- Email: [email protected]
Characterization of antigens adsorbed to anionic PLG microparticles by XPS and TOF‐SIMS
PHARMACEUTICS, PREFORMULATION AND DRUG DELIVERY Characterization of Antigens Adsorbed to Anionic PLG Microparticles by XPS and TOF-SIMS JAMES CHESKO,1 JINA KAZZAZ,1 MILDRED UGOZZOLI,1 MANMOHAN SINGH,1 DEREK T. O’HAGAN,1 CLAIRE MADDEN,2 MARK PERKINS,2 NIKIN PATEL2 1
Novartis Vaccines and Diagnostics Division, Emeryville, California
2
Molecular Profiles, Nottingham, UK
Received 3 October 2006; revised 16 April 2007; accepted 17 April 2007 Published online in Wiley InterScience (www.interscience.wiley.com). DOI 10.1002/jps.21040
ABSTRACT: The chemical composition of the surface of anionic PLG microparticles before and after adsorption of vaccine antigens was measured using X-ray photoelectron spectroscopy (XPS) and time-of-flight secondary ion mass spectrometry (TOF-SIMS). The interfacial distributions of components will reflect underlying interactions that govern properties such as adsorption, release, and stability of proteins in microparticle vaccine delivery systems. Poly(lactide-co-glycolide) microparticles were prepared by a w/o/w emulsification method in the presence of the anionic surfactant dioctyl sodium sulfosuccinate (DSS). Ovalbumin, lysozyme, a recombinant HIV envelope glyocoprotein and a Neisseria meningitidis B protein were adsorbed to the PLG microparticles, with XPS and time-of-flight secondary mass used to analyze elemental and molecular distributions of components of the surface of lyophilized products. Protein (antigen) binding to PLG microparticles was measured directly by distinct elemental and molecular spectroscopic signatures consistent with amino acids and excipient species. The surface sensitive composition of proteins also included counter ions that support the importance of electrostatic interactions being crucial in the mechanism of adsorptions. The protein binding capacity was consistent with the available surface area and the interpretation of previous electron and atomic force microscope images strengthened by the quantification possible by XPS and the qualitative identification possible with TOF-SIMS. Protein antigens were detected and quantified on the surface of anionic PLG microparticles with varying degrees of efficiency under different adsorption conditions such as surfactant level, pH, and ionic strength. Observable changes in elemental and molecular composition suggest an efficient electrostatic interaction creating a composite surface layer that mediates antigen binding and release. ß 2007 Wiley-Liss, Inc. and the American Pharmacists Association J Pharm Sci 97:1443–1453, 2008
Keywords: vaccine; anionic microparticle; antigen; XPS; TOF-SIMS; delivery system; PLG
Correspondence to: James Chesko (Vaccines and Diagnostics Division, Novartis Corporation, 4560 Horton St., Emeryville, CA 94608. Telephone: 510-923-3896; Fax: 510-923-2586; E-mail: [email protected]) Journal of Pharmaceutical Sciences, Vol. 97, 1443–1453 (2008) ß 2007 Wiley-Liss, Inc. and the American Pharmacists Association
INTRODUCTION The role of surface interactions is crucial to formulations with high surface areas such as microparticles and nanoparticles that interact
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 97, NO. 4, APRIL 2008
1443
1444
CHESKO ET AL.
with the immune system.1 The surface properties that modulate adsorption and release of biomolecules also may stabilize or denature the protein or biomolecule in question.2 Tomorrow’s generation of new vaccines with recombinant antigens will often require an optimized delivery system that carefully controls changes in protein molecular structure and disposition, factors that may change antigenic determinants of immunogenicity during uptake, processing, and presentation, thus affecting the ultimate potency of the vaccine.3,4 The interactions at the particle interface will ultimately determine the chemical composition localized at the surface. As processes such as lyophilization occur, the surface will undergo a dehydration process as an assortment of tightly bound molecules, ions, counter ions, surfactants, and adsorbed species consolidate on a solvent-depleted interface.5,6 The stability of the protein in this environment will depend upon the microenvironment and the various factors that influence its formation, evolution, and structure.7,8 A direct method to characterize the surface composition in the presence of bulk material is to utilize microspectral methods such as X-ray photoelectron spectroscopy (XPS)9 and time-offlight secondary ion mass spectrometry (TOFSIMS).10,11 These techniques are exquisitely sensitive to the outermost 5–10 nm even of a large 1000 nm particle, that is, the outer 1% of the total material.12 This measurement will provide specific information about the outermost surface of the particles, the dehydrated Stern layer, and the interfacial structure of the vaccine delivery system that is probed at the molecular level by receptors on the surface of antigen presenting cells of the immune system. Significant insights into the chemical composition and possible molecular players in the process of stabilizing or binding the protein (e.g., surfactants) may be gleaned from examination of this data.13 Previous research suggesting a surface monolayer of adsorbed protein and a large, Coulombic driving force accompanying the surface interactions14 may be tested by such results. Understanding the role of the various excipients, buffers, sugars, and salts in the system and during the lyophilization process may be possible following careful analysis. Through this understanding the formulation of vaccines with desirable adsorption, stability, and release properties for antigens may be possible.15,16 JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 97, NO. 4, APRIL 2008
MATERIALS AND METHODS Materials The random copolymer RG503, poly(D,L-lactideco-glycolide) 50:50 lactide to glycolide ratio (intrinsic viscosity 0.4, from manufacturers specifications) was obtained (Boehringer Ingelheim, Petersburg, VA). The surfactant dioctylsulfosuccinate (DSS), the proteins lysozyme, and ovalbumin came from Sigma-Aldrich Chemical (St. Louis, MO). The envelope protein gp120dV2, a CHO derived recombinant product, was expressed and purified in house (Chiron Vaccine Research in Emeryville, CA.) Escherichia coli (E. Coli)-derived recombinant meningococcal B protein MB1 was expressed in house (Novartis Vaccines, Siena, IRIS, Via Fiorentina 1, 53100 Siena, Italy) following isolation and purification as previously described.17 Phosphate buffered saline (pH 7) was made in house with USP grade 0.8% w/w NaCl, 0.02% KCl, 0.134% potassium dihydrogen phosphate, 0.002% w/w potassium dihydrogen phosphate. The citrate (pH 5) buffer was also made in house with USP grade 2.0% sodium citrate and 0.67% citric acid in autoclaved water for injection.
Methods Preparation of Microparticles Microparticles were prepared by a solvent evaporation technique. Briefly, microparticles were prepared by homogenizing 10 mL of 6% w/v polymer solution in methylene chloride (Reagent Grade ACS), with 2.5 mL phosphate buffered saline (1 PBS pH 7) or 10 mM citrate pH 5 buffer using a 10-mm probe (Ultra-Turrax T25 IKALabortechnik, Staufen, Germany) thus forming a water in oil emulsion which was then added to 50 mL of distilled water containing either 6 or 60 mg/mL DSS and homogenized at very high speed using a homogenizer with a 20-mm probe (ES-15 Omni International, Marietta, GA) for 25 min in an ice bath. This resulted in water in oil in water emulsion, which was stirred at 1000 rpm for 12 h at room temperature, and the methylene chloride was allowed to evaporate. The resulting microparticles had total surfactant levels of either 0.05% or 0.5% DSS wt/wt. The microparticle preparation was then freeze-dried (lyophilized.) DOI 10.1002/jps
CHARACTERIZATION OF ANTIGENS
X-ray Photoelectron Spectroscopy XPS measurements were taken on an Axis-Ultra instrument from Kratos Analytical (Manchester, UK) using monochromated Al Ka radiation (1486.6 eV) operated at 15 mA emission current and 10 kV anode potential (150 W). A low energy electron flood gun was used to compensate for insulator charging. Survey scans, from which quantification of the detected elements was obtained, were acquired with analyzer pass energy of 160 eV and a 1 eV step size. Highresolution scans of all detected elements were also acquired with a pass energy of 40 eV and a 0.1 eV step size. The area examined was approximately 700 mm 300 mm for the survey scans and a 110 mm diameter spot for the high-resolution scans. Time-of-flight Secondary Ion Mass Spectrometry TOF-SIMS was performed on a TOF-SIMS IV (IonTOF, Munster, Germany) using a 10 kV cesium ion source. Positive and negative ion spectra were acquired over a 500 500 mm area. In all experiments the primary ion dose density was less than 1012 ions per cm2 to ensure experiments were carried out under static conditions. To overcome sample charging a low energy electron flood gun (0.1 mA DC) was used. In each instance the lyophilized powder samples were mounted on a stub without additional modification and introduced into the vacuum chambers of appropriate instrument. Signals were averaged from several locations (typically N ¼ 3) of the sample to provide a representative average of quantities reported, with errors expressed as the standard deviation of the mean.
1445
the quantifiable range of 1.0–100 ug/mL. An alternate spectrophotometric method involving the enhanced partitioning of acridine orange into a toluene phase from water was also employed in attempts to lower the sensitivity limit, but gave basically identical results due to sensitivity to interferences at the lowest DSS concentrations (<1 ug/mL). Adsorption of Protein to Microparticles To prepare microparticles with adsorbed protein, a suspension containing 100 mg of PLG was incubated with 1 mg protein in 10 mL total volume of 10 mM of the appropriate buffer and left on a lab rocker (aliquot mixer, Miles Laboratories, Elkhart, IN) at 48C overnight. The suspension was then lyophilized. At the 1% loading level of protein to PLG weight ratio the particle surface is only covered at 0.5 monolayers and excess (free) protein and surfactant was only present in trace quantities to satisfy equilibrium partitioning.14 Earlier experiments that involved washing PLG microparticles with water to remove excess protein and surfactant showed no significant change in the amount of protein surface adsorbed, so the samples analyzed here are compositionally very similar to particles in the formulated dispersion, with excess (unbound or loosely bound) water removed during lyophilization. Some reorganization due to this solvent depletion was expected to occur, but we did not believe that surface species would migrate outside the 5–10 nm surface thickness composition probed by XPS and MS-TOF.
Microparticle Characterization
Surfactant (DSS) Adsorption to Microparticles
X-ray Photoelectron Spectroscopy
The fraction of DSS associated with the microparticles was measured as described earlier.18 Briefly, centrifugal separation of the solid (PLG) phase from the supernatant (aqueous) phase was followed by assaying both fractions by ion exchange chromatography with electrical conductivity detection across a proton exchange membrane (Waters 2695 Alliance chromatography system with Dionex AS-1 column.) The solid (PLG) phase sample was hydrolyzed in 1N NaOH to dissolve the matrix and then pH neutralized before injection. This analytical method was sensitive (LLOD) to 0.5–1 ppm (1 ug/mL) DSS, coefficient of determination typically 0.995 over
XPS was applied to quantitatively determine the surface elemental composition of the lyophilized powder samples. The technique uses X-ray irradiation to cause the emission of photoelectrons. These photoelectrons are emitted with specific binding energies that relate to the electron core level from which they were ejected and the peak intensity is proportional to the atomic concentration within the sampled volume. For example, carbon and oxygen abundance are distinguishable through the C(1s) and O(1s) photoelectrons near 280 and 530 eV, respectively. With the use of controls and standards the ratios of the observed elements and the binding energies
DOI 10.1002/jps
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 97, NO. 4, APRIL 2008
1446
CHESKO ET AL.
can be related to the molecular composition of a sample. The levels of detection in an XPS experiment are typically 0.1atomic % (1 atom in 1000). Time-of-flight Mass Spectrometry TOF-SIMS is a complementary technique to XPS providing molecular speciation from the surface under analysis via mass spectrometry.19 In SIMS, the sample, in ultra high vacuum (UHV), is bombarded with primary ions with energies in the range 2–25 keV. Energy transfer to the solid via a collision cascade results in ejection of material from the surface (sputtering). Approximately 95% of this material comes from the uppermost two monolayers (1 nm), hence SIMS is highly surfacespecific. Sputtered particles range in size from single atoms to large assemblies of atoms, including intact molecules. Typically, about 1% of emitted ions carries a single positive or negative charge, and therefore can be collected and analyzed by their mass to charge ratio. The peaks observed in the mass spectra were analyzed looking for fragments previously identified from common amino acids.20,21
RESULTS AND DISCUSSION XPS Elemental Quantification of Surface Species The XPS elemental quantifications from controls and lyophilized micro-particles are shown in Table 1. These data have been extracted from the integrated intensities of the main photoelectron lines observed in the survey spectra such as that shown in Figure 1. This survey spectrum is recorded from the lyophilized PLG particles alone and highlights that the only elements detected
Figure 1. A typical XPS wide scan from which elements are quantified. Wide scans were recorded from each sample that allow the atomic concentrations for each element present within the uppermost 5–10 nm to be determined.
here were carbon, oxygen with low levels of sodium. The oxygen:carbon ratio (O:C) was 0.7 and relates closely to the expected value from a pure PLG surface of 0.8 (from raw starting material). The excess carbon is not unexpected and probably relates to so called ‘‘adventitious’’ hydrocarbon. This represents low levels of CH found on the majority of polmeric samples analyzed by XPS.22 To verify the amount of DSS available to the PLG surface, ion exchange chromatography was performed on representative samples. Since the 0.05% DSS w/w PLG represented an average DSS concentration of 6 ug/mL, the DSS partitioning for this system was measurable to >90% on the PLG particles (5–6 ug DSS found in 1 mL of the PLG solid phase, with no more than 1 ug maximum DSS in the supernatant aqueous phase). For the
Table 1. Elemental Composition of Surface Layer for PLG Particles with Different Buffering Conditions and Levels of DSS Surfactant Element C O N Na S Cl P Si
PLG 60.1 39.7 0.0 0.1 0.0 0.0 0.0 0.0
(0.2) (0.1) (0.0) (0.1) (0.1) (0.0) (0.0) (0.0)
PLG þ 0.05% DSS 60.3 38.8 0.0 0.5 0.0 0.1 0.0 0.0
(0.1) (0.2) (0.0 (0.1) (0.0) (0.2 (0.0 (0.0)
PLG þ 0.5% DSS 60.8 38.3 0.0 0.5 0.1 0.0 0.0 0.0
(0.1) (0.2) (0.0) (0.3) (0.1) (0.0) (0.0) (0.0)
PLG þ 0.05% DSS (pH 5) 59.5 39.5 0.0 0.5 0.0 0.2 0.0 0.0
(0.1) (0.2) (0.0) (0.4) (0.0) (0.2) (0.0) (0.0)
PLG þ 0.5% DSS (pH 5) 62.7 36.3 0.0 1.2 0.2 0.0 0.0 0.0
(0.5) (0.3) (0.0) (0.1) (0.1) (0.0) (0.0) (0.0)
Data are the average of sampling from three areas for each preparation and are from integrated XPS band intensities, with standard deviations shown in parentheses. JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 97, NO. 4, APRIL 2008
DOI 10.1002/jps
1447
CHARACTERIZATION OF ANTIGENS
Table 2. XPS Elemental Composition of PLG and Protein Standards, Averaged from Three Distinct Sampling Regions
Element C O N Na S Cl P Si
PLG þ 0.05% DSS (pH 7) 55.6 40.6 0.0 2.7 0.0 0.8 0.7 0.0
(0.5) (0.5) (0.0) (0.2) (0.0) (0.1) (0.1) (0.0)
PLG þ 0.5% DSS (pH 7) Surfactant 58.8 37.4 0.0 2.1 0.2 0.4 0.8 0.0
(0.4) (1.1) (0.0) (0.5) (.01) (0.2) (0.1) (0.0)
Ovalbumin 66.4 17.4 15.5 0.3 0.1 0.2 0.5 0.0
(0.5) (0.1) (0.5) (0.1) (0.1) (0.1) (0.2) (0.1)
Lysozyme 65.3 15.6 17.0 0.0 0.4 0.0 0.0 0.0
(1.0) (0.6) (0.8) (0.0) (0.2) (0.0) (0.0) (0.0)
gp120 63.0 26.7 8.6 0.6 0.0 1.7 0.0 0.0
(0.5) (0.6) (0.5) (0.2) (0.0) (0.3) (0.0) (0.0)
MB1 61.4 20.2 12.5 3.0 0.0 2.4 0.0 0.6
(0.7) (0.4) (0.6) (0.2) (0.0) (0.3) (0.0) (0.3)
Data are the average of sampling from three areas for each preparation and are from integrated XPS band intensities, with standard deviations shown in parentheses.
0.5% DSS w/w PLG formulation, the DSS concentration typically saturated at 0.15– 0.20% w/w PLG, with mass balance measured to within 10% in the aqueous supernatant phase. The 0.5% DSS w/w PLG represents a microparticle with surface saturation of DSS (a limiting case to compare the intermediate 0.05% DSS concentration with). The XPS surface concentrations of sulfur show a clear increase with higher DSS loading (Tabs. 1 and 2), in addition to shifts in the carbon (upwards) and oxygen (downwards), all consistent with a higher surface presence of the molecular dioctylsulfosuccinate which has a higher C:O ratio than PLG. In samples that have PLG with the DSS surfactant (different pH and DSS concentrations) the oxygen to carbon ratios remain relatively constant. Sodium cations are present to form the sodium salt of the carboxyl terminal group of the lactic and glycolic acid residues. Anions present include chloride (typically present together with sodium, but in less than a 1:1 stoichoimetric ratio) sulfur from dioctylsulfosuccinate (DSS) in the case of high (0.5%) surfactant concentration, but below limits of detection for the low (0.05% DSS) loading. The presence of counter ions, especially cationic species such as Naþ or Kþ, are ubiquitous to the various samples and underscore the importance of charge balance and electrostatic interactions in the formation and structure of the microparticle surface. For example, the extent of dissociation from PLG ionizable groups (primarily lactide and glycolide carboxyls, with some hydroxyls) at pH 7.0 is greater than at pH 5.0, resulting in a greater charge density at the surface. This is reflected directly in a 20 mV negative shift in the surface zeta potential,14 and supports a more extensive DOI 10.1002/jps
layer of counter ions (and the possible formation of a weak dielectric stack structure). The most mobile ionic species with significant charge density include the group I alkali (Naþ, Kþ) and their counter ions group VII halogen (Cl) with some PO3 4 (or citrate anion) from the buffer. When these salt and buffer species are incorporated to a greater extent into the interfacial layer, their prevalence increases (from a combined Na þ P þ S þ Cl of 0.7 and 1.4% at pH 5 to 4.2 and 3.5% at pH 7) and the relative proportion of carbon and oxygen decreases. From the molecular formulas of DSS (C20H38O7SNa), the relative proportion on the surface increases from the S content (sulfosuccinate) from the 0.05% to 0.5% loading levels, indicating a surface enhanced adsorption process that would be expected for a surfactant species. The most pronounced case of DSS adsorption appears to be the 0.5% DSS pH 5 sample, where increased adsorption of DSS (dioctyl sulfosuccinate, sodium salt) to the PLG at the saturating surfactant concentration will sequester necessary counter cations (Naþ) to balance charge in the lyophilized powder. The relatively high sulfur content for this formulation (0.2%) brings a concomitant high amount of Naþ (1.2%) for charge balance. The surface composition of pH 7 buffered PLG and the powdered protein standards show some notable differences from the previous samples, as shown in Table 2. The surface composition of C and O has dropped slightly, and a richer diversity of ionic species (group I alkali metals, group V–VII nonmetal elements) is observed. The greater negative charge on the surface of the microparticles (from zeta potential measurements14) promotes the adsorption of ionic species of opposite JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 97, NO. 4, APRIL 2008
1448
CHESKO ET AL.
charge (sodium levels two to three times higher than the previous samples). Along with the sodium cations there are phosphorous (likely as hydrogen phosphate), chlorine (as chloride), and sulfur (as dioctylsulfosuccinate). Cationic proteins (e.g., basic proteins with high isoelectric points such as lysozyme) have less cationic sodium present than low isoelectric point (acidic) proteins such as MB1. Since the poly(lactide-co-glycolide) has significant proportions of carboxylic acid end groups that are ionized at physiological pH values, these provide a source of anionic species that can form salt bridges with the basic proteins, reducing the amount of other anions sequestered from solution to balance charge at the interface. The amount of cations, with Naþ the primary species considered, is expected to charge balance the total surface composition. By electrostatic considerations, the more acidic ovalbumin with lower pI will be less efficiently adsorbed on the PLG surface, as the nitrogen content values show very clearly. As a result, the counterbalancing concentration of cations (Naþ, Kþ) will also be less, as the data also reflect. The adsorption of efficiency of gp120 is much higher than MB1 and ovalbumin, so the cations required to form the charge neutral sodium salt will also be higher. The level of nitrogen present in the protein sample is indicative of posttranslational modifications such as glycosylation as well as proportion of basic (nitrogen containing) residues. For example, lysozyme is the most basic protein and has the highest nitrogen content at 17.8%. Contrast this value to that of 15.1% for ovalbumin, which has very little glycosylation of phosphorylation, the difference of 2.7% is attributable to the higher proportion of N containing residues. MB1 is a
more acidic protein and is also slightly glycosylated, having a nitrogen percentage of 12.0%. The envelope protein gp120 is less acidic but heavily glycosylated (approximately 50% sugar to total weight) and thus its nitrogen content of 8.2% is about half of lysozyme or albumin. Since previous data14 show the efficient adsorption of gp120 on PLG, the large difference in molecular composition due to extensive glycosylation is strongly reflected in the nitrogen content of the surface adsorbed species. When the proteins are adsorbed on to the PLG microparticles and then lyophilized, there is a noticeable variance in the outermost elemental composition. The presence of nitrogen is a unique contribution from the protein, and indicates the degree of incorporation of the protein into the interfacial region of the microparticles (see Tab. 3). The relative amount of nitrogen should scale with the concentration of protein adsorbed, showing the strongest adsorbing species (lysozyme, gp120, MB1) are more efficiently taken up than the weak adsorbers (albumin) as shown in Table 4. The differences seen in the strong adsorbers lysozyme versus gp120 are interpretable as the limiting cases of monolayer saturation in the case of the basic, monomeric lysozyme as compared to a propensity to form associated complexes and perhaps adsorb into a second layer for the more acidic, heavily glycosylated dimeric/ trimeric gp120. The MB1 protein falls into the intermediate range of these two extremes. The amount of chloride coadsorbed generally follows the extent of charge needed for electrostatic balance. For acidic proteins such as ovalbumin and MB1 the chloride levels are very low, whereas for the more basic gp120 and lysozyme it is higher.
Table 3. XPS Elemental Quantification of PLG Microparticles Containing the Protein and Either 0.05% or 0.5% DSS Surfactant
Element C O N Na S Cl P Si
gp120 0.05% DSS 53.7 35.2 4.1 3.4 0.0 1.9 0.0 0.0
(0.5) (0.6) (0.2) (0.4) (0.0) (1.1) (0.0) (0.0)
gp120 0.5% DSS 57.0 36.5 1.4 2.7 0.2 0.6 0.7 0.0
(0.8) (0.8) (0.3) (0.2) (0.0) (0.3) (0.2) (0.0)
MB1 0.05% DSS 60.6 32.8 4.1 2.5 0.0 0.4 0.0 0.0
(0.5) (0.5) (0.2) (0.1) (0.0) (0.1) (0.0) (0.0)
MB1 0.5% DSS 60.4 35.1 2.5 1.5 0.2 0.2 0.0 0.0
(0.7) (0.8) (0.3) (0.1) (0.1) (0.1) (0.0) (0.0)
Lysozyme 0.05% DSS 54.4 33.1 4.3 5.1 0.0 1.3 1.3 0.0
(0.6) (0.7) (0.1) (0.2) (0.3) (0.1) (0.0) (0.0)
Lysozyme 0.5% DSS
Ovalbumin 0.05% DSS
58.1 36.7 2.4 2.3 0.0 0.0 0.2 0.0
59.27 38.1 0.82 1.42 0.00 0.00 0.17 0.00
(0.9) (0.6) (0.4) (0.2) (0.2) (0.0) (0.3) (0.0)
(0.5) (1.0) (0.1) (0.2) (0.0) (0.0) (0.2) (0.0)
Ovalbumin 0.5% DSS 60.9 37.6 0.5 1.5 0.1 0.0 0.27 0.00
(0.9) (0.5) (0.1) (0.1) (0.1) (0.0) (0.2) (0.0)
Data are the average of sampling from three areas for each preparation and are from integrated XPS band intensities, with standard deviations shown in parentheses. JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 97, NO. 4, APRIL 2008
DOI 10.1002/jps
CHARACTERIZATION OF ANTIGENS
Table 4. XPS Quantification of Protein Concentrations in the Outermost 5–10 nm Shell of PLG Microparticle Delivery System, Taken from Two 700 mm 300 mm Surface Patches on the Loaded Microparticles
Sample gp120 MB1 Lysozyme Ovalbumin
Surfactant Concentration (%)
Nitrogen (Protein) Fraction of Adsorbed 5–10 nm Layer, as Compared to Protein Reference
0.05 0.5 0.05 0.5 0.05 0.5 0.05 0.5
47.7 16.3 32.8 20.0 25.3 14.1 5.3 3.0
These values are based upon the nitrogen content, considered as a unique elemental signature of the protein molecules.
Other anions such as phosphate (detected as phosphorous) can also be present to balance the positively charged functional groups. The presence of surfactant seems to compete for surface sites, perhaps carrying or accommodating molecules and ions other than the protein, and having a net effect of causing passivation toward efficient protein adsorption in all cases measured. The surfactant is a relatively small, primarily linear molecule and a relatively minor weight fraction adsorbed on the PLG surface could
1449
effectively modify a large fraction of the surface unfavorably against adsorption of proteins. This data suggest that DSS can generally be regarded as a species competing for surface adsorption sites, so at its highest (saturating) concentration it can effectively limit or reduce the capacity of PLG to adsorb protein antigens (in the four proteins shown). This less than desirable trait is counterbalanced by several benefits (increase in particle surface area, uniformity of formulation upon resuspension, modulation of release characteristics.) As the fraction of surface sites occupied by DSS increases (i.e., S content increases), the relative amount of proteins adsorbed (and N content) decreases. The sodium ions are still present at considerable levels, but seem to be preferentially displaced by the available Kþ in the PBS buffer, perhaps since potassium is a better counter cation (larger, more polarizable, more easily ionizable).
TOF-SIMS Qualitative Measurement of Amino Acids and Peptide Fragments, Leading toward the Identification of Proteins To further identify the presence of proteins and eliminate any reasonable uncertainty concerning the interpretation of elemental markers such as nitrogen for proteins and sulfur for DSS, TOFSIMS was undertaken to look at the molecular nature of the adsorbed layer. Due to very complicated factors influencing the production,
Figure 2. A positive ion spectra for the anionic surfactant DSS. DOI 10.1002/jps
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 97, NO. 4, APRIL 2008
1450
CHESKO ET AL.
Figure 3. TOF-SIMS of PLG with no DSS added. The labeled m/z peaks correspond to combinations of the glycolide (G) and lactide (L) polymer subunits as monomers (G or L), dyads (GG, GL, LL), and higher.
stability, suppression, and transmission of various ionic complexes, quantitation of the components through ion intensity is inherently difficult without special methods.23 In the case of DSS, the standard (pure) material yielded a strong m/z ¼ 467 peak, corresponding to a DSS molecule that has picked up a Naþ to yield a net plus one charge (Fig. 2). Some lighter m/z peaks show the typical pattern of fragmentation of an alkyl chain through successive loss of methylenes. The PLG microparticles without surfactant show a peak progression based upon combination of the lactide and glycolide monomer units, with the various permutations of glycolide (G) and lactide (L) expected in the dyads (GG, GL, and LL) and triads of a random copolymer (Fig. 3). This
type of fragmentation is expected and has been reported previously.24 The TOF-SIMS spectrum of the PLG microparticles with 0.5% w/w DSS added has bands that are attributable to both the dominant component, PLG, and the surfactant DSS (Fig. 4). This is clear evidence that the surfactant is present on the surface of the lyophilized microparticles and can influence the adsorption properties of the interface. The presence of parent ion peaks for DSS with both Naþ and Kþ adduct cations suggests that electrostatics are facilitating the association of different counter ions with the surfactant and surface polymer. The microparticles were assessed for protein content based on the identification of peaks known to be specific for amino acid residues20 using the
Figure 4. TOF-SIMS spectra of PLG with DSS (0.5%) added. The mass is a superposition of the dominant component, PLG, with its associated ion pattern, and a weaker but recognizable cluster of high m/z peaks near 467, attributable to DSS molecules that have picked up an alkali ion (Naþ, Kþ). JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 97, NO. 4, APRIL 2008
DOI 10.1002/jps
CHARACTERIZATION OF ANTIGENS
1451
Table 5. Typical Amino Acids Observed from Various Adsorbed Proteins, in the Case Lysozyme (from m/z Values Taken from TOF-SIMS of PLG/DSS with Adsorbed Protein Samples) Lysozyme (0.05% DSS) Gly þ Val þ Arg þ
Ala þ Thr þ Glu
Ser þ Cys Tyr þ
Met þ His þ Phe þ
Pro Asn þ Trp þ
Lysozyme (0.5% DSS) Gly þ Val þ Arg
Ala þ Thr Glu þ
Ser þ Cys þ Tyr
Met His þ Phe þ
Pro þ Asn þ Trp þ
The microparticles were assessed for protein content based on the identification of peaks known to be specific for amino acid residues. Typical positive ion fragments for common amino acids have previously been identified and illustrated in this table.
Figure 5. Typical mass fragments detected at the surface of the loaded microparticles. (a) NH4þ fragment ubiquitous to all proteins. (b) C2H6Nþ fragment from Alanine.
high mass discrimination of the spectrometer. For the PLG/DSS with adsorbed proteins, the TOFSIMS spectra were densely patterned peaks comprised as superpositions of the abundant PLG peaks, the DSS peaks, and numerous fragment peaks attributable to small peptides and amino acids that are derived from proteins.20 A comparison of the high mass resolution spectra from the blank and loaded microparticles is illustrated in Figure 5. This figure illustrates the typical nitrogen containing fragments found at the surface of the loaded microparticles. The examples shown are from lysozyme (Fig. 5a) and ovalbumin (Fig. 5b) loaded microparticles. The presence of a given amino acid for a particular protein did show variability for the different surfactant levels (Tab. 5), likely due to strong suppressive and associative effects of the chemical species present, and complex rearrangements and reactions that could occur in the ionization process. The presence of the amino acid fragments is strong evidence for the contribution of larger peptides and proteins to the TOF-SIMS spectrum, but the spectrum itself is a DOI 10.1002/jps
product of very complicated processes, and is too complex to give specific information about the identity of the protein and the loading levels on the particle surface. We believe that the mass spectral peaks did qualitatively establish the presence of proteins adsorbed on the surface layer that were more easily quantifiable by other means such as XPS.
CONCLUSION Proteins can be efficiently adsorbed on anionic PLG/DSS microparticle surfaces through controlling conditions such as solution pH and ionic strength. The surface of PLG microparticles has been investigated for the presence of surfactant (DSS) and protein by XPS and TOF-SIMS. In all instances, distinct mass spectral peaks specific for DSS and various amino acids were detected at the surface of the PLG microparticles. The XPS spectra allowed quantification of the DSS (through elemental sulfur) and protein (through nitrogen) and showed variation due to amount of surfactant present (more surfactant generally lowering adsorbed protein levels), the molecular nature of the protein (electrostatic adsorption strength, glycosylation), and various species coadsorbed such as counter ions and excipients. JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 97, NO. 4, APRIL 2008
1452
CHESKO ET AL.
Generally it was shown that protein concentration at the surface decreased with increasing DSS surfactant concentration. XPS and TOF-SIMS can identify and quantify the levels of protein and various other molecules such as surfactants, counter ions, and excipient species adsorbed to the PLG microparticle surface, allowing more detailed characterization of the structure, stability, and interactions of components of the functionalized layer of drug delivery or vaccine systems.
ACKNOWLEDGMENTS We thank Indresh Srivastava and Elaine Kan for providing the HIV-env (gp120) protein and providing helpful discussions concerning its properties.
REFERENCES 1. Vonarbourg A, Passirani C, Saulnier P, Benoit JP. 2006. Parameters influencing the stealthiness of colloidal drug delivery systems. Biomaterials 27: 4356–4373. 2. Calvo P, Remunan-Lopez C, Vila-Jato JL, Alonso MJ. 1997. Chitosan and chitosan/ethylene oxide block copolymer nanoparticles as novel carriers for proteins and vaccines. Pharm Res 14:1431– 1436. 3. Singh M, O’Hagan DT. 2002. Recent advances in vaccine adjuvants. J Pharm Res 19:715–728. 4. Singh M, Ugozzoli M, Kazzaz J, Chesko J, Soenawan E, Mannucci D, Titta F, Contorni M, Volpini G, Del Guidice G, O’Hagan DT. 2006. A preliminary evaluation of alternative adjuvants to a.u. using a range of established and new generation vaccine antigens. Vaccine 24:1680– 1686. [Epub ahead of print, 2005 Oct 6]. 5. Nail SL, Jiang S, Chongprasert S, Knopp SA. 2004. Fundamentals of Freeze drying. Pharm Biotechnol 14:281–360. 6. Roy I, Gupta MN. 2004. Freeze-drying of proteins: Some emerging concerns. Biotechnol Appl Biochem 39:165–177. 7. Hill JJ, Shalaev EY, Zografi G. 2005. Thermodynamic and dynamic factors involved in the stability of native protein structure in amorphous solids in relation to levels of hydration. J Pharm Sci 94:1636–1667. 8. Costantino HR, Schwendeman SP, Langer R, Klibanov AM. 1998. Deterioration of lyophilized pharmaceutical proteins. Biochemistry (Mosc) 63: 357–363. JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 97, NO. 4, APRIL 2008
9. Wagner MS, McArthur SL, Shen M, Horbett TA, Castner DG. 2002. Limits of detection for time of flight secondary ion mass spectrometry (TOFSIMS) and X-ray photoelectron spectroscopy (XPS): Detection of low amounts of adsorbed protein. J Biomater Sci Polym Ed 13:407–428. 10. Xia N, Castner DG. 2003. Preserving the structure of adsorbed protein films for time-of-flight secondary ion mass spectrometry analysis. J Biomed Mater Res A 67:179–190. 11. Delgado A, Lavelle EC, Hartshorne M, Davis SS. 1999. PLG microparticles stabilized using enteric coating polymers as oral vaccine delivery systems. Vaccine 17:2927–2938. 12. Lhoest JB, Wagner MS, Tidwell CD, Sastner DG. 2001. Characterization of adsorbed protein films by time of flight secondary ion mass spectrometry. J Biomed Mater Res 57:432–440. 13. Wang H, Castner DG, Ratner BD, Jiang S. 2004. Probing the orientation of surfaceimmobilized immunoglobulin G by time-of-flight secondary ion mass spectrometry. Langmuir 20: 1877–1887. 14. Chesko J, Kazzaz J, Ugozzoli M, O’Hagan DT, Singh M. 2005. An investigation into the factors controlling the adsorption of protein antigens to anionic PLG microparticles. J Pharm Sci 94: 2510–2519. 15. Singh M, Chesko J, Kazzaz J, Ugozzoli M, Kan E, Srivastava I, O’Hagan DT. 2004. Adsorption of a novel recombinant glycoprotein from HIV (Env gp120dV2 SF162) to anionic PLG microparticles retains the structural integrity of the protein, whereas encapsulation in PLG does not. Pharm Res 21:2148–2152. 16. Singh M, Kazzaz J, Ugozzoli M, Malyala P, Chesko J, O’Hagan DT. 2006. Polylactide-coglycolide microparticles with surface adsorbed antigens as vaccine delivery systems. Curr Drug Deliv 3:115–120. 17. Pizza M, Scarlato V, Masignani V, Guiliani M, Arico B, Comanducci M, Jenning GT, Baldi L, Bartolini E, Capecchi B, Galeotti CL, Luzzi E, Manetti R, Marchetti E, Mora M, Nuti S, Ratti G, Santini L, Savino S, Scarselli M, Storni E, Zuo P, Broeger M, Hundt E, Knapp B, Blair E, Mason T, Tettilin H, Hood DW, Jeffries AC, Saunders NJ, Granoff DM, Venter JC, Moxon ER, Grandi G, Rappouli R. 2000. Identification of vaccine candidates against serotype B Meningococcus by whole genome sequencing. Science 287: 1816–1820. 18. Singh M, Kazzaz J, Chesko J, Soenawan E, Ugozzoli M, Giulani M, Pizza M, Rappouli R, O’Hagan DT. 2004. Anionic microparticles are a potent delivery system for recombinant antigens from Neisseria meningitides serotype B. J Pharm Sci 93: 273–282.
DOI 10.1002/jps
CHARACTERIZATION OF ANTIGENS
19. Wagner MS, Horbett TA, Castner DG. 2003. Characterizing multicomponent adsorbed protein films using electron spectroscopy for chemical analysis, time-of-flight secondary ion mass spectrometry, and radiolabeling: Capabilities and limitations. Biomaterials 24: 1897–1908. 20. Wagner MS, Shen M, Horbett TA, Castner DG. 2003. Quantitative analysis of binary adsorbed protein films by time of flight secondary ion mass spectrometry. J Biomed Mater Res A 64911: 1–11. 21. Sanni OD, Wagner MS, Briggs D, Castner DG, Vickerman JC. 2002. Classification of adsorbed protein static TOF-SIMS spectra by principle com-
DOI 10.1002/jps
1453
ponents analysis and neural networks. Surf Interface Anal 33:715–728. 22. Briggs D, Grant JT. editors. 2003. Surface analysis by Auger and x-ray photoelectron spectroscopy. Chichester: IM Publications; Manchester: Surface Spectra. 23. Kim J, Shon HK, Jung D, Moon DW, Han SY, Lee TG. 2005. Quantitative chemical derivatization technique in time-of-flight secondary ion mass spectrometry for surface amine groups on plasma-polymerized ethylenediamine film. Anal Chem 7:4137–4141. 24. Shard AG, Clarke S, Davies M. 2002. Static SIMS analysis of random poly(lactide-co-glycolic) acid. Surf Interface Anal 33:528–532.
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 97, NO. 4, APRIL 2008
Novartis Vaccines and Diagnostics Division, Emeryville, California
2
Molecular Profiles, Nottingham, UK
Received 3 October 2006; revised 16 April 2007; accepted 17 April 2007 Published online in Wiley InterScience (www.interscience.wiley.com). DOI 10.1002/jps.21040
ABSTRACT: The chemical composition of the surface of anionic PLG microparticles before and after adsorption of vaccine antigens was measured using X-ray photoelectron spectroscopy (XPS) and time-of-flight secondary ion mass spectrometry (TOF-SIMS). The interfacial distributions of components will reflect underlying interactions that govern properties such as adsorption, release, and stability of proteins in microparticle vaccine delivery systems. Poly(lactide-co-glycolide) microparticles were prepared by a w/o/w emulsification method in the presence of the anionic surfactant dioctyl sodium sulfosuccinate (DSS). Ovalbumin, lysozyme, a recombinant HIV envelope glyocoprotein and a Neisseria meningitidis B protein were adsorbed to the PLG microparticles, with XPS and time-of-flight secondary mass used to analyze elemental and molecular distributions of components of the surface of lyophilized products. Protein (antigen) binding to PLG microparticles was measured directly by distinct elemental and molecular spectroscopic signatures consistent with amino acids and excipient species. The surface sensitive composition of proteins also included counter ions that support the importance of electrostatic interactions being crucial in the mechanism of adsorptions. The protein binding capacity was consistent with the available surface area and the interpretation of previous electron and atomic force microscope images strengthened by the quantification possible by XPS and the qualitative identification possible with TOF-SIMS. Protein antigens were detected and quantified on the surface of anionic PLG microparticles with varying degrees of efficiency under different adsorption conditions such as surfactant level, pH, and ionic strength. Observable changes in elemental and molecular composition suggest an efficient electrostatic interaction creating a composite surface layer that mediates antigen binding and release. ß 2007 Wiley-Liss, Inc. and the American Pharmacists Association J Pharm Sci 97:1443–1453, 2008
Keywords: vaccine; anionic microparticle; antigen; XPS; TOF-SIMS; delivery system; PLG
Correspondence to: James Chesko (Vaccines and Diagnostics Division, Novartis Corporation, 4560 Horton St., Emeryville, CA 94608. Telephone: 510-923-3896; Fax: 510-923-2586; E-mail: [email protected]) Journal of Pharmaceutical Sciences, Vol. 97, 1443–1453 (2008) ß 2007 Wiley-Liss, Inc. and the American Pharmacists Association
INTRODUCTION The role of surface interactions is crucial to formulations with high surface areas such as microparticles and nanoparticles that interact
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 97, NO. 4, APRIL 2008
1443
1444
CHESKO ET AL.
with the immune system.1 The surface properties that modulate adsorption and release of biomolecules also may stabilize or denature the protein or biomolecule in question.2 Tomorrow’s generation of new vaccines with recombinant antigens will often require an optimized delivery system that carefully controls changes in protein molecular structure and disposition, factors that may change antigenic determinants of immunogenicity during uptake, processing, and presentation, thus affecting the ultimate potency of the vaccine.3,4 The interactions at the particle interface will ultimately determine the chemical composition localized at the surface. As processes such as lyophilization occur, the surface will undergo a dehydration process as an assortment of tightly bound molecules, ions, counter ions, surfactants, and adsorbed species consolidate on a solvent-depleted interface.5,6 The stability of the protein in this environment will depend upon the microenvironment and the various factors that influence its formation, evolution, and structure.7,8 A direct method to characterize the surface composition in the presence of bulk material is to utilize microspectral methods such as X-ray photoelectron spectroscopy (XPS)9 and time-offlight secondary ion mass spectrometry (TOFSIMS).10,11 These techniques are exquisitely sensitive to the outermost 5–10 nm even of a large 1000 nm particle, that is, the outer 1% of the total material.12 This measurement will provide specific information about the outermost surface of the particles, the dehydrated Stern layer, and the interfacial structure of the vaccine delivery system that is probed at the molecular level by receptors on the surface of antigen presenting cells of the immune system. Significant insights into the chemical composition and possible molecular players in the process of stabilizing or binding the protein (e.g., surfactants) may be gleaned from examination of this data.13 Previous research suggesting a surface monolayer of adsorbed protein and a large, Coulombic driving force accompanying the surface interactions14 may be tested by such results. Understanding the role of the various excipients, buffers, sugars, and salts in the system and during the lyophilization process may be possible following careful analysis. Through this understanding the formulation of vaccines with desirable adsorption, stability, and release properties for antigens may be possible.15,16 JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 97, NO. 4, APRIL 2008
MATERIALS AND METHODS Materials The random copolymer RG503, poly(D,L-lactideco-glycolide) 50:50 lactide to glycolide ratio (intrinsic viscosity 0.4, from manufacturers specifications) was obtained (Boehringer Ingelheim, Petersburg, VA). The surfactant dioctylsulfosuccinate (DSS), the proteins lysozyme, and ovalbumin came from Sigma-Aldrich Chemical (St. Louis, MO). The envelope protein gp120dV2, a CHO derived recombinant product, was expressed and purified in house (Chiron Vaccine Research in Emeryville, CA.) Escherichia coli (E. Coli)-derived recombinant meningococcal B protein MB1 was expressed in house (Novartis Vaccines, Siena, IRIS, Via Fiorentina 1, 53100 Siena, Italy) following isolation and purification as previously described.17 Phosphate buffered saline (pH 7) was made in house with USP grade 0.8% w/w NaCl, 0.02% KCl, 0.134% potassium dihydrogen phosphate, 0.002% w/w potassium dihydrogen phosphate. The citrate (pH 5) buffer was also made in house with USP grade 2.0% sodium citrate and 0.67% citric acid in autoclaved water for injection.
Methods Preparation of Microparticles Microparticles were prepared by a solvent evaporation technique. Briefly, microparticles were prepared by homogenizing 10 mL of 6% w/v polymer solution in methylene chloride (Reagent Grade ACS), with 2.5 mL phosphate buffered saline (1 PBS pH 7) or 10 mM citrate pH 5 buffer using a 10-mm probe (Ultra-Turrax T25 IKALabortechnik, Staufen, Germany) thus forming a water in oil emulsion which was then added to 50 mL of distilled water containing either 6 or 60 mg/mL DSS and homogenized at very high speed using a homogenizer with a 20-mm probe (ES-15 Omni International, Marietta, GA) for 25 min in an ice bath. This resulted in water in oil in water emulsion, which was stirred at 1000 rpm for 12 h at room temperature, and the methylene chloride was allowed to evaporate. The resulting microparticles had total surfactant levels of either 0.05% or 0.5% DSS wt/wt. The microparticle preparation was then freeze-dried (lyophilized.) DOI 10.1002/jps
CHARACTERIZATION OF ANTIGENS
X-ray Photoelectron Spectroscopy XPS measurements were taken on an Axis-Ultra instrument from Kratos Analytical (Manchester, UK) using monochromated Al Ka radiation (1486.6 eV) operated at 15 mA emission current and 10 kV anode potential (150 W). A low energy electron flood gun was used to compensate for insulator charging. Survey scans, from which quantification of the detected elements was obtained, were acquired with analyzer pass energy of 160 eV and a 1 eV step size. Highresolution scans of all detected elements were also acquired with a pass energy of 40 eV and a 0.1 eV step size. The area examined was approximately 700 mm 300 mm for the survey scans and a 110 mm diameter spot for the high-resolution scans. Time-of-flight Secondary Ion Mass Spectrometry TOF-SIMS was performed on a TOF-SIMS IV (IonTOF, Munster, Germany) using a 10 kV cesium ion source. Positive and negative ion spectra were acquired over a 500 500 mm area. In all experiments the primary ion dose density was less than 1012 ions per cm2 to ensure experiments were carried out under static conditions. To overcome sample charging a low energy electron flood gun (0.1 mA DC) was used. In each instance the lyophilized powder samples were mounted on a stub without additional modification and introduced into the vacuum chambers of appropriate instrument. Signals were averaged from several locations (typically N ¼ 3) of the sample to provide a representative average of quantities reported, with errors expressed as the standard deviation of the mean.
1445
the quantifiable range of 1.0–100 ug/mL. An alternate spectrophotometric method involving the enhanced partitioning of acridine orange into a toluene phase from water was also employed in attempts to lower the sensitivity limit, but gave basically identical results due to sensitivity to interferences at the lowest DSS concentrations (<1 ug/mL). Adsorption of Protein to Microparticles To prepare microparticles with adsorbed protein, a suspension containing 100 mg of PLG was incubated with 1 mg protein in 10 mL total volume of 10 mM of the appropriate buffer and left on a lab rocker (aliquot mixer, Miles Laboratories, Elkhart, IN) at 48C overnight. The suspension was then lyophilized. At the 1% loading level of protein to PLG weight ratio the particle surface is only covered at 0.5 monolayers and excess (free) protein and surfactant was only present in trace quantities to satisfy equilibrium partitioning.14 Earlier experiments that involved washing PLG microparticles with water to remove excess protein and surfactant showed no significant change in the amount of protein surface adsorbed, so the samples analyzed here are compositionally very similar to particles in the formulated dispersion, with excess (unbound or loosely bound) water removed during lyophilization. Some reorganization due to this solvent depletion was expected to occur, but we did not believe that surface species would migrate outside the 5–10 nm surface thickness composition probed by XPS and MS-TOF.
Microparticle Characterization
Surfactant (DSS) Adsorption to Microparticles
X-ray Photoelectron Spectroscopy
The fraction of DSS associated with the microparticles was measured as described earlier.18 Briefly, centrifugal separation of the solid (PLG) phase from the supernatant (aqueous) phase was followed by assaying both fractions by ion exchange chromatography with electrical conductivity detection across a proton exchange membrane (Waters 2695 Alliance chromatography system with Dionex AS-1 column.) The solid (PLG) phase sample was hydrolyzed in 1N NaOH to dissolve the matrix and then pH neutralized before injection. This analytical method was sensitive (LLOD) to 0.5–1 ppm (1 ug/mL) DSS, coefficient of determination typically 0.995 over
XPS was applied to quantitatively determine the surface elemental composition of the lyophilized powder samples. The technique uses X-ray irradiation to cause the emission of photoelectrons. These photoelectrons are emitted with specific binding energies that relate to the electron core level from which they were ejected and the peak intensity is proportional to the atomic concentration within the sampled volume. For example, carbon and oxygen abundance are distinguishable through the C(1s) and O(1s) photoelectrons near 280 and 530 eV, respectively. With the use of controls and standards the ratios of the observed elements and the binding energies
DOI 10.1002/jps
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 97, NO. 4, APRIL 2008
1446
CHESKO ET AL.
can be related to the molecular composition of a sample. The levels of detection in an XPS experiment are typically 0.1atomic % (1 atom in 1000). Time-of-flight Mass Spectrometry TOF-SIMS is a complementary technique to XPS providing molecular speciation from the surface under analysis via mass spectrometry.19 In SIMS, the sample, in ultra high vacuum (UHV), is bombarded with primary ions with energies in the range 2–25 keV. Energy transfer to the solid via a collision cascade results in ejection of material from the surface (sputtering). Approximately 95% of this material comes from the uppermost two monolayers (1 nm), hence SIMS is highly surfacespecific. Sputtered particles range in size from single atoms to large assemblies of atoms, including intact molecules. Typically, about 1% of emitted ions carries a single positive or negative charge, and therefore can be collected and analyzed by their mass to charge ratio. The peaks observed in the mass spectra were analyzed looking for fragments previously identified from common amino acids.20,21
RESULTS AND DISCUSSION XPS Elemental Quantification of Surface Species The XPS elemental quantifications from controls and lyophilized micro-particles are shown in Table 1. These data have been extracted from the integrated intensities of the main photoelectron lines observed in the survey spectra such as that shown in Figure 1. This survey spectrum is recorded from the lyophilized PLG particles alone and highlights that the only elements detected
Figure 1. A typical XPS wide scan from which elements are quantified. Wide scans were recorded from each sample that allow the atomic concentrations for each element present within the uppermost 5–10 nm to be determined.
here were carbon, oxygen with low levels of sodium. The oxygen:carbon ratio (O:C) was 0.7 and relates closely to the expected value from a pure PLG surface of 0.8 (from raw starting material). The excess carbon is not unexpected and probably relates to so called ‘‘adventitious’’ hydrocarbon. This represents low levels of CH found on the majority of polmeric samples analyzed by XPS.22 To verify the amount of DSS available to the PLG surface, ion exchange chromatography was performed on representative samples. Since the 0.05% DSS w/w PLG represented an average DSS concentration of 6 ug/mL, the DSS partitioning for this system was measurable to >90% on the PLG particles (5–6 ug DSS found in 1 mL of the PLG solid phase, with no more than 1 ug maximum DSS in the supernatant aqueous phase). For the
Table 1. Elemental Composition of Surface Layer for PLG Particles with Different Buffering Conditions and Levels of DSS Surfactant Element C O N Na S Cl P Si
PLG 60.1 39.7 0.0 0.1 0.0 0.0 0.0 0.0
(0.2) (0.1) (0.0) (0.1) (0.1) (0.0) (0.0) (0.0)
PLG þ 0.05% DSS 60.3 38.8 0.0 0.5 0.0 0.1 0.0 0.0
(0.1) (0.2) (0.0 (0.1) (0.0) (0.2 (0.0 (0.0)
PLG þ 0.5% DSS 60.8 38.3 0.0 0.5 0.1 0.0 0.0 0.0
(0.1) (0.2) (0.0) (0.3) (0.1) (0.0) (0.0) (0.0)
PLG þ 0.05% DSS (pH 5) 59.5 39.5 0.0 0.5 0.0 0.2 0.0 0.0
(0.1) (0.2) (0.0) (0.4) (0.0) (0.2) (0.0) (0.0)
PLG þ 0.5% DSS (pH 5) 62.7 36.3 0.0 1.2 0.2 0.0 0.0 0.0
(0.5) (0.3) (0.0) (0.1) (0.1) (0.0) (0.0) (0.0)
Data are the average of sampling from three areas for each preparation and are from integrated XPS band intensities, with standard deviations shown in parentheses. JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 97, NO. 4, APRIL 2008
DOI 10.1002/jps
1447
CHARACTERIZATION OF ANTIGENS
Table 2. XPS Elemental Composition of PLG and Protein Standards, Averaged from Three Distinct Sampling Regions
Element C O N Na S Cl P Si
PLG þ 0.05% DSS (pH 7) 55.6 40.6 0.0 2.7 0.0 0.8 0.7 0.0
(0.5) (0.5) (0.0) (0.2) (0.0) (0.1) (0.1) (0.0)
PLG þ 0.5% DSS (pH 7) Surfactant 58.8 37.4 0.0 2.1 0.2 0.4 0.8 0.0
(0.4) (1.1) (0.0) (0.5) (.01) (0.2) (0.1) (0.0)
Ovalbumin 66.4 17.4 15.5 0.3 0.1 0.2 0.5 0.0
(0.5) (0.1) (0.5) (0.1) (0.1) (0.1) (0.2) (0.1)
Lysozyme 65.3 15.6 17.0 0.0 0.4 0.0 0.0 0.0
(1.0) (0.6) (0.8) (0.0) (0.2) (0.0) (0.0) (0.0)
gp120 63.0 26.7 8.6 0.6 0.0 1.7 0.0 0.0
(0.5) (0.6) (0.5) (0.2) (0.0) (0.3) (0.0) (0.0)
MB1 61.4 20.2 12.5 3.0 0.0 2.4 0.0 0.6
(0.7) (0.4) (0.6) (0.2) (0.0) (0.3) (0.0) (0.3)
Data are the average of sampling from three areas for each preparation and are from integrated XPS band intensities, with standard deviations shown in parentheses.
0.5% DSS w/w PLG formulation, the DSS concentration typically saturated at 0.15– 0.20% w/w PLG, with mass balance measured to within 10% in the aqueous supernatant phase. The 0.5% DSS w/w PLG represents a microparticle with surface saturation of DSS (a limiting case to compare the intermediate 0.05% DSS concentration with). The XPS surface concentrations of sulfur show a clear increase with higher DSS loading (Tabs. 1 and 2), in addition to shifts in the carbon (upwards) and oxygen (downwards), all consistent with a higher surface presence of the molecular dioctylsulfosuccinate which has a higher C:O ratio than PLG. In samples that have PLG with the DSS surfactant (different pH and DSS concentrations) the oxygen to carbon ratios remain relatively constant. Sodium cations are present to form the sodium salt of the carboxyl terminal group of the lactic and glycolic acid residues. Anions present include chloride (typically present together with sodium, but in less than a 1:1 stoichoimetric ratio) sulfur from dioctylsulfosuccinate (DSS) in the case of high (0.5%) surfactant concentration, but below limits of detection for the low (0.05% DSS) loading. The presence of counter ions, especially cationic species such as Naþ or Kþ, are ubiquitous to the various samples and underscore the importance of charge balance and electrostatic interactions in the formation and structure of the microparticle surface. For example, the extent of dissociation from PLG ionizable groups (primarily lactide and glycolide carboxyls, with some hydroxyls) at pH 7.0 is greater than at pH 5.0, resulting in a greater charge density at the surface. This is reflected directly in a 20 mV negative shift in the surface zeta potential,14 and supports a more extensive DOI 10.1002/jps
layer of counter ions (and the possible formation of a weak dielectric stack structure). The most mobile ionic species with significant charge density include the group I alkali (Naþ, Kþ) and their counter ions group VII halogen (Cl) with some PO3 4 (or citrate anion) from the buffer. When these salt and buffer species are incorporated to a greater extent into the interfacial layer, their prevalence increases (from a combined Na þ P þ S þ Cl of 0.7 and 1.4% at pH 5 to 4.2 and 3.5% at pH 7) and the relative proportion of carbon and oxygen decreases. From the molecular formulas of DSS (C20H38O7SNa), the relative proportion on the surface increases from the S content (sulfosuccinate) from the 0.05% to 0.5% loading levels, indicating a surface enhanced adsorption process that would be expected for a surfactant species. The most pronounced case of DSS adsorption appears to be the 0.5% DSS pH 5 sample, where increased adsorption of DSS (dioctyl sulfosuccinate, sodium salt) to the PLG at the saturating surfactant concentration will sequester necessary counter cations (Naþ) to balance charge in the lyophilized powder. The relatively high sulfur content for this formulation (0.2%) brings a concomitant high amount of Naþ (1.2%) for charge balance. The surface composition of pH 7 buffered PLG and the powdered protein standards show some notable differences from the previous samples, as shown in Table 2. The surface composition of C and O has dropped slightly, and a richer diversity of ionic species (group I alkali metals, group V–VII nonmetal elements) is observed. The greater negative charge on the surface of the microparticles (from zeta potential measurements14) promotes the adsorption of ionic species of opposite JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 97, NO. 4, APRIL 2008
1448
CHESKO ET AL.
charge (sodium levels two to three times higher than the previous samples). Along with the sodium cations there are phosphorous (likely as hydrogen phosphate), chlorine (as chloride), and sulfur (as dioctylsulfosuccinate). Cationic proteins (e.g., basic proteins with high isoelectric points such as lysozyme) have less cationic sodium present than low isoelectric point (acidic) proteins such as MB1. Since the poly(lactide-co-glycolide) has significant proportions of carboxylic acid end groups that are ionized at physiological pH values, these provide a source of anionic species that can form salt bridges with the basic proteins, reducing the amount of other anions sequestered from solution to balance charge at the interface. The amount of cations, with Naþ the primary species considered, is expected to charge balance the total surface composition. By electrostatic considerations, the more acidic ovalbumin with lower pI will be less efficiently adsorbed on the PLG surface, as the nitrogen content values show very clearly. As a result, the counterbalancing concentration of cations (Naþ, Kþ) will also be less, as the data also reflect. The adsorption of efficiency of gp120 is much higher than MB1 and ovalbumin, so the cations required to form the charge neutral sodium salt will also be higher. The level of nitrogen present in the protein sample is indicative of posttranslational modifications such as glycosylation as well as proportion of basic (nitrogen containing) residues. For example, lysozyme is the most basic protein and has the highest nitrogen content at 17.8%. Contrast this value to that of 15.1% for ovalbumin, which has very little glycosylation of phosphorylation, the difference of 2.7% is attributable to the higher proportion of N containing residues. MB1 is a
more acidic protein and is also slightly glycosylated, having a nitrogen percentage of 12.0%. The envelope protein gp120 is less acidic but heavily glycosylated (approximately 50% sugar to total weight) and thus its nitrogen content of 8.2% is about half of lysozyme or albumin. Since previous data14 show the efficient adsorption of gp120 on PLG, the large difference in molecular composition due to extensive glycosylation is strongly reflected in the nitrogen content of the surface adsorbed species. When the proteins are adsorbed on to the PLG microparticles and then lyophilized, there is a noticeable variance in the outermost elemental composition. The presence of nitrogen is a unique contribution from the protein, and indicates the degree of incorporation of the protein into the interfacial region of the microparticles (see Tab. 3). The relative amount of nitrogen should scale with the concentration of protein adsorbed, showing the strongest adsorbing species (lysozyme, gp120, MB1) are more efficiently taken up than the weak adsorbers (albumin) as shown in Table 4. The differences seen in the strong adsorbers lysozyme versus gp120 are interpretable as the limiting cases of monolayer saturation in the case of the basic, monomeric lysozyme as compared to a propensity to form associated complexes and perhaps adsorb into a second layer for the more acidic, heavily glycosylated dimeric/ trimeric gp120. The MB1 protein falls into the intermediate range of these two extremes. The amount of chloride coadsorbed generally follows the extent of charge needed for electrostatic balance. For acidic proteins such as ovalbumin and MB1 the chloride levels are very low, whereas for the more basic gp120 and lysozyme it is higher.
Table 3. XPS Elemental Quantification of PLG Microparticles Containing the Protein and Either 0.05% or 0.5% DSS Surfactant
Element C O N Na S Cl P Si
gp120 0.05% DSS 53.7 35.2 4.1 3.4 0.0 1.9 0.0 0.0
(0.5) (0.6) (0.2) (0.4) (0.0) (1.1) (0.0) (0.0)
gp120 0.5% DSS 57.0 36.5 1.4 2.7 0.2 0.6 0.7 0.0
(0.8) (0.8) (0.3) (0.2) (0.0) (0.3) (0.2) (0.0)
MB1 0.05% DSS 60.6 32.8 4.1 2.5 0.0 0.4 0.0 0.0
(0.5) (0.5) (0.2) (0.1) (0.0) (0.1) (0.0) (0.0)
MB1 0.5% DSS 60.4 35.1 2.5 1.5 0.2 0.2 0.0 0.0
(0.7) (0.8) (0.3) (0.1) (0.1) (0.1) (0.0) (0.0)
Lysozyme 0.05% DSS 54.4 33.1 4.3 5.1 0.0 1.3 1.3 0.0
(0.6) (0.7) (0.1) (0.2) (0.3) (0.1) (0.0) (0.0)
Lysozyme 0.5% DSS
Ovalbumin 0.05% DSS
58.1 36.7 2.4 2.3 0.0 0.0 0.2 0.0
59.27 38.1 0.82 1.42 0.00 0.00 0.17 0.00
(0.9) (0.6) (0.4) (0.2) (0.2) (0.0) (0.3) (0.0)
(0.5) (1.0) (0.1) (0.2) (0.0) (0.0) (0.2) (0.0)
Ovalbumin 0.5% DSS 60.9 37.6 0.5 1.5 0.1 0.0 0.27 0.00
(0.9) (0.5) (0.1) (0.1) (0.1) (0.0) (0.2) (0.0)
Data are the average of sampling from three areas for each preparation and are from integrated XPS band intensities, with standard deviations shown in parentheses. JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 97, NO. 4, APRIL 2008
DOI 10.1002/jps
CHARACTERIZATION OF ANTIGENS
Table 4. XPS Quantification of Protein Concentrations in the Outermost 5–10 nm Shell of PLG Microparticle Delivery System, Taken from Two 700 mm 300 mm Surface Patches on the Loaded Microparticles
Sample gp120 MB1 Lysozyme Ovalbumin
Surfactant Concentration (%)
Nitrogen (Protein) Fraction of Adsorbed 5–10 nm Layer, as Compared to Protein Reference
0.05 0.5 0.05 0.5 0.05 0.5 0.05 0.5
47.7 16.3 32.8 20.0 25.3 14.1 5.3 3.0
These values are based upon the nitrogen content, considered as a unique elemental signature of the protein molecules.
Other anions such as phosphate (detected as phosphorous) can also be present to balance the positively charged functional groups. The presence of surfactant seems to compete for surface sites, perhaps carrying or accommodating molecules and ions other than the protein, and having a net effect of causing passivation toward efficient protein adsorption in all cases measured. The surfactant is a relatively small, primarily linear molecule and a relatively minor weight fraction adsorbed on the PLG surface could
1449
effectively modify a large fraction of the surface unfavorably against adsorption of proteins. This data suggest that DSS can generally be regarded as a species competing for surface adsorption sites, so at its highest (saturating) concentration it can effectively limit or reduce the capacity of PLG to adsorb protein antigens (in the four proteins shown). This less than desirable trait is counterbalanced by several benefits (increase in particle surface area, uniformity of formulation upon resuspension, modulation of release characteristics.) As the fraction of surface sites occupied by DSS increases (i.e., S content increases), the relative amount of proteins adsorbed (and N content) decreases. The sodium ions are still present at considerable levels, but seem to be preferentially displaced by the available Kþ in the PBS buffer, perhaps since potassium is a better counter cation (larger, more polarizable, more easily ionizable).
TOF-SIMS Qualitative Measurement of Amino Acids and Peptide Fragments, Leading toward the Identification of Proteins To further identify the presence of proteins and eliminate any reasonable uncertainty concerning the interpretation of elemental markers such as nitrogen for proteins and sulfur for DSS, TOFSIMS was undertaken to look at the molecular nature of the adsorbed layer. Due to very complicated factors influencing the production,
Figure 2. A positive ion spectra for the anionic surfactant DSS. DOI 10.1002/jps
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 97, NO. 4, APRIL 2008
1450
CHESKO ET AL.
Figure 3. TOF-SIMS of PLG with no DSS added. The labeled m/z peaks correspond to combinations of the glycolide (G) and lactide (L) polymer subunits as monomers (G or L), dyads (GG, GL, LL), and higher.
stability, suppression, and transmission of various ionic complexes, quantitation of the components through ion intensity is inherently difficult without special methods.23 In the case of DSS, the standard (pure) material yielded a strong m/z ¼ 467 peak, corresponding to a DSS molecule that has picked up a Naþ to yield a net plus one charge (Fig. 2). Some lighter m/z peaks show the typical pattern of fragmentation of an alkyl chain through successive loss of methylenes. The PLG microparticles without surfactant show a peak progression based upon combination of the lactide and glycolide monomer units, with the various permutations of glycolide (G) and lactide (L) expected in the dyads (GG, GL, and LL) and triads of a random copolymer (Fig. 3). This
type of fragmentation is expected and has been reported previously.24 The TOF-SIMS spectrum of the PLG microparticles with 0.5% w/w DSS added has bands that are attributable to both the dominant component, PLG, and the surfactant DSS (Fig. 4). This is clear evidence that the surfactant is present on the surface of the lyophilized microparticles and can influence the adsorption properties of the interface. The presence of parent ion peaks for DSS with both Naþ and Kþ adduct cations suggests that electrostatics are facilitating the association of different counter ions with the surfactant and surface polymer. The microparticles were assessed for protein content based on the identification of peaks known to be specific for amino acid residues20 using the
Figure 4. TOF-SIMS spectra of PLG with DSS (0.5%) added. The mass is a superposition of the dominant component, PLG, with its associated ion pattern, and a weaker but recognizable cluster of high m/z peaks near 467, attributable to DSS molecules that have picked up an alkali ion (Naþ, Kþ). JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 97, NO. 4, APRIL 2008
DOI 10.1002/jps
CHARACTERIZATION OF ANTIGENS
1451
Table 5. Typical Amino Acids Observed from Various Adsorbed Proteins, in the Case Lysozyme (from m/z Values Taken from TOF-SIMS of PLG/DSS with Adsorbed Protein Samples) Lysozyme (0.05% DSS) Gly þ Val þ Arg þ
Ala þ Thr þ Glu
Ser þ Cys Tyr þ
Met þ His þ Phe þ
Pro Asn þ Trp þ
Lysozyme (0.5% DSS) Gly þ Val þ Arg
Ala þ Thr Glu þ
Ser þ Cys þ Tyr
Met His þ Phe þ
Pro þ Asn þ Trp þ
The microparticles were assessed for protein content based on the identification of peaks known to be specific for amino acid residues. Typical positive ion fragments for common amino acids have previously been identified and illustrated in this table.
Figure 5. Typical mass fragments detected at the surface of the loaded microparticles. (a) NH4þ fragment ubiquitous to all proteins. (b) C2H6Nþ fragment from Alanine.
high mass discrimination of the spectrometer. For the PLG/DSS with adsorbed proteins, the TOFSIMS spectra were densely patterned peaks comprised as superpositions of the abundant PLG peaks, the DSS peaks, and numerous fragment peaks attributable to small peptides and amino acids that are derived from proteins.20 A comparison of the high mass resolution spectra from the blank and loaded microparticles is illustrated in Figure 5. This figure illustrates the typical nitrogen containing fragments found at the surface of the loaded microparticles. The examples shown are from lysozyme (Fig. 5a) and ovalbumin (Fig. 5b) loaded microparticles. The presence of a given amino acid for a particular protein did show variability for the different surfactant levels (Tab. 5), likely due to strong suppressive and associative effects of the chemical species present, and complex rearrangements and reactions that could occur in the ionization process. The presence of the amino acid fragments is strong evidence for the contribution of larger peptides and proteins to the TOF-SIMS spectrum, but the spectrum itself is a DOI 10.1002/jps
product of very complicated processes, and is too complex to give specific information about the identity of the protein and the loading levels on the particle surface. We believe that the mass spectral peaks did qualitatively establish the presence of proteins adsorbed on the surface layer that were more easily quantifiable by other means such as XPS.
CONCLUSION Proteins can be efficiently adsorbed on anionic PLG/DSS microparticle surfaces through controlling conditions such as solution pH and ionic strength. The surface of PLG microparticles has been investigated for the presence of surfactant (DSS) and protein by XPS and TOF-SIMS. In all instances, distinct mass spectral peaks specific for DSS and various amino acids were detected at the surface of the PLG microparticles. The XPS spectra allowed quantification of the DSS (through elemental sulfur) and protein (through nitrogen) and showed variation due to amount of surfactant present (more surfactant generally lowering adsorbed protein levels), the molecular nature of the protein (electrostatic adsorption strength, glycosylation), and various species coadsorbed such as counter ions and excipients. JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 97, NO. 4, APRIL 2008
1452
CHESKO ET AL.
Generally it was shown that protein concentration at the surface decreased with increasing DSS surfactant concentration. XPS and TOF-SIMS can identify and quantify the levels of protein and various other molecules such as surfactants, counter ions, and excipient species adsorbed to the PLG microparticle surface, allowing more detailed characterization of the structure, stability, and interactions of components of the functionalized layer of drug delivery or vaccine systems.
ACKNOWLEDGMENTS We thank Indresh Srivastava and Elaine Kan for providing the HIV-env (gp120) protein and providing helpful discussions concerning its properties.
REFERENCES 1. Vonarbourg A, Passirani C, Saulnier P, Benoit JP. 2006. Parameters influencing the stealthiness of colloidal drug delivery systems. Biomaterials 27: 4356–4373. 2. Calvo P, Remunan-Lopez C, Vila-Jato JL, Alonso MJ. 1997. Chitosan and chitosan/ethylene oxide block copolymer nanoparticles as novel carriers for proteins and vaccines. Pharm Res 14:1431– 1436. 3. Singh M, O’Hagan DT. 2002. Recent advances in vaccine adjuvants. J Pharm Res 19:715–728. 4. Singh M, Ugozzoli M, Kazzaz J, Chesko J, Soenawan E, Mannucci D, Titta F, Contorni M, Volpini G, Del Guidice G, O’Hagan DT. 2006. A preliminary evaluation of alternative adjuvants to a.u. using a range of established and new generation vaccine antigens. Vaccine 24:1680– 1686. [Epub ahead of print, 2005 Oct 6]. 5. Nail SL, Jiang S, Chongprasert S, Knopp SA. 2004. Fundamentals of Freeze drying. Pharm Biotechnol 14:281–360. 6. Roy I, Gupta MN. 2004. Freeze-drying of proteins: Some emerging concerns. Biotechnol Appl Biochem 39:165–177. 7. Hill JJ, Shalaev EY, Zografi G. 2005. Thermodynamic and dynamic factors involved in the stability of native protein structure in amorphous solids in relation to levels of hydration. J Pharm Sci 94:1636–1667. 8. Costantino HR, Schwendeman SP, Langer R, Klibanov AM. 1998. Deterioration of lyophilized pharmaceutical proteins. Biochemistry (Mosc) 63: 357–363. JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 97, NO. 4, APRIL 2008
9. Wagner MS, McArthur SL, Shen M, Horbett TA, Castner DG. 2002. Limits of detection for time of flight secondary ion mass spectrometry (TOFSIMS) and X-ray photoelectron spectroscopy (XPS): Detection of low amounts of adsorbed protein. J Biomater Sci Polym Ed 13:407–428. 10. Xia N, Castner DG. 2003. Preserving the structure of adsorbed protein films for time-of-flight secondary ion mass spectrometry analysis. J Biomed Mater Res A 67:179–190. 11. Delgado A, Lavelle EC, Hartshorne M, Davis SS. 1999. PLG microparticles stabilized using enteric coating polymers as oral vaccine delivery systems. Vaccine 17:2927–2938. 12. Lhoest JB, Wagner MS, Tidwell CD, Sastner DG. 2001. Characterization of adsorbed protein films by time of flight secondary ion mass spectrometry. J Biomed Mater Res 57:432–440. 13. Wang H, Castner DG, Ratner BD, Jiang S. 2004. Probing the orientation of surfaceimmobilized immunoglobulin G by time-of-flight secondary ion mass spectrometry. Langmuir 20: 1877–1887. 14. Chesko J, Kazzaz J, Ugozzoli M, O’Hagan DT, Singh M. 2005. An investigation into the factors controlling the adsorption of protein antigens to anionic PLG microparticles. J Pharm Sci 94: 2510–2519. 15. Singh M, Chesko J, Kazzaz J, Ugozzoli M, Kan E, Srivastava I, O’Hagan DT. 2004. Adsorption of a novel recombinant glycoprotein from HIV (Env gp120dV2 SF162) to anionic PLG microparticles retains the structural integrity of the protein, whereas encapsulation in PLG does not. Pharm Res 21:2148–2152. 16. Singh M, Kazzaz J, Ugozzoli M, Malyala P, Chesko J, O’Hagan DT. 2006. Polylactide-coglycolide microparticles with surface adsorbed antigens as vaccine delivery systems. Curr Drug Deliv 3:115–120. 17. Pizza M, Scarlato V, Masignani V, Guiliani M, Arico B, Comanducci M, Jenning GT, Baldi L, Bartolini E, Capecchi B, Galeotti CL, Luzzi E, Manetti R, Marchetti E, Mora M, Nuti S, Ratti G, Santini L, Savino S, Scarselli M, Storni E, Zuo P, Broeger M, Hundt E, Knapp B, Blair E, Mason T, Tettilin H, Hood DW, Jeffries AC, Saunders NJ, Granoff DM, Venter JC, Moxon ER, Grandi G, Rappouli R. 2000. Identification of vaccine candidates against serotype B Meningococcus by whole genome sequencing. Science 287: 1816–1820. 18. Singh M, Kazzaz J, Chesko J, Soenawan E, Ugozzoli M, Giulani M, Pizza M, Rappouli R, O’Hagan DT. 2004. Anionic microparticles are a potent delivery system for recombinant antigens from Neisseria meningitides serotype B. J Pharm Sci 93: 273–282.
DOI 10.1002/jps
CHARACTERIZATION OF ANTIGENS
19. Wagner MS, Horbett TA, Castner DG. 2003. Characterizing multicomponent adsorbed protein films using electron spectroscopy for chemical analysis, time-of-flight secondary ion mass spectrometry, and radiolabeling: Capabilities and limitations. Biomaterials 24: 1897–1908. 20. Wagner MS, Shen M, Horbett TA, Castner DG. 2003. Quantitative analysis of binary adsorbed protein films by time of flight secondary ion mass spectrometry. J Biomed Mater Res A 64911: 1–11. 21. Sanni OD, Wagner MS, Briggs D, Castner DG, Vickerman JC. 2002. Classification of adsorbed protein static TOF-SIMS spectra by principle com-
DOI 10.1002/jps
1453
ponents analysis and neural networks. Surf Interface Anal 33:715–728. 22. Briggs D, Grant JT. editors. 2003. Surface analysis by Auger and x-ray photoelectron spectroscopy. Chichester: IM Publications; Manchester: Surface Spectra. 23. Kim J, Shon HK, Jung D, Moon DW, Han SY, Lee TG. 2005. Quantitative chemical derivatization technique in time-of-flight secondary ion mass spectrometry for surface amine groups on plasma-polymerized ethylenediamine film. Anal Chem 7:4137–4141. 24. Shard AG, Clarke S, Davies M. 2002. Static SIMS analysis of random poly(lactide-co-glycolic) acid. Surf Interface Anal 33:528–532.
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 97, NO. 4, APRIL 2008