Rational Design of Liquid Formulations of Proteins.

Twenty years ago, a number of eminent pharmaceutical scientists collaborated on an article describing a rational approach to developing stable lyophilized protein formulations (Carpenter, Pikal, Chang, & Randolph, 1997). Since that time, no corresponding document for rational development of liquid formulations of proteins has appeared. Certainly, many of the principles underpinning rational protein formulation have been known for some time, but no overarching scheme has ever been described in the literature. Now the time has come to provide a framework for the rational design of protein formulations as aqueous solutions. The objective of this review is to lay out four concepts that will guide one to obtaining a stable liquid protein formulation. Additionally, the aim will be to identify factors that are intrinsic to the stabilization of any protein, not just a particular class of proteins, such as monoclonal antibodies (Uchiyama, 2014; Wang, Singh, Zeng, King, & Nema, 2007) and to provide guidelines aiming to effect stabilization. Noting that all approaches to stabilization face validation that must be performed empirically, it is hoped that the rational strategies described here will help the formulation scientist in their daily tasks and inspire continued advancement of the science involved in protein formulation.

Use of the amide II infrared band of proteins for secondary structure determination and comparability of higher order structure.

Demonstrating comparability of secondary structure composition as part of higher order structure (HOS) in therapeutic proteins is a significant challenge. Previously, we showed that the variability of second derivative amide I Fourier transform infrared (FTIR) spectra were small enough that significant differences in secondary structures could be seen for a variety of model proteins. Those comparisons used spectral overlap and spectral correlation coefficients to quantify spectral differences. However, many of the excipients used in downstream purification process, drug substance, and drug product formulation, such as free amino acids and sugars, can interfere with the absorbance in the amide I region. In this study, analysis of amide II FTIR spectra is shown as an alternative to using spectral data from the amide I region to analyze protein secondary structure to assess their HOS. This research provided spectral overlap and spectral correlation coefficient mathematical approaches for analysis of amide II FTIR spectra to demonstrate comparability of protein secondary structure. Spectral overlap and spectral correlation coefficients results show strong correlations between changes in the second derivative of amide II and amide I FTIR spectra for various model proteins under different conditions, which demonstrate the applicability of using amide II FTIR spectra for the comparability of protein secondary structure. These results indicate that the analysis of the second derivative of amide II FTIR spectra may be used to monitor and demonstrate comparability of protein secondary structure during downstream process and formulation development of protein therapeutics.

Microparticle-based lung delivery of INH decreases INH metabolism and targets alveolar macrophages.

Microparticles prepared by the precipitation with a compressed antisolvent (PCA) process were evaluated for their potential in targeting an ionizable prodrug of isoniazid (INH), isoniazid methanesulfonate (INHMS), for sustained delivery of INH to alveolar macrophages (AMs). The charged prodrug was ion-paired with two different hydrophobic cations (tetrapentylammonium (TPA)- and tetraheptylammonium (THA)-bromide), and loaded separately into the poly(l-lactide) (PLA) microparticles. The drug/polymer particles were spherical in shape and between 1 and 3 mum in diameter. The choice of hydrophobic cations did not affect drug incorporation efficiencies or the release kinetics of INH from the microparticles. Using a sensitive liquid chromatographic tandem mass spectrometric (LC-MS/MS) assay developed for INH, high level of INH was detected in NR8383, a rat AM cell line, following exposure of these cells to drug-loaded microparticles. To confirm the microparticles can target AMs in vivo, we compared the INH levels in lavaged bronchoalveolar macrophages by LC-MS/MS after the Sprague-Dawley rats were administered either INHMS in PLA microparticles by intra-tracheal instillation or INH solution by gavage or intra-tracheal instillation. As expected, only microparticles provided sustained and targeted delivery of INH to AMs. Most importantly, this method of delivery led to substantial reduction in the blood levels of acetylisoniazid (AcINH), a major and potential toxic metabolite of INH.

Heterogeneous nucleation-controlled particulate formation of recombinant human platelet-activating factor acetylhydrolase in pharmaceutical formulation.

Clinical lots of recombinant human platelet-activating factor acetylhydrolase (rhPAF-AH) were prepared in a lyophilized formulation. After reconstitution with sterile water for injection to form an aqueous solution (10 mM sodium citrate, 7.5 w/v% sucrose, and 0.1 w/v% Pluronic-F68, pH 6.5), a few visible, slowly growing particles formed consistently within hours at room temperature. To investigate the mechanism of this phenomenon, immediately after reconstitution, all protein aggregates and exogenous particles were removed by filtration. During 20 days incubation at room temperature, no visible aggregates formed in these filtered samples. In contrast, when nano-sized hydrophilic silica particles were added, they seeded rapid and extensive aggregation of rhPAF-AH. This effect was exacerbated in solutions containing a lower Pluronic-F68 concentration at 0.01%. Aggregation occurred even under conditions where rhPAF-AH adsorption was reversible, and induced no detectable changes to protein secondary and tertiary structures. Decreasing the extent (e.g., adding Pluronic-F68) or affinity (e.g., increasing solution pH) of rhPAF-AH adsorption on nano-sized silica particles was found to be effective at reducing aggregation. Accelerated aggregation was not observed when rhPAF-AH formulation was seeded with aggregated rhPAF-AH. These results show that rhPAF-AH aggregation proceeds through a heterogeneous nucleation-controlled mechanism, where exogenous particles present in solution serve as seeds on which rhPAF-AH adsorb, nucleate, and grow into large aggregates.

Mechanism for benzyl alcohol-induced aggregation of recombinant human interleukin-1 receptor antagonist in aqueous solution.

Benzyl alcohol, an antimicrobial preservative, accelerates aggregation and precipitation of recombinant human interleukin-1 receptor antagonist (rhIL-1ra) in aqueous solution. The loss of native monomer during incubation at 37 degrees C was determined by analysis of sample aliquots with size exclusion high performance liquid chromatography (SE-HPLC). Benzyl alcohol caused minor perturbation of the tertiary structure of the protein without changing its secondary structure, documenting that the preservative caused a minor shift in the protein molecular population toward partially unfolded species. Consistent with this conclusion, in the presence of benzyl alcohol the rate of H-D exchange was accelerated and the fluorescence of 1-anilinonaphthalene-8-sulfonic acid in the presence of rhIL1ra was increased. Benzyl alcohol did not alter the free energy of unfolding based on unfolding experiments in urea or guanidine HCl. With differential scanning calorimetry it was determined that benzyl alcohol reduced the apparent Tm of rhIL-1ra, but this effect occurred because the preservative lowered the temperature at which the protein aggregated during heating. Isothermal calorimetry documented that the interaction of benzyl alcohol with rhIL-1ra is relatively weak and hydrophobically driven. Thus, benzyl alcohol accelerates protein aggregation by binding to the protein and favoring an increase in the level of partially unfolded, aggregation-competent species. Sucrose partially inhibited benzyl alcohol-induced aggregation and tertiary structural change. Sucrose is preferentially excluded from the surface of the protein, favoring most compact native state species over expanded aggregation-prone forms.

Improved potency of cisplatin by hydrophobic ion pairing.

Cisplatin is one of the most widely used and effective chemotherapeutic drugs ever discovered against certain forms of cancer. However, its use is limited by toxicity. A more potent form might allow lower doses to be used and would diminish the toxicity. A new analog of cisplatin has been synthesized by stoichiometric replacement of the chloride ligands with the anionic surfactant, Aerosol OT (AOT). The new compound has a very low aqueous solubility (about 2 mg/l) and a log P value of 2.17, which is more than 4 log units higher than cisplatin itself, indicating a dramatic increase in hydrophobicity. While hydrophobic cisplatin analogs have been synthesized previously, this is the first one with readily dissociable ligands replacing the chlorides. The resultant AOT complex is able to penetrate cellular membranes more efficiently, resulting in a threefold to fivefold increase in intracellular platinum levels. These increased intracellular concentrations correlate with lower IC50 values in a number of cancer and normal cell lines. These findings suggest that further development of the AOT complex as a chemotherapeutic agent is warranted, given its marked increase in potency over the parent compound.

Quantification of glycine crystallinity by near-infrared (NIR) spectroscopy.

The object of this investigation was to use near-infrared (NIR) spectroscopy for quantification of glycine crystallinity. Glycine samples, with different degrees of crystallinity, were obtained by physically mixing different proportions of crystalline beta-glycine with amorphous glycine. NIR spectra were obtained, directly from samples in glass vials, over the wavelength range of 1100-2500 nm. A partial least squares (PLS) model was developed to correlate the NIR spectral changes with the degree of crystallinity. Using this model, a standard error of calibration (SEC) of 2.1% was obtained with an r(2) value of 0.996. Cross validation was used to test the precision of the quantitative model, resulting in a standard error of prediction (SEP) of 3.2%. These results indicate that NIR spectroscopy is well suited to the measurement of glycine crystallinity in lyophilized products. Employing the PLS model, the crystallinity of glycine in freeze-dried sucrose-glycine mixtures was evaluated. At a sucrose to glycine ratio >4, glycine crystallization during lyophilization was inhibited. Conversely, at ratios < or =0.67, glycine remained substantially crystalline. At intermediate compositions, the glycine was partially crystalline.

Degradation kinetics of high molecular weight poly(L-lactide) microspheres and release mechanism of lipid:DNA complexes.

Plasmid DNA encoding the green lantern protein was ion-paired with 1,2-dioleoyl, 3-trimethylammonium propane (DOTAP) at a (+/-) charge ratio of (1:1) to form a hydrophobic ion-pair (HIP) complex using the Bligh and Dyer method, and transferred into methylene chloride. Precipitation with a compressed antisolvent (PCA) was then employed to encapsulate plasmid DNA into poly(L-lactide) (PLLA) microspheres. The hydrophobicity of DOTAP:DNA complexes allowed consistently high encapsulation efficiencies (>70%) to be achieved. Release of the DOTAP:DNA complex from PLLA microspheres exhibited minimal burst and a short (ca. 1 week) lag phase, followed by sustained release over a 20 week period. Release kinetics were consistent with a simple Fickian diffusion model. No correlation was identified between release rate of soluble poly(L-lactide) species (< or =10 lactate units) from PLLA and the DNA release kinetics. Only approximately 12% of the polymer was degraded into soluble poly(L-lactide) over the time frame where approximately 90% of the plasmid load had been released.

Retrospective statistical analysis of lyophilized protein formulations of progenipoietin using PLS: determination of the critical parameters for long-term storage stability.

Although certain criteria have become recognized as being essential for a stable lyophilized formulation, the relative importance of different stability criteria has not been demonstrated quantitatively. This study uses multivariate statistical methods to determine the relative importance of certain formulation variables that affect long-term storage stability of a therapeutic protein. Using the projection to latent structures (PLS) method, a retrospective analysis was conducted of 18 formulations of progenipoietin (ProGP), a potential protein therapeutic agent. The relative importance of composition, pH, maintenance of protein structure (as determined by infrared (IR) spectroscopy), and thermochemical properties of the glassy state (as measured by differential scanning calorimetry (DSC)) were evaluated. Various stability endpoints were assessed and validated models constructed for each using the PLS method. Retention of parent protein and the appearance of degradation products could be adequately modeled using PLS. The models demonstrate the importance of retention of native structure in the solid state and controlling the pH. The relative importance of T(g) in affecting storage stability was low, as all of the samples had T(g) values above the highest storage temperature (40 degrees C). However, other indicators of molecular mobility in the solid state, such as change in DeltaC(p) upon annealing, appear to be important, even for storage below T(g). For the first time, the relative importance of certain properties in controlling long-term storage stability could be assessed quantitatively. In general, the most important parameters appear to be pH and retention of native structure in the solid state. However, for some stability endpoints, the composition (concentration of protein or various excipients), as well as some DSC parameters, were found to be significant in predicting long-term stability.

Infrared spectroscopic studies of protein formulations containing glycine.

Glycine is extensively used as an excipient in protein formulations. However, it absorbs significant infrared (IR) radiation in the conformationally sensitive amide I region (1700-1600 cm(-1)) of proteins. Furthermore, glycine can form a number of polymorphs, as well as an amorphous phase. Each of these forms possibly exhibits a different IR absorption spectrum. Accurate subtraction of glycine signals, in order to obtain reliable amide I spectra, was found to be possible only if the protein-to-glycine ratio was >/=1:1. In those cases, the solid-state conformation of the protein could be determined. In addition, a new method for estimating the degree of crystallinity of freeze-dried glycine is described, using IR bands in the 1350-1300 cm(-1) region.

Effects of potassium bromide disk formation on the infrared spectra of dried model proteins.

The most common method of sample preparation for Fourier transform infrared spectroscopic analysis of proteins in the solid state is compression after mixing with potassium bromide (KBr). Recently, questions have arisen as to whether proteins are conformationally altered by this process. In this study, the amide I Fourier transform infrared spectra of two model proteins, lysozyme and alpha-chymotrypsinogen, were measured before and after compression in KBr, and the effects of moisture exposure and the ratio of KBr to protein were examined. Contrary to earlier reports, compaction of the KBr/protein mixtures did not foster aggregation, and only minor apparent structural alterations were observed.

In vitro and in vivo evaluation of the effects of PLA microparticle crystallinity on cellular response.

Previous research suggests that crystallinity of poly(L-lactide) P(L)LA microparticles can influence surface free energy, which in turn might influence biocompatibility. This work studies the cellular response to P(L)LA microparticles of different crystallinity both in vitro and in vivo. Following incubation with P(L)LA microparticles, the in vitro production of reactive oxygen intermediates (ROI) was measured as a marker of cellular response. In both fluorescence and chemiluminescence experiments to measure ROI, a small effect of microparticle crystallinity on NR8383 AM response was observed. Microparticles of higher crystallinity elicited a smaller inflammatory response compared to lower crystallinity particles. Compared to the elevated inflammatory response induced by zymosan, the response to all P(L)LA microparticles tested was practically negligible. Results from in vivo experiments further supported conclusions that P(L)LA microparticles elicit minimal inflammatory response. Following acute exposure to P(L)LA microparticles in guinea-pig lungs, the inflammatory response was not significantly different from the response observed when sterile saline was administered. In contrast to the in vitro experiments, there were not apparent differences in cellular responses to microparticles of different crystallinity.

The use of fluorescence resonance energy transfer to monitor dynamic changes of lipid-DNA interactions during lipoplex formation.

Fluorescence resonance energy transfer (FRET) was used to monitor interactions between Cy3-labeled plasmid DNA and NBD-labeled cationic liposomes. FRET data show that binding of cationic liposomes to DNA occurs immediately upon mixing (within 1 min), but FRET efficiencies do not stabilize for 1-5 h. The time allowed for complex formation has effects on in vitro luciferase transfection efficiencies of DOPE-based lipoplexes; i.e., lipoplexes prepared with a 1-h incubation have much higher transfection efficiencies than samples with 1-min or 5-h incubations. The molar charge ratio of DOTAP to negatively charged phosphates in the DNA (DOTAP+/DNA-) also affected the interaction between liposomes and plasmid DNA, and interactions stabilized more rapidly at higher charge ratios. Lipoplexes formulated with DOPE were more resistant to high ionic strength than complexes formulated with cholesterol. Taken together, our data demonstrate that lipid-DNA interactions and in vitro transfection efficiencies are strongly affected by the time allowed for complex formation. This effect is especially evident in DOPE-based lipoplexes, and suggests that the time allowed for lipoplex formation is a parameter that should be carefully controlled in future studies.

Effects of sucrose on conformational equilibria and fluctuations within the native-state ensemble of proteins.

Osmolytes increase the thermodynamic conformational stability of proteins, shifting the equilibrium between native and denatured states to favor the native state. However, their effects on conformational equilibria within native-state ensembles of proteins remain controversial. We investigated the effects of sucrose, a model osmolyte, on conformational equilibria and fluctuations within the native-state ensembles of bovine pancreatic ribonuclease A and S and horse heart cytochrome c. In the presence of sucrose, the far- and near-UV circular dichroism spectra of all three native proteins were slightly altered and indicated that the sugar shifted the native-state ensemble toward species with more ordered, compact conformations, without detectable changes in secondary structural contents. Thermodynamic stability of the proteins, as measured by guanidine HCl-induced unfolding, increased in proportion to sucrose concentration. Native-state hydrogen exchange (HX) studies monitored by infrared spectroscopy showed that addition of 1 M sucrose reduced average HX rate constants at all degrees of exchange of the proteins, for which comparison could be made in the presence and absence of sucrose. Sucrose also increased the exchange-resistant core regions of the proteins. A coupling factor analysis relating the free energy of HX to the free energy of unfolding showed that sucrose had greater effects on large-scale than on small-scale fluctuations. These results indicate that the presence of sucrose shifts the conformational equilibria toward the most compact protein species within native-state ensembles, which can be explained by preferential exclusion of sucrose from the protein surface.

Congo red populates partially unfolded states of an amyloidogenic protein to enhance aggregation and amyloid fibril formation.

Congo red (CR) has been reported to inhibit or enhance amyloid fibril formation by several proteins. To gain insight into the mechanism(s) for these apparently paradoxical effects, we studied as a model amyloidogenic protein, a dimeric immunoglobulin light chain variable domain. With a range of molar ratios of CR, i.e. r = [CR]/[protein dimer], we investigated the aggregation kinetics, conformation, hydrogen-deuterium exchange, and thermal stability of the protein. In addition, we used isothermal titration calorimetry to characterize the thermodynamics of CR binding to the protein. During incubation at 37 degrees C or during thermal scanning, with CR at r = 0.3, 1.3, and 4.8, protein aggregation was greatly accelerated compared with that measured in the absence of the dye. In contrast, with CR at r = 8.8, protein unfolding was favored over aggregation. The aggregates formed with CR at r = 0 or 0.3 were typical amyloid fibrils, but mixtures of amyloid fibrils and amorphous aggregates were formed at r = 1.3 and 4.8. CR decreased the apparent thermal unfolding temperature of the protein. Furthermore, CR perturbed the tertiary structure of the protein without significantly altering its secondary structure. Consistent with this result, CR also increased the rate of hydrogen-deuterium exchange by the protein. Isothermal titration calorimetry showed that CR binding to the protein was enthalpically driven, indicating that binding was mainly the result of electrostatic interactions. Overall, these results demonstrate that at low concentrations, CR binding to the protein favors a structurally perturbed, aggregation-competent species, resulting in acceleration of fibril formation. At high CR concentration, protein unfolding is favored over aggregation, and fibril formation is inhibited. Because low concentrations of CR can promote amyloid fibril formation, the therapeutic utility of this compound or its analogs to inhibit amyloidoses is questionable.

The stability factor: importance in formulation development.

Efficient development of stable formulations of protein pharmaceuticals requires an intimate knowledge of the protein and its chemical and physical properties. In particular, understanding the mechanisms by which a protein could degrade is critical for designing and testing formulations. This review describes the major pathways by which proteins can degrade, including denaturation, aggregation, oxidation, and interfacial damage. The methods to detect the degradation are covered, along with generalized strategies to retard or prevent each type of decomposition. Without an appreciation of the current best practices for devising stable formulations, the formulation process will be neither efficient nor optimal.

Lipid unsaturation determines the interaction of AFP type I with model membranes during thermotropic phase transitions.

We have previously shown that antifreeze protein (AFP) type I from winter flounder interacts with the acyl chains of lipids in model membranes containing a mixture of dimyristoylphosphatidylcholine (DMPC) and the plant thylakoid lipid digalactosyldiacylglycerol (DGDG), most likely through hydrophobic interactions. By contrast, in studies with pure phospholipid membranes, no such interaction was seen. DGDG is a highly unsaturated lipid, which renders these studies quite different from the previous studies of AFP-membrane interaction where the lipids were saturated or trans-unsaturated. Therefore, it seemed possible that either the digalactose headgroups or the unsaturated DGDG acyl chains, or both, may be important for interactions of membranes with AFP type I. To distinguish between these possibilities, we catalytically hydrogenated the DGDG to obtain a galactolipid with completely saturated fatty acyl chains. The results with the hydrogenated DGDG were strikingly different from those obtained previously with the unsaturated DGDG; the clear binding of AFPs to the bilayer appeared to be lost. Nevertheless, the temperature-dependent folding of AFP type I was inhibited in the presence of liposomes containing either the unsaturated or the hydrogenated DGDG. The results indicate that the liposomes and protein still interact, even following hydrogenation of the acyl chains, perhaps at the membrane-solution interface.

Encapsulating DNA within biodegradable polymeric microparticles.

In order for genetic medicines to become viable commercial products, the active form of the drug (e.g., DNA) must be able to reach the site of action and remain there long enough to accomplish its intended function. Encapsulation of plasmid DNA into biodegradable microspheres is one approach towards solving this challenge. This review describes the primary methods for satisfactorily entrapping intact DNA into biodegradable polymeric matrices. In particular, the materials, processes, and equipment required for each encapsulation method are described in detail. The resulting microspheres could be used for parenteral, oral, and inhalation therapy.

Hydrophobic ion pairing of isoniazid using a prodrug approach.

Inhalation therapy for infectious lung diseases, such as tuberculosis, is currently being explored, with microspheres being used to target alveolar macrophages. One method of drug encapsulation into polymeric microspheres to form hydrophobic ion-paired (HIP) complexes, and then coprecipitate the complex and polymer using supercritical fluid methodology. For the potent antituberculosis drug, isoniazid (isonicotinic acid hydrazide, INH), to be used in this fashion, it was modified into an ionizable form suitable for HIP. The charged prodrug, sodium isoniazid methanesulfonate (Na-INHMS), was then ion paired with hydrophobic cations, such as alkyltrimethylammonium or tetraalkylammonium. The logarithms of the apparent partition coefficients (log P') of various HIP complexes of INHMS display a roughly linear relationship with the numbers of carbon atoms in the organic counterions. The water solubility of the tetraheptylammonium-INHMS complex is about 220-fold lower than that of Na-INHMS, while the solubility in dichloromethane exceeds 10 mg/mL, which is sufficient for microencapsulation of the drug into poly(lactide) microspheres. The actual logarithm of the dichloromethane/water partition coefficient (log P) for tetraheptylammonium-INHMS is 1.55, compared to a value of - 1.8 for the sodium salt of INHMS. The dissolution kinetics of the tetraheptylammonium-INHMS complex in 0.9% aqueous solutions of NaCl was also investigated. Dissolution of tetraheptylammonium-INHMS exhibited a first-order time constant of about 0.28 min(-1), followed by a slower reverse ion exchange process to form Na-INHMS. The half-life of this HIP complex is on the order of 30 min, making the enhanced transport of the drug across biological barriers possible. This work represents the first use of a prodrug approach to introduce functionality that would allow HIP complex formation for a neutral molecule.