WO1989001628A1

WO1989001628A1 - Monoclonal antibodies reactive with cytomegalovirus glycoprotein a

Info

Publication number: WO1989001628A1
Application number: PCT/US1988/002644
Authority: WO
Inventors: Nancy O. Lussenhop; Bruce E. Kari; Richard C. Gehrz
Original assignee: The Children's Hospital, Incorporated
Priority date: 1987-08-07
Filing date: 1988-08-03
Publication date: 1989-02-23
Also published as: AU2324988A; JPH02504551A; EP0376973A1

Abstract

A composition capable of neutralizing the infectivity of human cytomegalovirus (HCMV) is provided comprising a mixture of (a) a first monoclonal antibody which binds to a first epitope on HCMV GPA, wherein said first monoclonal antibody neutralizes HCMV infectivity in vitro in the presence of complement; and (b) a second monoclonal antibody which binds to a second epitope on HCMV GPA, wherein said second monoclonal antibody does not neutralize HCMV infectivity in vitro in the presence of complement; wherein the neutralizing activity of the mixture is greater than the neutralizing activity of the first monoclonal antibody, and wherein said composition is optionally free of significant amounts of a monoclonal antibody which binds to a third epitope on HCMVGPA, and which interferes with the binding of the first and the second monoclonal antibodies.

Description

MONOCLONAL ANTIBODIES REACTIVE WITH CYTOMEGALOVIRUS GLYCOPROTEIN A

Cross-Reference to Related Application This application is a continuation-in-part of

U.S. patent application Serial No. 83,502, filed August 7, 1987, and a continuation-in-part of U.S. patent application Serial No. 83,818, filed August 10, 1987, which is a continuation of U.S. patent applica- tion Serial No. 83,502, filed August 7, 1987, which is a continuation-in-part of U.S. patent application Serial No. 933,789, filed November 24, 1986.

Background of the Invention Infection with human cytomegalovirus (HCMV) is usually asymptomatic and self-limiting in the normal human population. However, it can cause serious infec¬ tion if contracted i utero, leading to severe neuro¬ logical handicaps and/or deafness in a significant number of cases. Moreover, HCMV is among the most common causes of morbidity and mortality in immuno- suppressed patients and patients with immune deficiency diseases.

HCMV is a member of the family herpesvirdae, and is composed of a nuclear complex of nucleic acid and proteins surrounded by an external membrane envelope containing glycopeptides and glycolipids. Among the antigenically distinct envelope glycoproteins characterized by our laboratory and others, several are associated by disulfide bonds in high molecular weight complexes. Commonly assigned U.S. patent application Serial No. 933,789, filed November 24, 1986, entitled, "Immunogenic Glycopeptides of Human Cytomegalovirus" describes the isolation and characterization of anti- genically related glycopeptides of molecular weights 9X,000 daltons and 50,000-52,000 altons. These gly¬ coproteins-were obtained by the reduction of a set of glycoprotein complexes having molecular weights of _ 45.0,000 daltons and 130,000-180,000 daltons. Moreover, two? prototype murine monoclonal antibodies (mcAb) dis¬ closed in the above patent, 41C2 and 9B7, immunopreci- pcit^'ate these individual glycoproteins, as well as the antrgenically-related glycoprotein complexes. McAb 9B7 also exhibits neutralizing activity against HCMV ^n_ vitro in the presence of complement. Convalescent sera from patients who have recovered from HCMV infection also contain antibodies that immunoprecipitate these HCMV glycoproteins, suggesting that antibodies reactive with these glycoproteins may participate in neutraliza- tion of virus iji vivo.

Therefore, there is a need for HCMV-specific cAbs that have practical clinical utility, in the diagnosis and treatment of HCMV infections. A further needs exists for mcAbs that are useful both in direct detection of HCMV in clinical specimens and rapid identification of the virus in tissue culture. Such antibodies may be used either to capture and concen¬ trate a viral antigen, or alternatively, for immuno- logical detection of the virus using a variety of known immunoassay techniques. There is recent evidence suggesting that the administration of hyperimmune HCMV globulin to immunosuppressed patients prior to organ or bone marrow transplantation may possibly attenuate or prevent life-threatening HCMV infections in the post- transplant period. However, there is a continuing need for improved^" therapeutic methods to prevent, attenuate or cure HCMV infections in both healthy immuno- sjuppressed patients or in patients who already have clinical illness due to HCMV. Brief Description of the Invention The present invention provides 13 related murine mcAbs which can be generated by the immunization of mice with whole Towne HCMV virions and/or detergent extracted envelope glycoproteins of Towne strain HCMV. The spleen cells of the immunized mice are then fused with a murine myeloma cell line to generate antibody- producing murine B-cell hybridomas. The antibodies produced by these murine hybridomas immunoprecipitate glycoproteins of molecular weights 130,000, 93,000, and 50,000-52,000 daltons. These glycoproteins are asso¬ ciated with one or more high molecular weight glycopro¬ tein complexes in the viral envelope. These glycopro¬ tein complexes have been referred to as the gCI family by D. R. Gretch et al., 3. Virol., _62, 875 (1988). For example, the 130kD glycoprotein complex disclosed in patent application Serial No. 933,789 is referred to ^? herein as human cytomegalovirus glycoprotein A (HCMV GCA). Competitive enzyme-linked immunosorbent

(ELISA) assays were performed to^' study the relation¬ ships among the antigenic sites recognized by the various mcAbs. Surprisingly, either inhibition or augmentation in the binding of the antibodies to the virus was observed when a second antibody was present simultaneously. The mcAbs were assigned to at least 7 groups which recognize distinct antigenic sites within 3 domains of HCMV GCA.

The ability of these mcAbs to neutralize HCMV Towne strain in the presence of complement was tested in a plaque reduction assay. Certain of these mcAbs (i.e., 9B7) were highly neutralizing (2 μg/ml for 50% plaque reduction), whereas other mcAbs (i.e., 41C2) appeared to be non-neutralizing in the presence of complement. Binding of certain of the neutralizing mcAbs in the ELISA assay was increased two- to four-fold in the presence of a second non-neutralizing antibody. Moreover, the neutralizing activity of the combination of neutralizing and non-neutralizing antibodies was unexpectedly increased by 20-fold compared to that of the neutralizing antibody alone. Even combining two non-neutralizing mcAbs resulted in significant neutral¬ izing activity in certain instances, and such mixtures are also within the scope of the present invention.

Therefore, the present invention is directed to generation of an HCMV-specific murine mcAb produced by a process comprising: a. immunizing a mouse with a composition co pris- ing HCMV virions or purified HCMV envelope glycopeptides derived from HCMV GCA having molecular weights of about 130,000, 93,000, or 50,000-52,000 daltonsj

b. fusing spleen cells from said mouse with cells from a murine myeloma line to produce hybri¬ domas; and

c. selecting and clonally expanding a hybridoma which produces a mcAb reactive with HCMV GCA glycoproteins of molecular weights 130,000, 93,000, and 50,000-52,000 daltons, but not reactive with herpes simplex, adenovirus or varicella zoster virus.

The mcAbs of the present invention were characterized with respect to their specificity, binding affinity, and their ability to neutralize HCMV in the presence of complement, either alone or in combination with other murine mcAbs -of the invention which are also obtained from the hybridomas of step (c) above. The present invention also provides an effec¬ tive method for detecting HCMV in a sample of physio¬ logical material infected with HCMV. This method com¬ prises: a. reacting said physiological material with a first monoclonal antibody which binds to a first epitope of HCMV GPA to form a complex between said HCMV GPA and said first mono¬ clonal antibody; and

b. reacting said complex with a second monoclonal antibody to a second epitope of HCMV GPA, wherein the second monoclonal antibody enhances the binding of the first monoclonal antibody to the first epitope, and wherein said second monoclonal antibody comprises a detectable label or a binding site for said

I label, to yield a ternary complex; and

c. detecting the presence of said label, or alternatively, reacting said binding site with a detectable label prior to said detection. The concentration of said label provides a measure of the concentration of HCMV GPA pre- sent in said physiological material.

Preferably, the detectable label is a radio¬ active isotope or an enzyme which has been chemically linked to the second antibody. Preferably, the binding site for said label is provided by a molecule capable of binding a^* free enzyme or a radioisotope, e.g., biotin or deferoxamine. These molecules can be chemi¬ cally linked to mcAbs by methods known to the art. For example, see U.S. Patent No. 4,680,338. The enzyme preferably is linked to a substance which binds strongly to the binding site, for example, an avidin- enzyme conjugate will bind to biotin. Steps (a) and (b) can be performed simultaneously or the order of addition of the first and second mcAb can be inverted. Therefore, the pr.esent invention also provides a^"method for detecting HCMV in a sample of physiologi¬ cal material infected with HCMV comprising: a. reacting said physiological material with a first monoclonal antibody to a first epitope of HCMV GPA, said first monoclonal antibody

^■ comprises a detectable label or a binding site for said label, to yield a complex between HCMV GPA and said first monoclonal antibody;

b. reacting said complex with a second monoclonal antibody which binds to a second epitope of HCMV GPA, wherein the second monoclonal anti¬ body enhances the binding of the first mono¬ clonal antibody to the first epitope, to form a ternary complex; and

c. detecting the presence of said label, or alternatively, reacting said binding site with a detectable label prior to said detection. The concentration of said label provides a measure of the concentration of HCMV GPA pre¬ sent in said physiological material.

The- ability to neutralize HCMV infectivity _iι vitro with a composition which contains a high titer of monoclonal antibodies to HCMV GCA can provide the basis for assays of the effectiveness of some prototype HCMV vaccines, and will also be useful in basic investiga¬ tions into the mechanisms of HCMV infection, e.g., with respect to the effect of mutations on the products of viral gene expression. - 7-

The present invention is also directed to a method for raising plasma immunogenicity to HCMV GPA by administering a pharmaceutical unit dosage form com¬ prising one or more of the present antibodies to a patient who has been exposed to HCMV. It is believed that the antibodies of the present invention will have therapeutic potential, either as a prophylactic agent to prolong the life expectancy of infected and/or diseased patients by retarding the clinical progression of the disease, or possibly, as a curative agent which can act to eliminate the virulence of the virus. The present antibodies might also represent a valuable adjunct to chemical antiviral agents.

Therefore, the present invention is also directed to a composition consisting essentially of a mixture of:

(a) a first monoclonal antibody, e.g., 9B7, which binds to a first epitope on HCMV GPA, wherein said first monoclonal anti¬ body neutralizes HCMV infectivity in vitro in the presence of complement; and

(b) a second monoclonal antibody, e.g., 41C2, which binds to a second epitope on HCMV

GPA, wherein said second monoclonal anti¬ body does not neutralize HCMV infectivity in the presence of complement; wherein - the. neutralizing activity of the mixture is greater than the neutralizing activity of the first monoclonal antibody.

HCMV GCA-specific monoclonal antibodies were also isolated and characterized which possessed the unexpected ability to substantially inhibit or com¬ pletely block the ability of these therapeutically- useful first or second classes of HCMV GCA-specific mcAbs. Therefore, it is highly preferred that the HCMV GCA-neutralizing composition of the invention be pre¬ pared to be free of these interfering mcAbs, i.e., the composition should be free of a monoclonal antibody which binds to a third epitope on HCMV GPA, does not neutralize HCMV infectivity in the presence or absence of complement, and which inhibits the binding of the first antibody and the second antibody. Specific examples of such mcAbs are exemplified hereinbelow.

The isolated mcAbs preferably are diluted with a pharmaceutically-acceptable liquid carrier, such as an aqueous IV fluid, prior to being assayed for bioac- tivity or administered as a unit dosage form _i vivo. See Remington's Pharmaceutical Sciences', A. Osol", ed., ^'Mack Pub. Co., Easton, PA tl6th ed. 1980) at pages 1488-1496, the disclosure of which is incorporated by reference herein. The resultant solution is steril¬ ized, e.g., by filtration. Preservatives commonly employed with IgG preparations, such as maltose, gly- cine or thimerosal, may be added in pharmaceutically- acceptable amounts. The resulting solutions are preferably admin¬ istered parenterally, e.g., by intravenous infusion or injection. The amount of mcAb composition administered will vary widely, and will depend on the physique and physical condition of the im unosuppressed or HCMV- infected patient. Such factors are necessarily empiri¬ cal, and can^'be determined by the clinician, employing known HCMV staging criteria. In some clinical situa¬ tions, it may be necessary to administer a plurality of doses of the mcAb composition, in order to neutralize the infectivity of viral particles as they are released from infected target cells. Brief Description of the Drawings Figure 1 depicts immunoprecipitates formed by reacting unreduced (UR) or reduced (R) HCMV GCA pro¬ teins with GCA-specific monoclonal antibodies (mcAbs) and convalescent sera.

Figure 2 depicts a Western Blot of the HCMV glycoproteins recognized by GCA-specific mcAbs.

Figure 3 (panel A) is a diagrammatic represen¬ tation of the epitopes recognized by the GCA-specific mcAbs. Inhibition is represented by overlapping solid lines, while augmentation and the direction of augmen¬ tation are indicated by arrows. Domains I, II and III are shown to the right and left. Panel B schematically depicts a possible arrangement of the epitopes on the HCMV GCA peptide backbone.

Figure 4 is a graphic depiction of plaque reduction assays using Towne strain HCMV and combina¬ tions of antibodies.

Detailed Description of the Invention

Materials and Methods

A. Preparation of Viruses HCMV Towne strain and AD169 strain were grown with or without ³H-glucosamine on human skin fibroblast cultures, harvested and purified on sucrose gradients as described previously (Kari et ^al*» J- Virol., 60, 345-352 (1986)). The purified virus was resuspended in Tris-NaCl buffer (50 M Tris hydrochloride, 150 mM NaCl pH 7.4), and extracted with 1% Nonidet P-40

(NP-40, Sigma Chem. Co., St. Louis, MO) in Tris-NaCl buffer (50 M Tris hydrochloride, 10 M NaCl, 2 mM phenylmethyl sulfonylfluoride pH 7.5) as described by Kari et al., in _ι. Virol. , 60, 345-352 (1986). Uninfected skin fibroblasts were extracted in a similar fashion for use as negative controls. Reduc¬ tion and alkylation of HCMV Towne NP-40 crude extracted material was performed as described by Kari et al., J. Virol., 6Q_, 345-352 (1986). All detergent extracted viral or control fibroblast materials were passed over an Extracti-Gel D column and eluted with water to remove the detergent.

10

B. Generation of Murine Monoclonal Antibodies to

HCMV The production of mouse hybridomas secreting mcAbs to HCMV proteins was performed as previously described (Kari et al., 3^_

15 Virol., £0, 345-352 (1986)). The antibodies described in this patent application involved three separate fusion experiments using either AD169 or Towne strain purified HCMV virions. Balb/C mice were immunized for 2, 5 or 10

20 months. Spleen cells from immunized mice were fused with SP2-2-Agl4 myeloma cells (American Tissue Culture Collection) using polyethylene glycol as the fusing agent.

25 Resulting hybrid cells were screened for specific antibody production to HCMV using an enzyme-linked immunosorbent assay (ELISA). (See Section G below.) Antigens used in the ELISA assay were either purified HCMV Towne

30 Strain or AD169 strain whole virions or HCMV

Towne NP-40 extracted material or NP-40 extracted material from uninfected skin fibro- blasts. Ascites fluids from expanded clones were purified for IgG using high performance

35 hydroxyapatite chromatography (HPLC) (Juarez-

Salinas et al., Biotechniques, 2 , 164 (1984)). The titer of fractions collected from the column was measured with respect to HCMV- specific activity using the ELISA assay. The protein content of the fractions was deter- mined using the BioRad protein assay (BioRad) and purified mouse IgG as a standard. Only fractions with the highest titers from each run were used for subsequent experiments. The hybridomas 41C2 and 9B7 consistently produce 5-10 mg of monoclonal antibody per 1 ml of ascites.

The clonality of the mcAbs was established by twice subcloning of the hybridomas producing each antibody. Furthermore, each antibody had a unique isotype (i.e., 41C2-IgGl; 9B7 IgG-2b), and each appears as an individual antibody by isoelectric focusing.

C. Immunoprecipitation Monoclonal antibodies were allowed to imrnunoprecipitate ³H-glucos- amine-labelled HCMV Towne NP-40 extracted pro¬ teins which were unreduced or reduced and alkylated prior to immunoprecipitation. Proteins were solubilized in sodium dodecyl sulfate (SDS) and separated by SDS polyacryl- amide gel electrophoresis. The tritium- labelled glycoprotein bands in the gel were identified by fluorography as described (Kari et al., J. Virol., 60, 345-352 (1986)).

D. Western Blot For Western Blot assays, puri¬ fied HCMV Towne strain whole virus was solu¬ bilized with SDS and separated by 5-15% gradient polyacrylamide gel electrophoresis. The proteins on the gel were subsequently electroblotted onto nitrocellulose paper with a BioRad transblot -apparatus. The paper was blocked with 3% gelatin in Tris buffered saline (TBS, 20 mM Tris, 500 mM NaCl pH 7.5).

The mcAbs in ascites fluid were diluted 1/500 in 1% gelatin in TBS, and allowed to bind to the paper overnight at room temperature. The paper was washed with PBS-0.05% Tween 20 (polysorbate 20), and alkaline phosphatase- labelled goat anti-mouse IgG (KPL) diluted 1/2000 with 1% gelatin in TBS was added and allowed to incubate for one hour at room tem¬ perature. The paper was washed once again and the substrate 5-bromo-4-chloro-3-indolyl- phosphate/nitroblue tetrazolium in 0.1 M Tris buffer solution (KPL) was added. Reaction of the antigen-antibody complex with the sub¬ strate resulted in the formation of an insol- uble purple product. The reaction was stopped by immersing the paper in water.

E. Simultaneous Two Antibody Binding Assay HCMV antigens were coated onto 96 well microtiter Immulon II plates by incubating 200 ng of total protein in 50 μl of 0.05 M carbonate buffer (pH 9.5) per well for 18 hours at 4°C. The wells were then washed three times with pho"sphafee-buffered saline with 0.05% Tween 20 (PB$T), once with distilled water, and then were air dried. The plates were stored at 4°C with dessicant until use. The antigens used included purified HCMV Towne strain whole virion, unreduced and reduced proteins extracted from skin fibroblasts infected with - 13-

HCMV Towne strain for 8-12 days (See Section A, preparation of viruses), or NP-40 extracted proteins derived from uninfected skin fibro¬ blasts (See Section A, preparation of 5 viruses).

The mcAbs in ascites fluids were purified by HPLC and biotinylated using biotin N- hydroxysuccinimide ester. Twenty-five μl of a

10 fixed amount of biotinylated antibody and 25 μl of PBST or 25 μl containing varying amounts (50 μg to 10 pg) of a second unlabelled mcAb were added simultaneously to the antigen- coated wells and incubated for 90 minutes at

15 room temperature. The wells were washed 3 times with PBST and 50 μl/well of peroxidase- labelled streptavidiπ (KPL) diluted 1/2000 in , PBST was added and the mixture "incubated for 90 minutes at room temperature. After washing

20 • 3 times with PBST, the substrate o-phenylene- diamine (OPD) and H2O2 in citrate buffer was added and the reaction was stopped after 15 minutes with 2.5 N H2SO₄. Optical denisty (O.D.) of the product was read with a Dynatech

25 plate reader at 490 nm. The biotinylated antibody was titered in 10-fold dilutions beforehand to determine the amount that would give an O.D. of >0.5 and <1.5 in the absence of a second antibody.

30

The ^"unlabelled second antibody was shown to be active by titering against HCMV whole virus antigen. The amount of binding was quanti- tated by adding peroxidase-labelled goat anti-

35 mouse IgG and the substrate OPD. The assay was performed in duplicates and the data were compiled using the following formula:

Percent of augmentation or inhibition =

(OD¹*⁹⁰ in presence - (OD^1*90 in absence of 2nd antibody) of 2nd antibody) x 100

(OD*⁹.⁰ in absence of 2nd antibody)

F. Neutralization Plaque Reduction Assay Mono¬ clonal antibodies purified by HPLC and which titered positive against Towne strain virus in an ELISA assay (Section B, above) were tested for neutralizing activity aainst the same virus in a plaque reduction assay as described by Kari et al., supra. About 1,000-2,000 PFU of Towne strain virus were added to a single antibody at a protein content of 100 μg to 0.001 μg along with 2% guinea pig complement (Pel-Freez Biologicals, Rogers, AK) in a total volume of 0.05 μl of Dulbecco Modified Eagle Medium (DMEM). When two antibodies were used, an additional 10 μg of the second antibody was added to the test sample. Virus neutraliza- tion was expressed as % plaque reduction using the formula:

Percent plaque reduction =

^* (Average PFU with - (Average PFU no antibody) with antibody) x 100

(Average PFU with no antibody)

G. ELISA Detection of HCMV in^* Supernatants of Tissue Cultures Inoculated with Clinical

Specimens Monolayers of human skin fibro¬ blasts in either 24 well plates or shell vials - 15-

are inoculated with clinical specimens (urine or saliva) obtained from patients suspected to have HCMV infection. Uninoculated skin fibro¬ blasts are processed in the same fashion as a 5 negative control.

Five days after inoculation, the tissue culture media fluid is removed from the fibro- blast monolayer by vacuum suction. Two

10 hundred μl phosphate buffered saline (PBS) are added to each well and completely removed by suction. One hundred μl of solubilizing PBS-1% NP-40 are added to each well for five minutes. The bottom of each well is then

15 scraped with a pipette and cells are resus¬ pended thoroughly by repeated pipetting. This suspension of lysed cells is employed as the "patient sample" in Se'ction H, below.

20 H. ELISA Assay for Detection of HCMV in Tissue

Culture Supernatants Microtiter wells in 96- well microtiter plates are pre-coated with HCMV GCA-specific mcAb 41C2 (IVI-10119) as the capture antibody. Included in each plate are

25 a blank well, a well for an uninfected fibro- blast control and an HCMV positive control well in addition to wells for patient samples. After coating with the capture antibody, the micTotit-er wells are washed once with buffer

30 (PBS-0.05% Tween). Sample solutions are solu- bilized in PBS-1% NP-40 in each tissue culture well, and are then dispensed into individual microtiter wells. Blank, negative skin fibro- blast and HCMV positive control samples are

35 included with each clinical sample determina¬ tion. -16-

Five μl of biotinylated GCA-specific mcAb 9B7 (IVI-10117), which recognizes a different epi¬ tope on GCA than that recognized by 41C2, are then added to each microtiter well as a detec- 5 tion antibody. The microtiter plate is then placed in a plastic bag and incubated at 37°C for 1 hour. Wells are then washed 4 times with wash buffer (PBS-0.05% Tween), 100 μl of streptavidin-peroxidase is added and the 10 plate is incubated for 30 minutes at 37°C in a plastic bag. Wells are then washed 4 times with wash buffer and then one time with dis¬ tilled water. Ortho-pheylenediamine (OPD) substrate (100 μl) is added and reacted for 15 15 minutes at room temperature. The reaction is then stopped with 25 μl of 5 N H2SO . After ^" 10 minutes, the microtiter plate is read with a Dynatech plate reader at 490 nm or by visual colorometric reading. This method has been 20. shown to detect HCMV at a total protein con¬ centration of 10 nanograms.

Results

A. Immunoprecipitation of HCMV Envelope Glycopro- 25 teins by Murine Monoclonal Antibodies

1. Methodology

Purified ³H-glucosamine-labelled HCMV Towne strain virions were extracted with - 1% NP-40. The unreduced material, or 0 material reduced with dithiothreitol and

^"alkylated with diiodoacetamide, was immunoprecipitated with 13 murine mcAbs. After immunoprecipitation, all samples were further reduced with β-mercaptoetha- 5 nol and separated by SDS-PAGE. Fluoro¬ graphy showed that all the mcAbs precipi¬ tated antigenically related qlycooroteins - 17-

with molecular weights of 130,000, 93,000, and 50,000-52,000 daltons, whether or not the extract had been reduced prior to immunoprecipitation 5 (Representative results are shown in

Figure 1, where SP2 = control mcAb.).

Thus, the number and relative amounts of the individual proteins immunoprecipi-

10 tated were not affected by reduction prior to immunoprecipitation. Minor variations in the molecular weights of the species of glycoproteins precipitated between 50-52,000 may reflect slight

15 differences in charge and/or conformation following immunoprecipitation with mcAbs recognizing different epitopes on the same glycoproteins. The mcAbs also recognized the 52,000 dalton glycoprotein

20 of HCMV GCA by Western Blot, whereas the higher molecular weight species were not consistently identified by this method (Representative results are shown in Figure 2).

25 2. Discussion

We have isolated 2 antigenically related high molecular weight glycoprotein com¬ plexes having molecular weights of (a) - more than 450,000 daltons and (b)

30 _% 130,000-180,000 daltons, from the envel-

^" opes of HCMV virions by anion-exchange HPLC, which we have designated as HCMV GCA. Upon reduction of disulfide bonds, individual glycoproteins of molecular

35 weights 130,000, 93,000, and 50,000-

52,000 daltons were identified. As shown - 18-

in Figure 1, all three constituent glyco¬ proteins, as well as the disulfide-linked complexes, could be immunoprecipitated by the prototype mcAbs 41C2 and 9B7, 5 suggesting that these antibodies recog¬ nize a unique class of HCMV envelope gly¬ coproteins and glycoprotein complexes. Human HCMV positive convalescent sera also immunoprecipitated glycoproteins 10 similar to those recognized by these mcAbs.

A gene has been identified in the unique long (U_[_) segment of the HCMV genome,

15 which encodes a 95,000 dalton polypeptide which can be immunoprecipitated by these 2 mcAbs. Moreover, studies on the syn¬ thesis and processing of this group of HCMV glycoproteins using pulse-chase

20 experiments, metabolic inhibitors and endoglycosidases demonstrate that the antigenically related glycoproteins of molecular weights 93,000 and 50,000- 52,000 daltons are fully glycosylated

25 glycoproteins derived from a precursor glycoprotein of molecular weight 138,000 daltons found in infected cells but not in mature virions. The 138 kD glycopro- - teio in turn is derived from the 95,000

30 dalton polypeptide encoded by the gene for glycoprotein A. Thus, the immunopre¬ cipitation pattern recognized by the_. pro¬ totype mcAbs 41C2 and 9B7 using HPLC purified glycoproteins comprising the

35 HCMV GCA complex identified the major - 19-

glycoprotein A intermediate of 138,000 daltons as well as the three mature A-type glycoproteins of 130,000, 93,000 and 50,000-52,000 daltons. 5

Monoclonal antibodies 41C2, 9B7 and 11B4, and additional mcAbs have been shown to immunoprecipitate the same three glyco¬ proteins from detergent extracts of HCMV 10 virions. The presence of the precursor glycoprotein reflects contamination of virion preparations with cell-associated viral glycoprotein intermediates that result from the method used to obtain 15 large quantities of HCMV for these stu¬ dies. Although we have shown that HCMV contains additional glycoproteins of similar molecular weights to those comprising GCA, the pattern of 3 antige- 20 nically related glycoproteins recognized by the mcAbs described herein is unique to GCA and thus it can be concluded that all 13 monoclonals recognize the same glycoprotein complex, GCA. 25

B. Simultaneous Binding of Two Monoclonal Anti¬ bodies to HCMV Glycoproteins Expressed on the Surface of Purified Towne Strain HCMV As shown-in Figures 1 and 2, thirteen murine 30 mcAbs were identified to be specific for the anti^'genically-related 130,000, 93,000 and 50,000-52,000 dalton envelope glycoproteins of HCMV based on immunoprecipitation and Western Blot data. All mcAbs were IgG-purified by 35 hydroxyapatite HPLC and concentrated to the -20-

desired protein content using a multi-micro ultrafiltration system (Amicon) equipped with a YM10 filter. The envelope glycoproteins were presented as antigen on whole virions 5 adsorbed onto 96 well microtiter polystyrene plates.

Competitive enzyme-linked immunosorbent assays (ELISAs) were performed in order to study the

10 relationships among the antigenic -sites recog¬ nized by the various antibodies. In our ini¬ tial studies, fixed amounts of biotinylated and unlabelled mcAbs were added simultaneously to determine the relative inhibition or aug-

15 mentation of binding of the labelled antibody to the virus in the presence of the second unlabelled antibody. All thirteen mcAbs were tested for competitive'binding with two neu¬ tralizing (9B7, 18F9) and two non-neutralizing

20 (41C2, 34G7) biotinylated antibodies. As demonstrated by the data summarized on Table 1, below, the mcAbs exhibited 5 distinct patterns based on their ability to inhibit or augment binding of other mcAbs belonging to

25 the same or different groups.

30

35 00 TABLE 1 00

INHIBITION/AUGMENTATION OF BINDING OF A FIXED

AMOUNT OF BIOTINYLATED mcAb BY A FIXED AMOUNT

OF UNLABELLED mcAb υ Inhibition (-)

(X Percentage of Augmentation (+)

Unlabelled Antibody

15 μg 12.5 ng. 5 ng. 25 ng. 25 ng.

oo

ON oo

ND = not done

Values given are averages from two experiments with duplicates run in each experiment.

-22-

The following results were observed:

1. Monoclonal antibodies 12C11, 39E11 and 26B11 inhibited binding of bio-

5 tinylated 41C2, 9B7 and 18F9, but had little or no effect on the binding of biotinylated 34G7 (IVI-10142).

2. Monoclonal antibodies 41C2, 30C7,

10 38B11 and 43C8 inhibited the binding of biotinylated 41C2 and augmented binding of the other 3 biotinylated -mcAbs (34G7, 9B7, 18F9).

3. Monoclonal antibodies 9B7, 18F9 and 15 34G7 inhibited the binding of bio¬ tinylated 34G7, 9B7 and 18F9,

. whereas these same mcAbs either aug¬ mented or did not affect the binding of biotinylated 41C2. 20 4. Monoclonal antibodies 11B4 and 23B11 inhibited the binding of all four biotinylated mcAbs. 5. Monoclonal antibody 19D6 inhibited the binding of biotinylated 41C2 and 25 had little or no effect on the binding -of biotinylated 9B7, 34G7 and 18F9. There was no correlation between the pattern of inhibition or augmentation of binding of 30 the mcAbs with their neutralizing activity.

Moreover, it was not possible from these data to construct a definitive epitope map, since the inhibition or augmentation of binding of any particular mcAb in combination with other 35 mcAbs was likely to be dependent upon the -23-

binding affinity of each antibody at the par¬ ticular protein concentration used for these initial studies.

Therefore, topographical analysis of the anti¬ genic sites recognized by 7 of these mcAbs was carried out in greater detail by adding vary¬ ing amounts of unlabelled mcAb to the same or different biotinylated mcAbs to determine the

10 extent of inhibition and/or augmentation of binding. The simultaneous binding of all combinations of unlabelled and biotinylated mcAbs 26B11, 39E11, 34G7, 11B4, 18F9, 17B8, 41C2 and 9B7 to whole virus was studied, and

15 the results are summarized on Table 2, below.

20

25

30

35 -24-

TABLE 2

MAXIMAL INHIBITION/AUGMENTATION OF BINDING OF FIXED AMOUNT OF BIOTINYLATED mcAb BY SIMULTANEOUS ADDITION OF VARYING AMOUNTS OF UNLABELLED mcAb

Group II

34G7 +96% +98% +55% -80% -85% -75% -78%

18F9 +130% +77% +105% -77% -93% -90% -70%

9B7 +48% +21% ^" +2% -86% -95% -86% -66%

Group III

11B4 -52% -24% -33% -68% -55% -51% -54%

Unrelated mcAb

17B8 -13% - 5% - 9% -13% - 7% -14% -18% -25-

As can be seen from the data summarized on Table 2, addition of increasing amounts of unlabelled mcAb inhibited the binding of the same biotinylated antibody by 80% or more with 5 the exception of 11B4, which inhibited the homologous biotinylated antibody to a maximum of 54%. Adding increasing amounts of mcAb 17B8, which recognizes an unrelated non- glycosylated HCMV protein of molecular weight

10 24,000 daltons in Western Blot but does not immuπoprecipitate any ³H-glucosamine labelled material, did not significantly affect the binding of any of the biotinylated GCA- specific antibodies. Thus, the inhibition

15 and/or augmentation observed among the various combinations of GCA-specific mcAbs involves ' direct competition for the antigenic sites expressed on the GCA glycoproteins, rather than being due to the non-specific interaction

20 of i munoglobulin molecules.

Based on the inhibition/augmentation curves generated for all possible combinations of pairs of the 7 mcAbs, it was possible to

25 classify the antibodies into three major groups (Table 2). Group I consisted of anti¬ bodies 41C2, 26B11 and 39E11, and was consti¬ tuted oπ_the basis of the ability of each mcAb within this group to mutually inhibit binding

30 of other mcAbs within the group to their specific antigenic sites. Similarly, group II, comprising antibodies 34G7, 18F9 and 9B7, exhibited mutual inhibition among all anti¬ bodies within this group. Group III consisted

35 of a single, unique mcAb 11B4, which inhibited the binding of all of the other GCA-specific antibodies in groups I and II, and was itself inhibited in the presence of all of the other mcAbs.

Although individual mcAbs within the three major groups exhibited similar inhibitory patterns, significant differences were observed in their capacity to augment binding of mcAbs in unrelated groups. Antibody 41C2 in group I augmented the binding of antibodies 34G7, 18F9 and 9B7 in group II. Moreover, group II antibodies augmented the binding of biotinylated 41C2. This mutual augmentation of binding was also observed between mcAb 34G7

^■ . in group II and antibodies 41C2, 26B11 and 39E11 in group I. Both 26B11 and 39E11 in group I inhibited the binding of 18F9 and 9B7 in group II. However, 18F9 augmented the ^' binding of 26B11 and 39E11 while 9B7 enhanced the binding of 26B11 but not 39E11.

Inhibition of binding of the biotinylated mcAb was dependent on the protein concentration of the competing, unlabelled mcAb, and often reached a maximum which was maintained at higher concentrations^' of unlabelled antibody. Augmentation between two antibodies occurred in the presence of optimal concentrations of unlabelled antibody, whereas further increases in the concentration of the enhancing antibody resulting in no further enhancement and in some cases resulted in a decrease in augmen¬ tation or inhibition of binding of the^' bio- tinylated antibody. -27-

C. Simultaneous Binding of Two Antibodies to

Viral Glycoproteins in Unreduced and Reduced Detergent-Extracted Material from HCMV Towne Strain 5 Since the antigenic glycoproteins recognized by these mcAbs are normally present in the virus as glycoprotein complexes linked by disulfide bonds, similar competitive binding assays were performed using reduced or unre-

10 duced detergent-extracted material to deter¬ mine the role of confor ational structure in the binding of specific mcAbs to their respec¬ tive epitopes. Antibodies 41C2 and 39E11 in group I and antibodies 9B7 and 34G7 in group

15 II were tested. These pairs of antibodies contin-ued to mutually inhibit biTiding of other antibodies within the same group to GCA that had been extracted from whole virus, or to individual glycoproteins obtained by reduction

20 and alkylation of the detergent extract, to prevent re-association of the disulfide bonds.

Competitive binding between mcAb 41C2 (group I) and 9B7 (group II) and between antibodies

25 41C2 (group I) and 34G7 (group II) was then re-examined using unreduced or reduced detergent extracts of HCMV virions. When the antigen recognized was in the form of unre¬ duced extracted material, the ability of 9B7

30 and 34G7 to enhance binding of 41C2 to its epitope was diminished compared to whole virus. When reduced material was tested, 9B7 and 34G7 no longer augmented binding of 41C2. Conversely, 41C2 continued to augment the

35 binding of 9B7 to unreduced extracted material to the same extent as that to whole virus, whereas disruption of disulfide linkages by reduction dramatically reduced this augmen¬ tation. The enhancing effect of 41C2 on the ^' binding of 34G7 to GCA in unreduced and reduced virus extracts was very much dimin¬ ished compared to that in whole virus.

Antibody 18F9 (group II) enhanced the binding of 39E11 and 26B11 to GCA on the whole virus, whereas 39E11 and 26B11 reciprocally inhibited the binding of 18F9. When the reactivities of these antibodies were re-examined using unre¬ duced extracted material, 18F9 was noted to inhibit the binding of 39E11 and 26B11, while

39E11 -and 26B11. continued to inhibit the ^■ . binding of 18F9. With reduced material, they mutually inhibited each other except for 26B11, which had virtually no effect on the binding of 18F9. The antibody 11B4 (group

III) continued to mutually inhibit the binding of all other mcAbs, whether HCMV GCA was recognized on whole virus or in reduced or unreduced viral extracts.

Thus, it was apparent that the ability of cer¬ tain mcAbs to augment binding of certain other mcAbs to.unreduced detergent extract of HCMV was diminished compared to their ability to augment binding of the same mcAbs to native virus in many cases. Furthermore, reduction of disulfide bonds markedly diminished the augmenting effect of these mcAbs in most cases, suggesting that the conformatioπal structure of the GCA glycoprotein complexes played a significant role in the augmentation of binding.

D. Synerqistic Neutralizing Activity of Mono- 5 Clonal Antibodies

The biological significance of cooperative binding between mcAbs seen in the ELISA assay was demonstrated by examining their neutra¬ lizing activity against HCMV Towne strain.

10 ^' The neutralizing activity of individual mcAbs was first determined in the presence or absence of complement at IgG protein concen¬ trations of 0.001-100 μg/well. Greater than 50% plaque reduction was considered to be con-

15 sistent with specific neutralizing activity against Towne strain HCMV. The. results of these assays are summarized in Table 3, below,

20

25

30

35 - 30-

TABLE 3

NEUTRALIZING ACTIVITY OF INDIVIDUAL mcAbs IN THE PRESENT OF COMPLEMENT

Protein Concentra-

* At >50 μg/well, neutralizing activity may reflect non-specific effect of high IgG concentration.

Group I mcAb 41C2 did not exhibit neutralizing activity at any protein concentration, whereas 26B11 and 39E11 neutralized HCMV at protein concentrations of 6.5 μg/well and 0.5 μg/well, 5 respectively. Among group II mcAbs, 34G7 was non-neutralizing; 9B7 and 18F9 both neutra¬ lized HCMV in the presence of complement at concentrations of 0.55 μg/well and 5 μg/well, respectively. The group III mcAb 11B4 did not 10 neutralize Towne HCMV.

Since a number of the present monoclonal anti¬ bodies showed some degree of interaction in the simultaneous two antibody binding assay,

15 it was of interest to determine if these interactions could be detected in another assay. For these experiments, an i vitro plaque reduction assay was used. A number of these monoclonal antibodies also augmented the

20 binding of other antibodies in the simultane¬ ous two-antibody binding assay. Thus, it was of interest to determine if these antibodies would show a synergistic effect on neutraliza¬ tion in the plaque reduction assay. When

25 antibodies were used individually, neutraliz¬ ing activity of greater than 50% plaque reduc¬ tion was seen with antibody 9B7 at antibody concentrations of 0.5 to 1.1 μg/well, whereas antibodies 41C2 and 34G7 did not neutralize

30 HCMV^" Towne strain at antibody concentrations of 10 μg/well or less (Fig. 4 A and B). By combining mutually enhancing pairs of anti¬ bodies, namely 9B7 with 41C2 and 41C2 with 34G7, strong neutralizing effect was observed

35 for both pairs of antibodies (Fig. 4 A and B). The greatly enhanced percentage in plaque reduction achieved at a given concentration could not be explained by adding the percent¬ age of plaque reduction contributed by each antibody tested individually. Furthermore, the ability of the antibodies to neutralize HCMV when used in combination did not appear to depend on the subclass of antibody since all antibodies were of a different subclass. However, complement was required in all cases.

The ability of 11B4 (domain III) to inhibit the neutralizing activity of other GCA- specific antibodies was also examined. As determined by the simultaneous two-antibody binding assay, 11B4 was^" capable of inhibiting the binding of all other GCA-specific anti¬ bodies and, in addition, was non-neutralizing (Fig. 4C). The effect of 11B4 on the neutral- izing activity of 9B7 was then examined. In the presence of complement, 9B7 showed signif¬ icant neutralizing activity at 1 μg/well or greater (Fig. 4B). However, when 11B4 was present in the same well at concentrations of 1 μg/well or greater, the ability of 9B7 to neutralize HCMV was greatly reduced, even though 9B7 was present at a concentration of 10 μg/well (Fig. 4C).

E. Results of Detection of HCMV in Tissue Culture by Cytopathic Effect (CPE) vs. Direct Detec¬ tion of Virus in Supernatants by ELISA A total of 30 patient samples were examined simultaneously for development of cytopathic effect (CPE) in fibroblast cell cultures -33-

during a 4-week observation period, and viral protein was detected in tissue culture cell monolayer lysates by ELISA using a double antibody enzyme-linked immunosorbent assay after 5 days in culture. Twenty-one of the 30 samples were negative for HCMV by ELISA, whereas 20 of the 30 samples were negative for CPE. Nine of 30 samples were positive by ELISA_? 10 of 30 samples developed CPE. Thus, there was excellent agreement between the

ELISA and CPE results, with 2 false negative and 1 false positive ELISA tests out of 30 samples. The presence of HCMV in tissue culture as detected by CPE was confirmed by an indirect immunofluorescence assay using the same HCMV-specific mcAb antibody as that used for detection of HCMV in. the ELISA assay described hereinabove.

Discussion

We have previously isolated and characterized a major immunogenic envelope glycoprotein complex of HCMV which is made up of at least three species of antigenically-related glycoproteins linked by disulfide bonds. These glycoproteins have molecular weights of 130,000, 93,000 and 50,000-52,000 daltons. We have generated 13 murine B-cell hybridomas producing mcAbs reactive with this glycoprotein complex and its consti¬ tuent glycoproteins. The individual GCA-specific mcAbs were shown to' inhibit or augment the simultaneous binding of other of the GCA- cAbs in a manner which suggested at least 7 distinct antigenic sites in 3 separate domains expressed in close proximity on the tertiary structure of the glycoprotein complex. Of particular interest, certain of these mcAbs are shown to be reactive with neutralizing sites on the glycopro¬ tein complex in the presence of complement, whereas others are not neutralizing when assayed singly. How¬ ever, simultaneous binding of pairs of neutralizing and non-neutralizing antibodies markedly enhanced the overall neutralizing activity. In some cases, 2 non- neutralizing antibodies also exhibited significant neutralization activity in combination.

Based on all of the results disclosed in the present patent application, it appears that 7 fully characterized mcAbs recognized antigenic sites within 3 separate domains on HCMV GCA. Those monoclonals recog¬ nizing epitopes within a single domain appear to be in close physical proximity. This conclusion is based on their ability to mutually compete for binding sites, whereas those in separate domains appear to recognize more distant sites based on their ability to augment the binding of monoclonals to epitopes in different domains. (Figure 3, Panel A) The physical relation- ships between epitopes in different domains as well as those within individual domains appear to depend on conformational structure rather than the primary amino acid sequence, since reduction of disulfide bonds abro¬ gates augmentation of binding in many cases. Based on the patterns of competitive inhibition and/or augmenta¬ tion of binding among the mcAbs and their ability to neutralize HCMV in the presence of complement, either alone or in combination with other mcAbs, it is believed that the epitopes recognized by these mono- clonals associate with each other as part of a crypt in the peptide backbone, linked by one or more disulfide bridges (Figure 3B). Since the physical proximity of epitopes 41C2 and 26B11 in domain I and 34G7 in domain II is altered by reduction of tne disulfide bond(s), mutual augmentation of binding of monoclonals recogniz¬ ing these eDitopεs is also εliminatec DV reαuction. In -35-

contrast, epitope 39E11 of domain I and 9B7 and 18F9 of domain II mutually inhibit binding as a result of their close proximity on the linear structure of the peptide backbone, which is not altered by reduction. This hypothetical model suggests that the epitope(s) on GCA responsible for neutralization is located at the base of the crypt, since monoclonals recognizing these epitopes, while not recognizing other epitopes within the same domain, neutralize HCMV in the presence of complement. It also appears that augmen¬ tation of binding of a neutralizing antibody to its epitope may increase its biological activity. More¬ over, changes in conformation of so-called "non-neutralizing" epitopes brought about by simultane- ous binding of two non-neutralizing antibodies may result in significant viral neutralization associated with these same epitopes.

Antibody (11B4) was placed in domain III. This antibody had the ability to inhibit the binding of all other antibodies tested and visa versa. The ability of 11B4 to inhibit the binding of other antibodies was retained when GCA was extracted with a non-ionic deter¬ gent, but this ability appeared diminished after reduc¬ tion of disulfide bonds. The sensitivity of the epi- tope recognized by 11B4 to denaturation was also reflected in the lack of reactivity of 11B4 in the Western Blot system. Monoclonal antibody 11B4 was an IgG2b subclass. This subclass is capable of binding complement. However, 11B4 failed to neutralize HCMV Towne strain ^'even in the presence of complement. It is interesting to note that 11B4 is a non-neutralizing antibody which inhibited the binding of the neutraliz¬ ing antibodies 39E11, 18F9 and 9B7. Monoclonal anti¬ body 11B4 also inhibited the neutralizing activity of 9B7 in a plaque reduction assay. We have recently -36-

generated another monoclonal antibody (23B11) which also inhibits the binding of all other HCMV GCA- specific antibodies.

This would suggest that the generation of this type of antibody is not unusual, and may correlate with diminished protection against infection _ir^ vivo. Such, antibodies might, in fact, enhance infection. Whether or not the epitope recognized by 11B4 leads to the pro¬ duction of non-protective antibodies _in_ vivo is not known, but the fact that 11B4 is non-neutralizing, inhibits the binding of neutralizing antibodies, and prevents neutralization Lr vitro leads to that possi¬ bility.

Thus, the utility of the various aspects of this invention is supported by a topographical analysis of antigenic sites on a major immunogenic envelope gly- coprotein-of HCMV. Moreover, a panel of 13 mcAbs are provided in which particular combinations bind synergistically to the virus and enhance the biological activity over that of either mcAb used alone. This is of particular importance in determining the clinical utility of HCMV-specific murine mcAbs for diagnostic and therapeutic purposes, given that amounts required for any individual mcAb alone may either lack sen- sitivity or be required in greater quantity than that practical for commercial production.

These various mcAbs may play an important role in the development of viral diagnostic tests^* because different mcAtis can be used in a non-competing format. For example, one mcAb recognizing a unique epitope of a particular HCMV protein can be used to capture and con¬ centrate the virus, whereas a second mcAb which recog¬ nizes a different, non-competing epitope of the same protein can be used for immunologic detection of the viral antigen. Moreover, two mcAbs which mutually augment the binding of the other antibody to its par¬ ticular antigenic site may be used to increase the sen¬ sitivity of immunologic detection of the viral antigen. The utility of this method has been demonstrated using prototype diagnostic tests, in which nitrocellulose membranes and plastic plates have been used as the cap¬ tive matrix for HCMV antigen. Using the HCMV-specific mcAbs described in this application, these prototypes can detect HCMV in nanogra amounts. Monoclonal antibodies to HCMV may also be use¬ ful in the treatment of potentially life-threatening opportunistic HCMV infections in immunosuppressed patients. Among these patients are organ transplant patients, bone marrow transplant patients, patients with congenital or acquired immune deficiency diseases, patients on immunosuppressive drugs, and patients with congenital HCMV infection. The use of mcAbs in im unotherapy may be indicated in both prophylactic treatment of high-risk patients and specific therapy in patients with serious HCMV infections. There is recent evidence suggesting that administration of hyperimmune HCMV globulin to immunosuppressed patients prior to organ or bone marrow transplantation may possibly attenuate or prevent life-threatening HCMV infections in the post-transplant period. HCMV-specific mcAbs can be generated in large quantity at low cost, and thus represent a significant advantage over hyperimmune glo¬ bulin. Therapeutic administration of HCMV hyperimmune globulin to patients with active infection may also be useful, although success in this regard has been limited to date.

Monoclonal antibodies developed in our labora¬ tory have been shown to have high specific neutralizing activity against Towne strain HCMV in vitro. Moreover, synergistic neutralizing activity using two or more monoclonal antibodies directed against different epitopes of the same glycoprotein may allow for high therapeutic activity with relatively low quantities of mixtures of mcAbs compared to that required using indi- vidual mcAbs. Also, synergistic neutralizing activity is observed over a wide range of relative con¬ centrations of augmenting mcAbs. This should greatly simplify the selection of a pharmaceutical unit dose that will exhibit the desired therapeutic effect in patients.

It appears that certain non-neutralizing GCA- specific mcAbs (i.e., 11B4) inhibit the binding of all other mcAbs tested thus far, including those which neutralize HCMV. Therefore, HCMV hyperimmune globulin may contain certain HCMV-specific antibodies which com¬ pete with the neutralizing antibodies and^" alter their biological activity, thereby potentially promoting HCMV infection. Moreover, these data suggest that murine or human HCMV-specific mcAbs to be used for immunotherapy should be selected on the basis of their ability to augment the binding and neutralizing activity of other HCMV antibodies. Equally important, it is critical to exclude those mcAbs that may inhibit the binding and/or neutralizing activity of other HCMV-specific antibod- ies.

The HCMV monoclonal antibodies described herein are murine monoclonals. They potentially can elicit an allergic response in humans. However, it is likely that most patients with serious HCMV infections will require only one or two treatments with HCMV- specific mcAbs. Therefore, the chance of significant allergic response to the mcAb is lessened. In clinical trials using other murine mcAbs, hypersensitivity reac¬ tions have not presented a major problem. Furthermore, the FDA has already approved the use of murine 0KT3 mcAb (ORTHOCLONE, Ortho Pharmaceutical Co.) for cancer treatment, indicating that murine monoclonals will be an acceptable form of therapy. The following are rationales for use of murine versus human mcAbs for human therapeutics:

1. As previously mentioned, murine monoclonals have now been released by the FDA and are con¬ sidered effective treatment. The allergic reactions as a result of human-anti-mouse reactions do not appear to be a major problem.

2. The murine mcAbs of the present invention are well-characterized and exhibit the appropriate immunogenic responses. Thus, the present invention provides compositions which may be used in clinical trials.

3. Alternative methods to make human mcAbs include EBV hybridomas, human-human hybridomas and human-mouse immunoglobulin genetic chim¬ eras. In all cases, these methods do not appear to be commercially practical at the present time for the following reasons:

(a) very low frequency of stable human hybri¬ domas,

(b) lack of availability of human fusion systems; and

(c) very low levels of human mcAb production in the human system compared to that of the murine system (i.e., human hybridomas produce 1-10 μg/ml; murine hybridomas produce in excess of 1 mg/ l).

The present invention provides mixtures of murine anti-HCMV GPC mcAbs for prophylactic and thera¬ peutic treatment of patients with HCMV. It would not be practical or advisable to use human monoclonals because of the lack of stability of human hybridomas and lack of the necessary production levels for obtaining a therapeutic quantity of human mcAbs. It is becoming routine for transplant patients to receive immunoprophylaxis, both pre- and post-transplant.

Current therapy with HCMV globulin/plasma lacks speci¬ ficity of action and carries all the risks of blood products_! furthermore, this therapy is extremely expen¬ sive. Therefore, murine monoclonals would be more effective because of. their antigen specificity. They would have less potential side effects because they are not a blood product and would be much less expensive to produce.

Samples of hybridomas which produce mcAbs 9B7, 41C2 and 34G7 have been deposited with In Vitro Inter¬ national, Linthicu , MD, USA, in accord with the Draft Patent and Trademark Office Deposit Policy for Biologi¬ cal MateTials, BNA PTCJ, 32, 90 (1986), as indicated on Table I, below:

Table I - Deposit Information Antibody IVI Access Code Date of Deposit 9B7 IVI-10117 September 10, 1986

41C2 IVI-10119 September 10, 1986 34G7 IVI-10142 August 7, 1987

11B4 IVI-10181 July 31, 1988

Claims

WHAT IS CLAIMED IS:

1. A method for detecting HCMV in a sample of physio¬ logical material infected with HCMV, comprising:

(a) reacting said physiological material with a first monoclonal antibody which binds to a first epitope of HCMV GPA to form a complex between said HCMV GPA and said first mono¬ clonal antibody;

(b) reacting said complex with a second monoclonal antibody to a second epitope of HCMV GPA, wherein the second monoclonal antibody enhances the binding of the first monoclonal antibody to the first epitope, and wherein said second monoclonal antibody comprises a detectable label or a binding site for a detectable label, to yield a ternary complex;

(c) detecting the presence of said detectable label or reacting said binding site with said detectable label and detecting said label, wherein the concentration of said label provi¬ des a measure of the concentration of HCMV GPA present in said physiological material.

2. The method of claim 1 wherein said physiological material comprises lysed human fibroblasts.

3. The method of claim 1 wherein the first monoclonal antibody neutralizes HCMV infectivity _in_ vitro in the presence of complement.

4. The method of claim 3 wherein the first monoclonal antibody is 9B7.

5. The method of claim 3 wherein the second monoclonal antibody does not neutralize HCMV infectivity in vitro in the presence of complement.

6. The method of claim 5 wherein the second monoclonal antibody is 41C2.

7. A composition consisting essentially of a mixture of

(a) a first monoclonal antibody which binds to a first epitope on HCMV GPA, wherein said first monoclonal antibody neutralizes HCMV infec¬ tivity _in_ vitro in the presence of complement^ and

(b) a second monoclonal antibody which binds to a second epitope on HCMV GPA, wherein said second monoclonal antibody does not neutralize HCMV infectivity in the presence of comple¬ ment; wherein the neutralizing activity of the mixture is greater than the neutralizing acti¬ vity of the first monoclonal antibody.

8. The composition of claim 7 wherein the first monoclonal antibody is 9B7.

9. The composition of claim 8 wherein the second monoclonal antibody is 41C2.

10. The composition of claim 7 wherein said composition is free of a monoclonal antibody which binds to a third epitope on HCMV GPA, does not neutralize HCMV infectivity in the presence or absence of comple¬ ment, and which inhibits the binding of the first antibody and the second antibody.

11. A method for raising plasma immunogenicity to HCMV in human patients, comprising administering to a patient who has been exposed to HCMV, a pharmaceu¬ tical dosage form comprising an effective amount of the composition of claim 7, in combination with a pharmaceutically acceptable carrier.

12. The method of claim 11 wherein the infectivity of HCMV is substantially reduced.

13. The method of claim 11 wherein the dosage form is administered parenterally.

14. The method of claim 12 wherein the dosage form is

. administered by intravenous injection or infusion.

15. The method of claim 11 wherein a plurality of said dosage forms are administered to said patient.

16. A method for the _in_ vitro neutralization of HCMV comprising treating an amount of HCMV with an amount of the composition of claim 7 in a phar¬ maceutically acceptable carrier, which is effective to neutralize the infectivity of the HCMV.

17. A method for detecting HCMV in a sample of physio¬ logical material infected with HCMV comprising:

(a) reacting said physiological material with a first monoclonal antibody to a first epitope of HCMV GPA, wherein said first monoclonal antibody comprises a detectable label or a binding site for said label, to yield a complex betwen HCMV GPA and said monoclonal antibody; (b) reacting said complex with a second monoclonal antibody which binds to a second epitope of HCMV GPA, wherein the second monoclonal anti¬ body enhances the binding of the first mono¬ clonal antibody to the first epitope, to form a ternary complex; and

(c) detecting the presence of said label, or alternatively, reacting said binding site with a detectable label prior to said detection, wherein the concentration of said label pro¬ vides a measure of the concentration of HCMV GPA present in said physiological material.

18. The method of claim 17 wherein said physiological material comprises lysed human fibroblasts.

19. The method of claim 17 wherein the first monoclonal antibody neutralizes HCMV infectivity _in_ vitro in the presence of complement.

20. The method of claim 17 wherein the first monoclonal antibody is 9B7.

21. The method of claim 17 wherein the second monoclonal antibody does not neutralize HCMV infectivity in vitro in the presence of complement.

22. The method of claim 17 wherein the second monoclonal antibody is 41C2.

23. A method for detecting HCMV in a sample of physio¬ logical material infected with HCMV comprising:

(a) simultaneously reacting said physiological material with (i) a first monoclonal antibody to a first epitope of HCMV GPA, wherein said first monoclonal antibody comprises a detect¬ able label or a binding site for said label, and (ii) a second monoclonal antibody which- binds to a second epitope of HCMV GPA, wherein the second monoclonal antibody enhances the binding of the first monoclonal antibody to the first epitope, to form a ternary complex; and

(b) detecting the presence of said label-, or alternatively, reacting said binding site with a detectable label prior to said detection; wherein the concentration of said label pro¬ vides a measure of the concentration of HCMV GPA present in said physiological material.

24. The method of claim 23 wherein said physiological material comprises lysed human fibroblasts.

25. The method of claim 23 wherein the first monoclonal antibody neutralizes HCMV infectivity _in_ vitro in the presence of complement.

26. The metho^'d of claim 23 wherein the first monoclonal antibody is 9B7.

27. The method of claim 23 wherein the second monoclonal antibody does not neutralize HCMV infectivity in vitro in the presence, of complement.

28. The method of claim 23 wherein the second monoclonal antibody is 41C2.