WO2000032753A1 - Heat inactivated retrovirus preparations - Google Patents

Abstract

A process is provided for the use of heat to inactivate (abolish the infectivity of) retroviruses such as the human immunodeficiency virus, type 1 (HIV-1). At the same time that it abolishes the infectivity of retroviruses, this process is able to preserve a substantial fraction of the antigenicity, or to enhance the antigenicity, of the original infectious virus preparations. The retrovirus elements whose antigenicity is enhanced or preserved includes various immunologically-active elements (epitopes) that are highly conserved, i.e., present on substrains of the virus isolated at various times and/or geographical locations. Thus, this process is advantageous for applications that call for retrovirus preparations whose infectivity has been abolished but whose antigenicity has been enhanced, preserved, or substantially preserved, no matter when or where the retrovirus was isolated. Such applications would include, but would not be limited to, virus used as reference reagents, in diagnostic tests, or in preventive vaccines.

Description

HEAT INACTIVATED RETROVIRUS PREPARATIONS

CROSS REFERENCE TO RELATED APPLICATIONS

This application claims priority from Provisional Patent Application Ser. # 60/1 10,798, filed December 3, 1998.

BACKGROUND OF THE INVENTION

This invention relates to the use of heat to inactivate (abolish the infectivity of) retroviruses such as the human immunodeficiency virus, type 1 (HIV- 1 ) and to preserve or enhance the antigenicity, or a substantial fraction of the antigenicity, of the original infectious retrovirus preparation so that the retroviruses can be used as reference reagents, in diagnostic tests, or in preventive vaccines.

HIV-1 has been of great interest to the medical and research communities since it was determined to be the causative agent for the pandemic human disease known as Acquired

Immunodeficiency Syndrome (AIDS). [Barre-Sinoussi et al., 1983; Gallo et al., 1983] Due to the urgent need to eliminate infectious virus contaminating blood products such as antihemophilic factor concentrates, thermal and other means to inactivate HIV- 1 were studied extensively throughout the early years of the AIDS epidemic, [Spire et al., 1984; Levy et al., 1985; McDougal et al., 1985; Resnick et al., 1986; Piszkiewicz et al., 1986; Tersmette et al., 1986; Quinnan et al., 1986; Piszkiewicz et al., 1989; Winkelman et al., 1989; Horowitz, 1991]. These studies clearly demonstrate that HIV- 1 infectivity can be destroyed by various virucidal agents, including heat. They also show that HIV- 1 can be heat-inactivated under conditions that are sufficiently gentle to preserve the complex biological activity of large, labile proteins whose structure and function are notoriously easy to disrupt, such as coagulation factors 8 and 9. However, because HIV-1 was only an unwelcome contaminant in products made from blood plasma, the inactivated virus itself was not closely studied. No research has been reported previously on immunological properties of heat-inactivated HIV-1 in general, or on the use of heat to inactivate HIV-1 for vaccine applications in particular.

Inactivated Retrovirus Vaccine Applications

An important use for inactivated retrovirus preparations is in vaccines to prevent disease due to retroviral infection. A theoretical advantage of inactivated virus vaccines is that they could contain a complex mixture of viral antigens which might closely resemble antigens on the native virus particle ("virion"). An ideal inactivation procedure generally would be expected to preserve most or all of the immunologically-active elements ("epitopes") of the virus. Epitopes that might be preserved would include so-called "linear" epitopes consisting of contiguous amino acids from viral proteins or glycoproteins as well as so-called discontinuous or "conformational" epitopes incorporating noncontiguous amino acids. The spatial proximity and consequent ability of non-contiguous amino acids to function as a discrete immunologic element necessarily depends on the higher-order structure (i.e., the folding or conformation) of viral proteins or glycoproteins. The epitopes that survive inactivation might exhibit any degree of variability ranging from epitopes that are highly variable from one virus isolate to another, such as the hypervariable loops (e.g., VI, V2, V3) of gpl20 envelope, to epitopes that are broadly conserved, i.e., found on various strains and substrains of the virus isolated at various times and/or geographical locations, such as the primary host cell receptor (CD4) binding site (CD4BS). Included would be broadly conserved "virus-neutralizing" (VN) epitopes that serve as targets for immune responses that can reduce or destroy ("neutralize") virus infectivity. Also included are epitopes that might be capable of eliciting either antibody-based "B cell" responses characteristic of humoral immunity, "T cell" responses characteristic of cellular immunity, or other categories of immune responses.

Inactivated virus vaccines have worked successfully for a large number of viral diseases in the past, including some retroviral diseases; e.g., see the work of Yamamoto and collaborators with feline immunodeficiency virus (FIV) , Issel and colleagues with equine infectious anemia virus (EIAV). Also, work with feline lenti- and retrovirus vaccines has been reviewed. In most cases the immunity reported has not been "sterilizing immunity," the sort that can prevent infection entirely, and it usually has not been sufficient to withstand "massive" virus challenge with 10⁴ or 10⁵ infectious doses (I.D.). In the early 1990's efforts were made to develop an inactivated ("killed") SIV vaccine as a model for an HIV-1 vaccine [Stott, 1994]. Protection was obtained, but the interpretation of the results was confounded by the subsequent observation that much if not all the protection was conferred by an immune response against human cell antigens in the vaccine preparations [Arthur et al., 1995]. Preoccupation with failure of these trials to produce sterilizing immunity seems to have obscured the fact that some of these trials produced endpoints that potentially were clinically useful. Two cases in point: (1) Cranage et al. [1993] challenged intravenously (i.v.) ten vaccinated and four unvaccinated macaques with 10 "50% monkey infectious doses" (MID₅₀) of SIV_mac251 grown on monkey peripheral blood lymphocyte cells (MPBMC). All animals became infected, evidenced by the ability to recover virus as early as 2 weeks post challenge. Two of the four unvaccinated controls became ill and had to be sacrificed at 12 weeks, a 3rd at 21 weeks postchallenge. "In contrast," the authors report, "only 1 of 10 vaccinated animals showed a significant weight loss at 7 months after challenge and, despite a significant loss of peripheral CD4⁺ T cells and much reduced CD4/CD8 ratios in most animals, no overt disease has developed in these animals " (emphasis ours). (2) In a similar experiment, Hunsmann and colleagues [Stahl-Hennig et al. , 1993] vaccinated with a detergent-ether treated "split" virus preparation, and challenged four vaccinates and four unvaccinated control macaques by intravenous (i.v.) administration of 10 MID_J0 of SIV_mac25ι grown in MPBMC. All 8 animals became infected within 2 weeks, as judged by isolation of virus from MPBMC. However, measuring antigenaemia in blood sera using an SIV core antigen capture assay, the authors report: "The control animals displayed typical antigenaemia two weeks after challenge, whereas the vaccinees remained antigen-negative for the 24 weeks of follow-up (data not shown). " Despite the apparent clinical benefit from vaccination with inactivated SIV, no systematic efforts were made to characterize the inactivated SIV virions in regard to optimizing antigenic or immunogenic properties favorable for a vaccine.

Inactivated HIV-1 Prophylactic Vaccines

From experience with other diseases, public health officials recognize that a beneficial reduction of the HIV transmission rate in principal could be obtained from a vaccine able to reduce the extent and duration of viremia, the titer or concentration of virus found in blood or other bodily fluids and tissues,. And considerable benefit could be derived from a vaccine that would protect against even modest challenge doses (10-100 I.D.) of HIV-1. Inasmuch as endpoints such as these have been achieved with inactivated retroviral vaccines for animal diseases, as discussed above, ample reason exists to explore inactivated HIV-1 for possible use in vaccines to prevent HIV disease. However, work to date on whole inactivated ("killed") HIV-1 prophylactic vaccines has been remarkably limited, consisting of preliminary studies employing one or more chemical or photochemical agent for virus inactivation. Only a few research groups have reported any work on killed HIV-1 prophylactic vaccines over the past several years. Fultz et al. [1992] and Girard et al. [1991] collaborated on vaccine formulated with whole inactivated HIV-l_La„ recombinant gpl60_env/p25/pl8_gag , and synthetic peptides derived from viral envelope third hypervariable (V3) regions of several strains; these studies demonstrated protection against cell free and cell bound HIV- n c mpanzees. e r g e a . an regersen e a . co a ora e on stu es o w o e inactivated HIV-1 vaccine in rabbits and chimpanzees to evaluate the effectiveness of a new adjuvant formulation.

Salk and colleagues developed an inactivated HIV-1 vaccine based on the joint use of a chemical and a physical inactivation agent (gamma-ray irradiation) [Salk et al., 1993; Trauger et al., 1994]. However, their product retains only a small amount of the viral envelope and its major (120,000 dalton molecular weight) glycoprotein, gρl20, the bulk of which has been lost in the preparation process. Consequently, this "envelope-depleted" product bears little antigenic resemblance to the native virus and cannot be expected to stimulate an immune response that would protect against infection by wild-type virus. Therefore, the product has been classified as an "immunotherapeutic" agent designed for therapeutic treatment of already-infected individuals rather than for prophylactic (preventive) vaccination.

In addition to the physical and chemical methods for inactivation of retroviruses described in the above-cited research publications, U.S. patent 5,698,430 discloses that human retroviruses can be inactivated by use of biological enzymes such as proteinases to release viral nucleic acids from the viral coat proteins, followed by treatments designed to disrupt or remove sufficient viral nucleic acids to render the material non-infectious. However, the use of enzymes that attack proteins, such as proteinases, is likely to modify the antigenic nature of the viral coat .proteins so that the inactivated material, if used in a vaccine, no longer would resemble native virions to a degree that would stimulate immunity relevant to infectious retrovirus.

U.S. patent 5,639,730 discloses that the use of tensides (i.e., detergents) followed by heating is able to inactivate various viruses present as contaminants in biological preparations such as blood coagulation factor concentrates. U.S. patent 5,610,147 discloses that viruses (other than retroviruses) can be inactivated by heating without detergent treatment. In neither of these cases is consideration given to whether the inactivated virus might have attributes useful as a vaccine.

Heat is shown to be capable of inactivating poliovirus type I in the presence of various biological substances in U.S. patent 4,687,664 and hepatitis B virus in U.S. patent 4,490,361. U.S. patent 4,424,206 also teaches that hepatitis virus present as contaminant in cold insoluble globulin ("fibronectin") can be inactivated by heat. In all three instances the virus is considered to be a contaminant and further use of the virus for vaccination or any other purpose is not mentioned.

Several examples of using heat to produce inactivated vaccines for viruses other than retroviruses have been disclosed previously. U.S. patent 3,060,094 describes the use of heating for short times, about one-thirtieth of one second (0.033 seconds), at temperatures ranging from about 71 egrees cent gra e to egrees cen gra e or preparat on o vacc nes or newcast e seases v rus and influenza A virus. This work was done more than fifteen years .before the AIDS epidemic.

Accordingly, no consideration was given to possible application of the method to HIV-1 or other retroviruses. But use of temperatures in excess of 70° C. would be expected to affect the antigenicity of retrovirus preparations more rapidly than treatment at or below 70° C. and might be expected to limit subsequent use of the inactivated virus for immunologic purposes such as vaccines.

U.S. patent 4,438,098 describes the use of heating for about ten hours at temperatures of about sixty degrees centigrade to prepare a vaccine used to immunize chimpanzees against non-A, non-B hepatitis virus (now called hepatitis C); U.S. patent 5,639,461 describes a heat inactivated influenza vaccine made from heating the chick egg allantoic fluid containing influenza virus at temperatures of about 45 degrees centigrade to 59 degrees centigrade for about twenty-five to one hundred eighty minutes. Again, in these cases, no consideration was addressed to applying the method to retroviruses such as HIV- 1.

The most comprehensive inactivated HIV- 1 vaccine development has been reported by Race et al., working in a group led by J. S. Oxford. As described in the cited references and in U.S. patent 5,698,432, Oxford's work teaches that the main issue that has limited investigation of inactivated HIV vaccines is safety. Some workers have believed that "it is difficult to ensure that virus is completely inactivated" and have expressed concern that an inactivated virion vaccine might contain "some infectious virus which evaded the inactivation process." Oxford reminds us that such concerns can be traced to the 1950's "when an incompletely inactivated batch of the polio vaccine was released for use." In order to obtain higher safety margins and suitable preservation of native viral antigen, Oxford and colleagues employ a "multistage" procedure involving sequential use of multiple agents, viz. β-propiolactone (BPL), sodium cholate, binary ethylenimine (BEI), and formaldehyde. The procedure must be carried out in a specific order, requires long treatment times, utilizes reagents that are unstable or toxic or both and that need to be eliminated before the inactivated retrovirus preparation can be used in people, and offers certain other disadvantages (see below). But it is said to be particularly useful for certain types of viruses, such as viruses that are resistant to inactivation by heat. Oxford apparently considers this procedure to be not suitable for HIV and/or was not aware that heat-inactivated HIV antigens might survive heat inactivation, because U.S. Patent 5,698,432 states: "where a virus.can efficiently be inactivated by an alternative one-step procedure, then it would not generally be appropriate to use the [multistage] method of the present invention."

It should be noted that other workers also have reported results that would appear to disqualify heat as a suitable method for obtaining inactivated retroviruses with useful antigenic properties. Lifson and coworkers [Rossio et al., 1998] have described a new approach to chemical inactivation of HIV-1 (and SIV), utilizing the compound 2,2'-dithiodipyridine, which is active against highly conserved retroviral elements called "zinc finger motifs." They report that proteins on the virion surfaces of 2,2'-dithiodipyridine-inactivated virus "retain conformational and functional integrity" and that the virus behaves similarly to native infectious virus in a variety of bioassays. In contrast, virus inactivated by use of formaldehyde or by heating for 2 hours at 56° C. fails to perform similarly in the same assays. Eibl and colleagues explored a variety of inactivation methods and reported that heat inactivation destroys HIV-1 antigenicity. [Sheets and Goldenthal, 1998]

Additional features of Oxford's invention, U.S. Patent 5,698,432 may be cited here to illustrate prominent disadvantages and limitations of all prior art for inactivation of retroviruses.

(a) One assumption underlying prior work with inactivated virus vaccines is that it is necessary, or at least desirable, to maintain the structural integrity of the "virion" (virus particle). For example, a preferred objective of Oxford is to "maintain the structural integrity of the virus and hence the major antigenic determinants of both externally-situated glycoproteins and core proteins." Clearly, Oxford views maintainance of structural integrity as a precondition for maintaining antigenic characteristics suitable for a vaccine. Therefore, he cites a warning that use of glutaraldehyde as an inactivating agent "can severely impair structural integrity." Unfortunately, evidence exists that many common viral inactivants, including those used by Oxford, can impair structural integrity or modify antigenic structures, or both. It would be advantageous, therefore, to have a method that could maintain essential immunologic properties even if structural integrity of the virion were impaired to one degree or another.

(b) Another problem often cited as an impediment to AIDS vaccine development is the

"antigenic variability" of HIV. [Starcich et al., 1986; Moore et al., 1994; Nya bi et al., 1996; Letvin, 1998] Accordingly, a preferred objective of Oxford's invention is "to incorporate a 'cocktail', or selection, of viruses selected to match circulating wild type or 'street' viruses on the basis of antigenic and nucleotide sequence analysis of gpl20 loop regions. That objective derives from concern about the hypervariability of these loop structures. Oxford expected that a vaccine made according to his invention might need to incorporate a "cocktail" of virus strains, and would need to be reformulated from time to time and/or from place to place, as is done annually for current influenza vaccines. But a more satisfactory solution would be to identify the "group-specific" antigen(s) carried by all virus within the group of HIV-1 strains that cause infections, at least within a given geographic domain, and to develop a vaccine with potential to direct the immune response to such group-specific antigen(s) [Hilleman, 1998]. For example, it would be far less work, less labor intensive, more economical, and more reliable to produce a single vaccine that could stimulate protective immunity directed to broadly- conserved virus epitopes such as the host cell receptor binding site (CD4BS) present on all or nearly all virus isolates. However, no prior art that meets this objective has been reported.

Accordingly, several objects and advantages of the invention are:

(a) to obtain retrovirus preparations in which every single virion has been inactivated (rendered non-infectious) but which retains antigenic properties that are consistent with subsequent use of the inactivated virus preparations as vaccine antigens, as reference material, or in diagnostic assays.

(b) to obtain antigenic inactivated retrovirus preparations whose safety can be assured with high confidence through use of a simple, rapid, reliable, robust, and economical inactivation process.

(c) to obtain safe, antigenic inactivated retrovirus preparations whose safety can be enhanced by very large treatment margins, i.e., inactivation treatments that can be applied with intensities or for times in large excess of the treatment levels required to abolish infectivity, without destroying antigenic properties of the retrovirus.

(d) to obtain antigenic, enhanced-safety, inactivated retrovirus preparations using inactivation agents that (i) are not biohazardous, (ii) need not be replenished to counteract their intrinsic instability, and (iii) need not be removed or neutralized after treatment has been completed.

(e) to obtain antigenic, enhanced safety, inactivated retrovirus preparations whose antigenicity is retained, even if the structural integrity of the virions is degraded or destroyed.

(f) to obtain antigenic, enhanced safety, inactivated retrovirus preparations using an inactivation process that is effective on every type and subtype of retrovirus, every strain and substrain, and every isolate, no matter where or when it was isolated.

(g) to obtain antigenic, enhanced safety, inactivated retrovirus preparations in which antigenicity of broadly-conserved epitopes is preserved, thereby providing a solution to the problem of antigenic variability of other epitopes. Further objects and advantages are to obtain inactivated retrovirus preparations with potential to elicit either B-cell or T-cell immunity depending on how they are formulated and administered, to obtain inactivated retrovirus preparations in which the relative immunologic potency of different epitopes or groups of epitopes may have been modified as compared to native retrovirus, and to obtain inactivated retrovirus preparations that can stimulate immune responses distinct from typical responses to infection, thereby providing a means to distinguish vaccinated individuals from infected individuals Still further objects and advantages of my invention will become apparent from consideration of the ensuing description. ■

SUMMARY OF THE INVENTION

We have discovered that the disadvantages of the prior art can be overcome by the use of heat to inactivate retroviruses. In a first aspect the present invention provides a process for inactivation of retroviruses consisting of heating preparations of retroviruses in liquid suspension for times ranging from about 3 to 300 minutes, preferably 30 to 60 minutes, at temperatures ranging from about 50° C. to 70° C, preferably 56° C. to 70° C. It has been found that these heat treatments preserve some or all or enhance the antigenicity of the inactivated ("killed") retrovirus preparations, making such preparations useful as a source of safe (non- infectious) vaccine antigen, as a reference material, or in diagnostic assays.

DETAILED DESCRIPTION OF THE INVENTION

The following terms are used as understood in the art and as defined herein:

Retrovirus: A member of the Retroviridae family characterized by "reverse transcription, " the process by which genetic information of the virus, encoded in molecules of RNA, is transcribed (copied) into molecules of DNA in a reaction catalyzed by the "reverse transcriptase" (RT) enzyme. Inasmuch as every other biological entity transcribes genetic information in the opposite direction, from DNA to RNA, the RT-dependent reverse (or backward) transcription scheme of Retroviridae is the distinguishing characteristic of the family and has given rise to the name retrovirus

("backward virus"). Examples include, but are not limited to human immunodeficiency virus (HIV), simian immunodeficiency virus (SIV), hepatitis C virus, equine infectious anemia virus, human T-cell leukemia virus, various animal leukemia viruses and the like. Infectivity: Capacity of a microorganism to cause infection in a suitable host organism or target cell under suitable conditions. Infectious and infective both mean "possessing the quality of infectivity." Noninfectious means lacking that quality. To abolish infectivity means to destroy or annul completely the capacity to infect, based upon a suitable assay of infectivity or extrapolation therefrom to a desired confidence level.

Antigenicity: Capacity to function as an antigen, particularly as a target for binding of existing antibodies. [Note: in recent decades this definition has supplanted earlier usage which classified as antigen a substance which caused (or had the capacity to cause) the production of antibodies (or another immune response). The latter capacity in current usage usually is termed "immunogenicity," and is the defining property attributed to an "immunogen." An inactivated virus preparation is deemed to be preserved or retained if its ability to bind an antibody to a virus epitope is at least 1 % of non- inactivated virus.

Inactivation: Act or process of reducing or destroying capacity to carry out a biological activity such as infection.

Requirements for inactivation of retroviruses.

It will be appreciated that the most crucial prerequisite of any inactivated ("killed") virus vaccine made from infectious and highly pathogenic retroviruses such as HIV-1 is that the inactivation process must be able to abolish the infectivity of every single HIV virion. In the argot of vaccine manufacture, the inactivation process must be able to provide any needed level of confidence that the virus is dead, and any needed margin of safety. In more precise terms, the probability that an infectious virus particle has survived the inactivation processing must be reducible to an acceptably low level.

To achieve the requisite confidence, it is preferable to have an inactivation process whose mechanism is well understood, whose mode of action targets and destroys at least one feature (e.g., physical integrity of the virion or the viral RNA) that is indispensable to the infectivity of every virus particle— even the rarest mutant, and whose application is intrinsically so simple and reliable that no virus particle can escape on account of a rare genetic event, epi-genetic factor, unwanted physical process (e.g., aggregation), mechanical or reagent failure, nor even on account of human error. As a practical matter, a process is needed whose inactivation kinetics are known, are first order vs time (and whose first order rate constant is suitably high), and which can be applied with enough intensity over sufficient time to produce whatever margin of safety is required.

SUBSTITUTE SHEET ( ULE 26) The inactivation process needs to accomplish all of the above while preserving, or at least without destroying, key immunologic properties of the virus- properties that may be able to elicit a protective response in vaccinated individuals. Historically, therefore, when relying on inactivation agents that react readily with protein, such as formaldehyde, one is required to confront an intrinsic tradeoff between safety and efficacy. As treatment time, temperature, or agent concentration are increased, protein damage goes up and, in general, efficacy falls. This problem alluded to by Jonas

Salk in his very first report [Salk, 1953] on inactivated poliovirus vaccine when he said: "The basic prerequisites are simple enough. These are first, a rich source of virus reasonably free of extraneous antigenic material and second, a method for destroying pathogenicity without completely destroying antigenic capacity."

To insure that a product has been adequately inactivated, one needs to understand the intrinsically statistical nature of the problem, as represented in Figure 1 adapted from Shadle et al. [1995] Both the risk and the efficacy of inactivation are subject to variation around mean values, and the breadth of both distributions must be well characterized before an accurate assessment of safety can be obtained. Risk can never be abolished entirely. But the probability that an infectious organism will survive tretment can be reduced to an acceptably low level if the treatment margin can be made sufficiently large. At a workshop on HIV Inactivated Vaccines convened by the National Institute of Allergy and Infectious Diseases "it was generally agreed that an inactivation factor of 10¹⁵ (15 logs) would be needed for an acceptable margin of safety" for HIV-1 [Schultz et al., 1990].

Advantages of heat inactivation.

Inactivation of HIV-1 by heat exhibits several properties that can contribute to vaccine safety. Heat is simple and reliable to apply. Unlike chemical inactivants, it is not subject to exhaustion or depletion, is not diffusion limited, produces negligible pH change, and unreacted reagent need not be removed from the reaction mixture. Most important of all, it has been shown to exhibit first-order reaction kinetics whose rate constant is high and relatively independent of the surrounding milieu. For a first order inactivation process starting with initial infectivity I₀, the residual infectivity, I, after any time, t, is given by the relationship log (1/ 1₀)= -(k/2.303) t. This linear dependence on time permits reliable calculation of the predicted treatment clearance margin by extrapolation from the observable loss of virus infectivity to longer treatment times. So, for example, the observation that heating at 60° C. abolishes one log of HIV-1 infectivity per 25 seconds, translates to a treatment margin of greater than FIFTY logs for 30 minutes of heat treatment, and greater than ONE HUNDRED logs for 60 minutes, Figure 2. That represents an unprecedentedly large treatment margin compared to prior methods for chemical and physical inactivation. In comparison, Oxford's multistage procedure theoretically offers "in excess of 20" logs of virus inactivation. [Race, 1995b]

Description of Specific Embodiments.

Cells and viruses: Various isolates of HIV-1 may be grown using standard methods in cells routinely employed for the propagation of human retroviruses, such as phytohemagglutinin (PHA)- stimulated normal human lymphocytes or established tissue culture cell lines. Typically, virus would be inoculated at about 0.1 multiplicity of infection (m.o.i) into about 5x10⁶ PHA (5 mg/ml)-stimulated cells in RPMI 1640 growth medium containing interleukin 2 and 10%-20% fetal bovine serum. After about 48 to 72 hours, cells are washed and changed into serum free medium. When virus-containing culture supematants are harvested about five hours later, they typically yield about lOOng HIV-1 p24 core antigen per ml of culture fluid.

Bulk virus culture fluids can be clarified and concentrated by tandem cross flow microfiltration (5 micron, 1 micron, 0.45 micron, 0.22 micron) and ultrafiltration (using membranes with exclusion limits generally between 100,000 d and 1,000,000 d). For other retrovirus, particularly feline leukemia virus (FeLV), filtration methods have been found to be more successful for recovery of intact virus, as measured by infectivity in culture. Virus preparations can be equilibrated with media suitable for subsequent inactivation and/or chemical modification procedures by diafiltration. Virus may be further purified by sucrose or glycerol density gradient centrifugation. Crude culture fluids may be inactivated or partially inactivated prior to processing to reduce the risk of exposure to viable virus and to enhance the retention of viral envelope glycoprotein. [Race 1995b] Sonication of virus materials in sealed containers may be performed for some preparations using a sonicator with a cup horn attachment. Samples may be collected at each step for infectivity determination and biochemical analysis. Titration of infectious virus when needed can be performed using permissive host cells (PHA and IL-2 treated PBMC from HIV seronegative donors) by measuring post-infection supernatant reverse transcriptase activity and/or supernatant HIV p24 ELISA. The virus dose sufficient to kill 50% of tissue culture cells (TCID₅₀) can be calculated using standard methods ). Purity of virus samples can be defined by quantitative densitometric measurement of a silver stained SDS-PAGE and an SDS-PAGE immunoblot on the same sample using standardized HIV-1 antigen specific serum.

Heat Inactivation process: Most typically, heat inactivation is performed on crude cell culture supematants after clarification by filtration through a 0.22 micron (pore size) filtration membrane to eliminate large particulates including debris from tissue culture cells. In some instances, the clarified supernatant is also treated by ultrafiltration with a commercial 300,000 dalton(d) cutoff filter device

(Pall-Filtron) to generate filtrate (F) and retentate (R) fractions. Retentate may be reconstituted to the original sample volume with serum-free tissue culture medium or phosphate buffered saline (PBS). Ultrafiltration fractions or unfractionated clarified culture supematants generally are held in sealed tubes at 4° C. in an ice bath before and after heat treatment. Tubes are immersed in a temperature- regulated water bath set at selected temperatures ranging from about 45° C. to 80° C, usually at about 60° C. to 62° C. After various times ranging typically from 3 minutes to 3 hours, usually 30 to 60 minutes, sample tubes are removed from the heating bath and stored at 4° C. until assayed.

Assay of Immunologic Properties: Virus preparations (0.22μ filtrates of virus culture supematants) and control material such as soluble recombinant HIV-1 envelope protein (s-rgpl20) were centrifuged at low speed through 300 kD (or sometimes lOOkD) exclusion filters. Material partitioned into high-molecular weight retentate fraction (R) or low molecular weight filtrate fraction (F) were assayed. Any conventional assay of antigenicity could be used, especially an assay based on use of monoclonal antibodies or other "epitope-specific" ligands. It has been preferred to use a gpl20 two-sided sandwich capture ELISA adapted from Moore and colleagues [Moore et al., 1990; McKeating et al., 1991]. In brief, this assay involves incubation of the (heated or unheated) vims preparation with primary antibody (ligand) for 45 minutes at 37° C, addition of one-tenth volume 10% Triton X-100 (1% final concentration) for 15 minutes at room temperature. Sample solutions are then transferred to a microtiter plate coated with sheep anti gpl20 C-terminus peptide capture antibody; microtiter plate is washed with room temperature tris-buffered saline (TBS)-containing Tween 80 detergent prior to sample addition. Capture plates are incubated 1 hour at 37° C, then washed 5 times with TBS at room temperature to remove unbound sample. Horseradish peroxidase (HRP)-labeled goat anti human immunoglobulin G second antibody is added and incubated in the microtiter plate for 45 minutes at room temperature. The plate again is washed 5 times with TBS. Tetramethyl benzidine (TMB) in acetate buffer is added as substrate for HRP and incubated 15 to 30 minutes at room temperature. Enzyme action is stopped with addition of 2N sulfuric acid (1 :4 v/v) to a final concentratin of 0.4N. Optical density of the reaction product is read in an ELISA plate reader.

Antigen dilution series using virus or s-rgpl20 were run for validation of the filter fractionation (and the gpl20 ELISA). Also refiltration of R and F fractions gave highly reproducible results for either fraction. At least 85% of material retained or filtered in the first filter fractionation was found in the same fraction on the second pass.

Partitioning of gp!20 from virus strain HIV-l_Sχ. grown in PBMC (prep Vac 7) was found to be stable with storage over 24 hours at temperatures ranging from 4° C. to 37° C; the portion in the 300 kD retentate remained similar (although absolute values fell appreciably) after 7 days @ 4° C.

Thermal stability to heating at a nominal temperature of 60° C. (actual recorded temperatures = 61° C. or 62° C.) for times up to 60 minutes was determined for rgpl20 and three virus preps including Vac

7, Vac 2 (HTVsx vims grown in U-87 cell line) and HIV- 1_NL4-₃ (T-cell tropic vims grown in U-87 cells).

Example 1 :

Antigenicity of inactivated virion preparations initially was tested using monoclonal antibodies directed against three discontinuous, broadly conserved vims-neutralizing neutralizing (VN) epitopes found in separate regions of the gpl20 glycoprotein: the CD4 binding site (CD4BS) recognized by human monoclonal antibodies IgGlbl2 [Burton et al., 1994], 205-46-9 (also designated HT-7 [Moore et al., 1995]; a partially-occluded site which overlaps the chemokine receptor binding site and is induced (i.e., better exposed) upon CD4 binding (CD4i), recognized by monoclonals 17b and 48d [Wyatt et al., 1995]; and a final site on the heavily glycosylated surface of gpl20 recognized by monoclonal antibody 2G12 [Trkola et al., 1996]. The gpl20 amino acids that constitute each of these epitopes and their location in the gpl20 core crystal structure recently have been identified [Kwong et al., 1998; Wyatt et al., 1998]. Assay results (Table 2) suggest that all of the highly conserved epitopes tested remain antigenic after inactivation with heat. Further examples demonstrate that additional retrovirus epitopes are preserved and that soluble CD4 (sCD4) can still bind efficiently to retrovirus preparations after heat inactivation.

Some epitopes (viz., those defined by monoclonal antibodies 2G12 and 205-43-1 [also designated HT5; see Moore et al., 1995] reproducibly exhibit greater reactivity after heat treatment of viral preparations (Table 2). In addition, the "CD4 induced" epitope defined by monoclonal 17b appears to be more accessible following heat treatment. The reactivity of 17B monoclonal antibody with heat killed HIV-1_SX is similar with and without sCD4, although in some experiments the magnitude of the 17b reaction for heat-treated virus (with or without CD4 present) is lower than that observed with non-treated vims in the presence of sCD4. One conclusion consistent with these data is that heat inactivation of HIV-1 not only can abolish infectivity of the HIV-1 but also may generate potentially improved retrovirus vaccine antigens through enhanced exposure of otherwise occluded epitopes.

Example 2:

HIVςy is inactivated at 60° C. at the rate of 1 log every 25 to 31 seconds. Prior studies show that the rate of HIV-1 inactivation by heat depends on the physical state of the sample (lyophilized or liquid) and may or may not be sensitive to the milieu (e.g., concentration of sucrose). In tissue culture medium, HIV-1 inactivation kinetics are reported to be first order with a rate constant of one log_!0 loss of infectivity per 121 seconds at 56° C, one log₁₀ loss of infectivity per 24 seconds at 60° C.

[McDougal et al., 1985] Comparable sensitivity to heating (within a factor of 2 or 3) has been seen in at least one other lab whose data are consistent with a first order rate constant of one logio loss of HIV-1 (LAV) infectivity per 5 minutes at 56° C. [Spire et al., 1984], However, as much as a ten-fold lower rate of thermal inactivation also has been reported for HTLV-III in tissue culture medium [Quinnan et al., 1986] or in 50% human semm [Resnick et al., 1986]. We measured the rate of inactivation at 60° C. for preparations of the infectious molecular clone HTVsx, an R5-tropic (CD4 /CCR5 full length infectious molecular chimera incorporating the VI to V3 region of HTV_JR. _FL in a background of HIV_M ^ as described previously [Adachi et al., 1990; O'Brien et al., 1990; Koyanagi et al., 1987]. The highest-titer fflV_sx virus stock tested, starting titer (TCID₅₀) = 10⁶'²⁵ for a sample mock-inactivated at 4°C, when heated at 60°C for various times, lost infectivity at the rate of 1 logio over the first minute, 1 logio every 25 seconds in the interval between 1 and 3 minutes, and 1 logio every 31 seconds over the entire first three minute- interval (Figure 3). No infectious titer was detected for the subsequent (10 minute) time point; it is not known how long beyond 3 minutes any infectious virus may have survived. The lower rate of inactivation in the initial minute, 1 logio per 60 seconds, presumably reflects time needed for the 4°C sample to equilibrate at 60°C. Because the 1- and 3-minute data points produced the only two non-zero titers for samples unambiguously at 60°C, this experiment neither validates nor contradicts the cited report that inactivation by heat at 60° C. exhibits first-order kinetics; however the loss of infectivity measured over minutes 2 and 3, reduction of one logio pe^r 25 seconds, is nearly identical to the first-order 60° C. inactivation rate constant, one logio per 24 seconds, reported for the HTLV-III/LAV isolate [McDougal et al., 1985]. The observation that heating at 60° C. abolishes one log of HIV-1 infectivity per 25 seconds, translates to a treatment margin greater than FIFTY logs for 30 minutes of heat treatment, and greater than ONE HUNDRED logs for 60 minutes. This represents an unprecedentedly large treatment margin compared to prior methods for chemical and physical inactivation (see Figure 2).

Example 3:

Heating at 60° C. dismpts virions and leads to loss of p24 antigenicity. RT activity, and RNA titer. It may be concluded from preliminary studies that heating HIV-1 to 60° C. dismpts the physical integrity of HIV-1 virions, coincident with its demonstrated ability to abolish infectivity, based on the following evidence: (I) intact virions can not be visualized in cryo e.m. or whole-mount negative- stained e.m.preparations of heat- inactivated vims preparations (data not shown); (2) gpl20 antigen can be found in the ultracentrifuge pellet (1 hour at 50,000 x g) of unheated virus but not of 60° C. heat-inactivated virus; (3) the ratio of p24 or viral RNA to gpl20 drops profoundly and RT activity disappears entirely upon heating a vims preparation to 60° C. as may be seen from data in Table 3. In contrast, the same assays (see legend to Table 3) detected rather similar ratios (±30%) of gp!20, p24 and RNA as vims was taken through various stages of vims purification prior to heat inactivation, including ultrafiltration, ultracentrifugation (Table 4) and density gradient ultracentrifugation (Figure

4). Western blots of heat-inactivated vims preparations (not shown) suggest that p24 may not be destroyed by heating, even though standard p24 ELISA values drop by 2 logs when virus preparations are heated to 60° C. The discordant ratio of RNA (and p24) to gpl20 in heated preparations corroborates the other cited indications that HIV-1 is disrupted by heating at 60° C.

Parenthetically, Lifson and colleagues have reported that heating HIV-1 _MM for 2 hours at 56° C. destroys the ability to precipitate p24 (!) with any of several envelope-directed antisera and the ability of heated vims to participate in functional aspects of virus infection such as binding target cells, membrane fusion, etc. [Rossio et al., 1998] However all of their assays presumably require antigenically-intact p24, intact virions, or both, neither of which may be present in heat-inactivated virus. In all probability heating at 56° C. is sufficient to dismpt virions, given that heating at 56° C. for 1 hour is sufficient to abolish HIV-1 infectivity and that heating for 3 minutes at 56° C. or at 60° C. produces very similar effects on binding of gpl20 ligands like CD4-IgG (see below).

Note that all gpl20 antigenicity data discussed in the following examples, based on the ELISA assay described above, employ a 45 minute incubation of ligand with antigen at 37°C. When effects of 60°C heat on ligand binding are discussed, terms like "preheated" antigen are used to emphasize that the ligand binding assay itself was not done at 60°C.

Example 4: .

Effects of heating on antigenicity of HIV. Antigenicity was assessed in terms of ability of vims preparations to bind CD4 cell receptor coupled to Immunoglobulin G (CD4IgG). As seen in Table 5, heating at 60° C. for times as brief as three minutes produces a two-fold to three-fold increase in the observed level of CD4-IgG binding for two different preparations of HIV≤x, but not for recombinant gpl20 monomer. Observed CD4-IgG binding levels to samples preheated for times as long as 90 minutes remain about 20% to 50% higher than unheated control samples.

Example 5:

Effects of heating on partitioning of gp!20 antigen between monomer-size and larger-size aggregates. 50% to 90% of gpl20 antigenicity detected by ELISA on live or heat-inactivated HIV-1 remains in a particulate fraction retained by 300kD cutoff exclusion membranes that do not retain monomeric gpl20. In order to distinguish and fractionate gpl20 in association with virions or other high molecular- weight aggregates (e.g., subvirion fragments, gpl20 multimers) from free gpl20 or rgpl20 monomer, the method employed was centrifugal ultrafiltration through a commercial high- molecular weight (MW) cutoff filter device (Pall-Filtron; NanoSep™) with mean molecular weight cutoff of 300,000 daltons (300kD), sometimes lOOkD, to generate retentate (R) and filtrate (F) fractions. Ultrafiltration has been employed by others for HIV-1 [Wiessenhom et al., 1996;

Wiessenhorn et al., 1997; Chan et al., 1997] and it is widely-used in commercial production of killed virus vaccines, e.g., for polio [Van Wesel et al., 1984; Montagnon et al., 1984] and hepatitis A

[Peetermans, 1992]. With this method as much as 96% of s-rgpl20 and 80-90% of gpl20 from detergent-disrupted HIV-1 passes into the 300 kD filtrate, but not into the 100 kD filtrate (data not shown). In contrast, 80% or more of the total (R+F) gpl20 ELISA assay signal for infectious vims preparations may be found in the 300 kD retentate (R), provided vims preparations have not been detergent treated, even after preheating those preparations for 30 minutes at 60° C; values as low as 44%) also have been observed, depending on the vims isolate, cell substrate, and perhaps other factors that are not yet well understood. Although the absolute amount of particulate gpl20 obtained from this assay is seen to vary, that variation is not so large as to obscure the general principle that a firm association of viral gpl20 with high molecular weight particulates survives heating at 60° C.

A number of HIV isolates were tested; the substantial majority of the tested HIV isolates exhibit the same useful properties (Table 6): antigenic gp!20 survives conditions able to inactivate high titers of HIV-1, virions are dismpted in the process, and the gpl20 antigenicity remains associated and with particles larger than gpl20 monomers. The great majority of gpl20 antigen found in cmde culture supematants is not shed gpl20 monomer, and apparently is not transformed into shed monomer in the course of heat inactivation. Even HIV_N .₃ , the t-tropic isolate whose supematants display the highest proportion (-50%) of monomeric gpl20, when heated for up to 60 minutes at 60°C, shows only a 10% drop of antigenicity for both the particulate (R) and monomeric (F) fractions (Table 7). These data carry no suggestion that heating strips gpl20 monomer from virions as might have been presumed from previous work. Perhaps either envelope gpl20 enjoys a more stable association than has been believed, or heat inactivation at 60°C is "fixing" a more stable association than exists for live vims.

Example 6:

Effects of heating on gp!20 antigenicity depends on the source of the viral envelope. Analysis of CD4-IgG binding to the t-tropic H _ML4.₃ isolate after preheating the vims to 60°C for various times (see Table 7 above), showed only a 10% drop in bound ligand over one hour of heating. A different picture emerges, especially when the analysis includes effects of shorter treatment times, for HTVsx, a molecular chimera whose envelope is derived from HIV_NL4.₃ except for a portion of envelope from fflVj_R._a [Adachi et al., 1990; O'Brien et al., 1990; Koyanagi et al., 1987] (Figure 5, panel a). As compared with virus that has not been heated, preexposure to 60° C. for just one minute produces a 70% increase in binding of CD4-IgG to FflV_Sχ; three minutes of preheating produces greater than a 3-fold increase. Longer heating periods result in lower CD4-IgG binding, until at 90 minutes, the ligand binds at a level just slightly higher than to an unheated control. In contrast, heating monomeric gpl20, whether vims-derived or recombinant protein derived from any of three

HIV strains (IIIB, N=MN, or SF2), produces the same near-negligible effect as preheating HIV_N -₃.

(Figure 5, panel b). Whether or not it is coincidental that the antigenic impact and infectivity impact both are maximal for virus heated about 3 minutes, this heat-induced "antigenicity jump" of HIV_SX can be employed as a surrogate marker for other effects (see below).

Example 7: Temperature dependence: Preheating HIV-1 for a 3-minute duration has little or no effect on gpl20 binding site antigenicity, measured in terms of ability to bind CD4-IgG, unless the temperature reaches a critical threshold above 45° C (Fig. 6). The observed impact on CD4-IgG binding (to SX) is negligible for virus heated to 37° C. or 45° C. Approximately a 3-fold jump is observed for virus preheated to 56° C. (!) or 60° C.

Example 8:

Substrate dependence: Effects of heat treatment on gpl20 binding site antigenicity of HIV-1 depends upon whether the vims is propagated in primary peripheral blood mononuclear cells (PBMC) or in cultured T cells (Table 8). Propagated in PBMC, HIVsx displays its characteristic antigenicity jump, with characteristic temperature dependence. Propagated in T-cell line U 87, it responds like HIVNU.3!

Example 9:

Isolate (donor) dependence was examined by comparing, for vims of 8 long-term survivors (LTS) either preheated 5 in. at 60°C or held at 4° C, the binding of CD4-IgG and human monoclonal antibody (huMAb) 2G12, and, for heated and not-heated vims specimens of an additional six donors, the binding of 2G12 (Table 9). Six of the eight LTS HIV isolates exhibit the "antigenicity jump" phenotype with respect to CD4-IgG binding. But only one of eight with respect to binding 2G12. In contrast, five of the additional six donors display a jump (of 21 to 99%) in 2G12 binding to preheated virus. HuMAb 2G12 recognizes a unique carbohydrate-dependent epitope that has been localized to a face of the gpl20 molecule opposite and 25 A removed from the CD4 binding site [Wyatt, 1998]. The fact that these two sites respond so differently to heating suggests that heating may exert effects specific to given epitopes, or to specific molecular domains, or both. Example 10:

Epitope/do ain dependence. To test the preceding conjecture, evaluation was made of the treatment response profile, a side by side comparison of the change (gain or loss) in binding of a panel of gpl20-binding ligands after treatment of vims by heat or another agent, expressed as a percentage of the binding observed with mock treatment of live vims. The treatment response profile is shown (Figure 7) for three preparations of HIV_SX (panels A, B, E) and two other HIV isolates (panel C, D) treated by heat at 60° C. for 30 minutes and, for comparative purposes, another HIV_Sχ preparation treated with 0.1% formaldehyde (HCHO) for 2 hours at 37° C. (Panel F). Treatment response is evaluated for binding of CD4-IgG and five huMAbs: IgGlbl2 (CD4BS); 447-52D and 694-98D; 2G12 (which defines its own epitope; see above); and 670-30D (C5). A quantitatively different impact from heating 30'@ 60° C. is seen on the binding of each ligand to each isolate. In most cases, for selected combinations of virus and cell substrate, impact is similar for the ligand pairs that share a given binding domain. For example, for the HrV_sx Vacl 1 preparation (panel A), the impact of heating on each of the two CD4 binding site (CD4BS) ligands is an antigenicity jump of -200%; the impact on each of two V3 ligands is about 60%. That preparation would appear to offer the most promising treatment response profile for antigenicity tilted away from V3 in favor of the CD4 binding site. In contrast, response of isolate PMBC-14 (panel D) enhances the V3 epitopes and treatment with 0.1% formaldehyde (2 hours @ 37° C.) destroys 91% to 94% of binding, for each of the four tested ligands irrespective of the binding domain; neither of those treatments appears to be promising.

Example 11 :

Heat in combination with other inactivants/modifiers. A number of studies have suggested the possibility that certain HIV-1 epitopes, especially epitopes prominent on the V3 loop of monomeric gpl20 shed by virions, may serve to decoy the immune response in a direction that carries more benefit to the virus than the host [Nara, 1991 ; Nara & Goudsmit, 1991 ; Zwart et al., 1991 ; reviewed Moore & Ho, 1995]. From this perspective, the ability of HIV-1 (and other retrovirus) to elicit, then to mutate away from, the dominant host immune response, may contribute to the weakening, and the ultimate defeat of host defenses. It may be advantageous to use vaccine immunogens that have been modified so as to diminish or eliminate the potential immunodominanance of prominent variable, strain-specific (PVSS) epitopes. An approach suited to killed virus vaccines is to treat heat killed vaccine virus before or after heating with an agent able to reduce the immunogencity of unwanted epitopes. Formaldehyde, glutaraldehyde, binary ethyleneimine, and perhaps other common viral inactivants can be used. The discovery that several major conserved gpl20 VN epitopes survive exposure to heat inactivation means that it is unnecessary to rely on chemical agents to kill virus. Therefore it is possible to use formaldehyde or other chemicals at sub-lethal doses. An example of such a procedure is shown in Table 3, where heat inactivation was proceded by treatment with 0.02% formaldehyde for 1 hour at 37° C.

Considerable precedent exists for the idea that epitope-specific modification is possible. The use of formalin to inactivate poliovims was shown to abolish reactivity with 7 or 9 of 19 mouse monoclonal antibodies for two different inactivated poliovims vaccines; the sites most sensitive to formalin disruption were highly-exposed loop structures containing lysine and arginine residues whose side chains are highly reactive with formaldehyde [Ferguson et al., 1993]. Sattentau [1995] has shown that treatment with formaldehyde at appropriate doses can improve the binding of monoclonal antibodies for several HIV-1 envelope epitopes. Common viral vaccine inactivants including formaldehyde have been observed to impact specific epitopes to different degrees on other viruses [Blackburn and Besselaar, 1991]. Although such differential inactivation has been observed, it has not previously been used deliberately to modify or modulate the hierarchy of immunodominance of selected viral epitopes.

The preceding examples show that heat treatment of HIV_Sχ at 60° C. produces a 10-fold drop in infectivity approximately every 30 seconds. At that rate, a 30 to 60 minute treatment of HIV-1 would be sufficient to generate treatment safety margins several-fold larger than those attainable with other methods. Heat inactivated HIV-1 preparations also displays a number of additional characteristics that serve to distinguish them from live vims. Such ^'distinguishing properties notably include the loss of physical integrity, loss of p24 antigenicity, loss of RT activity, and substantial reduction of RNA titer. Nonetheless, the retrovirus envelope in heat- inactivated HIV-1 preparations preserves a particulate nature distinct from gpl20 monomer and can be fractionated away from monomer by ultrafiltration. Moreover key gpl20 epitopes, including several conserved VN epitopes remain highly antigenic and can function well in immunodiagnostic assays such as the ability to bind monoclonal antibodies and to bind the CD4 cell receptor. Some preparations exhibit a pronounced enhancementof conserved epitopes at or near the CD4 binding site relative to epitopes on the V3 loop, as measured by binding of epitope-defining ligands. Evaluation of the treatment response profile for a variety of HIV-1 primary isolates, grown in PBMC or another cell substrate, after inactivation by heat and other methods, may allow for selection of candidate immunogens able to stimulate immune responses that differ quantitatively and qualitatively from those of native vims.

Conclusions, Ramifications, and Scope

Accordingly, it will be appreciated that the simple, rapid, effective, and reliable heat inactivation process of the invention can inactivate preparations of retrovimses such as the human immunodeficiency vims, type 1 with very high treatment safety margins and in so doing can preserve a substantial fraction of the antigenicity, or enhance the antigenicity, of the native infectious vims preparations. It further will be appreciated that the retrovims elements whose antigenicity has been enhanced or preserved includes various immunologically-active epitopes that are highly conserved, i.e., present on most or all strains and substrains of the vims isolated at various times and geographical locations. Thus it will be appreciated that the process of the invention is advantageous for applications that call for retrovirus whose infectivity has been abolished with very high confidence, but whose antigenicity has been enhanced, preserved, or substantially preserved, no matter when or where the retrovims was isolated. Such applications would include, but would not be limited to, retrovims used as reference reagents, in diagnostic tests, or in prophylactic vaccines. Furthermore, the use of heat to inactivate retroviruses for various immunological applications has the additional advantages that:

• it permits inactivated preparations to be used to elicit either B-cell or T-cell immunity;

• it permits additional biological, chemical, or physical agents to be used at "sub-lethal" doses;

• it permits modification of the relative strength of epitopes in different antigenic domains; and • it permits immune response to inactivated retrovirus to be distinguished from vims infections.

While the above description contains many specificities, these should not be construed as limitations on the scope of the invention, but rather as an exemplification of one preferred embodiment thereof. Many other variations are possible. For example, the heat inactivation process can be applied to smaller and larger volumes, ranging from fractions of a microliter to hundreds or thousands of liters. Heat may be supplied by any convenient means, including conduction heating, convection heating, radiation, etc. Heat may be applied during a single interval of time or during multiple intervals, and may be applied continuously or in pulses. If heati is applied in multiple intervals or multiple pulses, the retrovims preparations may or may not be cooled or heated in the time between heating intervals, or prior to or subsequent to the inactivation process. Vims preparations may be heat- inactivated in static containers, with or without mixing or stirring, or in flow systems. The thermal or other forms of energy supplied to heat the retrovirus preparations, and the temperature achieved, may be constant or variable. The purity and concentration of the vims preparations to be inactivated may range independently from very low purity to very high purity and from very low concentration material as in crude cultured-cell supematants to highly concentrated material. The vims preparations may be either liquid or solid state preparations, including lypholized (freeze dried) preparations, although the kinetics of retrovims inactivation are found to be considerably lower for solid-state preparations. If it is in liquid state, the retrovims preparation may be in an aqueous medium or in a non-aqueous medium. Any and all combinations of the variations described above represent possible embodiments within the scope of the invention. Accordingly, the scope of the invention should be determined not by the embodiment(s) described but' by the appended claims and their legal equivalents.

Table 2: Retention of broadly-conserved HIV-1 gp120 epitope antigenicity after heat inactivation.

gp120 epitope Gp120

Monoclonal antibody specificity (tvpe) Antiqen and treatment conditions ELISA

2G12(400ng/ml) V4, carbohydrate HIV-1 rgp120 20 ng/ml 1.767 (neutralizing) SX/4°C 1.992

SX 30" @ 62°C 2.348

205-43-1 (400ng/ml) CD4 binding site HIV-1 rgp120, 20ng/mi 1.536 (non-neutralizing) SX/4°C 1.976

SX/30' @ 62°C 2.417 lgG1b12(1000ng/ml) CD4 binding site HIV-1 rgp120, 20ng/ml 0.643 (neutralizing) SX/4°C 0.494

SX/60' @ 62°C 0.437

17b (500ng/ml) CD4 induced HIV-1 rgp120, 50ng/ml 0.240 (neutralizing) SX/4°C 0.897

SX 60' @ 62°C 1.347

HIV-1 rgp120, 50ng/w, 1.3 mg/ml sCD4 0.476

SX/4°C, 1.3 mg/ml sCD4 1.689

SX/60' @ 62°C, 1.3 mg/mi sCD4 1.346

48d (500ng/ml) CD4 induced HIV-1 rgp120, 50ng/ml <0.1 (non-neutralizing) SX/4°C <0.1

SX/60' @ 62°C <0.1

HIV-1 rgp120, 50ng/ml, 1.3 mg/ml sCD4 <0.1

SX/4°C, 1.3 mg/ml sCD4 1.390

SX 60' @ 62°C, 1.3 mg/ml sCD4 0.723

1° antibody omitted HIV-1 IIIB (2° antibody only) (Bacculovirus) SX/60' @ 62°C <0.1

Aliquots of HIV-1 SX (130ng p24/ml) harvested in serum free media were held at 4°C or heated to higher temperatures for times specified. Virus preps or controls were incubated with specified human monoclonal antibodies or controls for 45 min at 37°C in 200 μl reaction volume. Retention of epitope antigenicity was assessed by a modified HIV-1 gρl20 capture ELISA (62). Control reactivity without primary (1°) antigen was <0.1 O.D. units for each antibody. Antibody concentrations were determined by titration. 2G12, IgGlbl2, and 17b have been reported to neutralize diverse strains of FQV-l. 2G12 was a gift from H. Katinger, 205-43-1 was a gift from M. Fung, IgGlbl2 was a gift from D. Burton, 17b and 48d are gifts from J. Robinson. The recombinant gpl20 (rgpl20) was obtained commercially(Intraceι). Table 3: Effects of heat on retrovirus components

Table 4: HIV-1 sx Composition vs. Concentration / Inactivation

Material Assayed Volume³ Concen. Gp120^c p24^c RNA^d

Fraction Recovered (ml) Factor^b [ng/ml] [ng/ml] [molecule/ml] |

Pooled Culture Supernatant⁶ 2000 1X 33 83 3.5 x 10⁸ Fraction of Starting Material (1.0) (1.0) (1.0)

300 kD MW Filter Retentate 200 10X 205 812 2.5 x 10^s

Fraction Recovered (Starting Mat'l) (0.62) (0.97) (0.71)

OFFSCALE

Ultracentrifuge Pellet' 10 200X 3900 > 5000 4.8 x 10'0

Fraction Recovered (Starting Mat'l) [This step] (0.59)[0.95] (?)[?] (0.69) [0.97]

Inactivated Preparation⁹ 20 100X 2240 76 5.5 x 10⁹

Fraction Recovered (Starting Mat'l) [This step] (0.68)[1.15f (0.01)[<0.03 f (0.16)[0.23 f

Nominal (approximate) volume.

Concentration Factor = 2000 ml / Volume. gp120 ELISA O.D. converted to ng/ml by measured calibration curve based on HIVw_B rgp120.

RT-PCR converted to HIV-1 molecule (half-genome) equivalents/ml by calibration curve.

Infected culture supematants clarified by passage through 0.22μ filter.

Recovered after centrifugation for 1 hour at 50,000g mean RCF.

Treated with 0.02% (6 mM) formaldehyde 1 hour at 37°C, followed by buffer exchange (diafiltration with two-fold dilution) and heating at 62°C for three 10-min. intervals in separate tubes (30 min. total).

Recovery from inactivation step compared to predicted value for two-fold volume increase is: gp120 = 115%; p24 = < 3%; RNA = 23%.

Table 5: Binding of CD4-lgG to preheated antigen

Time³ (minutes) @ 60°C

Antigen 0 1 3 10 30 90

HIVsx (VAC II) 0.752 — 1.497 1.367 1.302 1.126

Rgp120 (IIIB) 1.345 — No Data 1.019 No Data 0.509

Mock supernatant 0.009 — No Data 0.016 No Data 0.025

HIV_Sχ(Pool w/ FBS) 0.550 0.942 1.806 L712 1.342 0.667

Samples placed in 60°C water bath at time 0, removed after 1 , 3, 10, 30, 90 min.

Table 6: Partitioning of HIV envelope before and after preheating at 60°C.

Isolate or antiαen % αo120 Retained

4°C 60°C

HIV(SX) 83 88

HIV(NL4-3) 50 53

LTS Γ 96 69

LTS 3 80 . 66

RP 1 99 87

RP 2 100 81

SC 13^C 75 71 rgp120 HIV(lllB) 4 4

Percent of total gpl20 retained by 300 kD cutoff filter device (Pall-Filtron NanoSep 300) for each isolate/antigen with and without treatment at 60°C for 30 minutes, after which preheated samples were returned to 4°C. Virus input adjusted to 26 ng p24 in 200 μl. Retentates reconstituted to the original volume with PBS. Assay used CD4-IgG as primary ligand in two-sided gpl20 sandwich ELISA (see above). Calculated percent retained, [R/(R+F)] x 100%, was based on mean ELISA value of triplicate wells. Representative data from 2 independent experiments. Recombinant HTV(iπB) gpl20 (rgpl20; Intracell) was produced in baculovirus. ^aLong Term Survivor, "Rapid Progressor, ^cSeroconvertor. Retentate usually was reconstituted to the original sample volume (analytical samples), or to tenfold- smaller volume (preparative samples), with serum-free tissue culture medium or phosphate buffered saline (PBS). gpl20 in each fraction was measured by the two-sided sandwich ELISA described above. When previously-filtered fractions were refiltered, typically about 70% and 90% of gpl20 in the retentate and filtrate respectively partitioned into the same fraction on the second pass. Table 7: Binding of CD4-lgG to HIV(NL4-3) Fractions

Fraction Unheated Heated 4°C 3 min 30 min 60 min

300 kD Retentate 2.161 2.118 2.096 1.966

300 kD Filtrate 2.113 2.177 1.870 1.875

Table 8: Binding of CD4-lgG to Preheated or not heated HIV-1

Source of HIV_Sχ 4°C 5'@60°C

PBMC Untractionated 0.606 2.334

PBMC > 300 kD Ret. 1.160 2.776 PBMC < 300 kD Fil. 0.218 . 0.286

U 87 > 300 kD Ret. 1.448 1.333 U 87 < 300 kD Fil. 0.800 0.710

U 87 < 100 kD Fil. 0.062 0.008

Table 9: Donor/Epitope Dependence

CD4-lαG 2G12

Donor 4°C Δ 5'@60°C 4°C Δ 5'@60°C

LTS 1 0.109 +++ 1.115 0.032 — 0.020

LTS 2 0.019 — 0.012 0.017 — 0.004

LTS 3 0.622 + 1.153 3.422 — 3.459

LTS 4 0.090 +++ 2.135 0.007 — 0.004

LTS 5 1.850 + 2.232 3.028 — 3.000

LTS 6 0.014 - 0.005 0.012 — 0.009

LTS 8 1.214 ++ 2.656 0.066 — 0.049

LTS 13 0.929 ++ 3.084 1.487 + 2.817

50836 0.799 + 1.546

50638 0.413 + 0.667

50107 No Data 0.181 + 0.262

50495 0.012 — 0.016

JR-FLC 0.821 + 1.189

JR FLV3 1.211 + 1.472

Δ = change upon heating:

+ = 1 to 2-fold

++ = > 2 to 10-fold

+++ = > 10-fold DESCRIPTION OF THE DRAWINGS

Figure 1 shows probability distribution around risk and clearance values (in logs).

Figure 2 shows abolition of HIV-1 Infectivity by heating.

Figure 3 effect of Heating on HIV_Sχ Infectivity. Virus stocks from plasmid DNA were made by electroporation of 25 μg of DNA into a donor pool of 5 x 10⁶ phytohemagglutinin (PHA) - stimulated peripheral blood mononuclear cells (PBMC) as described previously [Cann et al., 1988]. Virus stocks were propagated in a 3-donor pool of PHA stimulated PBMC and harvested in serum- free medium. Clarified culture supernatant containing about 160 ng/ml of HIV p24 were aliquoted for single use. Aliquots were loaded into 0.5 ml thin-walled tubes, heated for the indicated times in a metal block heated by 60°C circulating water, and returned to 4°C promptly at the end of the heating period. Endpoint dilution studies were performed with half-log dilutions of heated samples, and unheated controls that had been held at 4°C, incubated with PMBC from pools of 3 allogeneic donors prescreened for susceptibility to HTVsx as indicators cells for assay of (residual) infectivity. Quadruplicate assays performed for each dilution of each time point. RPMI 1640 medium replenished at day 7 and 14, and virus infectivity assessed by p24 ELISA on culture supematants harvested at day 21. Fifty percent tissue culture infectious dose (TCID₅₀) calculated by the method of Reed and Muench [1938].

Figure 4 shows ELISA values for gp 120 (circles) and p24 (heavy dashes) in fractions collected from gradients of colloidal silica (Percoll; solid line) and iodixanol (Optiprep; light dashes) formed in situ by ultracentrifugation (2 hour at 120,000 x g). RNA values determined by RT-PCR for Percoll gradient fractions 0 through 8 (pooled), 13, and 15 were 8 x 10⁸, 5.8 x 10⁹ and 2.6 x 10⁸.

Figure 5 shows antigenicity effects vs. source of gpl20.

Figure 6 shows antigenicity effects vs. temperature.

Figure 7 shows treatment response profile for (A) SX Vac 11, (B) SX 990814, (C) Ba-L, (D) PBMC-14, (E) SX Vac 7 to treatment with a 60° C heat for 30 minutes (A-E) or with 0.1% formaldehyde for 2 hours at 37° C (F). Data based on ELISA for CD4-domain ligands (white bars) CD4-IgG (left) and huMAb IgGlbl2 (right), V3-domain huMAb (black bars) 447-52D and 694- 98D, huMAb 2G12 (shaded bar) and C5 domain huMAb 670-30D (dotted bar).

References

1. Adachi et al. 1986 J Virology 59:284-291.

2. Arthur et al. 1995 J Virol 69:3117-3124. 3. Barre-Sinoussi et al. 1983 Science 220:868-871.

4. Blackburn and Besselaar 1991 J Virol Meth 33:367-374.

5. Burton et al. 1994 Science 266: 1024-1027.

6. Cann et al. 1988 Oncogene 3:123-128.

7. Chan et al. 1997 Cell 89:263-273. 8. Cranage et al. 1993 AIDS Research and Human Retroviruses 9: 13-31

9. Essex 1995 Annual Review of Immunology 8:557-578.

10. Fergusen et al. 1993 J Gen Virol 74:685-690.

1 1. Fultz et al. 1992 Science 256(5064): 1687- 1690.

12. Gallo et al. 1983 Science 220:865-867. 13. Girard et al. 1991 Proc Nat Acad Sci 88(2):542-546.

14. Gregersen et al. 1989 Int Conf AIDS 5:549.

15. Hilleman 1998 Vaccine: 16:778-790.

16. Horowitz et al. 1985 Transfusion 25:523-527.

17. Issel et al. 1992 J Virol 66(6):3398-3408. 18. Koyanagi et al. 1987 Science 236:819-822.

19. Krogstad et al. 1996 AIDS Research and Human Retroviruses 12(1 1):977-983.

20. Kwong et al. 1998 Nature 393:648-659.

21. Letvin 1998 Science 280:1875-1880.

22. Levy et al. 1985 The Lancet 1(8443):1456-1457. 23. McDougal et al. 1985 J Clin Invest 76:875-877.

24. McKeating et al. 1991 J Virology 65(2):852-860.

25. Montagnon et al. 1984 Rev Infec Dis 6 Suppl 2:5341-5344.

26. Moore 1990 AIDS 4(4):297-305.

27. Moore and Ho. 1995 AIDS 9 (suppl A):S1 17-S136. 28. Moore and Jarrett. 1988 AIDS Research and Human Retroviruses 4(5):369-379.

29. Moore et al. 1989 AIDS 3(3): 155-163.

30. Moore et al. 1990 Science 250(4984): 1139-1 142.

31. Moore et al. 1994 J Virology 68:8350-8364.

32. Moore et al. 1995 J Virology 69:101-109. 33. Nara and Goudsmit. 1991 in Vaccine 91: Modern Approaches to New Vaccines including the prevention of AIDS (Brown et al.) ppl-8.

34. Nara et al. 1991 FASEB J 5:2437-2455.

35. Nyambi et al. 1996 J Virology 70(9):6235-6243.

36. O'Brien et al. 1990 Nature 348:69-73. 37. Peetermans 1992 Vaccine 10 Suppl 1 : 5099-5101.

38. Piszkiewicz et al. 1986 Thrombosis Research 44:701-707.

39. Piszkiewicz et al. 1989 Curr Stud Hematol Blood Transfus 56:44-54.

40. Quinnan et al. 1986 Transfusion 26(5):481-483.

41. Race et al. 1991 Int Conf AIDS 7:180. 42. Race et al. 1995 Vaccine 13(1): 54-60.

43. Race et al. 1995b Vaccine 13:1567-1575.

44. Reed and Muench 1938 Am J Hygiene 27:493-497.

45. Resnick et al. 1986 JAMA 255:1887-1891.

46. Rossio et al. 1998 J Virol 72:7992-8001. 47. Salk 1953 JAMA 151 : 1081-1098.

48. Salk et al. 1993 Science 260:1270-1272. 49. Sattentau 1995 AIDS 9:1383-1385.

50. Schultz et al. 1990 Vaccine 8:516-517.

51. Schwiebert et al. 1994 J Virology 68(9): 5538-5547.

52. Shadle et al. 1995 Vytotechnology 18:21-25. 53. Sheets and Goldenthal. 1998 AIDS Res Hum Retroviruses 14(7):627-633.

54. Spire et al. 1984 The Lancet 899-901.

55. Stahl-Hennig et al. 1993 AIDS 7:787-795

56. Starcich et al. 1986 Cell 45:637-648.

57. Stott 1994 Current Topics in Microbiology and Immunology 188:221-237. 58. Tersmette et al. 1986 Vox Sang 51:239-243.

59. Trauger et al. 1994 J Infect Dis 169:1256-1264.

60. Trkola et al. 1996 70:1 100-1108.

61. Van Wezel 1984 Rev Infect Disease 6 Suppl 2:5335-5340.

62. Wiessenhorn et al., 1996 EMBO J 15: 1507-1514. 63. Wiessenhorn et al., 1997 PNAS 94:6065-6069.

64. Winkelman et al. 1989 Curr Stud Hematol Blood Transfus 56:55-69.

65. Wyatt et al. 1995 J Virology 69:5723-5733.

66. Wyatt et al. 1998 Nature 393:705-711.

67. Yamamoto et al. 1993 J Virol 67(l):601-605. 68. Zwart et al. 1991 J Virologyl81 :481-489.

The above references are incorporated by reference herein.

Claims

CLAIMS:What is claimed is:

1. A method for inactivation of a retrovirus preparation comprising the steps of:

a. heating said preparation until infectivity of said preparation is abolished, and

b. cooling said preparation before antigenicity of said preparation is abolished,

whereby a non-infectious but antigenic preparation is provided.

2. A method for heat inactivation of a retrovirus preparation comprising the steps of:

a. supplying energy to said retrovirus preparation until said preparation reaches specified elevated temperature sufficient to destroy infectivity of said preparation, said elevated temperature in the range from about 50° C. to about 70° C, preferably 56° C. to 70° C,

b. maintaining said retrovirus preparation at said elevated temperature for an initial period of time sufficient to abolish infectivity of said retrovirus preparation, said initial period ranging from about three minutes to about thirty minutes, preferably about ten to about thirty minutes,

c. maintaining said retrovirus preparation at specified elevated temperature for one or more additional periods of time sufficient to reduce to a specified level the probability that infectivity of retrovirus in said preparation will be destroyed, and

d. reducing temperature of said retrovirus preparation so that antigenicity of said retrovirus preparation is largely preserved or enhanced,

whereby a safe, non-infectious, antigenic retrovirus preparation will be provided.

3. A method of treating a retrovirus preparation to render it incapable of causing infection without destroying its capability for immunologic functions, consisting of heating said preparation for about to about 300 minutes, preferably about 30 to 60 minutes, to temperatures in the range from about ° C. to about 70° C, preferably about 56° C. to 70° C.