2. The method of claim 6, wherein the membrane-free truncate comprises the N-terminal region of gD polypeptide up to about amino acid residue 300. A method according to any one of the preceding claims, further comprising formulating a vaccine with the polypeptide.