WO2000057906A1

WO2000057906A1 - Human papillomavirus vaccine with disassembled and reassembled virus-like particles

Info

Publication number: WO2000057906A1
Application number: PCT/US2000/007595
Authority: WO
Inventors: David B. Volkin; Henryk Mach; Li Shi
Original assignee: Merck & Co., Inc.
Priority date: 1999-03-26
Filing date: 2000-03-22
Publication date: 2000-10-05
Also published as: US20010021385A1; LU91326I9; FR07C0026I1; NL300270I1; DE122007000033I1; LU91328I9; LU91327I9; CY2007010I1; LU91326I2; US6245568B1; AR023166A1; CY2007010I2; KR20020013516A; CA2368391A1; CY2007008I1; EP1165126A1; DK1165126T3; DE60030698T2; DE122007000034I1; DE122007000032I1

Abstract

Human Papillomavirus vaccine formulations which contain virus-like particles (VLPs) can be made more stable and have an enhanced shelf-life, by treating the VLPs to a disassembly and reassembly process. Also provided are formulation buffers to long term stable storage of VLPs.

Description

HUMAN PAPILLOMA VIRUS VACCINE WITH DISASSEMBLED AND REASSEMBLED VIRUS-LIKE PARTICLES

BRIEF DESCRIPTION OF THE INVENTION

This invention relates to a human papillomavirus (HPV) vaccine which contains virus-like particles (VLPs) which have been disassembled into capsomeres and then reassembled into VLPs This invention also relates to processes of making this vaccine resulting in more homogeneous HPV VLPs and greatly improved storage stability.

BACKGROUND OF THE INVENTION

Human Papillomavirus (HPV) infects the genital tract and has been associated with various dysplasias, cancers, and other diseases. These diseases are currently targets for vaccine development and vaccines containing virus-like particles (VLPs) which contain LI or the combination of L1+L2 proteins are currently in clinical tπals.

It has been found, however, that recombmant LI protein HPV VLPs puπfied from yeast are not stable duπng long-term storage, either in solution or when adsorbed onto aluminum adjuvant particles.

In the past, vaπous researchers have investigated the conditions of HPV VLP assembly and disassembly. For example, McCarthy et al, 1998 "Quantitative Disassembly and Reassembly of Human Papillomavirus Type 11 Viruslike Particles in Vitro" J Virology 72(1):32-41, descπbes the disassembly and reassembly of recombmant LI HPV 11 VLPs puπfied from insect cells in order to obtain a homogeneous preparation of VLPs. A prolonged incubation (about 16 hours at 4°C) with a relatively high concentration of reducing agent at physiological ionic strength was used to disassemble the VLPs, and removal of the reducing agent at a higher ionic strength was used to reassemble the VLPs. This method is quite time- consuming, however.

In order to develop a commercially useful vaccine, a storage stable formulation is needed. It would be desirable to have a relatively simple, quick and quantitative treatment procedure for making a storage stable HPV VLP formulation DETAILED DESCRIPTION OF THE INVENTION

This invention relates to a process for making stable human papillomavirus (HPV) virus-hke particles (VLPs), the process compπsing the steps of:

(a) incubating VLPs in a dissociation mixture comprising a relatively low concentration of a reducing agent, a salt present in a range from physiological ionic strength up to 1.25 M , a non-ionic surfactant, a metal chelatmg agent and a buffer until disassembled VLPs are produced;

(b) removing the reducing agent from the dissociation mixture; and

(c) reassembling the disassembled VLPs into VLPs.

In some embodiments, the reassembled VLPs are adsorbed onto an aluminum adjuvant.

This invention also relates to VLPs made by a process compπsing the steps of:

(a) incubating puπfied VLPs in a high salt dissociation mixture compπsing a relatively low concentration of a reducing agent, a salt present in a range of 0.5 M to 1.25 M salt, a non-ionic surfactant, a metal chelatmg agent, and a buffer until disassembled VLPs are produced;

(b) removing the reducing agent from the dissociation mixture; and

(c) reassembling the disassembled VLPs; and

(d) optionally adsorbing the reassembled HPV VLPs to aluminum adjuvant.

Another embodiment of this invention relates to VLPs made by a process comprising the steps of:

(a) incubating puπfied VLPs in a low salt dissociation mixture compπsing a relatively low concentration of a reducing agent, a salt present at approximately physiological ionic strength, a non-ionic surfactant, a metal chelatmg agent, and a buffer until disassembled VLPs are produced;

(b) removing the reducing agent from the dissociation mixture; and

(c) reassembling the disassembled VLPs; and

(d) optionally adsorbing the reassembled HPV VLPs to aluminum adjuvant

Another aspect of this invention relates to vaccines made from the VLPs produced by any of the above processes. This invention also relates to vaccine formulations compnsing VLPs made by the above processes, and which are stored in a formulation buffer The formulation buffer compπses a salt, a histid e buffer, and a nonionic surfactant

BRIEF DESCRIPTION OF THE FIGURES

FIGURE 1 is a graph showing disassembly and reassembly of HPV 16 VLPs as determined by analytical ultracentπfugation analysis.

FIGURE 2 is a graph showing particle size distπbutions of dis/ reassembled HPV 16 VLPs (dashed line) and untreated HPV 16 VLPs (solid line) as determined by analytical ultracentπfugation analysis.

FIGURE 3 is a graph showing accelerated storage stability (37°C) of HPV 16-alumιnum vaccine formulations as determined by in vitro antigemcity via BIAcore analysis

FIGURE 4 shows the accelerated storage stability of HPV 16 VLPs in physiological salt solution as determined by in vitro antigemcity via BIAcore analysis.

FIGURE 5 shows the effect of dis/reassembly treatment on the thermal stability of different lots of HPV 16 VLPs as determined by momtonng thermal- induced aggregation by UV turbidity (cloud point analysis).

FIGURES 6A-D show the accelerated storage stability (37°C) of four different types (HPV 16, 11, 6a, 18) of dis/reassembled HPV VLPs as well as untreated control HPV VLPs in solution as determined by in vitro antigemcity via BIAcore analysis. FIGURE 6A is HPV 6A; FIGURE 6B is HPV 11; FIGURE 6C is HPV 16; and FIGURE 6D is HPV 18.

FIGURES 7A-D show the accelerated storage stability of four different types (HPV 16, 11, 6a, 18) of dis/reassembled HPV VLPs as well as untreated control HPV VLPs absorbed on aluminum adjuvant. The samples were assayed for in vitro antige city via BIAcore analysis after treated in citrate solution to release the antigen from aluminum adjuvant. FIGURES 7 A and 7B are HPV 6 A; FIGURES 7C and 7D are HPV 11; FIGURES 7E and 7F are HPV 16; and FIGURES 7G and 7H are HPV 18 VLPs

FIGURES 8A-F are compansons of thermal stability and hydrodynamic size distπbution of disassembled/reassembled HPV 6a, HPV 11 and HPV 16 VLPs, prepared from a disassembly process either near physiological conditions (low salt process) or at higher salt concentrations (high salt process). These physical properties of the HPV VLPs were determined by cloud point analysis and analytical ultracentπfugation analysis. In the low salt process, HPV VLPs are disassembled in 0.166 M NaCl, 10 mM TRIS solution (pH 8.2) while in the high salt process, HPV VLPs are disassembled in 0 63 M NaCl, 35 mM Phosphate, and 100 M TRIS solution (pH 8 2) Both solutions also contain EDTA and polysorbate 80 FIGURES 8A and 8B show the cloud point and analytical ultracentπfugation data of HPV 6A VLPs; FIGURES 8C and 8D show the same analysis with HPV 11 VLPs, FIGURES 8E and 8F show the same analysis with HPV 16 VLPs

FIGURES 9A-D are the particle size distπbution and surface affinity of untreated and dis/reassembled HPV VLPs as measured by SEC-HPLC analysis. FIGURE 9A is HPV 6a; FIGURE 9B is HPV 1, FIGURE 9C is HPV 16; and FIGURE 9D is HPV 18.

This invention relates to a process for producing more stable HPV VLPs which have been found to contribute significantly to the overall stability of an HPV vaccine formulation. The puπfied HPV VLPs are produced, disassembled into presumably capsomeres, reassembled into VLPs, and then used as the active ingredient in a vaccine formulation.

In accordance with this invention, any type of HPV VLPs may be used as the antigemc portion of the vaccine formulation. The VLPs may contain only LI protein, or may contain both LI and L2 protein. The proteins may be of a wild-type ammo acid composition, or they may contain mutations VLPs containing only LI protein which is of a wild-type ammo acid composition are preferred.

The HPVs which are preferred are those associated with disease, including but not limited to. HPV 1, HPV 2, HPV 3, HPV 4, HPV 6a, HPV 6b, HPV 7, HPV 10, HPVl l, HPV 16, and HPV 18, HPV 34, HPV 39, HPV 41-44 and HPV 51-55 Preferred HPVs include HPV 6a, HPV 6b, HPV 11, HPV 16, and HPV 18. In addition, the formulations of this invention are suited for combinations of HPV types, including multivalent vaccines containing a plurality of HPV antigens, such as a combination of HPV 6, 11, 16 and 18

It is preferred that the VLPs be made by recombmant techniques, as is known in the art. It is particularly preferred that the host cell used to make the VLPs is a yeast cell, although other cell types, such as bacteπal, insect and mammalian cells are known and currently used as hosts While not wishing to be bound by theory, it appears that the status of disulfide bonds in the LI protein in a VLP may differ depending on the host cell used to express the recombmant LI as well as the subsequent puπfication method McCarthy et al 1998, supra descπbes some disulfide cross nked LI tπmers in recombmant VLPs produced insect cells On the other hand, in accordance with this invention, the disulfide bond status in the LI protein of VLPs produced in yeast may differ and thus require different conditions for optimal disassembly and reassembly.

For some formulations, an aluminum absorbed product is desired. Generally in this case, the concentration of HPV VLPs which are adsorbed onto aluminum is from about 10-200 mcg/mL for each HPV VLP type. This may be adjusted, depending on such factors as antige city of the particular type of HPV, and the presence of multiple types of HPVs in a "cocktaιl"-type of vaccine.

Disassembly

After the VLPs are produced in the recombmant host cell and puπfied, they are disassembled by incubating them in a dissociation mixture. Disassembly is pπmaπly dπven by the reduction of disulfide bonds with a reducing agent such as dithiothreitol (DTT) in a relatively wide range of ionic strength environments varying from physiological (also referred to as the "low salt process") up to 1.25 M NaCl (also referred to as the "high salt process"). The dissociation mixture compπses a reducing agent, a salt, a nonionic surfactant, a metal chelatmg agent and a buffer.

The reducing agent is preferably dithiothreitol (DTT), although other reducing agents are known and can be used. In the past, others have used other reducing agents (such as beta-mercaptoethanol, β-ME) to disassemble HPV particles (see, e.g McCarthy et al, supra), but they have employed a relatively high concentration of reducing agent (5%, or 713 mM of β-ME). In contrast, one aspect of the present invention is the use of a relatively low concentration of the reducing agent

For purposes of the specification and claims, the term "relatively low concentration of a reducing agent" means from about 2 mM to 20 mM if using DTT as the reducing agent, or if an alternate reducing agent is employed, the amount of the alternate reducing agent which would have approximately the same effect as employing 2-20 mM DTT. A preferred amount of reducing agent is about 10 mM DTT

Another important component of the dissociation mixture is the salt. Generally, the ionic strength of the dissociation mixture is maintained by the presence of salts. Almost any salt which can contπbute to the control of the ionic strength may be used Preferred salts which can be used to adjust ionic strength are any physiologically acceptable salts, such as NaCl, KC1, Na2SO4, (NFLi^SOzi., sodium phosphate, sodium citrate and mixtures thereof Particularly preferred salts are. NaCl, KC1, and Na2SO4_, and especially NaCl. In a low salt embodiment of this invention, the dissociation mixture should have an ionic strength which is approximately physiological. For purposes of this specification and claims, the term "approximately physiological" when referπng to a salt, means 0.10-0.20 M, and preferably about 0.15-0.18 M, and more preferably about 0 16 M when employing NaCl as the salt. For other salts which may be used, it is within the skill of the artisan to calculate molanty which would be the equivalent of an approximately physiological NaCl solution.

Alternatively, the dissociation can take place in a high salt environment. As used herein, "high salt" means at least 0.5M salt, and at least 10 mM buffenng agents. A preferred range for high salt is 0.5 to 1.25 M, if the salt used is NaCl, and a more preferred range is 0.60 to 0.1 M. For other salts, it is withm the skill of the artisan to determine molanty which would be the equivalent of a 0.5M and 1.25 M NaCl solution.

The successful use of a high salt disassembly procedure was quite unexpected — literature reports such as Brady et al, 1977 Virology 23: 717-724, Belnap et al 1995 J. Mol. Biol. 259: 249-263 and published patent application WO 99/13056 limit salt concentration to less than 0.5M NaCl. Nonetheless, it has been found that the use of a high salt disassembly is advantageous in that the high salt limits possible protein aggregation, results in improved protein mass recovery, and also allows the dis/reassembly treatment to be initiated at the final purification chromatography step, thereby reducing the number of processing steps needed duπng puπfication.

It has been surpnsingly found that a preferred high salt solution (such as 1.0 M NaCl, 10 mM TRIS buffer, pH 8.2, 2 mM EDTA, 0.03% polysorbate 80 and 2-20 mM DTT) results in approximately 100% HPV protein mass recovery in the process. Another high salt solution (0.63M NaCl, 35 mM Phosphate buffer, 100 mM TRIS buffer, pH 8.2, 2 mM EDTA, 0.03% polysorbate 80 and 2-20 mM DTT) is also preferred.

In order to better understand and characteπze the effect of the high salt concentration on the effectiveness of HPV VLP disassembly, comparative studies under both high (0.5-1.0M NaCl) and low (physiological) salt disassembly conditions were carried out. Changes in static light scatteπng intensity were monitored to evaluate the VLP disassembly. It was observed that the results were quite similar for both high salt and physiological salt conditions — complete VLP disassembly occurred within about two minutes of (20 mM) DTT addition. Further the kinetics of disassembly seems controlled by the concentration of reducing agent rather than by salt concentration, with the higher concentration of reducing agent correlating with a faster disassembly at a given temperature and pH

Other biophysical properties of VLPs disassembled and reassembled in either physiological or high salt were examined Thermal stability and hydrodynamic size distributions for disassembly/reassembly VLPs generated under both high and low salt conditions are similar. However, the total protein mass recovery was relatively higher (close to 100%) for the disassembled/reassembled HPV VLPs generated with high salt disassembly process.

Another component of the dissociation mixture is a non-ionic surfactant The non-ionic surfactant may be selected from the group consisting of Polysorbate 80, Polysorbate 20, polyoxyethylene alkyl ethers, Tπton X-100®, Tnton X-l 14®, NP-40®, Span 85 and a member of the Pluronic seπes of non-ionic surfactants Polysorbate 80 is particularly preferred A preferred concentration of Polysorbate 80 is from about 0.01-0.50%, and preferably about 0.01-0.10%, and especially about 0.03%.

The dissociation mixture should be pH controlled by the presence of a buffer The disassembly of HPV VLPs can take place at a wide range of basic pH such as from above 7.0 to 10. A preferred pH is 8.0-8.5, and preferably about 8.2. A preferred buffer is TRIS at 5-100 mM, preferably 5-15 mM, although other buffers such as phosphate may be used. TRIS is generally preferred, as it provides good pH and ionic strength control in the dissociation mixture conditions Optionally phosphate buffer may be added at 0-50 mM

A metal chelatmg agents is also present in the dissociation mixture to ensure complete disassembly of VLPs. A preferred metal chelatmg agent is EDTA at 0.5 to 5 mM, and preferably about 2 mM, although other known chelatmg agents may be used.

A particularly preferred dissociation mixture compπses approximately 300 mcg/mL HPV VLP protein m a 10 mM TRIS buffer, pH 8.2, additionally containing 0.16M-0.18M NaCl, 10 or 20 mM DTT, 2 mM EDTA and 0.03% Polysorbate 80. Alternatively, the dissociation mixture contains the same amount of protein in 0.63M NaCl, 35 mM phosphate, 100 mM TRIS buffer, pH 8.2, 10 or 20 mM DTT, 2 mM EDTA and 0 03% Polysorbate 80.

One advantage to the process of this invention is that the disassembly step is fairly rapid, and only requires incubation of the VLPs in the dissociation mixture for less than about one hour at room temperature, and times as short as 30-40 minutes can be used If desired, the dissociation step (as well as other steps of the disassembly/ reassembly process) may be performed under steπle conditions. For example, the buffer components may be sterile filtered, and used with sterile protein solutions. In addition, the dis/reassembled HPV VLPs in solution may be steπle filtered pπor to use or pnor to adsorption to aluminum adjuvant.

After disassembly of VLPs is complete, the reducing agent should be removed This may be accomplished through the use of a dialysis step or a diafiltration/ultrafiltration step. The dialysis step comprises a dialysis against a solution of salt (physiological strength or, if the high salt process was used, higher concentration of salt such as 1M), non-ionic surfactant, and buffer similar to that present in the dissociation mixture, but at a lower pH than is present m the dissociation mixture. Recommended pH ranges in this dialysis step are from about 6.5-7.5, and preferably about 7.0 One preferred dialysis solution compnses 0.16-0.18 M NaCl, 0.01-0.03% polysorbate 80, 10 mM phosphate at pH 7 0 Another example of a recommended dialysis is a 1: 100 dialysis (three changes, 30 minutes each) against a solution of 0.166M NaCl, 10 mM phosphate buffer, pH 7.0 Alternatively, a preferred solution is 1.0 M NaCl, 10 mM phosphate and 0.03% polysorbate 80, pH 7.0

If needed, additional non-ionic detergent such as polysorbate 80 can be added to the buffer used to remove DTT in order to maintain a sufficient level of nonionic detergent with the protein throughout the process.

Alternatively, the reassembly solution descnbed below could be used to remove the DTT.

With larger volumes, a scalable diafiltration or ultrafiltration setup can be used in place of dialysis procedure

Reassembly

The next step is to reassemble VLPs. This may be accomplished dunng the dialysis step descnbed above, dunng the diafiltration/ultrafiltration step described above, or dunng a separate, or additional dialysis step, and may take several hours up to about 24 hours. If using a separate dialysis, the dialysis is accomplished using a reassembly buffer comprising: ionic strength salt in the range of 0.5-1.35M NaCl, a metal ion source, a buffer at pH 6.0-6.5.

The salt is preferably the same salt as has been used in the previous steps, although its concentration is preferably higher. For example, if NaCl is used, it should be present at a concentration of 0.5-1.35M, and preferably about 1.0 M. The source of metal ions should be a Ca+2 or a Mg+2 source, such as CaCl2, or MgCl2 and is present presumably to provide stability to the reassembled

VLPs. Zinc may also be used, but typically is less advantageous than the other ions. CaCl2 is preferred The amount of metal ion should be present is a concentration of about 1-10 mM, preferably about 2 mM.

Glycine or sodium citrate are added as a reassembly buffer component and/or a pH controller. Amounts range from about 20-70 mM, preferably about 50 mM.

The buffer can be the combination of glyc e and phosphate, or citrate and phosphate, or citrate alone. The use of phosphate buffer alone is not recommended A preferred reassembly buffer is 1M NaCl, 2 mM Ca+2, 50 mM citrate, pH 6.2 and 0.03% polysorbate 80

If needed, additional non-ionic detergent such as polysorbate 80 can be added to the reassembly buffer to maintain a sufficient level of nonionic detergent with the protein throughout the process

Buffer Exchange

Finally, any remaining reagents may be removed by dialysis using a final dialysis buffer. One example contains salt (preferably NaCl), and non-ionic surfactant, and optionally histidme. The salt, if NaCl, is preferably present at about 0.25M-1 M, more preferably about 0.5M Histidme may be present at 2-50 mM, preferably about 5-20 mM with a pH 6.0-6.5, and preferably about pH 6.2. The nonionic surfactant, such as polysorbate 80 or polysorbate 20 may be present at 0.01- 0.03%. An alternative preferred buffer exchange is 0.5M NaCl and 0.03% polysorbate 80.

These multiple dialysis steps used to dis/reassemble the HPV VLPs may be also performed with larger scale equipment such as diafiltration and/or ultrafiltration systems. If needed, additional non-ionic detergent such as polysorbate 80 can be added to any of the dis/reassembly buffers to maintain a sufficient level of nonionic detergent with the protein throughout the process.

If desired, the dis/reassembled HPV VLPs may be adsorbed onto an aluminum adjuvant using known techniques. Optionally, the stability of the reassembled VLP formulation may be further enhanced by the addition of polyamonic, polymeπc excipients. As used throughout the specification and claims, the term "polya onic polymer" is meant to refer to compounds which have either a single long chain, or those with multiple cross linked chains; either type possessing multiple negative charges along the chaιn(s) when in solution. Examples of polyamonic polymers include: proteins, polyamons, peptides and poly-nucleic acids Specific stabilizers may be selected from the group consisting of: carboxymethyl cellulose (particularly 10-800cps),hepaπn (6-30 kDa), poly-ammo acids (2-100 kDa) such as poly(Glu), Poly(Asp), and Poly(Glu, Phe), oxidized glutathione [Glu-Cys-Gly]2 (613 Da), poly-nucleotides such as polycytidylic acid (200-700 kDa), and polyadenyhc acid (200-700 kDa), RNA, DNA, and serum albumins

The concentration of the polyamonic polymenc excipient, when present, is from about 0.01% to about 0.5%, particularly about 0.05-0.1% (by weight/volume), although the addition of even a ten fold lower amount of polyamonic excipients (for example, 0.01% albumin, DNA or hepaπn) still provides enhanced stability to untreated HPV VLP-aluminum formulations.

A typical vaccine formulation includes:

1) aluminum absorbed product containing 10-200 mcg/mL of each HPV VLP type and 0.15-0.32M saline; 2) 5-10 mM histidme, pH 6.2; and 3) 0.005- 0.03% non-ionic surfactant.

The resulting HPV vaccine formulations of this process are stable at 2- 8°C to room temperature for at least 6 months (study ongoing) as illustrated in FIGURE 7.

Another aspect of this invention is the use of a formulation buffer to provide for stable, long-terms storage of vaccines made with the disassembled/ reassembled VLPs. The formulation buffer compnses 0.15-0.32M NaCl, 5-10 mM histidme, pH 6.2, and 0.005-0.015% polysorbate 80. These vaccine formulations are stable at various temperature ranges (from 4-30°C) for up to at least six months. The following non-limiting Examples are presented to better illustrate the invention.

EXAMPLE 1 A frozen solution of yeast denved recombmant LI protein HPV 16 VLP (greater than 95% protein puπty) in 0.5M NaCl, 0.003% Polysorbate 80, pH targeted at approximately 6.2, was used for most of these expenments. HPV 6a VLPs and HPV 11 VLPs were in 0.5M NaCl, 0.03% Polysorbate 80 The HPV 18 VLPs were at 0.5 M NaCl, 0.01% Polysorbate 80. Some of the studies utilized HPV 6a, 11, and 16 VLPs in 75 mM phosphate buffer at pH 7 containing 1.25 M NaCl.

Release of HPV VLPs from aluminum adiuvant for subsequent in vitro analysis. To release HPV VLPs from the aluminum adjuvant, an aluminum dissolution method was developed which included dilution of HPV-alum um formulation into a high salt solution containing citrate and polysorbate 80 The HPV VLP samples from the aluminum dissolution method are directly subjected to an in vitro antigemcity assay using BIAcore analysis

In vitro antigemcity assay. The HPV VLP samples from the aluminum dissolution method were in a stabilizing solution of pH 6-6.5 and high ionic strength (about 0.5M- 1M NaCl, 0.1M sodium citrate) containing 0.02% Polysorbate 80. The samples are directly subjected to BIAcore analysis without further dilution (utilizing HPV VLP type specific neutralizing antibody). Solution samples (no aluminum adjuvant) were diluted to a target concentration in about 0.5- 1M NaCl before BIAcore analysis. All in vitro antigemcity data are referenced to a frozen HPV VLP control sample (not aluminum adsorbed).

Protein concentration analysis. The protein concentration of HPV VLPs in both bulk solutions and formulated samples (after aluminum dissolution) were determined by UV absorbance spectra measurement at ambient temperature using a HP 8452A Diode Array spectrophotometer and a cuvette with a path length of 1 cm. The sample volumes used were 100-250 microhters The protein concentration was calculated using a multi-component second deπvative analysis technique. For some expenments, protein concentration was also determined by BCA coloπmetnc analysis.

Hydrodynamic size analysis. The hydrodynamic size of untreated and dis/reassembled HPV VLPs were determined by dynamic light scatteπng, analytical ultracentπfugation, and SEC-HPLC. Dynamic light scatteπng measurements were performed at ambient temperature using a Malvern 4700 Light Scatteπng System at 90°. The apparent hydrodynamic size of HPV VLPs was determined as Z-average hydrodynamic diameter. Each of the values represents the mean of five measurements of the same sample.

Sedimentation velocity expenments were performed using Beckman XLI analytical ultracentπfuge with UV detection at 280 nm Two different methods were used to determine the sedimentation coefficient including fixed and vaπable speed modes.

SEC-HPLC measurements were performed at ambient temperature using a Waters/Millennium 2690 system with UV and fluorescence detectors. A PL- GFC 4000 A column (preconditioned with HPV VLPs) was used at a flow rate of 0.4 ml/minute using 750 mM NaCl, 25 mM phosphate buffer ( pH 7) as mobile phase

Electron microscopic analysis. Transmission Electron Microscopy (TEM) was performed using negative staining. Samples were fixed and stained with phosphotungstic acid and examined in a JEOL 1200 EX Transmission Electron Microscope. Micrographs were taken of random areas with samples prepared multiple times at a magnification between 30,000x and 40,000x. An additional 3-fold magnification was introduced m developing the pπnts from the negatives.

Real time and accelerated storage stability studies. HPV-alummum formulation storage stability studies were earned out. The temperature of stability studies was generally 2-8, 15, 25, 30 and 37°C. Previous conformational integnty data (not shown) with circular dichroism and fluorescence spectroscopy has shown that increasing the temperature above 40-45°C induces significant conformational changes in the LI protein of HPV VLP, a condition which definitely needs to be avoided in accelerated storage stability studies.

HPV VLP thermal stability was also evaluated with turbidity assays which were earned out using a HP 8452A Diode Array spectrophotometer equipped with a HP 845X UV-Visible system software and a temperature control system. The light scatteπng of the solutions (due to protein aggregation) was followed at 350 nm under the kinetic mode of the program by increasing the temperature from 24°C to 74° C at a controlled rate. Mouse potency testing- In vivo immunogenicity of HPV VLPs was evaluated in BALB c mice. HPV VLPs samples were adsorbed onto aluminum adjuvant, diluted to several target concentrations with aluminum adjuvant and injected into mice Sera were collected and analyzed for antibody levels by ELISA assay

EXAMPLE 2 Disassembly and reassembly of HPV 16 VLPs as determined by analytical ultracentπfugation analysis.

As seen in FIGURE 1, HPV 16 VLPs are disassembled after incubation in a pH 8.2 buffer containing 10 mM TRIS or phosphate, 0.166 M NaCl, 2 mM EDTA, and 2 mM DTT at room temperature for 20 minutes (middle curve) The disassembled HPV 16 sample shows a peak at smaller S value expected for LI capsomeres, suggesting a size of HPV LI protein pentamers. The top curve shows that the disassembled HPV 16 (capsomeres) was then reassembled into VLPs by dialyzmg into a pH 6.2 buffer containing 10 mM sodium phosphate, 0.5 to 1M NaCl, 2 mM calcium chloride in the presence of either glycine or citrate. The bottom curve shows the untreated HPV 16 VLPs

EXAMPLE 3 Particle size distnbutions of dis/reassembled HPV 16 VLPs

FIGURE 2 shows particle size distributions of HPV 16 VLPs which were disassembled and reassembled as descnbed in Example 2 (dashed line) and untreated HPV 16 VLPs (solid line) as determined by analytical ultracentnfugation analysis The data reveal that dis/reassembled HPV 16 VLPs are larger and have a more homogenous distribution, as judged from the positions and widths of the peaks, respectively. Electron microscopy (EM) analysis confirms the analytical ultracentπfugation analysis; it shows that the dis/reassembled VLPs are more homogenous and exist as rounder particles of uniform size with mean diameter of approximately 80-85 nm (EM data not shown) Untreated samples appear to be smaller and with more heterogeneous shape and size distnbution with an mean size around 50-68 nm. Similarly, a later set of data shows untreated HPV 16 VLPs to have a mean size of approximately 40 nm, while the reassembled HPV 16 VLPs have a mean size of approximately 60 nm

EXAMPLE 4 Accelerated stability (37°C) of HPV 16 VLP-aluminum vaccine formulations as

π determined by BIAcore analysis.

The dis/reassembled HPV 16 VLPs as well as untreated HPV 16 VLPs were adsorbed onto aluminum adjuvant (at 160 mcg/mL protein and 450 mcg/mL Al) and formulated in 0.32M NaCl, lOmM histidine, 0.015% polysorbate 80, pH 6.2 The in vitro antigemcity of the samples were assayed at identified times by BIAcore analysis.

In FIGURE 3, disassembled and reassembled HPV 16 VLPs are shown with filled circles and open squares; and untreated are filled diamonds. Run 1 contained citrate and Run 2 has glycme in the reassembly buffer. The data demonstrate the dis/reassemble process results in a significant and dramatic enhancement in accelerated storage stability of aluminum adjuvant adsorbed HPV 16 VLPs as compared to untreated HPV VLPs No loss in in vitro antigemcity of aluminum adsorbed dis/reassembled HPV VLPs was observed in similar expenments performed at 25°C and 4°C after three months.

EXAMPLE 5 Accelerated stability (37°C) of HPV 16 VLPs in solution in physiological salt solution as determined by BIAcore analysis.

80 mcg/ml of dis/reassembled HPV 16 VLPs as well as untreated HPV 16 VLPs were incubated in 0.15M NaCl, lOmM histidme, 0.015% polysorbate 80, pH 6.2 at 37°C. The in vitro antigemcity of the samples were assayed at identified times by BIAcore analysis. The data demonstrate the dis/reassemble process results in a significant enhancement in accelerated storage stability of HPV 16 VLPs as compared to untreated HPV VLPs This is shown in FIGURE 4, where disassembled and reassembled VLPs are in open circles and open squares, untreated VLPs are shown in filled diamonds. Run 1 contains citrate and Run 2 has glycine in the reassembly buffer.

EXAMPLE 6

Effect of dis/reassembly process treatment on the thermal stability of different lots of HPV 16 VLPs as determined by momtonng thermal-induced turbidity formation (aggregation) by UV spectroscopy (cloud point analysis).

A standard cloud point protocol was applied to the samples: heat from 24°C to 74°C at a controlled rate with optical density of the solution being monitored at 350 nm As shown in FIGURE 5, three different lots of HPV 16 VLPs (lots 1,2,3) were analyzed before (dash lines and open symbols) and after (solid lines and filled symbols) dis/reassembly treatment. Among these samples HPV 16 VLPs (lot 1) was tested three times. The data demonstrate that dis/reassembly process treatment results in a significant enhancement of the intnnsic stability of HPV 16 VLPs against heat- mduced aggregation. In addition to the thermal stability enhancements, the dis/reassembly process treatment also results in a more consistent and homogeneous stability profile

EXAMPLE 7

The in vitro antigemcity of untreated and dis/reassembled was further evaluated with a lot of HPV 16 VLPs which both untreated and dis/reassembled VLPs were evaluated before and after adsorption to aluminum adjuvant. The in vitro antigemcity was measured by BIAcore analysis and the protein concentration by UV spectroscopy and BCA coloπmetnc assay. The antigen to protein ratio was then determined for both untreated and dis/reassembled HPV 16 VLPs. The in vitro antigemcity of the dis/reassembled HPV 16 VLPs was enhanced by about 50% For example, the antigen to protein ratio for dis/reassembled vs. untreated HPV 16 VLPs was 1.7 with a range of 1.5-1.9. This observation was confirmed by EIA analysis (mean of 1.6 and range of 1.4-1.8) and IVRP analysis (mean 1.4 and range of 1.3-1.5).

The in vivo immunogenicity of untreated and dis/reassembled aluminum adsorbed HPV 16 VLPs was evaluated with a mouse potency test. This test generates an ED50 value representing the dose (meg) of HPV VLPs in which more than 50% of the mice seroconvert. The in vivo immunogenicity of the dis/reassembled HPV 16 VLPs was equivalent to, or better than, the untreated HPV 16 VLPs One experiment shows an ED50 value of 0.034-0.062 for two dis/reassembled samples vs. an ED50 value of 0.146 for the untreated sample. In a second mouse expeπment, an ED50 of <0.0125 was obtained for three dis/reassembled HPV VLP samples vs. an ED50 value of 0.074 for the untreated sample.

EXAMPLE 8 Accelerated stability (37°C) of untreated and dis/reassembled HPV VLPs in solution and on aluminum adjuvant as determined by BIAcore analysis.

80 mcg/mL HPV VLPs were incubated in solution containing 0.32M NaCl, lOmM histidme, 0.015 % Polysorbate 80, pH 6.2 at 37°C. Samples were assayed for in vitro antigemcity by BIAcore at times indicated on FIGURE 6 The data indicate that the dis/reassembly treatment results in a dramatic stability enhancement for HPV VLP types 6a, 11, and 16 in solution duπng accelerated storage stability testing. The untreated HPV 18 VLPs show a stability profile similar to the dis/reassembled VLPs for the other three types No significant stability enhancement is observed with treatment of the HPV 18 VLPs probably due to the inability of the dis/reassembly treatment to affect HPV 18 VLPs (data not shown). The samples used for this study are generated with low salt disassembly process. The protein mass recovery for the four types across the dis/reassembly treatment was approximately 85- 95% for HPV 11, 16 and 18 VLPs and approximately 60-70% HPV 6a VLPs The yield across dis/reassembly treatment seems to be affected by the quality of the specific lot of HPV VLP in terms of VLP aggregation. The protein mass yield across dis/reassembly treatment increases to nearly 100% when the disassembly process take places under high salt conditions. Analysis of the dis/reassembled HPV VLPs by analytical ultracentπfugation suggests that the dis/reassembly treatment results in nearly quantitative reassembly of capsomeres into VLPs.

FIGURE 7 shows the storage and accelerated stability of HPV VLPs absorbed on aluminum adjuvant in which 400 mcg/mL HPV VLPs were adsorbed on 450 mcg/mL aluminum adjuvant and incubated in a solution containing 0.32M NaCl, lOmM histid e, 0.015 % Polysorbate 80, pH 6.2 at 4°C, 15°C, 25°C, 30°C and 37°C. Samples were assayed for in vitro antigemcity by BIAcore at the times indicated on the figure after treatment in a citrate solution to release the antigen from the aluminum adjuvant The data indicate that the dis/reassembly treatment results in a dramatic stability enhancement for aluminum adsorbed HPV VLP types 6a, 11, and 16 in accelerated storage stability testing. The untreated HPV 18 VLPs show a stability profile similar to the dis/reassembled VLPs for the other three types. No significant stability enhancement is observed with disassembly/reassembly treatment of the HPV 18 VLPs.

EXAMPLE 9 Thermal stability and hydrodynamic size distribution of reassembled HPV 6a, HPV 11 and HPV 16 VLPs, prepared by both the low salt disassembly process and high salt disassembly process. Samples were evaluated by cloud point and analytical ultracentnfugation analysis. In the low salt process, HPV VLPs are disassembled in 0.166 M NaCl, 10 mM TRIS solution (pH 8.2) while in the high salt process, HPV VLPs are disassembled in 0 63 M NaCl, 35 mM Phosphate, and 100 mM TRIS solution (pH 8.2). Both solutions also contain EDTA and polysorbate 80. The data, as seen in FIGURE 8 show that the disassembled/reassembled HPV VLPs prepared by the low and high salt processes are similar regarding their thermal stability and hydrodynamic size distπbution for HPV VLPs types 11, 6a and 16.

EXAMPLE 10 Particle size distnbution and surface affinity of HPV VLPs as measured by SEC- HPLC analysis

Dis/reassembled (FIGURE 9, solid line) and untreated (dash line) HPV VLPs (Types 11, 16 and 6a) were analyzed on 4000A GPC size exclusion column. Single peak was obtained from dis/reassembled HPV VLPs compared to a more heterogeneous distribution of untreated HPV VLPs. The monomer peaks of dis/reassembled HPV VLPs have a smaller elution volume than the monomer peaks of untreated HPV VLPs suggesting a relatively larger particle size for the dis/reassembled VLPs. The relatively larger total peak areas of dis/reassembled HPV VLPs indicate a higher recovery rate of dis/reassembled VLPs suggesting that dis/reassembled VLPs have lower affinity to the column than untreated VLPs. Untreated HPV 18 VLPs show a SEC HPLC profile similar to the dis/reassembled VLPs for the other three types.

Claims

WHAT IS CLAIMED IS-

1 A process for making a stable human papillomavirus (HPV) vaccine compnsmg HPV virus-like particles (VLPs), the process compnsmg the steps of:

(a) incubating VLPs in a dissociation mixture compnsmg a relatively low concentration of a reducing agent, a salt present in a range of approximately physiological ionic strength up to 1.25 M , a non-ionic surfactant, a metal chelatmg agent and a buffer until disassembled VLPs are produced,

(b) removing the reducing agent from the dissociation mixture; and

(c) reassembling the disassembled VLPs

2. A process according to Claim 1 wherein the reducing agent in the dissociation mixture is about 2-20 mM dithiothreitol (DTT).

3. A process according to Claim 1 wherein the salt in the dissociation mixture is selected from the group consisting of: NaCl, KC1, Na2SO4, (NH4)2SO4 sodium phosphate, sodium citrate and mixtures thereof.

4. A process according to Claim 1 wherein the non-ionic surfactant in the dissociation mixture is selected from the group consisting of: Polysorbate 80, Polysorbate 20, polyoxyethylene alkyl ethers, Tπton X-100®, Tπton X-l 14®, NP-40®, Span 85 and a Pluromc non-ionic surfactant.

5. A process according to Claim 4 wherein the non-ionic surfactant is 0.01-0.5% Polysorbate 80.

6. A process according to Claim 1 wherein the buffer in the dissociation mixture is TRIS or phosphate buffer, at pH 7-10.

7. A process according to Claim 1 wherein the metal chelatmg agent in the dissociation mixture is EDTA at 0.5-5 mM

8. A process according to Claim 1 wherein step (a) takes place for less than about one hour at room temperature.

9 A process according to Claim 8 wherein step (a) takes place for

30-40 minutes at room temperature

10. A process according to Claim 3 wherein the salt is NaCl.

11. A process according to Claim 10 wherein the salt in the dissociation mixture is 0.10M to-0.2M NaCl

12. A process according to Claim 10 wherein the salt in the dissociation mixture is 0.50 - 1.25M NaCl

13 A process according to Claim 1 wherein step (b) comprises a dialysis or diafiltration/ultrafiltration step.

14 A process according to Claim 13 wherein the dialysis step compnses a dialysis or diafiltration/ultrafiltration against a solution of physiological salt or a higher salt concentration, non-io c surfactant, and buffer at a lower pH than is present in the dissociation mixture.

15. A process according to Claim 13 wherein the dialysis step compπses a dialysis or diafiltration/ultrafiltration in a reassembly buffer, said reassembly buffer compnsmg- ionic strength salt in the range of 0.5-1.35M NaCl. a metal ion source, and a buffer at pH 6.0-6.5.

16. A process according to Claim 1 wherein step (c) compnses a dialysis or diafiltration/ultrafiltration step with a reassembly buffer, the reassembly buffer compπsing: ionic strength salt in the range of 0.5-1.35M NaCl a metal ion source, a buffer at pH 6.0-6.5.

17 A process according to Claim 16 wherein the metal ion source is a Ca+2 ₀r a Mg+2 source.

18 A process according to Claim 17 wherein the metal ion source is CaCl2 or MgCb at 0.5-5.0 mM.

19 A process according to Claim 16 wherein the buffer is selected from the group consisting of: a) glycine and phosphate, b) citrate and phosphate; and c) citrate

20 A process according to Claim 16 wherein the reassembly buffer compπses 1 M NaCl, 2 mM Ca+2, 50 _mM citrate, pH 6.2 and 0.03% polysorbate 80

21 A process according to Claim 1 further compnsmg-

(d) further purifying with a dialysis step compπsing a final buffer compnsmg 0.25- 1.25 M NaCl, a nonionic detergent and optionally, a histid e buffer at pH 6 0-6.5

22 A process according to Claim 21 wherein the nonionic detergent is polysorbate 80 or polysorbate 20

23 A process according to Claim 1, further compnsmg adsorbing the dis/reassembled VLPs onto aluminum adjuvant

24 A process according to Claim 1 wherein the VLPs are selected from the group consisting of HPV 6a, HPV 6b, HPV 11, HPV 16, HPV 18, and combinations thereof

25 A process for making a storage stable human papillomavirus (HPV) vaccine comprising HPV virus-like particles (VLPs), the process comprising the steps of

(a) incubating VLPs m a low salt dissociation mixture for 30-40 minutes, the dissociation mixture compnsmg 0.16-0.18 NaCl, 2-20 mM DTT, 0.01- 0.03% Polysorbate 80, 0.5-5 mM EDTA and 5-15 mM TRIS buffer, at pH 8.2 to produce dissembled VLPs;

(b) optionally removing the DTT from the dissociation mixture using a dialysis against a buffer compnsmg 0 16-0 18 M NaCl, 0.01-0.03% polysorbate 80, and 10 mM phosphate at pH 7.0,

(c) reassembling the disassembled VLPs using dialysis against a reassembly buffer, the reassembly buffer comprising 1 0 M NaCl, 2 mM CaCl2; and a pH 6.2 buffer selected from the group consisting of. 50 mM sodium citrate; 50 mM glycine and phosphate; and 50 mM citrate; and (d) further purifying the reassembled VLPs using dialysis against a final buffer, the final buffer compnsmg 0.5M NaCl, and 10 mM histidine, pH 6.2

26. A process according to Claim 25, further compnsmg

(e) adsorbing the reassembled VLPs from step (d) onto aluminum adjuvant

27. A process for making a storage stable human papillomavirus (HPV) vaccine compnsmg HPV virus-like particles (VLPs), the process compnsmg the steps of

(a) incubating VLPs in a high salt dissociation mixture for 30-40 minutes, the dissociation mixture compnsmg: 0.5-1.25 M NaCl, 2-20 mM DTT, 0.01- 0.03% polysorbate 80, 0 5-5 mM EDTA and 5-100 TRIS buffer, and 0-50 mM phosphate at pH 8.2 to produce disassembled VLPs;

(b) optionally removing the DTT from the dissociation mixture using a dialysis against a buffer compnsmg at least 1 M NaCl, 0.01-0.03% polysorbate 80, and 10 mM phosphate at pH 7.0;.

(c) reassembling the disassembled VLPs using dialysis against a reassembly buffer, the reassembly buffer compnsmg 1.0-1.35 M NaCl, 2 mM CaCl2; and a pH 6.2 buffer selected from the group consisting of: 50 mM sodium citrate; 50 mM glycine and phosphate; and 50 mM citrate; and

(d) further punfymg the reassembled VLPs using dialysis against a final buffer, the final buffer compπsing 0.5M NaCl, and 10 mM histidme, pH 6.2

28. A process according to Claim 27, further comprising

(e) adsorbing the reassembled VLPs from step (d) onto aluminum adjuvant.

29 A vaccine made by the process of Claim 1.

30. A vaccine made by the process of Claim 23.

31. A vaccine made by the process of Claim 25

32 A vaccine made by the process of Claim 28

33 A vaccine formulation comprising VLPs and a formulation buffer, said formulation buffer compπsing- 0.15-0.32M NaCl, 5-10 mM histidine, pH 6.2, and 0.005-0.015% polysorbate 80

34. A vaccine formulation compπsing aluminum adsorbed VLPs and a formulation buffer, said formulation buffer comprising: 0.15-0.32M NaCl, 5-10 mM histidine, pH 6.2, and 0.005-0.015% polysorbate 80.