Vaccine 24 (2006) 5839–5851
A systematic approach to stabilizing EBA-175 RII-NG for use as a malaria vaccine Laura J. Peek, Duane T. Brandau, LaToya S. Jones 1 , Sangeeta B. Joshi, C. Russell Middaugh ∗ Department of Pharmaceutical Chemistry, University of Kansas, 2095 Constant Avenue, Lawrence, KS 66047, USA Received 9 February 2006; received in revised form 26 April 2006; accepted 27 April 2006 Available online 12 May 2006
Abstract Region II of the erythrocyte-binding antigen (EBA-175 RII) has been identified as a promising target for a malaria vaccine. A systematic approach to identify optimal preformulation conditions of a non-glycosylated (NG) antigen, EBA-175 RII-NG, has been developed. This approach consists of development of an empirical temperature/pH phase diagram, high throughput stabilizer screening and aluminum salt adjuvant adsorption studies. Using these physical methods, we developed a stable formulation for EBA-175 RII-NG at pH 6.0 with sucrose and Brij® 35 as stabilizers and Adju-Phos® as an adjuvant. This approach should be generally applicable to guiding the development of stable vaccine formulations. © 2006 Elsevier Ltd. All rights reserved. Keywords: Malaria; Stability; Protein
1. Introduction More than 3 billion people are threatened each year by malaria, a parasite that kills over 1 million people annually [1]. For decades, the challenge of developing a malaria vaccine has been a primary focus in many research laboratories throughout the world. Many of the problems are due in part to the complexity of the life cycle of parasites that cause human malaria, most commonly Plasmodium vivax and Plasmodium falciparum [2,3]. P. falciparum has been the focus of many of these vaccines since this species is responsible for the majority of malaria infections and deaths seen worldwide [4,5]. Potential malaria vaccine antigens have been identified in an attempt to inhibit hepatocyte invasion by sporozoites, erythrocyte invasion by merozoites, as well as various other stages of growth and development throughout the parasite’s life cycle [2,6]. ∗
Corresponding author. Tel.: +1 785 864 5813; fax: +1 785 864 5814. E-mail address:
[email protected] (C.R. Middaugh). 1 Present address: Department of Pharmaceutical Sciences, University of Colorado, 4200 E. 9th Avenue, C238, Denver, CO 80262, USA. 0264-410X/$ – see front matter © 2006 Elsevier Ltd. All rights reserved. doi:10.1016/j.vaccine.2006.04.067
A 175 kDa antigen was found to be involved in facilitating the invasion of erythrocytes by merozoites in 1985 [7]. This finding suggested that a vaccine antigen capable of blocking the organism’s red blood cell attack might be a promising vaccine candidate. Certain regions in the gene of this antigen, specifically Region II (RII), are conserved among strains of P. falciparum and are responsible for binding of the organism to erythrocytes [8–10]. For these reasons, a non-glycosylated (NG) erythrocyte-binding antigen, EBA175 RII-NG, is being pursued as a vaccine candidate. A major problem in the development of any vaccine based on a recombinant protein is the creation of a stable, effectively adjuvanted formulation that will permit the vaccine to be stored and delivered anywhere in the world, even under adverse environmental conditions. We have, therefore, developed a systematic approach to guide the identification of optimal stabilizing conditions for EBA-175 RII-NG for use in a liquid injectable formulation. This approach involves the use of high-resolution second derivative absorbance spectroscopy, circular dichroism (CD), and both intrinsic and extrinsic fluorescence spectroscopies to monitor structural changes of the protein while undergoing thermal stress. Data
5840
L.J. Peek et al. / Vaccine 24 (2006) 5839–5851
from these physical techniques are combined to develop empirical phase diagrams, which define regions of similar physical states of the protein across a variety of pH and temperature conditions [11–15]. Based on the phase diagram, conditions for a high throughput screening of a library of generally regarded as safe (GRAS) excipients are selected. In this case, a turbidity assay was used for this purpose, and the ability of each excipient to inhibit protein aggregation was evaluated. Several excipients exhibiting significant inhibition of aggregation are then selected, and their effect on the conformational stability of the protein is evaluated using CD or fluorescence spectroscopies, or both. Conformational stabilizers are then analyzed in various combinations to determine if additional stabilization can be attained. It is critical that optimal stability is achieved so that the activity of the vaccine can be maintained even when a cold chain is not available [16] since malaria presents an especially difficult problem in tropical regions. Once the antigen’s stability is optimized, adsorption isotherms are constructed to determine the amount of antigen that can be adsorbed to an aluminum salt adjuvant over a range of conditions.
2. Materials and methods 2.1. Materials Citrate–phosphate (25 mM) buffers (pH 3.0, 4.0, 5.0, 6.0, 7.0 and 8.0) containing 100 mM NaCl were prepared using citric acid monohydrate (Fisher, Pittsburgh, PA) and sodium phosphate dibasic, anhydrous (Sigma, St. Louis, MO). Sodium phosphate (10 mM) buffers (pH 6.0 and 7.2) were prepared using sodium phosphate dibasic, anhydrous, and sodium phosphate monobasic, monohydrate (Sigma, St. Louis, MO) and contained 150 mM NaCl. All excipients were purchased from Sigma (St. Louis, MO), except for sucrose octasulfate and Pluronic F68, which were purchased from Toronto Research Chemicals (Ontario, Canada) and Spectrum Chemical & Laboratory Products, Inc. (Gardena, CA), respectively. Sodium chloride was purchased from J.T. Baker (Phillipsburg, NJ). Purified EBA-175 RII-NG was produced by Cambrex BioScience, Baltimore (Baltimore, MD) in collaboration with Protein Potential (Rockville, MD) and provided by Science Applications International Corporation under contract with the National Institute of Allergy and Infectious Diseases, National Institutes of Health. The protein produced a single band on SDS-PAGE at approximately 2 mg/mL in 6 mM phosphate buffer, pH 7.4, containing 5% sucrose and 154 mM NaCl. Sucrose was used to inhibit dimerization detectable by size exclusion chromatography (SEC) upon thawing after freezing at −20 ◦ C. This dimerization was not observed upon freezing at −70 ◦ C (Protein Potential, personal communication). Protein concentration was determined by UV absorbance at 280 nm (A280 ), using an extinction coefficient of 1.8 mL/mg cm, which was calculated using the Protein Cal-
culator [17]. This tool employs Gill and von Hippel’s method [18] to calculate the extinction coefficient from the number of tryptophan (Trp), tyrosine and cysteine residues in the protein. Most protein dialyses were performed at 4 ◦ C using SlideA-Lyzer® Dialysis Cassettes, 10,000 MWCO (Pierce, Rockford, IL). In some cases, Spectra/Por® Cellulose Ester Sterile DispoDialyzers (2000 MWCO, 2 mL sample volume) (Spectrum Laboratories, Inc.; Rancho Dominguez, CA) were employed. 2.2. Biophysical characterization of EBA-175 RII-NG The biophysical properties of EBA-175 RII-NG were evaluated at six pH conditions from pH 3–8.0 and over the temperature range of 10–85 or 90 ◦ C, depending on the technique. In addition, the studies compared the properties of the antigen in the presence and absence of 5% sucrose, which was previously shown by size exclusion chromatography at Cambrex BioScience to prevent dimerization at −20 ◦ C [19]. 2.2.1. High-resolution absorbance spectroscopy High-resolution absorbance spectra were obtained employing an Agilent 8453 UV–visible diode array spectrophotometer equipped with a Peltier temperature controller (Palo Alto, CA). Spectra were collected every 2.5 ◦ C from 10 to 90 ◦ C with a 5 min equilibration at each temperature. An integration time of 25 s was used to obtain spectra with a high degree of precision. The protein concentration used for each experiment was approximately 0.11 mg/mL. Each sample was evaluated in duplicate and analyzed from 200 to 400 nm. Turbidity traces were also generated for each sample by plotting the optical density at 360 nm (OD360 ) as a function of temperature to monitor protein aggregation. Second derivatives of the absorbance spectra were obtained by ChemStationTM software (Agilent) using a nine point data filter and the Savitsky-Golay smoothing function to fit data to a third order polynomial. The spectra were smoothed using 99 interpolated points between each data point, which resulted in highly resolved spectra on the order of ±0.01 nm [20]. Peak positions were selected by utilizing the Pick Peaks Tool in Microcal OriginTM 6.0. A width of 1.00 and a height of 10−4 were the dimensions of the box used to locate the negative peaks in each spectrum; a minimum height value was set to 5.00. 2.2.1.1. Empirical phase diagrams using second derivative absorbance spectroscopy data. Second derivative absorbance spectroscopy peak position data were used to generate empirical phase diagrams. These multi-colored plots display changes in the physical state of the protein as a function of temperature and pH. All calculations for construction of the phase diagrams were performed using Mathematica (Wolfram Research, Champaign, IL). The multi-component vector approach uses n-dimensional vectors composed of data at each combination of pH and temperature to generate
L.J. Peek et al. / Vaccine 24 (2006) 5839–5851
an n × n density matrix, where n represents the number of variables employed in the calculation (for example, when the absorbance spectrum’s second derivative displays six peaks, n = 6). From the density matrix, n sets of eigenvalues and eigenvectors are derived. The three largest eigenvectors are then re-expanded into three dimensions and each is assigned a color – red, green or blue – to create a single, multi-colored plot. Regions of like color represent similar physical states of the protein. Thus, each final color in the diagram is representative of a different physical state of the protein. Further explanation of this approach is described in Kueltzo et al. [14]. 2.2.2. Intrinsic tryptophan fluorescence Intrinsic Trp fluorescence spectra were acquired using a Photon Technology International (PTI) spectrofluorometer (Lawrenceville, NJ) equipped with a turreted four-position Peltier-controlled cell holder and a xenon lamp. The 17 tryptophan residues in EBA-175 RII-NG [21] were used to monitor changes in the tertiary structure of the protein as it was subjected to thermal stress. An excitation wavelength of 295 nm was used (>95% Trp emission), and spectra were collected from 305 to 440 nm. Spectra were obtained every 2.5 ◦ C from 10 to 85 ◦ C at an EBA-175 RII-NG concentration of 0.1 mg/mL in the citrate–phosphate buffers at pH 3–8.0 in the presence and absence of 5% sucrose. A buffer spectrum was subtracted from each protein spectrum prior to data analysis using Felix 32TM software (PTI). Peak positions were determined using a mean spectral center of mass method. This method locates peaks at approximately 13 nm higher wavelength than their actual position but results in significantly improved precision. 2.2.3. Far-UV circular dichroism Circular dichroism spectra of EBA-175 RII-NG in the citrate–phosphate buffer at all six pH conditions (3–8.0) with and without sucrose (5%) were recorded utilizing a Jasco J-810 spectropolarimeter equipped with a six-position Peltier temperature controller (Easton, MD). Spectra were collected at 10 ◦ C from 260 to 205 nm with a scanning speed of 100 nm/min and 1.0 nm resolution. Thermal melts were acquired for each sample by monitoring the CD signal at 222 nm over the temperature range from 10 to 90 ◦ C at intervals of 0.5 ◦ C. The temperature was increased gradually at a rate of 15 ◦ C/h. The CD signal was converted to molar ellipticity using Spectra ManagerTM software (Jasco). 2.2.4. Extrinsic 8-anilino-1-naphthalene sulfonate (ANS) fluorescence 8-Anilino-1-naphthalene sulfonate (Sigma, St. Louis, MO) was used as an extrinsic fluorescence probe to gain information regarding changes in the accessibility of apolar sites in the protein. Fluorescence thermal melts utilizing ANS were performed in the presence of 5% sucrose for EBA175 RII-NG at pH 3–8.0 in the citrate–phosphate buffer. Each sample contained an optimized 10-fold molar excess of ANS
5841
to protein. An excitation wavelength of 375 nm was used, and spectra were collected from 400 to 600 nm. Spectra were obtained from 10 to 85 ◦ C, every 2.5 ◦ C. A spectrum of buffer containing ANS was subtracted from each spectrum prior to data analysis using Felix 32TM software (PTI). Maximum fluorescence intensity was plotted as a function of temperature to evaluate binding of ANS to the protein. 2.2.5. Empirical phase diagrams using CD, intrinsic fluorescence and ANS fluorescence Empirical phase diagrams were constructed by the method previously described except that CD molar ellipticity, intrinsic fluorescence spectral center of mass and ANS fluorescence intensity data replaced the near-UV absorption second derivative peaks initially employed [11–13,15]. In this case, a three-dimensional vector was used to generate a 3 × 3 density matrix, from which three eigenvectors were derived and assigned red, blue and green colors to represent the different physical states of the protein. This approach was used to generate an empirical phase diagram that reflects changes in protein structure at all levels—secondary and tertiary structure as well as the accessibility of apolar binding sites in the protein. 2.3. Identification of stabilizers 2.3.1. Screening of GRAS compounds A series of GRAS compounds from a library that we have previously found to be of optimal utility were screened for their ability to inhibit the aggregation of EBA-175 RIING. Among those compounds evaluated in the screening were a variety of carbohydrates, nonionic surfactants, amino acids, proteins, polyols and cyclodextrins (Table 1). Concentrated stock solutions of each GRAS excipient were prepared in advance by dissolving the compound in the pH 6.0 sodium phosphate buffer and adjusting the solution to pH 6.0 using NaOH or HCl. A turbidity assay was developed based on monitoring the OD360 of the solution with time using a BMG Technologies FLUOstar Galaxy 96-well plate reader (Durham, NC). Based on the phase diagram and turbidity traces generated from data acquired during the high-resolution absorbance spectroscopy studies, pH 6.0 was chosen for the stabilizer screening since it is well within the physiological range but still produces detectable aggregation at a low enough temperature that can be easily attained by the instrument (see results). A variety of protein concentrations at pH 6.0 in the presence and absence of the reducing agent dithiothreitol (DTT) (Sigma, St. Louis, MO) were screened to develop this assay. DTT was included in each sample to reduce some or all of the thirteen intra-molecular disulfide bonds present in EBA-175 RII-NG. This promotes protein aggregation [21]. By perturbing the protein’s structure using a reducing agent, the effect of various compounds on the extent of protein aggregation can be more easily monitored. Several different temperatures and incubation times were also examined to optimize aggregation for this assay. Con-
5842
L.J. Peek et al. / Vaccine 24 (2006) 5839–5851
Table 1 Effect of generally regarded as safe (GRAS) compounds on the aggregation of EBA-175 RII-NG at 45 ◦ C GRAS compound
Guanidine HCl (100,000×) Sorbitol (20%) Brij® 35 (0.05%) Sucrose (20%) Arginine (0.15 M) Dextrose (20%) Trehalose (20%) Tween 20 (0.1%) Brij® 35 (0.1%) Lactose (20%) Glycerol (20%) Sorbitol (10%) Brij® 35 (0.01%) Dextran sulfate (0.1×) Phytic acid (1×) Tween 20 (0.01%) Proline (0.3 M) Glycerol (10%) Tween 20 (0.05%) Sucrose (10%) Mannitol (10%) Dextrose (10%) Diethanolamine (0.3 M) 2-Hydroxypropyl-␥-cyclodextrin (10%) Trehalose (10%) Tween 80 (0.05%) Tween 80 (0.1%) Tween 80 (0.01%) ␣-Cyclodextrin (2.5%) Dextran sulfate (0.01×) Histidine (1000×) Proline (1000×) Glycine (1000×) Gelatin (2.5%) Dextran T40 (1×) Lactic acid (1000×) Arginine (1000×) Guanidine (1000×) Diethanolamine (1000×) Ascorbic acid (1000×) ␣-Cyclodextrin (5%) Lactose (10%) Dextran sulfate (0.001×) Lysine (1000×) 2-Hydroxypropyl--cyclodextrin (10%) Malic acid (1000×) Glutamic acid (0.03 M) Dextran T40 (0.1×) Histidine (0.03 M) 2-Hydroxypropyl-␥-cyclodextrin (5%) Aspartic acid (1000×) Lactic acid (0.3 M) Dextran T40 (2.5×) Lysine (0.1 M) Glutamic acid (1000×) Pluronic F-68 (0.01%) Ascorbic acid (0.1 M) Gelatin (5%) Glycine (0.3 M) Malic acid (0.1 M)
% Inhibition of aggregation ± standard deviation 62 ± 1 55 ± 1 54 ± 15 53 ± 0 52 ± 3 51 ± 2 50 ± 7 47 ± 17 46 ± 19 41 ± 3 39 ± 7 39 ± 4 37 ± 11 36 ± 8 36 ± 8 33 ± 16 31 ± 18 31 ± 2 26 ± 15 26 ± 1 24 ± 2 20 ± 10 17 ± 11 17 ± 7 13 ± 13 11 ± 3 11 ± 2 10 ± 12 10 ± 6 9 ± 10 1 ± 12 1±9 0±5 −1 ± 31 −2 ± 7 −2 ± 5 −3 ± 11 −3 ± 3 −4 ± 14 −4 ± 8 −5 ± 10 −6 ± 18 −6 ± 6 −6 ± 3 −7 ± 13 −7 ± 4 −7 ± 9 −7 ± 7 −7 ± 6 −9 ± 18 −11 ± 5 −11 ± 3 −12 ± 4 −14 ± 18 −14 ± 7 −19 ± 27 −26 ± 8 −28 ± 10 −31 ± 3 −44 ± 20
Table 1 (Continued ) GRAS compound
% Inhibition of aggregation ± standard deviation
Pluronic F-68 (0.1%) Pluronic F-68 (0.05%) Albumin (1.0%) Albumin (2.5%) Albumin (5%) Sucrose octasulfate (1×) Dextran sulfate (1×) Dextran sulfate (2.5×)
−70 ± 12 −84 ± 18 −580 ± 26 −660 ± 22 −792 ± 94 i.i.a. i.i.a. i.i.a.
Turbidity was evaluated by monitoring the optical density at 360 nm as a function of time. The effect of each compound on the aggregation of the protein is represented as percent inhibition of aggregation. Compounds that inhibit aggregation have large, positive values for inhibition, while compounds that induce aggregation are represented by negative values. Compounds that immediately induced aggregation (i.i.a.) of EBA-175 RII-NG are noted accordingly.
ditions selected for the final high throughput screening were 0.2 mg/mL EBA-175 RII-NG in the presence of 30 mM DTT at 45 ◦ C for 4 h. The OD360 was measured every 5 min, and each potential excipient was screened in triplicate. The percent inhibition of aggregation was calculated to determine the extent to which a compound affects aggregation of the protein. This value is calculated by subtracting from 100 the quotient of the change in optical density at 360 nm (OD360 ) of the protein in the presence of the excipient (E), multiplied by 100, over the change in OD360 of the protein without excipient (C). OD360 (E) % inhibition of aggregation = 100 − × 100 OD360 (C) (1) 2.3.2. Effects of inhibitors of aggregation on protein conformation Once inhibitors of aggregation were identified, the effect of these potential excipients (in combination and alone) on the stability of the tertiary structure of the protein in the phosphate buffer (pH 6.0) was studied using intrinsic Trp fluorescence spectroscopy as described above. Far-UV CD spectroscopy was also employed to look at the effects of representative compounds on the stability of the secondary structure of the protein using the same parameters indicated previously; however, fluorescence spectroscopy was the preferred technique due to the unfavorable influence of several of the compounds on the CD signal. To determine the optimum minimum concentration of sucrose and Brij® 35 (see results), intrinsic Trp fluorescence and CD spectroscopies were also used to monitor conformational stability while the turbidity assay developed previously was employed to evaluate the effects on aggregation. Concentrations of sucrose evaluated were 0–20%, every 2.5%, while Brij® 35 was monitored at 0.005, 0.01, 0.03, 0.05 and 0.07%. Once the concentration of Brij® 35 was optimized,
L.J. Peek et al. / Vaccine 24 (2006) 5839–5851
these studies were repeated maintaining a constant Brij® 35 concentration and varying the sucrose concentration. 2.4. Adjuvant studies 2.4.1. Adsorption studies Two aluminum salt adjuvants, Adju-Phos® (aluminum phosphate) and Alhydrogel® (aluminum hydroxide) (Brenntag Biosector, Frederickssund, Denmark), were studied for their ability to adsorb EBA-175 RII-NG in the presence and absence of stabilizers in the sodium phosphate buffer (pH 6.0). These experiments were performed using 0.25 mg aluminum in 0.5 mL, a pharmaceutically acceptable amount of the adjuvant [22]. The amount of protein added to each sample was determined by absorbance spectroscopy prior to addition. The samples were prepared by adding stabilizer(s) to protein, followed by adjuvant. The order of addition, however, was determined to be insignificant. The samples were then tumbled on an end-over-end tube rotator at 4 ◦ C for 5 min (previously determined by a kinetics study to give a steady state of adsorption). Samples were centrifuged at 14,000 × g for 30 s to pellet the adjuvant. An absorbance reading at 280 nm of the supernatant was obtained from which the concentration of protein not adsorbed to the adjuvant was determined. Knowing the amount of protein added, the amount adsorbed to the adjuvant was calculated, and adsorption isotherms were created. Adsorption of at least 75–80% of the antigen is generally desired for optimal efficacy [23]. 2.4.2. Desorption studies The ability of EBA-175 RII-NG to be desorbed from the adjuvants was studied using 0.5, 0.75, 1, and 2 M NaCl. The salt solutions were prepared by dissolving the respective amounts of NaCl in the 10 mM sodium phosphate buffer, pH 6.0 and adjusting the solution to pH 6.0. Samples of protein adsorbed to adjuvant were prepared as described previously, centrifuged, and the supernatant discarded. The pellet was resuspended in the appropriate salt solution, immediately centrifuged, and an absorbance reading at 280 nm of the supernatant was obtained to determine the amount of EBA175 RII-NG desorbed.
5843
An additional study monitored the effect of time on the desorption of EBA-175 RII-NG. Samples of the protein adsorbed to Alhydrogel® and Adju-Phos® in the presence of stabilizers were prepared as described previously and stored at 4 ◦ C. Once a week for 1 month, samples were removed, centrifuged and an absorbance reading at 280 nm of the supernatant obtained to determine if the protein had desorbed over time. 3. Results 3.1. Empirical phase diagrams using second derivative absorbance spectroscopy data Information regarding the average environments of a protein’s Phe, Tyr, and Trp residues, and thus its tertiary structure, can be obtained by evaluating the shifts of the UV absorption spectrum’s second derivative negative peak positions as a function of alterations in solution conditions [20]. Five negative peaks are clearly observed in the second derivative absorbance spectrum of EBA-175 RII-NG. At 10 ◦ C, the peaks are positioned at the following approximate wavelengths: 253 nm (peak 1—Phe), 259 nm (peak 2—Phe), 275 nm (peak 3—Tyr), 284 nm (peak 4—Tyr/Trp), and 292 nm (peak 5—Trp). A sixth peak, which is usually observed in second derivative spectra of proteins but was not observed in spectra of EBA-175 RII-NG, often occurs between 266 and 269 nm and corresponds to Phe/Tyr. This peak became identifiable in spectra of EBA-175 RII-NG, pH 6–8.0 only at temperatures greater than approximately 50 ◦ C. Using second derivative absorbance data, empirical phase diagrams were created to display changes in the physical state of the protein as a function of temperature and pH in the presence and absence of 5% sucrose (Fig. 1(a and b)). The presence of 5% sucrose, despite its frequently stabilizing properties, has very little effect on the stability of the tertiary structure of EBA-175 RII-NG. The empirical phase boundaries displayed in these phase diagrams suggest the presence of a major stable state in the range of approximately pH 5–8 and up to about 50 ◦ C for samples both with and without sucrose. At temperatures greater than 50 ◦ C, a
Fig. 1. Empirical phase diagrams created using high-resolution second derivative absorbance spectroscopy data for EBA-175 RII-NG (a) in the presence of 5% sucrose and (b) in the absence of sucrose. Each area of similar color represents a different physical state of the protein.
5844
L.J. Peek et al. / Vaccine 24 (2006) 5839–5851
Fig. 2. The effect of temperature and pH on the aggregation of EBA-175 RII-NG (a) in the absence of sucrose and (b) in the presence of 5% sucrose. The turbidity traces were generated by plotting the optical density at 360 nm as a function of temperature. The six pH conditions evaluated were pH 3.0 (), pH 4.0 (), pH 5.0 (×), pH 6.0 (), pH 7.0 (+) and pH 8.0 (䊉). Error bars representing standard error of the mean (N = 2) are plotted in one direction only to make the data easier to interpret.
major structurally disruptive transition occurs. Another phase exists at pH 3.0 throughout the entire temperature range studied and also encompasses pH 4.0, at temperatures greater than 20 ◦ C, and pH 5.0, at temperatures greater than ∼45 ◦ C. Other phases exist at higher temperatures for pH 5–8.0 in the presence of sucrose, and for pH 6–8.0 in the absence of the sugar. Note that the absolute colors in the phase diagrams have no meaning because of the data normalization process employed. Rather, it is the changes in color that provide useful information. The susceptibility of the protein to aggregation processes can be evaluated from plots of OD360 versus temperature (Fig. 2). These turbidity traces suggest that EBA-175 RIING has a high propensity to aggregate at higher temperatures both with and without sucrose. In the absence of sucrose, the protein at pH 6 and 7 begins to aggregate at slightly reduced temperatures, around 65 ◦ C. At pH 5 and 8, the protein shows an increase in OD360 at approximately 73 ◦ C. The protein does not seem to aggregate at pH 3 and 4 until much higher temperatures. In the presence of sucrose (5%), aggregation of the protein at pH 6 begins gradually around 53 ◦ C and rapidly begins to aggregate at temperatures above 68 ◦ C. At pH 4 and 5, aggregates slowly begin to form at approximately 30 and 35 ◦ C, respectively. At pH 7, aggregation processes are initiated around 60 ◦ C while aggregation at pH 8 gradually increases from ∼35 ◦ C. At pH 3, the protein does not begin to aggregate until much higher temperatures, ∼75 ◦ C. Based on the turbidity traces, it appears that the presence of 5% sucrose does not have a significant stabilizing effect when studied at these temperature and pH conditions.
at pH 4.0, the protein begins to unfold around 15 ◦ C and becomes more extensively unfolded as it reaches 35 ◦ C, at which point it maintains a partially unfolded conformation (see below for further comments). At pH 5.0, unfolding begins around 30 ◦ C. For pH 6.0, 7.0 and 8.0, the transitions behave very similarly with events between 45 and 55 ◦ C. In contrast, the protein appears highly structurally disrupted at all temperatures at pH 3.0. These results suggest that the most stable environment for the protein is within the pH range 6–8. The CD spectrum of EBA-175 RII-NG at 10 ◦ C displays double minima at 208 and 222 nm, indicating significant ␣-helix structure within the protein (Fig. 4(a)). This is in agreement with the X-ray crystal structure of the protein [21]. Circular dichroism thermal melt data provide evidence for a loss of ␣-helical content upon thermal unfolding since the negative molar ellipticity decreases with increasing temperature (Fig. 4(a and b)). At pH 3.0, the protein has less
3.2. Characterization by intrinsic Trp fluorescence, CD, and ANS fluorescence The transitions observed in plots of intrinsic fluorescence spectral center of mass versus temperature (Fig. 3) confirm the boundaries indicated in the phase diagrams generated from the second derivative absorbance data. For example,
Fig. 3. The effect of temperature and pH on the intrinsic Trp fluorescence spectral center of mass of EBA-175 RII-NG. The six pH conditions evaluated were pH 3.0 (), pH 4.0 (), pH 5.0 (×), pH 6.0 (), pH 7.0 (+) and pH 8.0 (䊉). Transitions occurring at higher temperatures indicate more stable conditions for the protein. (The results are shifted +13 nm from actual peak maxima.)
L.J. Peek et al. / Vaccine 24 (2006) 5839–5851
5845
Fig. 5. The effect of temperature and pH on the ANS fluorescence intensity of EBA-175 RII-NG in the presence of 5% sucrose. Six pH conditions were evaluated: pH 3.0 (), pH 4.0 (), pH 5.0 (×), pH 6.0 (), pH 7.0 (+) and pH 8.0 (䊉). As the protein is disrupted, ANS binds apolar regions in the protein, resulting in an increase in fluorescence quantum yield. Transitions are identified as large increases in fluorescence intensity. Transitions that occur at higher temperatures indicate more stable conditions for the protein.
Fig. 4. Circular dichroism spectrum at 10 ◦ C before () and after () thermal denaturation (a) and thermal melt plots (b) of EBA-175 RII-NG in the presence of 5% sucrose. Molar ellipticity at 222 nm was monitored as a function of temperature for six pH conditions—pH 3.0 (), pH 4.0 (), pH 5.0 (×), pH 6.0 (), pH 7.0 (+) and pH 8.0 (䊉). Transitions at higher temperatures indicate more stable conditions for the protein.
␣-helical structure at 10 ◦ C than at any other pH value studied. As the temperature increases, the protein at pH 3–5 rapidly loses most of its native structure, as suggested by the low-temperature transitions. At pH 4.0, significant secondary structure is lost from 30 to 40 ◦ C. The protein at pH 5.0 begins to lose its native structure at ∼48 ◦ C with the transition completed near 53 ◦ C. At pH 6–8 the transition occurs between 50 and 60 ◦ C. A second transition occurs for EBA-175 RII-NG at pH 6–8 at higher temperatures. In the presence of sucrose, this is observed at ∼78 ◦ C for pH 6–7 and ∼75 ◦ C for pH 8. Without sucrose, the second transition occurs at ∼75 ◦ C for pH 6–7 and ∼70 ◦ C for pH 8 (data not shown). Results from the CD and intrinsic fluorescence studies again indicate that the presence of 5% sucrose has little-to-no effect on the conformational stability of the protein under the conditions of this study (data not shown). The transitions observed by CD occur at significantly higher temperatures than those found by intrinsic fluorescence. This suggests that EBA-175 RII-NG maintains secondary structure after loss of its native tertiary structure, suggesting the presence of a molten-globule-like state
[24,25]. This non-native state of a protein is highly susceptible to aggregation due to its loosened, more dynamic structure [26,27]. ANS is known to bind to apolar regions in proteins, resulting in a large increase in its fluorescence quantum yield [28]. In general, as a protein begins to unfold and/or form molten-globule states, apolar regions of the protein are exposed resulting in increased ANS binding, and subsequently increased fluorescence intensity. It is thought, however, that an electrostatic component may also be involved in some forms of binding due to the negative charge on the ANS [29]. Nevertheless, we can postulate that ANS is binding to apolar sites under the lower pH conditions since repulsive forces would be expected to dominate. Sharp transitions in the plots of ANS fluorescence intensity versus temperature (Fig. 5) closely resemble those seen by the previously described techniques. At pH 3.0 and 10 ◦ C, the ANS shows significant fluorescence suggesting that the tertiary structure of the protein is at least partially structurally disrupted at this temperature. Less fluorescence at pH 4.0 would seem to suggest that EBA-175 RII-NG also has some apolar character at 10 ◦ C, but less than at pH 3.0. A large transition occurs at pH 4.0 between 30 and 38 ◦ C, indicating an even greater change in structure that involves further binding of the dye. In contrast, the results at pH 5.0 suggest a decreased interaction with ANS, even after the transition near 50 ◦ C. At pH 6–8, a relatively small transition occurs from 53 to 58 ◦ C. Thus, under relatively neutral pH conditions, the protein’s structure seems to be disrupted to a lesser extent as a function of increasing temperature, at least in terms of ANS binding. In general, the transitions observed by ANS fluorescence occur ∼15–20 ◦ C higher than seen by intrinsic fluorescence and ∼5 ◦ C higher than by second derivative absorbance spectroscopy. This suggests that ANS may be
5846
L.J. Peek et al. / Vaccine 24 (2006) 5839–5851
tures greater than ∼70 ◦ C. Additionally, the turbidity traces reveal significant aggregation of the protein under such conditions. The regions of extended color transition near the phase boundaries may reflect the presence of aggregationprone molten-globule states as supported by differences in Tm from CD and fluorescence data (TmCD > Tmfl ), significant aggregation of the protein and enhanced binding of ANS to EBA-175 RII-NG under intermediate conditions. 3.3. Stabilizer screening
Fig. 6. Empirical phase diagram generated for EBA-175 RII-NG in the presence of 5% sucrose using ANS fluorescence intensity, CD molar ellipticity at 222 nm and intrinsic fluorescence spectral center of mass data. The labels indicate the state of the protein within the same region of color based on observations concluded from transition temperatures obtained by these techniques. The region of greatest stability lies within pH 5.0–8.0 at lower temperatures.
interacting with a more extensively unfolded form of the protein than that detected by the other methods. 3.2.1. Empirical phase diagrams using CD, intrinsic fluorescence and ANS fluorescence A more comprehensive and potentially higher resolution phase diagram was created by combining the data from Figs. 3, 4(b) and 5 to reflect secondary and tertiary structural changes as well as the accessibility of apolar binding sites in the protein as a function of pH and temperature (Fig. 6). Each color can be assigned a specific physical state of the protein by comparing conditions at phase boundaries with transitions observed in the actual data. The orange-colored region at the lower, right-hand corner of the diagram can be considered the phase of maximum stability. Intrinsic and ANS fluorescence, CD and high-resolution second derivative absorbance spectroscopy data all confirm this. Another phase is present at pH 4–8 at elevated temperatures. This represents a physical state in which the protein is severely structurally altered and tends to aggregate. Under such temperatures, CD and intrinsic fluorescence data suggest that the protein is significantly unfolded and plots of OD360 indicate accompanying protein aggregation. A third phase (green) is apparent at pH 3–4 and represents a state of the protein that is at least partially unfolded, but remains in solution. As the temperature is increased above 10 ◦ C at pH 3, the protein assumes the same physical state that is present at very high temperatures between pH 4 and 8. Again, this is supported by CD and intrinsic fluorescence data. The molar ellipticity at 222 nm and the fluorescence peak positions at all temperatures for pH 3 are very similar to that resulting after the transition occurs for pH 4–8. At elevated temperatures, the protein at pH 3 takes on yet another physical state (blue). At this point, the protein is presumably extensively unfolded and aggregates rapidly. This is confirmed by the change in slope of the CD and fluorescence data curves at tempera-
Since the conformational changes and subsequent aggregation of the protein seem to manifest the major degradation pathway of EBA-175 RII-NG (at least under stress conditions), a turbidity assay was developed based on light scattering (see Section 2) to screen for stabilizers of the protein. The conditions for this assay were selected according to the turbidity traces and the boundaries observed in the phase diagram. DTT was included in each sample to further promote aggregation of the protein. The turbidity results show significant aggregation of EBA-175 RII-NG at moderate temperatures (∼65 ◦ C) and pH 6.0. Additionally, the phase diagram displays a marked color transition under the same conditions. Using this aggregation-based approach, a screening of 32 GRAS compounds identified several inhibitors of EBA175 RII-NG aggregation at 45 ◦ C. These included nonionic surfactants, several sugars, polyols, guanidine hydrochloride and arginine (Table 1). Sorbitol (20%) produced 55% inhibition while Brij® 35 (0.05%) and sucrose (20%) yielded ∼54 and 53% inhibition of aggregation, respectively. Arginine (0.15 M) inhibited aggregation 52% while dextrose (20%) produced 51% inhibition. Trehalose (20%), Tween 20 (0.1%) and Brij® 35 (0.1%) provided 50, 47 and 46% inhibition, respectively. Guanidine HCl (100,000×) also showed good inhibition of aggregation, but this compound was not pursued in further studies due to its toxicity at high concentrations in injectable formulations. Using selected compounds, various combinations were created and the screening was repeated (Table 2). Increased stabilization was achieved using Brij® 35 (0.05%) and arginine (0.2 M). This combination showed complete inhibition of aggregation of EBA-175 RII-NG at 45 ◦ C. Several Table 2 Percent inhibition of aggregation of EBA-175 RII-NG obtained by using combinations of GRAS compounds Combination (concentration)
% Inhibition
Brij® 35 (0.05%) + Arg (0.2 M) Tween 20 (0.1%) + Arg (0.2 M) Brij® 35 (0.05%) + Dextrose (20%) Brij® 35 (0.05%) + Trehalose (20%) Brij® 35 (0.05%) + Sucrose (20%) Tween 20 (0.1%) + Sucrose (20%) Brij® 35 (0.05%) + Sorbitol (20%) Sucrose (20%) + Arg (0.2 M) Tween 20 (0.1%) + Sorbitol (20%) Sorbitol (20%) + Arg (0.2 M)
100 88 85 81 81 79 75 68 67 64
± ± ± ± ± ± ± ± ± ±
2 2 4 7 6 6 1 5 18 1
L.J. Peek et al. / Vaccine 24 (2006) 5839–5851
other combinations also significantly inhibited aggregation of EBA-175 RII-NG (Table 2). In all cases, these binary mixtures provided enhanced stabilization over each compound alone. Note that the observed stabilization could be due to either an effect on structural stability or direct inhibition of protein association and aggregation. 3.4. Effect of inhibitors of aggregation on protein conformation Fluorescence spectroscopy was employed to evaluate possible conformational stabilization by selected compounds (Table 3). The thermal transition temperature (Tm , the temperature at which half the conformational change is complete) was determined directly from plots of fluorescence peak position versus temperature by fitting a sigmoidal curve to the data and calculating the midpoint as the Tm . Stabilization of a protein’s conformation is considered to occur when the Tm of the protein is increased in the presence of a compound. Conversely, a structurally destabilizing compound is associated with a decrease in the Tm . On this basis, while some of the compounds selected by the turbidity-based stabilizer screening method stabilized EBA-175 RII-NG conformationally, several compounds had destabilizing effects (Table 3). For example, arginine
5847
Table 3 Effect of several GRAS compounds and combinations of these on the conformational stability (Tm ) of EBA-175 RII-NG as determined by intrinsic fluorescence spectroscopy Excipient(s) (concentration)
Tm (◦ C)
Sorbitol (20%) Sucrose (20%) Sorbitol (20%) + Brij® 35 (0.05%) Dextrose (20%) + Brij® 35 (0.05%) Sucrose (20%) + Brij® 35 (0.05%) Control—no excipient Dextrose (20%) Brij® 35 (0.01%) Brij® 35 (0.05%) Arginine (0.2 M) + Brij® 35 (0.05%) Arginine (0.2 M)
55.8 54.6 54.4 54.4 53.0 51.6 51.4 51.3 50.9 47.2 44.2
± ± ± ± ± ± ± ± ± ± ±
0.2 0.2 0.2 0.4 0.2 0.2 0.5 0.2 0.3 0.3 0.4
When the Tm is greater than that of the control, the excipient is stabilizing. When the Tm is below that of the control, the excipient is destabilizing.
exhibited such a conformational destabilizing effect. This is perhaps not surprising given the presence of the guanidinium group in this amino acid. Among those that had stabilizing effects on the protein are sorbitol (20%) and sucrose (20%), as well as combinations of each of these with Brij® 35 (0.05%). Dextrose (20%) combined with Brij® 35 (0.05%) also provided conformational stabilization.
Fig. 7. Effect of sucrose concentration on the thermal transition temperature (Tm ) of EBA-175 RII-NG in the presence of 0.01% Brij® 35 by (a) circular dichroism; (b) turbidity assay; and (c) intrinsic Trp fluorescence.
5848
L.J. Peek et al. / Vaccine 24 (2006) 5839–5851
3.5. Effects of sucrose and Brij® 35 on EBA-175 RII-NG stability Based on the fluorescence spectroscopy (Table 3) and turbidity results (Tables 1 and 2), sucrose and Brij® 35 were selected as potential excipients for an initial vaccine formulation. Sucrose at high concentration (e.g. 20%) was preferred for its substantial increase in the conformational stability of EBA-175 RII-NG as well as its widespread use in marketed injectable formulations. Brij® 35 was chosen as a second excipient for its ability to inhibit the aggregation of the protein at 45 ◦ C. Intrinsic fluorescence and CD spectroscopies as well as turbidity measurements were employed to identify the optimal minimum concentrations of Brij® 35 and sucrose. For Brij® 35, 0.01% provided optimal stabilization (data not shown). Sucrose was also found to progressively stabilize the conformational structure of the protein as the concentration was increased from 0 to 20% (data not shown). To determine the least amount of sucrose that could provide the greatest stabilization of EBA-175 RII-NG in the presence of 0.01% Brij® 35, the sucrose concentration was increased from 0 to 20%, every 2.5%, and the samples were evaluated by CD, fluorescence and a turbidity assay (Fig. 7(a–c)). A relatively steady increase in stability of the secondary structure and in aggregation inhibition with increasing sucrose concentration is observed (Fig. 7(a and b)). In contrast, increased stabilization of tertiary structure with increasing sucrose concentration is not as regular (Fig. 7(c)). A large decrease in stability occurs in the presence of 5–7.5% sucrose and 0.01% Brij® 35. There is also a hint of this effect in the CD and aggregation data (Fig. 7(a and b)) suggesting that the protein becomes conformationally destabilized in the presence of 0.01% Brij® 35 and 5–7.5% sucrose. Significant stabilization is observed, however, for EBA-175 RII-NG in the presence of 0.01% Brij® 35 and at sucrose concentrations of 10% and greater. A possible explanation for these results is that specific interactions between the protein and sugar result in some destabilization, an effect which is eventually overridden by the well known stabilization effect of sucrose through a preferential hydration mechanism. 3.6. Adjuvant studies For decades, aluminum salt adjuvants have been widely used in human vaccine formulations [23,30–34] to generate an enhanced immune response. Studies have shown that this effect is dependent upon the antigen being adsorbed to the adjuvant [30,32–37]. Mechanisms of protein adsorption can be quite complex and may include forces such as electrostatic attraction, hydrogen bonding, apolar interactions, ligand exchange and van der Waals’ forces [23,32,35,37–39]. The two aluminum salt adjuvants evaluated have quite different physicochemical properties, which often influence the mechanism by which proteins are adsorbed. Alhydrogel® typically exists in a crystalline state and has a point of zero charge (PZC, analogous to the isoelectric point (pI) of a pro-
Fig. 8. Adsorption isotherm of EBA-175 RII-NG to Adju-Phos® (closed symbols) and Alhydrogel® (open symbols) vaccine adjuvants (0.25 mg Aluminum) in the presence of various excipients—control, no excipient (square); sucrose (upside down triangle); sorbitol (circle); sorbitol and Brij® 35 (triangle); and sucrose and Brij® 35 (diamond). Adsorption of the antigen in the presence of sucrose and Brij® 35 was evaluated only at 80 g EBA175 RII-NG. Error bars representing standard error of the mean (N = 2) are included for each point; however, due to their small magnitude, they are not visible in most cases.
tein) of approximately 11. Adju-Phos® , on the other hand, is an amorphous solid and has a point of zero charge of 4.0–5.5 [40]. To evaluate adsorption of EBA-175 RII-NG to the aluminum salts, binding isotherms were generated in the presence and absence of sucrose, sorbitol and each of these combined with Brij® 35. These studies reveal that for amounts of EBA-175 RII-NG less than ∼150 g, the antigen is completely adsorbed to both adjuvants in the presence and absence of excipients (Fig. 8). In the absence of excipients, the binding of EBA-175 RII-NG saturates after addition of ∼250 g to Adju-Phos® and ∼270 g to Alhydrogel® . This suggests that excipients have a small inhibitory effect on adsorption, resulting in saturation of both adjuvants after ∼200 g EBA-175 RII-NG was added to 0.25 mg of aluminum. Desorption of the antigen was studied using four concentrations of NaCl to investigate the mechanism of adsorption (Fig. 9(a and b)). For Adju-Phos® , 2 M NaCl desorbed ∼94% of the antigen; however, due to residual buffer in the pellet that could not be completely discarded before adding the NaCl solution (which resulted in dilution of the desorbed protein), we expect that all of EBA-175 RII-NG was actually desorbed. For Alhydrogel® , less than 10% of the protein was desorbed using 2 M NaCl, suggesting a different type of physical interaction between the protein and adjuvant than seen with Adju-Phos® . Since the isoelectric point of EBA-175 RIING was calculated to be 8.62 [17], it is not surprising that high concentrations of NaCl were able to compete the antigen off of the surface of Adju-Phos® (PZC, 4.0–5.5). At pH 6.0, the protein would be positively-charged and Adju-Phos® negatively-charged, facilitating an electrostatic mechanism for adsorption. Since Alhydrogel® (PZC, ∼11) would also
L.J. Peek et al. / Vaccine 24 (2006) 5839–5851
5849
Fig. 9. Effect of NaCl concentration on the desorption of EBA-175 RII-NG from (a) Adju-Phos® and (b) Alhydrogel® vaccine adjuvants (0.5 mg aluminum) in the presence and absence of excipients. Four concentrations of NaCl were evaluated—0.5 M (vertical stripe), 0.75 M (diagonal stripe), 1.0 M (gridlines) and 2.0 M (solid).
be positively-charged at pH 6.0, it was not too surprising that NaCl did not desorb the antigen from Alhydrogel® . This suggests that some other type of interaction (not electrostatic) is responsible for adsorption of EBA-175 RII-NG to Alhydrogel® . Some conversion of Alhydrogel® to an AdjuPhos® like state may be occurring because of the use of a phosphate buffer complicating any simple interpretation of these observations. The inability to compete the protein off the surface with salt, however, argues that the majority of the adjuvant remains in the hydroxyl form. To evaluate if gradual desorption of the antigen from Alhydrogel® or Adju-Phos® occurred with time, samples were stored at 4 ◦ C. Once a week for 1 month, a sample was centrifuged and a UV absorption spectrum of the supernatant was acquired. No detectable desorption of EBA-175 RII-NG was observed over the course of this study.
4. Discussion Since malaria kills more than 1 million people each year (primarily children), an efficacious vaccine is urgently needed. Complicated by the complex life cycle of P. falciparum and P. vivax, the development of such a vaccine has proven to be difficult. An additional complication lies in the ultimate stability of any such a vaccine. Appropriate cold storage conditions are often not available in regions where malaria poses the greatest threat. Without an effective cold chain, an unstable vaccine will rapidly lose its activity. Using the approach discussed herein, optimal stabilizing conditions were identified for EBA-175 RII-NG. This antigen shows stability up to approximately 50–55 ◦ C at pH 5–8 in the absence of any stabilizers. Formulating the vaccine in a slightly acidic environment should also reduce the
chances of chemical degradation of the protein by deamidation. Additionally, several stabilizers of EBA-175 RII-NG were identified. Although sorbitol (20%) provides slightly greater stabilization, preference was given to sucrose due to its common use in approved formulation’s subsequent ideal safety profile. Brij® 35 (0.01%) combined with sucrose provides enhanced inhibition of aggregation over sucrose alone. Therefore, our proposal to include Brij® 35 in the formulation may also be advantageous despite its inability to conformationally stabilize the protein. Since the adjuvant adsorbed completely to both Adju-Phos® and Alhydrogel® , either of these adjuvants might be acceptable but Adju-Phos® is probably preferred due to its stability in phosphate buffer. Since the interaction between EBA-175 RII-NG and AdjuPhos® is understood to be primarily electrostatic based on our ability to desorb the antigen using NaCl, Adju-Phos® was selected as the adjuvant for formulation. The ability to desorb the antigen will be crucial when evaluating the longterm chemical stability of the vaccine. Following desorption of the protein from the adjuvant, techniques such as HPLC and mass spectrometry can be employed to monitor changes in the chemical stability of the protein that may occur during storage. The systematic approach described here has guided the selection of optimal stabilizing conditions for a recombinant protein vaccine, EBA-175 RII-NG based on accelerated techniques. A stable vaccine formulation for EBA-175 RII-NG was developed containing 0.01% Brij® 35 and at least 10% sucrose in the phosphate buffer (pH 6.0) employing AdjuPhos® as the adjuvant. We think a similar approach may become a valuable tool during the preformulation process of many recombinant protein vaccine candidates. It will permit fast (1–2 months for all analyses) and comprehensive conclusions regarding the relative physical stability of the protein
5850
L.J. Peek et al. / Vaccine 24 (2006) 5839–5851
and create a more truly optimized vaccine formulation. Further studies will be necessary, however, to confirm that these stabilizers are capable of providing long-term stability for the vaccine. Additionally, this systematic method reduces the chances of potentially erroneous conclusions by analyzing multiple levels of protein structure, instead of only examining one aspect of a protein’s structure. Other studies are currently underway to develop a similar approach for evaluating the chemical stability (e.g. deamidation, oxidation, etc.) of protein antigens as well as analyze protein stability on the surface of the aluminum salt adjuvants [41]. Implementation of this approach should result in a very straightforward, step-by-step preformulation routine, which may also help to reduce the time it takes to get protein-based therapeutics to patients.
Acknowledgements This project has been funded in part with Federal funds from the National Institute of Allergy and Infectious Diseases, National Institutes of Health, under Contract No. AI05421. Financial support was also provided by the Madison & Lila Self Graduate Fellowship for L. Peek and the PhRMA Foundation for L. Jones as well as grants from the DOD (DAMD17-03-C-0086) and the Thrasher Foundation.
References [1] Korenromp E, Miller J, Nahlen B, Wardlaw T, Young M. World Malaria Report 2005; 2005. [2] Richie TL, Saul A. Progress and challenges for malaria vaccines. Nature 2002;415(6872):694–701. [3] Xiao L, Rafi-Janajreh A, Patterson P, Zhou Z, Lal AA. Adjuvants and malaria vaccine development. Chem Immunol 2002;80:343–65. [4] Gilberger TW, Thompson JK, Reed MB, Good RT, Cowman AF. The cytoplasmic domain of the Plasmodium falciparum ligand EBA175 is essential for invasion but not protein trafficking. J Cell Biol 2003;162(2):317–27. [5] Webster D, Hill AV. Progress with new malaria vaccines. Bull World Health Organ 2003;81(12):902–9. [6] Miller LH, Howard RJ, Carter R, Good MF, Nussenzweig V, Nussenzweig RS. Research toward malaria vaccines. Science 1986; 234(4782):1349–56. [7] Camus D, Hadley TJ. A Plasmodium falciparum antigen that binds to host erythrocytes and merozoites. Science 1985;230(4725):553–6. [8] Sim BK, Chitnis CE, Wasniowska K, Hadley TJ, Miller LH. Receptor and ligand domains for invasion of erythrocytes by Plasmodium falciparum. Science 1994;264(5167):1941–4. [9] Sim BK, Orlandi PA, Haynes JD, Klotz FW, Carter JM, Camus D, et al. Primary structure of the 175 K Plasmodium falciparum erythrocyte binding antigen and identification of a peptide which elicits antibodies that inhibit malaria merozoite invasion. J Cell Biol 1990;111(5 Pt. 1):1877–84. [10] Liang H, Sim BKL. Conservation of structure and function of the erythrocyte binding domain of Plasmodium falciparum EBA-175. Mol Biochem Parasitol 1997;84:241–5. [11] Ausar SF, Rexroad J, Frolov VG, Look JL, Konar N, Middaugh CR. Analysis of the thermal and pH stability of human respiratory syncytial virus. Mol Pharm 2005;2(6):491–9.
[12] Fan H, Ralston J, Dibiase M, Faulkner E, Middaugh CR. Solution behavior of IFN-beta-1[alpha]: an empirical phase diagram based approach. J Pharm Sci 2005;94(9):1893–911. [13] Jiang G, Joshi SB, Peek LJ, Brandau DT, Huang J, Ferriter MS, et al. Anthrax vaccine powder formulations for nasal mucosal delivery. J Pharm Sci 2006;95(1):80–96. [14] Kueltzo LA, Ersoy B, Ralston JP, Middaugh CR. Derivative absorbance spectroscopy and protein phase diagrams as tools for comprehensive protein characterization: a bGCSF case study. J Pharm Sci 2003;92(9):1805–20. [15] Rexroad J, Evans RK, Middaugh CR. Effect of pH and ionic strength on the physical stability of adenovirus type 5. J Pharm Sci 2006;95(2):237–47. [16] Brandau DT, Jones LS, Wiethoff CM, Rexroad J, Middaugh CR. Thermal stability of vaccines. J Pharm Sci 2003;92(2):218–31. [17] Putnam C. Protein Calculator; 2005 http://www.scripps.edu/∼ cdputnam/protcalc.html. [18] Gill SC, von Hippel PH. Calculation of protein extinction coefficients from amino acid sequence data. Anal Biochem 1989;182(2):319–26. [19] Science Applications International Corporation (SAIC). EBA-175 RII-NG Malaria Vaccine Milestone Meeting—Readiness for GMP Production; 2003. [20] Mach H, Middaugh CR. Simultaneous monitoring of the environment of tryptophan, tyrosine, and phenylalanine residues in proteins by near-ultraviolet second-derivative spectroscopy. Anal Biochem 1994;222(2):323–31. [21] Tolia NH, Enemark EJ, Sim BK, Joshua-Tor L. Structural basis for the EBA-175 erythrocyte invasion pathway of the malaria parasite Plasmodium falciparum. Cell 2005;122(2):183–93. [22] United States Code of Federal Regulations. 21 CFR § 610.15(a); 2001. p. 69. [23] Gupta RK, Rost BE, Relyveld E, Siber GR. Adjuvant properties of aluminum and calcium compounds. In: Powell MF, Newman MJ, editors. Vaccine Design: The Subunit and Adjuvant Approach. New York: Plenum Press; 1995. [24] Dolgikh DA, Gilmanshin RI, Brazhnikov EV, Bychkova VE, Semisotnov GV, Venyaminov SY, et al. [Alpha]-lactalbumin: compact state with fluctuating tertiary structure? FEBS Lett 1981;136(2): 311. [25] Ohgushi M, Wada A. ‘Molten-globule state’: a compact form of globular proteins with mobile side-chains. FEBS Lett 1983; 164(1):21–4. [26] Mach H, Ryan JA, Burke CJ, Volkin DB, Middaugh CR. Partially structured self-associating states of acidic fibroblast growth factor. Biochemistry 1993;32(30):7703–11. [27] Kotik M, Zuber H. Mutations that significantly change the stability, flexibility and quaternary structure of the l-lactate dehydrogenase from Bacillus megaterium. Eur J Biochem 1993;211(1–2):267–80. [28] Stryer L. The interaction of a naphthalene dye with apomyoglobin and apohemoglobin. A fluorescent probe of non-polar binding sites. J Mol Biol 1965;13(2):482–95. [29] Matulis D, Lovrien R. 1-Anilino-8-naphthalene sulfonate anionprotein binding depends primarily on ion pair formation. Biophys J 1998;74(1):422–9. [30] Baylor NW, Egan W, Richman P. Aluminum salts in vaccines—US perspective. Vaccine 2002;20(Suppl. 3):S18–23. [31] HogenEsch H. Mechanisms of stimulation of the immune response by aluminum adjuvants. Vaccine 2002;20(Suppl. 3):S34–9. [32] Lindblad EB. Aluminium compounds for use in vaccines. Immunol Cell Biol 2004;82(5):497–505. [33] Morefield GL, Sokolovska A, Jiang D, HogenEsch H, Robinson JP, Hem SL. Role of aluminum-containing adjuvants in antigen internalization by dendritic cells in vitro. Vaccine 2005;23(13):1588–95. [34] WHO. Immunological Adjuvants. Geneva: World Health Organization; 1976. [35] Gupta RK. Aluminum compounds as vaccine adjuvants. Adv Drug Deliv Rev 1998;32(3):155–72.
L.J. Peek et al. / Vaccine 24 (2006) 5839–5851 [36] Mannhalter JW, Neychev HO, Zlabinger GJ, Ahmad R, Eibl MM. Modulation of the human immune response by the non-toxic and non-pyrogenic adjuvant aluminium hydroxide: effect on antigen uptake and antigen presentation. Clin Exp Immunol 1985;61(1): 143–51. [37] Seeber SJ, White JL, Hem SL. Predicting the adsorption of proteins by aluminium-containing adjuvants. Vaccine 1991;9(3):201– 3. [38] Al-Shakhshir RH, Regnier FE, White JL, Hem SL. Contribution of electrostatic and hydrophobic interactions to the adsorption of proteins by aluminium-containing adjuvants. Vaccine 1995;13(1):41–4.
5851
[39] Iyer S, Robinett RS, HogenEsch H, Hem SL. Mechanism of adsorption of hepatitis B surface antigen by aluminum hydroxide adjuvant. Vaccine 2004;22(11–12):1475–9. [40] Vogel FR, Powell MF, Alving CR. A compendium of vaccine adjuvants and excipients. In: Powell MF, Newman MJ, editors. Vaccine Design: The Subunit and Adjuvant Approach. New York: Plenum Press; 1995. [41] Jones LS, Peek LJ, Power J, Markham A, Yazzie B, Middaugh CR. Effects of adsorption to aluminum salt adjuvants on the structure and stability of model protein antigens. J Biol Chem 2005;280(14): 13406–14.