WO1990006771A1

WO1990006771A1 - Proteolytic fragments obtained from human cytomegalovirus glycoprotein complex i

Info

Publication number: WO1990006771A1
Application number: PCT/US1989/005254
Authority: WO
Inventors: Richard C. Gehrz; Bruce E. Kari
Original assignee: The Children's Hospital, Incorporated
Priority date: 1988-12-12
Filing date: 1989-11-15
Publication date: 1990-06-28
Also published as: IL92447A0; AU4664689A; CA2003943A1

Abstract

The present invention provides antigenic and immunogenic peptides prepared by the proteolysis of human cytomegalovirus (HCMV) glycoprotein complex I (gc-I) which peptides can be used as the active ingredient in anti-HCMV vaccines, to detect anti-HCMV antibodies or to proliferate HCMV-specific T cells.

Description

PROTEOLYTIC FRAGMENTS OBTAINED FROM HUMAN

CYTOMEGALOVIRUS GLYCOPROTEIN COMPLEX I

Background of the Invention

Human cytomegalovirus contains several disulfidelinked glycoprotein complexes (Britt, Virology, 135, 369 (1984); Rasmussen et al., J. Virol., 55, 274 (1985);

Farrar and Greenway, J. Gen. Virol., 67, 1469 (1986);

Kari et al., J. Virol., 60. 345 (1986); Gretch et al., J. Virol., 62, 875 (1988)). One of these complexes contains a glycoprotein related to. HSV glycoprotein gB (Cranage et al., EMBO J, 5, 3057 (1986)). This complex has been designated p130/55 or gC-I (Rasmussen et al., ibid., 1985; Lussenhop et al., Virology, 164, 362, (1988)) and is the most well characterized complex of those found in HCMV. Complex gC-I isolated from whole cells or

extracellular virus has been reported to contain at least three glycoproteins with molecular weights of 160- 130,000, 92-93,000 and 50-55,000 (Britt, ibid., 1984); Farrer and Greenway, ibid., Kari et al., ibid., 1986;

Rasmussen, Virology, 163. 308 (1988)). These

glycoproteins are held together by disulfide bonds and contain largely N-linked oligosaccharides. Complex gC-I can be immunoprecipitated by human serum positive for HCMV and will also stimulate human T cells (Liu et al., J. Virol., 62, 1066, (1988)).

Several murine monoclonal antibodies which recognize gC-I have been disclosed. Based on a simultaneous two-antibody-binding assay, these antibodies could be grouped into three domains (Lussenhop et al., ibid., (1988)).

Epitopes in domains I and II are recognized by antibodies which can neutralize Towne strain HCMV in vitro. Antibodies within a single domain were effective at

inhibiting each other's binding. Furthermore, as

determined by an ELISA assay, antibodies in domain I were found to augment the binding of antibodies in domain II and vice versa. This augmenting activity was also observed in a plaque reduction assay (Lussenhop et al, ibid., (1988)). Only one antibody (11B4) was placed in a third domain (domain III) based on its unique ability to inhibit the binding of all other antibodies tested.

Antibody 11B4 was also observed to be non-neutralizing and had the ability to inhibit the neutralizing activity of other antibodies in a plaque-reduction assay. These gC-I specific murine monoclonal antibodies were found to recognize several clinical isolates of HCMV as determined by immunofluoresence. This suggested that the epitopes recognized by these antibodies are conserved (Lussenhop et al., ibid., (1988)).

Since all the murine monoclonal antibodies which we originally characterized showed some type of interaction, i.e., augmentation or inhibition, it seemed possible that the three domains containing the epitopes recognized by these antibodies might be physically close on gC-I. Proteolysis has been one approach used to determine if epitopes are physically close or distant on viral

glycoproteins. For example, proteolysis has been used to determine the location of antigenic determinants on glycoproteins from tick-borne encephalitis virus (Heinz et al., Virology. 130, 485 (1983)), murine leukemia virus (Pinter et al., Virology. 116, 499 (1982)) and glycoprotein D from HSV (Eisenberg et al., J. Virol., 41, 478 (1982)).

Summary of the Invention

The present invention provides antigenic and immunogenie proteolytic fragments of human cytomegalovirus (HCMV) gC-I which contain all three immunoreactive domains of gC-I. Using proteases such as trypsin or chymotrypsin, a set of fragments could be generated from Towne strain HCMV which were recognized by antibodies from all three domains. These fragments were also recognized by several human sera positive for HCMV. Trypsin and chymotrypsin fragments were also examined for T cell reactivity. Lymphocyte proliferative responses were detected with fragments made with either enzyme in the case of individuals who reacted strongly with whole gC-I, but higher responses were consistently obtained with trypsin fragments. Cells obtained from an

individual who had low responses to whole gC-I had low responses to the fragments as well. A T cell clone from this individual was observed to react with trypsin fragments, but not chymotrypsin fragments.

More specifically, the HCMV envelope glycoprotein complex gC-I was digested with chymotrypsin and fragments immunoaffinity purified. Two major glycosylated peptides were obtained which had molecular weights (MWs) of 34,000 and 43,000 under non-reducing conditions. After

reduction, one major glycosylated fragment, with a MW of 34,000, was observed in addition to at least two other peptides with MWs of 30,000 and 28,000. Under non- reducing conditions, monoclonal antibodies (MoAbs) which bind to all three domains of gC-I immunoprecipitated the

34,000 and 43,000 MW fragments and reacted with them in Western blot. After reduction, the MoAb assigned to domain III and one MoAb in domain I were non-reactive while all other MoAbs reacted with the 34,000, 30,000 and 28,000 MW peptides in Western blot.

Five positive human sera were reactive with whole purified gC-I in Western blot under non-reducing conditions. After reduction, nine MoAbs and one serum reacted with 130,000 and 52,000 MW proteins and four sera reacted with these plus a 93,000 MW protein. The five sera were also reactive with the chymotrypsin fragments in Western blot under non-reducing conditions. After reduction, four sera reacted strongly with the 34,000, 30,000 and 28,000 MW peptides and one reacted weakly. Thus, the gC-I domains recognized by murine MoAbs may be important in human immune recognition of HCMV, and all of the immunogenic gC-I subunit peptides described hereinabove are within the scope of the present invention.

The present invention is also directed to a method comprising increasing the proliferative response of T cells derived from an individual seropositive for HCMV exposure comprising contacting said cells in vitro with an effective amount of these reduced or unreduced gC-I proteolytic fragments.

This method is useful in methods to clone and expand populations of HCMV antigen-specific T helper lymphocytes (T_h cells) or T cytotoxic lymphocytes (T_c cells). These populations can be used as therapeutic agents to boost the titer of T_h or T_c cells which recognize HCMV antigens in individuals at risk for, or afflicted with, HCMV.

It is also possible that these reduced or unreduced HCMV gC-I fragments may be useful to increase the

proliferative response of T cells in vivo in an

individual seropositive for HCMV. An effective amount of one or more of these peptides would be administered to an individual, e.g., parenternally, in combination with a pharmaceutically acceptable carrier.

The reduced or unreduced gC-I fragments disclosed herein can also provide the basis for an HCMV subunit vaccine of high immunogenicity. Such a vaccine can be employed to raise the serum titer of anti-HCMV

antibodies, to provide a protective effect against HCMV in susceptible individuals. General methods of

preparation of anti-HCMV vaccines using gC-I ("gA") glycoproteins are disclosed in Pereira (U.S. Patent No. 4,689,225).

The vaccines of the present invention will

incorporate one or more of the gC-I peptide forms, and may include other immunogenic substances as desired to provide immunity against other infectious agents. The gC-I peptides and polypeptides will normally be

incorporated in a physiologically-acceptable liquid medium, such as water, saline, phosphate-buffered saline, and may be administered intravenously, intramuscularly, subcutaneously, or intraarterially. Often, an adjuvant will be incorporated into the vaccine to provide for release of the gC-I over a prolonged period. Suitable adjuvants include various inorganic gels, such as alum, aluminum hydroxide, aluminum phosphate, and the like.

The amount of immunogenic peptide employed per dose will generally be in the range from about 0.5 to 150 g in a liquid volume of about 0.25 to 2 ml. The vaccines may be administered repeatedly at from 2 to 4 week intervals, usually for a total of 2 to 4 times.

Inoculation with the vaccine of the present

invention is particularly suitable for individuals about to undergo immunosuppression, such as those undergoing skin grafting, bone marrow or organ transplantation.

Such individuals, if they lack immunity (antibodies), are at great risk to HCMV infection. The vaccine is also of great benefit to women of childbearing years, preferably at or before the early stages of pregnancy. The primary threat from cytomegalovirus infection arises from

transmission or reactivation of HCMV during the gestation period. Thus, if the infective cycle of HCMV can be inhibited by stimulation of the mother's immune system, the likelihood of infection of the infant can be reduced or eliminated.

Additionally, the present gC-I peptide fragments can be immobilized, e.g., on a flexible substrate or

microtiter well, and used to assay a liquid sample, such as a physiological fluid, for the presence of anti-HCMV antibodies. For example, it was demonstrated by Western blot analysis that human anti-HCMV antibodies recognize (bind to) epitopes on the present peptides. After a suitable incubation period, the anti-HCMV antibodies bound to the peptide-coated substrate can be detected by a suitable direct or indirect method. In a direct assay, a second antibody, comprising a detectable label, is used which binds to the first, bound anti-HCMV antibody, and the amount of bound label is quantified. In an indirect assay, an exogeneous portion of an anti-HCMV antibody comprising a detectable label is added to, and allowed to react with, the peptide (antigen) -anti-HCMV antibody complex. The amount of labelled antibody which is able to bind to the immobilized HCMV antigen is inversely proportional to the amount of anti-HCMV antibody in the fluid under assay.

Detailed Description of the Invention The invention will be further described by reference to the following detailed Example.

EXAMPLE I.

A. Materials and Methods

1. Preparation of Towne and AD169 HCMV strains. Both strains of HCMV were grown on human skin fibroblasts, harvested and purified as described (Kari et al., J.

Virol., 60, 345 (1986)). Virus was labeled by growing it with either [³H]arginine, [¹⁴C]GlcN, or [³H]G1cN

(Amersham).

Generation, characterization and purification of

monoclonal antibodies. Monoclonal antibodies were generated and characterized as previously described

(Lussenhop et al., Virology, 164, 362 (1988)) and in application Serial No. 144,760, filed January 19, 1988, the disclosures of which are incorporated by reference herein. Monoclonal antibodies were purified from mouse ascities fluid by high performance liquid chromatography using a hydroxyapatite column (Bio Rad) as previously described (Juarez-Salinas et al., Biotechniques, 2, 164 (1984)). The properties of representative monoclonal antibodies (MoAbs) reactive with the three gC-I domains is summarized in Table I, below. Table I

Monoclonal Antibodies Reactive with gC-I Domains

Augments

Binding

Neutralizes to gC-I IVI IVI

Reactive With Towne Strain of MCA AccesDeposit

MoAb Domain HCMV^a to Domain sion No. Date^b

41C2 I + II 10119 10 Sept. 198626B11 I + II - - - - - - - - - -

9B7 II + I 10117 10 Sept. 1986

34G7 II + I 10142 7 Aug. 1987 11B4 III - Inhibits 10181 31 July 1988

binding and

neutralization

a With complement.

b In Vitro International, Inc., Linthicum, MD 21090 USA. The deposits were made in accord with MPEP 608.01(p) (C) (5th ed. 1983, rev'd Dec. 1987).

Proteolysis. Either purified Towne strain HCMV or whole cells at 7 to 14 days post infection were solubilized with 1.0% NP-40 in 50 mM Tris buffer (pH 7.4) containing 150 mM NaCl. Insoluble material was removed by

centrifugation at 16,000 x g for 30 min. The extract was collected and protein content determined using the BCA protein assay (Pierce). The extract was digested with either TPCK-trypsin (Worthington) or TLCK-chymotrypsin (Sigma) at an enzyme-to-protein ratio of 1:50 for 24 to 48 hours at room temperature. Proteolysis was terminated with either enzyme by addition of phenylmethyl-sulfonyl fluoride (PMSF). Extracts were also digested with pronase (Sigma) using the same conditions. Pronase digestion was stopped by addition of BSA.

Purification of whole gC-I or its proteolytic fragments. Whole gC-I or its fragments were isolated by a

modification of an immunoaffinity method using

biotinylated monoclonal antibodies and streptavidin agarose (Gretch et al., Anal. Biochem., 163 , 270 (1987)). Briefly, a biotinylated monoclonal antibody was added to 1.0% NP-40 extracts containing whole gC-I or its

fragments. The antibody was incubated with the extracts for 30 min. before adding streptavidin agarose. This mixture was allowed to react for an additional 45 min. with constant mixing. The agarose beads were pelleted by centrifugation and then washed twice with PBS containing 0.1% NP-40 and twice with PBS. Proteins and peptides were eluted from the bound antibody by heating at 100°C in a Tris buffer (0.2 M Tris, pH 6.8, containing 4% SDS) for 3 min.

SDS-PAGE and Fluorography. Radioactively labeled

glycoproteins or glycopeptides which had been

immunoprecipitated were separated by SDS-PAGE in 5-15% polyacrylamide gradient gels. Radioactivity in these gels was detected by fluorography using Enhance (New England Nuclear).

Western blot analysis. A mini-gel apparatus (Bio Rad) was used for Western blot. Glycoproteins were separated in straight 10% polyacrylamide gels. Undigested

glycoproteins were electroblotted onto nitrocellulose membrane with a .45 micron pore size. In order to retain the small glycopeptides obtained by proteolysis, it was necessary to use a nitrocellulose membrane with a 0.2 micron pore size. After electroblotting, the paper was blocked with 3% gelatin in Tris buffered saline (TBS, 20 mM Tris, 500 mM NaCl, pH 7.5) Monoclonal antibodies (MoAbs) in asities fluid were diluted 1/500 and human serum was diluted 1/30 in 1% gelatin in TBS. Diluted

MoAbs or human serum was allowed to bind to blotted paper overnight at room termperature. The paper was washed with TBS containing 0.05% Tween 20. Phosphate labeled goat anti-mouse IgG or goat anti-human IgG (Kirkegaard and Perry), diluted 1/1000 with 1% gelatin in TBS, was added and allowed to react at room temperature for one hour. The paper was washed and the substrate 5-bromo-4-chloro-3-indolyl phosphate/ tetrazolium in 0.1 M Tris buffer (Kirkegaard and Perry) was added. After

visualization of bands, the reaction was stopped by immersing the paper in water.

Deglvcosylation of glycopeptides. Chymotrypsin fragments were digested with the enzyme N-glycanase (Genzyme Corp., Boston, MA) which hydrolyzes asparagine-linked oligosaccharides from glycopeptides to give free oligosaccharide and peptide-containing aspartic acid at the glycosylation site. Fragments obtained with chymotrypsin were eluted from streptavadin agarose affinity columns with 1.0% SDS (w/v). Eluted fragments were reduced with beta-mercaptoethanol and diluted with phosphate-buffered saline (pH 8.6) containing 1.0% NP-40 so that the final SDS

concentration was 0.1%. The protease inhibitor 1,10-phenanthroline hydrate was added according to the manufacturer's instructions. The reaction was done at room temperature for 24 hours with constant mixing. At the end of this time, SDS was added to bring the SDS concentration back to 1.0% and the reaction mixture was dialyzed against 0.1% SDS overnight. Samples were concentrated prior to SDS-PAGE.

Lymphocyte proliferation assays. Lymphocyte

proliferation assays were performed as previously described (Liu et al., 1988). Briefly, proteolytic fragments used for T cell analysis were extensively dialyzed against PBS to remove toxic substances. A dialysis membrane with a molecular weight cut off of 10- 12,000 was used. Because of this, fragments with molecular weights of less than 12,000 were lost. The T cell clone used in this study was generated and

characterized as previously described in Liu et al. J. Virol., 62, 1066 (1988), the disclosure of which is incorporated by reference herein.

Results

Trypsin and chymotrypsin digestion of gC-I. These studies were done to determine which enzyme would most effectively degrade gC-I to smaller fragments. To determine the end point of proteolysis, a kinetic study was done. An NP-40 detergent extract was obtained from infected whole cells which were collected from culture media at 7 and 11 days post infection by low speed centrifugation. Proteins were labeled with [³H] arginine and glycoproteins with [¹⁴C]GlcN. A portion of the extract was immunoprecipitated immediately with a gC-I specific monoclonal antibody (41C2, domain I) to

establish the nature of the starting material. The remaining extract was divided into several equal

aliquotes. These were subjected to proteolysis using either TPCK-trypsin or TLCK-chymotrypsin. Proteolysis was stopped at 0.5, 2, and 24 hours by addition of PMSF and the gC-I fragments were immunoprecipitated with 41C2. In addition, an aliquot was allowed to remain at room temperature for 24 hours without exposure to proteolysis. Proteins and glycoproteins immunoprecipitated with 41C2 were examined by SDS-PAGE with and without reduction of disulfide bonds.

The gC-I complexes immunoprecipitated by 41C2 from the initial extract or from the same extract after 24 hours at room temperature were similar if not identical regardless of the label used. This demonstrated that gC- I in the absence of trypsin or chymotrypsin was stable in the extract for at least 24 hours. In the absence of proteolysis, complexes typical of gC-I were obtained which had molecular weights from 130,000 to greater than 200,000. The most abundant glycoproteins obtained from these complexes after reduction had molecular weights of 130,000, 93,000 and 50-52,000 regardless of the

radioactive label used. In addition, two [³H]arginine- labeled peptides were detected when gC-I was examined without reduction which had molecular weights of 35,000 and 20,000. These peptides were present regardless of proteolysis and were only clearly detected with

[³H] arginine, suggesting that they contained little or no carbohydrate.

Between 30 min. and 24 hrs. of proteolysis with both enzymes, complexes with molecular weights of 106,000 to 130,000 were detected with either radioactive label.

When these complexes were reduced, glycoproteins with molecular weights ranging from 35,000 to 93,000 were most abundant. It appears that the degradation of complexes in the 106,000 to 130,000 molecular weight range is slow compared to the initial proteolysis of intact gC-I. After 24 hours of proteolysis with chymotrypsin, one major [³H]arginine-labeled peptide was detected under non-reducing conditions, as well as less abundant peptides with molecular weights of between 43,000 and 34,000 and one at 20,000. With [¹⁴C]GlcN labeling, only the peptides between 43,000 and 34,000 could be clearly detected.

After reduction of disulfide bonds, these chymotrypsin fragments generated one major 34,000 molecular weight glycopeptide, but a number of less abundant lower

molecular weight peptides were also observed.

Similar, but not identical, results were obtained with trypsin. There was a major glycosylated fragment with a molecular weight of 44,000. However, some of the 106,000 molecular weight glycopeptide remained after 24 hours and there was a lack of glycosylated peptides having molecular weights between 44,000-34,000.

After reduction of peptides obtained with trypsin, two major glycosylated peptides with molecular weights of 47,000 and 35,000 were detected. In addition, a smear of glycosylated material was detected above the 47,000 molecular weight glycopeptide. A number of minor lower molecular weight peptides similar to those detected with chymotrypsin were also present. Finally, the banding pattern observed with either enzyme was not changed by extending the reaction time to 48 hrs or by adding additional enzyme.

Immunoprecipitation of chymotrypsin. trypsin and pronase fragments with antibodies from domains I. II and III.

For these experiments, gC-I was obtained from purified extracellular virus. Immunoprecipitations were done with monoclonal antibodies selected from domains I (39E11), II (34G7), and III (11B4) to determine if the proteolytic fragments contained epitopes from all three domains.

Other antibodies identified as representative of domains I and II can also be used (see U.S. Patent application Serial No. 144,760, filed January 19, 1988, the

disclosure of which is incorporated by reference herein.) The chymotrypsin and trypsin glycopeptide fragments obtained from gC-I isolated from extracellular virus were the same as those immunoprecipitated when gC-I was obtained from infected cells. In addition, antibodies from all three domains were capable of

immunoprecipitating the same fragments, suggesting that they contained the three domains previously described in application Serial No. 144,760, filed January 19, 1988. A similar experiment was done with pronase (a non- specific protease). Pronase fragments were similar to those obtained with chymotrypsin. However, with pronase, little of the 43,000 molecular weight peptides remained and most of the glycopeptides formed a smear with a molecular weight from 30-34,000. This further demonstrates the resistance of this portion of gC-I to proteolysis. Western blot analysis of whole gC-I and chymotrypsin fragments of gC-I reacted with monoclonal antibodies and human serum positive for HCMV. Western blot analysis was done for three reasons. First, with immunoprecipitation, several peptides were detected regardless of reduction of disulfide bonds and it was of interest to determine which of these were recognized by the monoclonal antibodies. Secondly, we wished to extend our study to additional gC-I-specific monoclonal antibodies to see if they would all react with the fragments generated. In this regard, we focused on chymotrypsin fragments since they were less heterogeneous and smaller in molecular weight. Thirdly, it was of interest to determine whether or not human antibodies would recognize the same fragments as the murine antibodies. Six human serum were selected for this study. Five were determined to be positive and one negative for HCMV by complement fixation and indirect immunofluoresence assays (Table II).

Table II

Sera CMV Titers* HSV Titers**

A,1 <4/<10 (-) <4 (-)

A,2 32/40 (+) <4 (-)

A,3 256/160 (+) 16 (+)

A,4 32/80 (+) ELISA (-) I.1 128/640 (+) <4 (-)

I.2 32/160 (+) <4 (-)

* CMV titers were determined by a latex aggultination assay (LA) or immunofluoresences (IF). The first number represents the LA titer and the second the IF titer.

** HSV titers were determined by immunofluoresence.

In addition, since gC-I contains glycoproteins which have homology with gB from HSV, these sera were tested for their reactivity with HSV by immunofluoresence. Of those tested, only one (A, 3) was positive for HSV (Table II). However, the reactivity of this serum was similar to the others. For Western blot analysis, both whole gC-I and chymotrypsin fragments were immunoaffinity purified with a gC-I specific monoclonal antibody. Because of this, a small amount of monoclonal antibody sometimes

contaminated the preparations and could be detected in Western blot as a small band at the top of the

nitrocellulose membrane or as heavy and light chains.

Under non-reducing conditions, two monoclonal antibodies (11B4 and 26B11) failed to react with whole gC-I in Western blot. This occurred even though these

antibodies were capable of immunoprecipitating gC-I. The pattern obtained with all other monoclonal antibodies was very similar to that obtained with the positive human serum, including reactivity with the 34,000 and 20,000 molecular weight peptides which were also observed by immunoprecipitation of whole gC-I. After reduction of disulfide bonds, monoclonal antibodies 11B4 and 26B11 still failed to react. All other monoclonal antibodies reacted strongly with the 130,000 and 52,000 molecular weight proteins and weakly with proteins slightly below the 130,000 molecular weight protein and a 50,000

molecular weight protein. Again, the human sera positive for HCMV reacted with these proteins. However, while human serum A, 2 gave almost identical results to the monoclonal antibodies, the other positive sera reacted in varying degrees with the 93,000 molecular weight protein. In addition, the human sera reacted weakly with a 26,000 molecular weight protein which was not recognized by the monoclonal antibodies. None of the proteins recognized by the human sera or the monoclonal antibodies were reactive with the negative human serum (A,l) or with the negative murine antibody control (SP2).

Examination of chymotrypsin fragments under non- reducing conditions showed that all monoclonal antibodies reacted with 43,000 and 34,000 molecular weight peptides. These included 11B4 and 26B11, which failed to react with whole gC-I under non-reducing conditions. The reactivity of 11B4 and 26B11 was less than that of the other monoclonal antibodies. There were weaker reactions with peptides having an apparent molecular weight of 63,000.

This may represent a small amount of incompletely

digested peptide. These peptides appear to represent only a small portion of the protein present since they could not be detected by Coomassie blue staining while the 43,000 and 34,000 molecular weight peptides were clearly visible. Again, the negative murine antibody control was not reactive with the peptides recognized by the monoclonal antibodies. Under non-reducing

conditions, the pattern obtained with all positive human serum was identical to the monoclonal antibodies, and the negative human sera (A,1) was non-reactive. After reduction of disulfide bonds, monoclonal antibodies 11B4 and 26B11 were not reactive with the

chymotrypsin fragments of gC-I. All other monoclonal antibodies reacted strongly with the 34,000 and 30,000 molecular weight peptides and, with the exception of 41C2, reacted strongly with a 28,000 molecular weight peptide. Of the human positive sera tested, four (A, 2 A,3, I,1 and I,2) reacted strongly with the 34,000,

30,000 and 28,000 molecular weight peptides. The human positive serum A,4 was only weakly reactive after

reduction of disulfide bonds. There were also weaker bands detected at an apparent molecular weight of 50,000 with the monoclonal antibodies, human sera and the negative controls. One of these was probably the heavy chain contributed by the antibody used to initially purify the fragments since this band was detected in the SP2 negative control. Nonetheless, since these bands were detected by the human negative sera, it is likely that reactivity with them in Western blot was not

specific. Finally, the reactivity of the human sera with either whole gC-I glycoproteins or chymotrypsin fragments of gC-I did not appear to depend on their HCMV titer since a serum with a titer of 32/40 (sera A, 2) reacted as well as sera with titers greater than 100 (sera A, 3 1,1 and I,2).

Finally, for comparison, gC-I from HCMV strain AD169 was also digested with chymotrypsin, fragments of

immunoaffinity purified and examined by Western blot under non-reducing conditions to determine whether or not all three domains would be present. When this was done, all monoclonal antibodies reacted with the fragments in Western blot, demonstrating the presence of these domains in AD169. However, the pattern was slightly different. A weak band was detected at 63,000 molecular weight, but the lower molecular weight peptides appeared more diffuse than they did with peptides from Towne strain HCMV. A broad band covering molecular weights from 40,000 to 35,000 was detected along with a band at 31,000 and another diffuse band with a molecular weight of 23,000. Deglvcosylation of chymotrypsin fragments. After

reduction of chymotrypsin fragments, the monoclonal antibodies recognized three peptides in Western blot. The heterogenity in molecular weight may have been due to the extent of glycosylation. To examine this

possibility, glycopeptides obtained with chymotrypsin were reduced and digested with N-glycanase.

Glycopeptides labeled with either [³H]arginine or

[¹⁴C]GlcN were used. Since N-glycanase action generates peptides free of carbohydrate, all label should have been lost in the [¹⁴C]GlcN labeled peptides, which provided a way to assay the extent of carbohydrate removal. After digestion with N-glycanase, the major 34,000 molecular weight [³H] arginine labeled peptide was lost and peptides with molecular weights of 30,000 and 28,000 were

observed. Comparable bands were not detected when

[^1AC]GlcN labeled peptides were digested, suggesting that carbohydrate removal was complete.

T cell recognition of' unreduced trypsin and chymotrypsin fragments. Peripheral blood mononuclear cells (PBMC) were obtained from three seropositive donors. As we previously showed, cells from these individuals were reactive with whole gC-I (Liu et al., J. Virol., 62, 1066 (1988)). Cells from individuals which were seronegative to HCMV did not proliferate when exposed to purified gC-I. Proliferative responses were obtained with PBMCs obtained from seropositive individuals with either trypsin or chymotrypsin fragments (Table III). Table III

A6 (-) A4 (+) A5 (+) Al (+)

MNC* 2,359 1,303 60 314

MNC+HCMV** 232 42,481 9,020 81,689

MNC Neg Con*** 1,111 7,039 231 1,272

MNC Try Frag.**** 1,099 74,642 586 84,327

MNC Chy. Frag.***** 654 22,224 208 6,260

* MNC Mononuclear cell culture only.

** MNC plus whole virus.

*** MNC plus negative control containing buffer but no virus or fragment.

**** MNC plus trypsin fragments.

***** MNC plus chymotrypsin fragments.

Table III summarizes the proliferative responses of mononuclear cells (MNC) from HCMV negative (-) and HCMV positive (+) individuals. Cultures containing 10⁶ cells per well were exposed to antigens for six days prior to being pulsed with [³H]thymidine for 24 hours before harvesting cultures. Numbers represent average counts per min. from triplicate cultures.

However, with the same amount of protein present, responses were greater with trypsin fragments than with chymotrypsin fragments. This suggested that some T cell epitopes may have been lost by the more complete proteolysis when chymotrypsin was used. In addition, one individual which had a low response to whole gC-I (A5) was observed to have very low or no response to fragments obtained with either enzyme.

However, a T cell clone from this individual was observed to have a very strong response to the trypsin fragments, but not to the chymotrypsin fragments. This further demonstrated the loss of T cell epitopes from the chymotrypsin fragments and suggested that the number of T cells recognizing the epitopes on the trypsin fragments was low in the population from which this clone was obtained. Discussion

A small portion of gC-I has been demonstrated to contain the three domains which were detected by

Lussenhop et al., ibid., (1988), using a simultaneous two-antibody-binding assay. While both trypsin- and chymotrypsin-generated fragments were recognized by antibodies from all three domains, chymotrypsin was capable of generating the smallest fragments. Because of this, further studies were focused on the chymotrypsin fragments. After proteolysis of [³H] arginine labelled gC- I with chymotrypsin, there appeared to be only one major peptide present under non-reducing conditions which had a molecular weight of 43,000. With Coomassie blue

staining, the 43,000 molecular weight peptide was still most abundant, but a 34,000 molecular weight peptide could be more clearly detected. Both of these unreduced peptides were glycosylated as determined by incorporation of [^1AC]GlcN and were recognized by all monoclonal

antibodies and human sera positive for HCMV. There are probably other B cell epitopes in gC-I which are bound by human antibodies, e.g., those which appear to be on the 93,000 molecular weight glycoprotein. The chymotrypsin fragments also appear to contain sites for neutralization since all antibodies in domains I and II neutralized

Towne strain HCMV either individually or in combination. However, with all monoclonal antibodies, complement was required for neutralization (Lussenhop et al. ibid., (1988)). The chymotrypsin fragments also contained T helper cell epitopes; however, more of these appear to remain in the trypsin fragments. That T cell epitopes were deleted by chymotrypsin was clearly demonstrated by the positive reactivity of a T cell clone with the

trypsin fragments, but not with chymotrypsin fragments. Furthermore, the individual from which the T cell clone was obtained showed a very low response to either trypsin or chymotrypsin fragments when mononuclear cell cultures were used. Thus, the population of T cells present in the mononuclear cell cultures from this individual recognizing the T cell epitope present in the trypsin fragments must have been low.

Another interesting observation was that the

chymotrypsin fragments from gC-I also contained the epitope recognized by antibody 11B4 which was placed in domain III. This was established by the ability of 11B4 to immunoprecipitate the chymotrypsin fragments, and its reactivity in Western blot under non-reducing conditions. Antibody 11B4 is a non-neutralizing antibody which blocks binding of neutralizing antibodies. From this

perspective, antibodies directed toward this epitope would have benefit for the virus, but not the host. It would seem that the chymotrypsin fragments could be used as part of a subunit vaccine, although inclusion of the epitope recognized by 11B4 should be avoided in any attempt to make a synthetic vaccine. However, 11B4 appears to recognize a conformational epitope since its reactivity with the chymotrypsin fragments was lost after reduction of disulfide bonds. Unlike 11B4, those

antibodies which neutralized Towne strain HCMV and the human serum were reactive after reduction of disulfide bonds. Furthermore, after reduction, three peptides were detected by these monoclonal antibodies and human serum. One of the major differences between these three peptides was the extent of glycosylation. The 34,000 molecular weight peptide was most heavily glycosylated, but could be reduced to at least 30,000 molecular weight by action of N-glycanase. The presence of polysaccharides on the 34,000 molecular weight peptide could have some impact on antibody binding as has been demonstrated with other glycoproteins (Alexander and Elder, Science, 226, 1328 (1984); Caust et al., Arch. Virol., 96, 123 (1987)).

However, preliminary results (data not shown) show that removal of carbohydrate did not prevent binding of neutralizing monoclonal antibodies and in some cases binding appeared to increase. Moreover, the 28,000 and 30,000 molecular weight peptides were under- or non- glycosylated as compared to the 34,000 molecular weight peptide, but they still bound antibody in Western blot. Thus, a synthetic vaccine might include only the linear amino acid sequences of the peptides in the chymotrypsin fragments. Such subunit vaccines are within the scope of this invention.

Several observations suggest that the fragments of gC-I obtained by digestion with chymotrypsin are

important to the function of gC-I. First, the monoclonal antibodies

reactive with these fragments also recognize several laboratory and wild type strains of HCMV (Lussenhop et al., ibid., (1988)). In fact, conformational and non- conformational epitopes were shown to be present in HCMV strains AD169 and Towne. These results suggest that the epitopes are conserved. Secondly, a blocking epitope in domain III was present in these fragments. This epitope may help survival of the virus by preventing other antibodies from binding to important sites for gC-I function. Thirdly, the gC-I chymotrypsin fragments were glycosylated. One important function of glycosylation is to prevent the action of proteases (Iwase, Int. J.

Biochem., 20, 479 (1988)). In our experiments, even pronase (a non-specific protease) was prevented from generating fragments smaller than those obtained with chymotrypsin. These results suggest that this part of gC-I is well protected from proteolysis. Finally, the maintenance of the higher order structure of the

chymotrypsin fragments seems to be independent of the rest of the gC-I molecule and resistant to denaturation with detergents. This indicates a very stable structure. CITED LITERATURE

The disclosures of the papers cited below are

incorporated by reference herein.

I. S. Alexander et al., Science, 226, 1328-1330 (1984). 2. J. B. Britt et al., Virology, 135. 369-378 (1984).

3. J. Caust et al., Arch. Virol., 96. 123-134 (1987).

4. M. Cranage et al., EMBO J., 5, 3057-3063 (1986).

5. R. J. Eisenberg et al., J. Virol., 41, 478-488

(1982).

6. G. H. Farrar et al., J. Gen. Virol., 67, 1469-1473

(1986).

7. D. R. Gretch et al., Anal. Biochem., 163, 270-277 (1987).

8. D. R. Gretch et al., J. Virol., 62. 875-881 (1988). 9. F. X. Heinz et al., Virology, 130, 485-501 (1983). 10. H. Iwase, Int. J. Biochem., 20, 479-491 (1988).

11. H. Juarez-Salinas, Biotechniques, 2, 164-169 (1984). 12. B. Kari et al., J. Virol., 60, 345-352 (1986).

13. Y.-N. Liu, J. Virol., 62, 1066-1070 (1988).

14. N. O. Lussenhop et al., Virology, 164, 362-372

(1988).

15. A. Pinter et al., Virology, 116, 499-516 (1982).

16. L. Rasmussen et al., Virology, 163, 308-318 (1988a). 17. L. Rasmussen et al., J. Virol., 55, 274-280 (1985b).

The invention has been described with reference to various specific and preferred embodiments and

techniques. However, it should be understood that many variations and modifications may be made while remaining within the spirit and scope of the invention.

Claims

WHAT IS CLAIMED IS:

1. An immunogenic peptide which is prepared by a

process comprising:

(a) digesting Towne strain human cytomegalovirus (HCMV) envelope glycoprotein complex I (gC-I) with chymotrypsin to yield a glycosylated peptide fragment having a molecular weight of about 34,000 or 43,000.

2. The immunogenic peptide of claim 1 wherein the

peptide fragment reacts with monoclonal antibodies reactive with antigenic domains I, II and III of HCMV gC-I.

3. An immunogenic peptide which is prepared by a

process comprising reducing the disulfide bonds of the peptide fragment of claim 1 to yield a reduced peptide having a molecular weight of about 28,000, 30,000 or 34,000.

4. The immunogenic peptide of claim 1 wherein the

reduced peptide reacts with a monoclonal antibody reactive with domains I and II but not with domain III of HCMV gC-I.

5. The immunogenic peptide of claims 1 or 2 which has a molecular weight of about 34,000.

6. The immunogenic peptide of claim 2 which has a

molecular weight of about 28,000 or 30,000.

7. An immunogenic peptide which is prepared by a

process comprising:

(a) digesting Towne strain HCMV envelope

glycoprotein complex I (gC-I) with trypsin to yield a glycosylated peptide fragment having a molecular weight of about 44,000.

8. An immunogenic peptide which is prepared by a process comprising reducing the disulfide bonds of the peptide fragment of claim 7 to yield a reduced peptide having a molecular weight of about 35,000 or 47,000.

9. A method comprising increasing the proliferative

response of T cells derived from an individual seropositive for HCMV comprising contacting said cells in vitro with an effective amount of one or more of the glycoproteins of claims 7 or 8.

10. A method comprising increasing the proliferative

response of T cells in an individual seropositive for HCMV by administering to said individual an effective amount of one or more of the glycoproteins of claims 2 or 8.

11. A vaccine against HCMV comprising an effective

immunogenic amount of the peptide of claims 1, 3, 7 or 8 in combination with a physiologically- acceptable liquid carrier.

12. The vaccine of claim 11 comprising the peptide of claim 6.

13. A method comprising detecting human anti-HCMV antibodies in a physiological sample comprising

contacting said sample with an immobilized

glycoprotein of claims 1, 3, 5 or 6 so that said anti-HCMV antibodies bond to said glycoprotein, and detecting (a) the amount of bound anti-HCMV

antibodies or (b) the amount of unbound immobilized glycoprotein.

14. The method of claim 13 wherein the sample is blood serum.