An Improved Manufacturing Process for a Recombinant Polyprotein Vaccine

ABSTRACT

A new recombinant protein vaccine against leishmaniasis has been produced by fusing portions of three leishmania-derived proteins. To simplify manufacturing, process improvements included removal of a 6-histidine sequence near the amino terminus, and the mutation of a proteolytic hot spot. The improved manufacturing method consists of fermenting an engineered version of the HMS-174 strain of E. coli following current good manufacturing practice regulations at the 30-L scale. A purification scheme yields purified protein product at greater than 100 mg/L. The purified protein then yields a stable cake that can safely be sent to clinical trials. The product was tested with the adjuvant monophosphoryl lipid A in stable emulsion (MPL-SE) in a murine potency assay. The Leish-110f recombinant protein product reported here is an improved version of the original leishmaniasis vaccine.

Leishmaniasis is a devastating disease that causes untold human suffering and numerous deaths worldwide.¹ Leishmaniasis is a spectrum of diseases ranging from a fatal visceral form to a localized self-healing cutaneous lesion. The global prevalence of leishmaniasis, caused by the intracellular protozoan Leishmania species, is estimated to be approximately 12,000,000 cases total, with 500,000 new reports of visceral leishmaniasis and 1,500,000 million new reports of cutaneous leishmaniasis each year.² Although much progress has been made in developing drug therapies for leishmaniasis, a number of clinical problems, such as length of treatment and prohibitive expense, are associated with these therapies. Therefore, new vaccines and improvements to existing clinical products are valuable additions to the fight against this terrible disease.

The first-line clinical therapy, antimonial drug, was developed almost a century ago. This clinical regimen requires multiple injections (four weeks of daily injections for the visceral form of leishmaniasis), is costly, often associated with unpleasant side effects, and is becoming ineffective in many endemic locales. With resistance to therapeutic antimonials increasing to 60% in certain parts of India, the mortality rate from visceral leishmaniasis has been very high.³ Recent progress has been made by an oral anti-leishmaniasis drug, miltefosine, with the current clinical recommendation to use this drug as frontline treatment for some forms of leishmaniasis, including cutaneous leishmaniasis. Miltefosine, however, is teratogenic and requires an extended therapeutic regimen. Another therapeutic, liposomal amphotericin B, also has considerable efficacy against clinical leishmaniasis, but the expense of the drug renders it impractical in all but severe cases of the disease. Both miltefosine and amphotericin B function independently of the immune system, therefore, when the drug is withdrawn before completion of the therapeutic regimen, parasite loads re-emerge.⁴ Control of the phlebotomine sand fly vectors is also not feasible given the prohibitive expense associated with insecticide use.

The development of a safe, effective, and economical vaccine product is possible, given that controlled vaccination with live parasites has been practiced for centuries in the Middle East, and individuals who recover from clinical leishmaniasis develop immunity against re-infection. Additionally, recent trials with killed parasites demonstrate good efficacy of such a vaccine in post-infection settings of cutaneous leishmaniasis,^5–8 and numerous studies have demonstrated the effectiveness of such vaccines in experimental models of leishmaniasis.^9–14

The Infectious Disease Research Institute (IDRI) has developed a prophylactic vaccine for cutaneous leishmaniasis by constructing a recombinant fusion protein consisting of portions of three Leishmania proteins: thiol-specific antioxidant (TSA),¹⁵ Leishmaniamajor (L. major) stress-inducible protein 1 (LmSTI1),¹⁶ and Leishmania elongation initiation factor (LeIF).¹⁷ The schematic of this construct is shown in Figure 1. The recombinant approach was preferable to live vaccine or a killed parasite vaccine. Live vaccination is being discontinued because of undesirable side effects,¹⁸ and the manufacturing process of a killed parasite vaccine presents many difficulties relating to reproducibility, the need for animal-derived products in the culture medium, and variability in potency. The recombinant approach lends itself to the production of an inexpensive and effective vaccine at large scale, and if efficacious, would drastically reduce the morbidity and mortality caused by Leishmania throughout the world.

Figure 1

The original form of this vaccine was manufactured under current good manufacturing practice (cGMP) conditions, is currently undergoing human clinical testing, and will ideally provide a lasting protective immunological response. To simplify manufacturing, several subtle process improvements included removal of a 6-histidine sequence near the amino terminus, and the mutation of a proteolytic hot spot. The improved manufacturing method consists of fermenting an engineered version of the HMS-174 strain of Escherichia coli (E. coli) following cGMP regulations at the 30-L scale. A purification scheme combining anion exchange chromatography with hydroxyapatite chromatography was developed that yielded purified protein product at greater than 100 mg/L. The purified protein was formulated and lyophilized to yield a stable cake that can safely be sent to clinical trials worldwide. The cGMP-manufactured product was tested with the adjuvant monophosphoryl lipid A in stable emulsion (MPL-SE) in a murine potency assay. The Leish-110f recombinant protein product reported here is an improved version of the original leishmaniasis vaccine.

Materials and Methods

Master and Working Cell Banks

Master and working cell banks were manufactured under cGMP conditions at Charles River Laboratories (Malvern, PA). The cell banks were tested for purity, viability, and identity and are stored as frozen stocks in glycerol at –80° C. The flow chart of fermentation and purification is shown in Figure 2.

Figure 2

Fermentation

Both the inoculums and the fermentation used 2XYS as the growth media, with dextrorotatory glucose (D-glucose) as the primary carbon source, to grow the recombinant E. coli in a fed-batch fermentation. The fermentor was programmed to control the pH at 7.0 and dissolved oxygen at 30%. When the optical density at 600 nanometers had increased to 8 + 1, isopropyl-β-D-thiogalactopyranoside (IPTG) was added to the fermentor to a final concentration of 1 mM to induce expression of the recombinant gene encoding the Leish-110f protein. At three hours post-induction, the culture was harvested by centrifugation.

Inclusion Body Isolation

Following fermentation harvest, the wet cell paste was resuspended in lysis buffer (150 mM NaCl, 20 mM Tris pH 8.0, 5 mM EDTA) at a defined volume-to-mass ratio of 5 mL of lysis buffer to 1 g of cell paste. The resuspended E. coli were disrupted with an M-110S microfluidizer (Microflluidics, Newton, MA). The resulting lysate was centrifuged at 18,000 grams for 30 min at 4° C, which pelleted the insoluble inclusion bodies (IBs). The IBs were washed with a 2% CHAPS (3[(3 cholamidopropyl) di-methylammonio]-propanesulfonic acid) wash, followed by an isopropanol wash. The washed IB pellets were stored at –80° C until needed.

Purification

The inclusion body pellets were solubilized with 8 M urea, 50 mM Tris pH8, and 25 mM dithiothreitol at 150 mL/g of inclusion body pellet. The solubilized inclusion bodies were clarified using centrifugation. The first column, Amersham Biosciences' Q Sepharose Fast Flow resin (QFF), was equilibrated using buffer A (8 M urea, 50 mM Tris pH8). The clarified inclusion body preparation was loaded onto the anion exchange column at a linear velocity of approximately 120 cm/h. The full-length Leish-110f protein was eluted in a single peak from the resin, using buffer B (8 M urea, 50 mM Tris pH8, 150 mM NaCl). The second column was a BioRad Macro-Prep ceramic hydroxyapatite (CHT), type I, 40 mM. The CHT column was equilibrated using buffer B at a linear velocity of 120 cm/h. The reduced QFF elution peak was loaded onto the CHT column and washed with buffer B for 10 column volumes. The Leish-110f protein was eluted using buffer C (8 M urea, 100 mM potassium phosphate, pH 7.5, 150 mM NaCl).

Buffer Exchange

Diafiltration was performed on the CHT elution peak to remove the urea and replace the buffer with 20 mM Tris pH 8.0. The diafiltration was performed using a tangential flow filtration device with a 30 kDalton molecular weight cut off (MWCO) polyethersulfone ultra filtration membrane, and with a final protein concentration of 0.5 mg/mL as determined by A280. The purified Leish-110f bulk protein was then filter-sterilized using a 0.2 µm filter. The final purified bulk protein was stored at –80° C until formulation.

Fill and Finish

The final configuration was a lyophilized cake, containing sucrose mannitol and polysorbate 80. The formulated product was then loaded into a Virtis lyophilizer and lyophilized.

Testing for Process-Related Contaminants

General safety and residual amounts of process-related contaminants, such as kanamycin, urea, and endotoxins, were measured using standard procedures. In the case of general safety and residual kanamycin, United States Pharmacopeia (USP) methods were applied. For urea, a modification of the method of Knorst, Neubert, and Wohlrab was used, adapting the test to a 96-well format.¹⁹ For endotoxins, the limulus amebocyte assay was used.

Gel Electrophoresis

SDS-PAGE (sodium dodecyl sulfate polyacrylamide gel electrophoresis) was performed according to the Laemmli method.²⁰ Samples were diluted 1:1 in reducing 2x Laemmli sample buffer (Sigma) and boiled for 5 min. Proteins were then loaded on 4–20% acrylamide gradient gels (Bio-Rad Laboratories, Hercules, CA) and were run at 30 milliamperes (mA) constant for 60 min. The gels were stained using Coommassie blue, with standard techniques.

Immunogenicity Assay

Female BALB/c (Bagg albino color locus genotype c/c) mice were purchased from Charles River Laboratories (Wilmington, MA) and were maintained in specific-pathogen-free conditions. Eight- to 12-week-old mice at the beginning of experiments were used. Three mice per group were immunized with 10 µg of Leish-110f or Leish-111f formulated with 20 µg of an adjuvant MPL-SE (GlaxoSmithKline Biologicals, Rixensant, Belgium) in a volume of 0.1 mL. The mice were immunized subcutaneously once at the base of the tail. Blood was taken from the mice one week after the immunization to determine immunoglobulin G1 (IgG1) and immunoglobulin G2a (IgG2a) antibody titers. Plates were coated with 200 ng per well of Leish-110f or Leish-111f diluted in coating buffer, followed by blocking with phosphate-buffered saline containing 0.05% Tween 20 and 1% bovine serum albumin. Mouse serum samples were diluted to 1:100 and applied to the plates in two-fold serial dilutions. The plates were incubated with horseradish peroxidase (HRP) conjugated goat anti-mouse IgG1 or IgG2a (SouthernBiotech, Birmingham, AL), were developed with 3,3',5,5'-Tetramethylbenzidine (TMB) substrate, and were read by a microplate reader at a 450 nanometer wavelength. Endpoint titers were calculated with GraphPad Prism software using an optical density (OD) value of 0.1 as a cutoff.

Figure 3

Results

Protein Production

The fed-batch fermentation of the HMS-174/L110f recombinant system was consistent in its growth profile and production of the vaccine product Leish-110f. The Leish-110f protein was produced in the inclusion bodies. A two-step orthogonal purification procedure post-inclusion body washing achieved significant purification of the vaccine candidate. A chromatogram of the QFF anion exchange column separation (Figure 3) shows a single major peak that contained the Leish-110f polyprotein. The second step in the purification used ceramic hydroxyapatite, with the Leish-110f protein eluted as a single major peak (Figure 4) at low concentrations of phosphate. The majority of the contaminants remained bound to the resins used. The in-process SDS-PAGE gel shown in Figure 5 illustrates purification as additional purification steps were added. The utility of coupling these methods is that the eluted protein from the first step need not be conditioned between chromatographies, as the protein binds to the hydroxyapatite even in the presence of NaCl. SDS-PAGE of the eluted material illustrates the purity of the final Leish-110f protein product (Figure 6). Yield of purified protein is more than 150 mg/L and is a tremendous improvement over the original L-111f protein yield of 30 mg/L.

Figure 4

Following purification, urea was removed by diafiltration, and the protein was filled and lyophilized as described in the materials and methods section, above. The protein formed a stable white cake in the vials after lyophilization, and that quickly redissolved upon addition of water. Real-time stability studies at several temperatures are ongoing, but preliminary data suggests that the purified bulk protein and the lyophilized cake are stable at all temperatures tested (data not shown).

Figure 5

Characterization of the Protein

The final product was tested for endotoxins, residual moisture, pH, concentration, physical appearance, identity, and quality as measured by SDS-PAGE. These data are shown in Table 1.

Figure 6

Immunogenicity of the cGMP Material in a Murine Model

We compared the Leish-111f protein with the improved version Leish-110f in a murine potency assay. Mice were immunized with a single dose of 10 µg of Leish-111f or Leish-110f adjuvanted with MPL-SE. Seven days post-injection, mice were bled and antigen-specific IgG1 and IgG2a antibody titers were measured by enzyme-linked immunosorbent assay (ELISA) (Figure 7). The two forms of the protein induced comparable titers of both IgG1 and IgG2a antigen cross-reactive antibody. These results suggest that Leish-110f made using our new process and construct induced antibody responses comparable to that obtained with Leish-111f in mice.

Table 1. Summary of in-process contaminant clearance using CEX for capture and HCI for polish

Discussion

Researchers at IDRI have chosen to develop a Leishmania vaccine that is a fusion of three distinct, conserved Leishmania antigens. These three antigens—TSA, LmSTI1, and LeIF—were selected for the development of a subunit vaccine based on their demonstrated abilities to induce at least partial protection in the BALB/c mouse model of L. major in either prophylactic (TSA and LmSTI1) or therapeutic (LeIF) applications.¹³ All three antigens are present in the amastigote and the promastigote forms of the parasite and are highly conserved among Leishmania species, a requisite for ensuring cross-species protection. Immunization with a single recombinant antigen, in addition to having a simplified manufacturing process, may insure equivalent uptake of the components by individual antigen-presenting cells; in turn, it may generate an immune response that is broadly specific for all the immunogenic epitopes spanning the polyprotein.

Figure 7

Creating a vaccine composed of three or more independently produced recombinant proteins would be prohibitively expensive to manufacture and formulate, compared to the single polypeptide vaccine described here. The manufacturing cost of a vaccine is particularly important in developing countries, and it was one of the considerations in producing Leish-110f. The original Leish-111f fermentation was different from that used in the new process. Conditions were developed for Leish-110f to attain a higher cell density before induction and thus, a higher product yield. This outcome was accomplished by using a richer, semi-defined media, and a controlled feeding strategy to manage the growth of the culture. These two changes allowed for a substantial increase in recombinant protein production per liter of culture.

The purification procedures for the two recombinant products were similar, in that both processes used an anion exchange resin followed by the hydroxyapatite column purification steps. However, the previous process used Amersham Pharmacia Biotech's Source 30Q resin, and linear gradients were used to elute the product. In the new process the Q-Sepharose Fast Flow resin was used, and step-wise elutions were used for ease in scale-up and manufacturing. The process described above makes a consistent product at acceptable yields for the current stage of development. IDRI is producing nearly 200,000 doses per lot of vaccine based on the highest dose level currently envisioned. This level is satisfactory for the amounts needed to conduct current clinical trials, but scale-up of the process to the 300-liter scale is already ongoing. Assuming no loss in unit production, this scale would deliver 2,000,000 doses per lot and could enable large-scale clinical trials of the vaccine even without further process improvements.

An additional encouraging fact is the stability of the lyophilized protein. When performing trials or vaccinations in field settings where a cold chain cannot be guaranteed, stability of the product is crucial to successful deployment of the trial and the final product. This lyophilized vaccine product has been shown to be stable for more than three years.

IDRI has developed a process to manufacture a new recombinant protein product for the treatment and prophylaxis of leishmaniasis. This candidate vaccine consisting of Leish-110f formulated in MPL-SE will be tested in a spectrum of diseases caused by the various Leishmania species. A robust, scalable procedure will allow rapid progress in bringing this vaccine to patients and will be a step in the fight against leishmaniasis.

THOMAS S. VEDVICK, PHD, is director of process sciences, 206.330.2530, tvedvick@idri.orgLAUREN CARTER is a research associate, GARNER MOULTON is a senior research associate, YASUYUKI GOTO, PHD, is a scientist, SYLVIE BERTHOLET, PHD, is a senior scientist, STEVEN G. REED, PHD, is founder and head of research and development, and DARRICK CARTER, PHD, is director of formulations, all at at the Infectious Disease Research Institute, Seattle, WA. Dr. Carter is also the vice president of research and development at Dharma Therapeutics, Seattle, WA.

References

1. World Health Organization. The leishmaniases and leishmania/HIV: Co-infections (Fact Sheet No 116). Geneva; 2000.

2. Desjeux P. Leishmaniasis. Nat Rev Microbiol. 2004;2:692.

3. Bryceson A. Visceral leishmaniasis in India. Lancet. 2000;356:1933.

4. Murray HW. Clinical and experimental advances in treatment of visceral leishmaniasis. Antimicrob Agents Chemother. 2001;45:2185–97.

5. De Luca PM, et al. Evaluation of the stability and immunogenicity of autoclaved and nonautoclaved preparations of a vaccine against American tegumentary leishmaniasis. Vaccine. 1999;17:1179–85.

6. Genaro O, de Toledo VP, da Costa CA, Hermeto MV, Afonso LC, Mayrink W. Vaccine for prophylaxis and immunotherapy, Brazil. Clinics in Dermatology. 1996;14:503–12.

7. Machado-Pinto J, et al. Immunochemotherapy for cutaneous leishmaniasis: a controlled trial using killed Leishmania (Leishmania) amazonensis vaccine plus antimonial. Int J Dermatol. 2002;41:73–8.

8. Mayrink W, et al. Immunotherapy as a treatment of American cutaneous leishmaniasis: preliminary studies in Brazil. Parassitologia. 1992;34:159–65.

9. Barbosa AJ, Raso P, Genaro O, da Costa CA, Mayrink W. On leishmanial antigen detection in tissue sections of Montenegro's reaction [letter]. Revista do Instituto de Medicina Tropical de Sao Paulo. 1994;36:95–6.

10. Mayrink W, et al. Phase 1 and 2 open clinical trials of a vaccine against Leishmania chagasi infections in dogs. Memorias Do Instituto Oswaldo Cruz. 1996;91:695–7.

11. Mora AM, Mayrink W, Costa RT, Costa CA, Genaro O, Nascimento E. Protection of C57BL/10 mice by vaccination with association of purified proteins from Leishmania (Leishmania) amazonensis. Revista do Instituto de Medicina Tropical de Sao Paulo. 1999;41:243–8.

12. Badaro R, Lobo I, Nakatani M, Muinos A, Netto EM, Coler RN, Reed SG. Successful use of a defined antigen/GM-CSF adjuvant vaccine to treat mucosal leishmaniasis refractory to antimony: A case report. Braz J Infect Dis. 2001;5:223–32.

13. Campos-Neto A, et al. Protection against cutaneous leishmaniasis induced by recombinant antigens in murine and nonhuman primate models of the human disease. Infect and Immun. 2001;69:4103–8.

14. Campos-Neto A, Webb JR, Greeson K, Coler RN, Skeiky YA, Reed SG. Vaccination with plasmid DNA encoding TSA/LmSTI1 leishmanial fusion proteins confers protection against Leishmania major infection in susceptible BALB/c mice. Infect and Immun. 2002;70:2828–36.

15.Webb JR, Kaufmann D, Campos-Neto A, Reed SG. Molecular cloning of a novel protein antigen of Leishmania major that elicits a potent immune response in experimental murine leishmaniasis. J Immunol. 1996;157:5034–41.

16. Webb JR, et al. Human and murine immune responses to a novel Leishmania major recombinant protein encoded by members of a multicopy gene family. Infection and Immunity. 1998;66:3279–89.

17. Skeiky YA, et al. A recombinant Leishmania antigen that stimulates human peripheral blood mononuclear cells to express a Th1-type cytokine profile and to produce interleukin 12. J Exp Med. 1995;181:1527–37.

18. Titus RG, Gueiros-Filho FJ, De Freitas LA, Beverley SM. Development of a safe live Leishmania vaccine line by gene replacement. Proc Natl Acad Sci USA. 1995;92:10267–71.

19. Knorst MT, Neubert R, Wohlrab W. Analytical methods for measuring urea in pharmaceutical formulations. J Pharm Biomed Analysis. 1997;15:1627–32.

20. Laemmli UK. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature. 1970;227:680–5.