Chemical degradation of proteins in the solid state with a focus on photochemical reactions

ADR-12712; No of Pages 12 Advanced Drug Delivery Reviews xxx (2014) xxx–xxx

Contents lists available at ScienceDirect

Advanced Drug Delivery Reviews journal homepage: www.elsevier.com/locate/addr

Chemical degradation of proteins in the solid state with a focus on photochemical reactions☆ Olivier Mozziconacci, Christian Schöneich ⁎ Department of Pharmaceutical Chemistry, 2095 Constant Avenue, University of Kansas, Lawrence, KS 66047, USA

a r t i c l e

i n f o

Available online xxxx Keywords: Protein Solid Formulation Stress Light Photochemistry Stability

a b s t r a c t Protein pharmaceuticals comprise an increasing fraction of marketed products but the limited solution stability of proteins requires considerable research effort to prepare stable formulations. An alternative is solid formulation, as proteins in the solid state are thermodynamically less susceptible to degradation. Nevertheless, within the time of storage a large panel of kinetically controlled degradation reactions can occur such as, e.g., hydrolysis reactions, the formation of diketopiperazine, condensation and aggregation reactions. These mechanisms of degradation in protein solids are relatively well covered by the literature. Considerably less is known about oxidative and photochemical reactions of solid proteins. This review will provide an overview over photolytic and non-photolytic degradation reactions, and specially emphasize mechanistic details on how solid structure may affect the interaction of protein solids with light. © 2014 Elsevier B.V. All rights reserved.

Contents 1. 2.

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . Photochemical reactions. . . . . . . . . . . . . . . . . . . . . . . 2.1. Introduction to photochemical processes in protein solids . . . . 2.1.1. Excipient and surfactants . . . . . . . . . . . . . . . 2.1.2. A mathematical model of the protein solid . . . . . . . 2.1.3. Solid-liquid interfaces . . . . . . . . . . . . . . . . 2.1.4. Solid protein on metal . . . . . . . . . . . . . . . . 2.1.5. Colloids and nanoparticles . . . . . . . . . . . . . . 2.1.6. Influence of packing material . . . . . . . . . . . . . 2.2. The mechanisms of photodegradation of proteins in the solid state 3. Non photochemical oxidation reactions . . . . . . . . . . . . . . . . 3.1. Oxidation of Met and deamination of amino-phenylalanine . . . 3.2. Condensation reactions . . . . . . . . . . . . . . . . . . . . 3.3. β-Elimination reactions . . . . . . . . . . . . . . . . . . . . 3.4. Covalent aggregation of proteins. . . . . . . . . . . . . . . . 4. Hydrolytic reactions . . . . . . . . . . . . . . . . . . . . . . . . 4.1. Deamidation. . . . . . . . . . . . . . . . . . . . . . . . . 4.2. Peptide bond cleavage . . . . . . . . . . . . . . . . . . . . 4.3. Diketopiperazine formation . . . . . . . . . . . . . . . . . . 5. Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Acknowledgment . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

☆ This review is part of the Advanced Drug Delivery Reviews theme issue on "Protein stability in drug delivery applications". ⁎ Corresponding author at: Department of Pharmaceutical Chemistry, 2095 Constant Avenue, University of Kansas, Lawrence, KS 66047. E-mail address: [email protected] (C. Schöneich).

. . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . .

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

1. Introduction In the future, pharmaceutical development will be increasingly dominated by biologics produced in organisms. Biologics developed by the biotechnology industry encompass a wide range of products

http://dx.doi.org/10.1016/j.addr.2014.11.016 0169-409X/© 2014 Elsevier B.V. All rights reserved.

Please cite this article as: O. Mozziconacci, C. Schöneich, Chemical degradation of proteins in the solid state with a focus on photochemical reactions, Adv. Drug Deliv. Rev. (2014), http://dx.doi.org/10.1016/j.addr.2014.11.016

2

O. Mozziconacci, C. Schöneich / Advanced Drug Delivery Reviews xxx (2014) xxx–xxx

including monoclonal antibodies, blood products, and vaccines [1]. Most of the biologics currently under development target the treatment of cancer, followed by cardiovascular diseases, autoimmune and hormone disorders [2]. The development cost for a biologic is estimated to $1–2 billion but this cost does not discourage pharmaceutical companies to expect profits from biologics currently in late-stage development. In the long-term, the development of biosimilars may save costs, and their production will be guided by regulatory agencies, and will depend on the degree to which cost saving measures are required by national health systems and medical insurers. Economic studies show that biologics should drive the market’s growth for the next few years [3]. The fragile nature of protein structure represents a major limiting factor for the development of protein therapeutics. The instability of proteins renders them susceptible to multiple degradation routes during manufacturing, storage, and handling. A prerequisite for the production of safe protein drugs is to avoid any chemical and physical degradation processes that may reduce potency, limit shelf life, and could increase the potential for immunogenic side effects. In a first approximation, the internal energy of a protein in a solid, in comparison to that in solution, is close to zero, making lyophilized solids products of high interest to prevent the chemical and physical degradation of proteins. The physico-chemical properties of protein solids are essentially determined by their thermal history [4]. Drying conditions (e.g. temperature, vacuum, time) affect the thermal history of the solid. Because the shell of water around a protein affects its structure, the removal of water can irreversibly change protein structure. Therefore, making the freezedrying process of proteins is a determinant step which controls the nature of the solid, and ultimately its final physico-chemical properties, that will limit the long-term stability of the final product [5]. Mechanistic studies of protein degradation in solids need to take into account the types of solids eventually produced during the drying process. Because of the importance of thermal processes during the transformation of a liquid protein formulation to its equivalent in the solid state, most stability studies of protein drugs in the solid state have essentially focused on the impact of variation of temperature, pH, and the presence of surfactants or excipients during the drying process and the subsequent storage of the products. Mechanistic studies of degradation reactions such as deamidation [6–8], the Maillard reaction [9,10], hydrolysis [6,11,12], diketopiperazine formation [13–15], and β-elimination at cysteine (Cys) [16], were extensively performed and reviewed. Costantino's thesis presents an extensive description of the most important oxidation reactions [17]. In the present review, we will cover degradation reactions in general and specifically emphasize photochemical reactions encountered by proteins in solids. For many years the photostability of proteins has not been of great concern. However, the number of photo-labile drugs, and protein drugs, is increasing, resulting in guidance from the European Pharmacopeia to protect more than 250 different drugs from light [18,19]. In this regard, the underlying photochemical processes must be clearly identified to develop better guidelines for the testing of protein photostability [20,21]. More importantly, photochemical stability studies offer a versatile toolbox to investigate the effects of formulation on protein degradation: photochemical reactions can be initiated after preparation of the respective solid, the duration of primary photochemical processes can be controlled through the duration of light exposure, and quantum yields for solid state photochemical reactions can be determined by means of an integrating sphere. 2. Photochemical reactions Most drug substances are formulated as white powders, minimizing the absorption of visible light by these formulations. However, all lamps, even incandescent ones, emit some radiation in the ultraviolet (UV) region of the spectrum. Scheidegger et al. observed that proteins in whole and skim milk underwent severe oxidative damage (e.g., formation of dityrosine, N-formylkynurenine, fragmentation) after the exposure to

fluorescent light [22]. Thus, the effect of UV-light exposure needs to be evaluated. From production to delivery, a protein is exposed to light from various light sources [19]. The photochemistry of tryptophan (Trp) [23,24], tyrosine (Tyr) [24–26], phenylalanine (Phe) [24,27], and cystine [24,28] has been well documented, predominantly for the individual amino acids but also for these amino acid residues located in peptides and proteins [29–33]. Protein conformation plays an important role in protein photo-degradation processes [19,34]. Variation of protein conformation can be achieved by modification of pH, ionic strength, and the presence of ligands, which may change sites for energy transfer(s), which ultimately generate reactive species [35,36]. For example, the structural properties of α-crystalline are modified when the protein is subjected to UV-C irradiation. The primary degradation reactions, which correspond to the oxidation of the methionine residue (Met1) and the racemization of aspartate (Asp151), contribute to the alteration of secondary structure of α-crystalline [37]. While the photo-degradation of a protein is initiated by the exposure to light, final product formation can depend on a series of complex processes such as the formation and reaction of excited states, radical species, and energy transfer [38]. The respective extents of these processes depend on the wavelengths, the intensity of light, the time of light exposure, as well as the geometry of the photo-irradiated sample (e.g. the nature of the container, the distance between the light source and the sample, and the orientation of the sample towards the flux of light). All these parameters ultimately control the dose of photoirradiation, which can be determined by actinometry [39]. However, at any given dose the processes taking place in the solid upon light exposure need to be defined. To address this issue, we will briefly introduce the photophysics of proteins in solids, and we will review the nature of photoproducts observed during light exposure of solid protein formulations, together with a mechanistic rationale for product formation. A comparison of the photochemical behavior of proteins in solution and in the solid state needs to take into account multiple variables. In solution, the temperature range for the thermal degradation of a protein is limited to the physical properties of the solvent itself, and, as we will see later, the thermal processes are essentially but not exclusively represented by hydrolysis and deamidation reactions. During photo-degradation the energy of the photons usually exceeds the thermal activation energy [39,40]. Therefore, primary photochemical reactions are rather independent of temperature. However, secondary processes may depend on temperature. The rates of photo-degradation in solution and in the solid state are different since the probability of photon absorption by solid matter is lower than in solution. The latter is essentially rationalized by the radiative nature of light and the lack of transparency of most solids, which is, in part, related to the reflection of photons at the surface of the solid. According to the Beer-Lambert law, the intensity of an electromagnetic wave inside a material (I) decreases exponentially from the surface as described in Eq. 1, where I0 is the intensity of the incident light, z the depth of the material, and κ a constant relative to the nature of the material. That is why the crystal structures of solids can be identified solely from the diffraction [41] of a wave front and light scattering [41] of incident X-ray photons, for which the penetration depth is optimal. The latter raises the question of how does penetration depth vary with the wavelength of incident light? For X-rays, in a first approximation, the penetration depth and the absorption of the radiation increase as the wavelength of incident light decreases. The latter is only true if the material has a constant conductivity over a given bandwidth. In such case the absorption of the radiation within the material is exponentially dependent of the penetration depth in terms of wavelengths. X-ray penetration increases with decreasing wavelength because the cross-section of the material increases by λ3. Now, if photons are absorbed and not refracted or diffracted by the solid, the atomic lattice starts to vibrate, leading to a net displacement of electric charges. However, it is impossible for the positive charge lattice (the nuclei) to vibrate in unison with the negative charges (the electrons). Thus, layers of charge density will appear along with the vibrations. The displacement of charges generates a local electric

Please cite this article as: O. Mozziconacci, C. Schöneich, Chemical degradation of proteins in the solid state with a focus on photochemical reactions, Adv. Drug Deliv. Rev. (2014), http://dx.doi.org/10.1016/j.addr.2014.11.016

O. Mozziconacci, C. Schöneich / Advanced Drug Delivery Reviews xxx (2014) xxx–xxx

field E. When the frequency, ω, of the vibrations equals the frequency of the incident photons, all the energy of the incident photons is absorbed, resulting in the conversion of energy of the photons into mechanical vibration. For a given atom, the time dependent wave function for a two-state system, which interacts with an incidental wave of which the frequency is ω, is described by Eq. 2, where H is the hamiltonian of the time-dependent wave function, and μ the dipole moment of the atom interacting with the wave. Knowing that the energy content of vibrations is proportional to the square of their amplitude, (μE)2, we can, therefore, estimate that the energy contained in the vibrations is proportional to the displacement of the charges, i.e the electric field E, and to the dipole of the solid. −KZ

I ¼ Io e

ð1Þ

H ¼ μE cosðwt Þ

ð2Þ

In solution, in a first approximation, the energy of the incident photons cannot be contained in the vibrational states, since there is no permanent dipole moment (μ = 0, therefore the amplitude of the wave is zero). The rigidity of a solid lattice, in comparison to a liquid, reduces the number of vibrational states. The latter ultimately decreases the spectrum of the light, which can be absorbed by the vibrational states. In solution, the number of photons absorbed by the matter and leading to the photolysis of the molecules being photo-irradiated (η) is determined by Eq. 3, where N is the number of absorbed photons and ϕ is the quantum yield of the reaction [39]. η ¼ Nϕ

ð3Þ

In a homogenous solution, to which the Beer-Lambert law can be applied, and where almost all the incident photons are absorbed, the rate of the reaction follows zero-order kinetics [42]. When only a small fraction of the incident photons is absorbed (b10%) the rate of the reaction follows first-order kinetics (Eq. 4), where C is the concentration of the absorbing molecule at a given time, and C0 is the initial concentration of the absorbing molecule. The factor k represents an apparent rate coefficient, which contains several parameters depending on the type of reaction. The term (1-α) in Eq. 4 represents the remaining fraction of the absorbing molecule, which is not degraded. During the photo-irradiation of a solid most of the photons are reflected on the surface. We can, therefore, in a first approximation, consider that Eq. 4 describes the kinetics of photo-degradation of a solid. Here, the factor k can be expressed in different ways, depending on the mathematical model used to describe the solid. An example of a mathematical model will be presented below. n¼k

C ¼ kð1−aÞ Co

ð4Þ

2.1. Introduction to photochemical processes in protein solids The fundamental law of photochemistry states that only light absorbed by a system, either in solution or in the solid state, can bring about reaction. We must, therefore, understand the mechanisms of absorption of electromagnetic radiation by solids. Although we have already introduced certain concepts above, this review does not intend to provide an exhaustive description of the photophysics of interaction between light and matter. However, some principles need to be further developed if we want to understand the mechanisms of photoproduct formation. A detailed description of the photochemical mechanisms occurring in protein solids would require a precise knowledge of the physical state of the protein in the solid. When protein formulations are dried using different methods, such as spray drying, lyophilization, or vacuum drying, the resulting solids exhibit, in general, a wide variety of physico-chemical properties,

3

characterized by a specific surface area (SSA), surface composition, and molecular mobility [43]. These properties can have significant impact on in-process and storage stability [44], and, presumably, on photo-stability. It is, for example, well known that powders produced during spray drying have a higher SSA compared to corresponding products formulated by lyophilization [43]. Because the amount of light absorbed by molecules in solids depends on the SSA and the size of the particles, the drying methods used to produce the protein solids may impact the photo-stability of the final products. 2.1.1. Excipient and surfactants A study of the effect of formulation on the stability of spray freezedried bovine serum albumin (BSA) shows that the loss of monomer of BSA increases with the SSA [45]. When β-carotene is dried using different processes such as spray-drying, drum-drying, and freeze-drying, the amount of β-carotene oxidized varies between 8% and 14%, depending on the drying process [46]. The nature of the surfactant and excipients present during the drying process should impact the stability of proteins in solids [4,47–53]. The photo-stability in the solid state of nonprotein drugs, such as amlodipine, was studied: Ragno et al. observed that when amlodipine is exposed to light (300 nm b λ b 800 nm) the photo-stability of amlodipine increases in the following order, when the drug is formulated in powder b included within liposomes b formulated in tablets b included within cyclodextrin b formulated as microspheres [54]. The photo-stability of amlodipine was also investigated by photo-irradiation with light between 280 nm and 360 nm and showed a similar stability pattern [55]. In spray-dried products, in the presence of some surfactants, more protein may be distributed near the surface of the particle. For example, a spray-dried formulation of lactate dehydrogenase (LDH) has a 10-fold higher surface content in dried trehalose than expected for a homogenous distribution [56]. The presence of a high content of a photosensitive protein such as LDH [57,58] at the surface of a solid is likely to result in the formation of a solid with higher probability to absorb light and to photo-induce protein damage. The presence of excipients can either increase or decrease the stability of proteins or have no effect, depending on the potential photosensitizing property of the molecule. An example is provided by the photo-stability study of sixty nifedipine preparations, which showed differences depending on the nature of excipients present and the dosage forms [59]. The nature of the solid, i.e. amorphous, crystalline, or crystalline polymorphs, influences the way the light is absorbed by the solid, as observed by Matsuda et al. during a photo-stability study of solid carbamazepine [60]. While carbamazepine is not a protein, the underlying principles likely apply to solid proteins. During manufacturing of protein solids heterogeneous solid cakes can be generated with different physico-chemical properties (e.g. SSA, repartition of the protein in the bulk, presence of photosensitizing excipient molecules). It, therefore, appears that the mechanistic evaluation of photodegradation processes of proteins in solids requires the construction of a model, which could be used to describe the interaction of light with matter. 2.1.2. A mathematical model of the protein solid The studies of surface photochemistry of solids provide different models to explain photochemical processes occurring with molecules adsorbed on metals and semiconductors [61–63]. For pharmaceutical purposes, Sande developed a model to study photochemical degradation reactions in the solid state [64]. This model is based on a powder bed of irradiated surface area, S, with a depth, B. The powder bed is seen as consisting of n thin layers of thickness, b, so that a single layer contains a fraction b/B of the total amount of reactant. Assuming that the radiation, Im, is uniform through each layer, the rate of reaction in layer n is given by Eq. 5 where ϕ is the quantum yield of the reaction, Nn is the number of reactant molecules photolyzed, and Ia,n is the

Please cite this article as: O. Mozziconacci, C. Schöneich, Chemical degradation of proteins in the solid state with a focus on photochemical reactions, Adv. Drug Deliv. Rev. (2014), http://dx.doi.org/10.1016/j.addr.2014.11.016

4

O. Mozziconacci, C. Schöneich / Advanced Drug Delivery Reviews xxx (2014) xxx–xxx

number of photons absorbed in the nth layer per unit of time. Ia,n is defined as I in Eq. 1, which defines the penetration of light in the solid. η ¼ Sla;n Φ η¼

ð5Þ

la;nΦ dan ¼ dt LA bC o

ð6Þ

Division of Eq. 5 by the Avogadro number (LA) and the volume of the nth layer, i.e the product of S x b, gives the rate of reaction per time unit of change of concentration of degraded material. Division of Eq. 5 by the initial concentration C0 of the reactant gives the dimensionless coefficient αn,t, which represents the contribution of the nth layer to the degradation rate at the time t. Thus, the factor 1-αn,t, represents the fraction of solid remaining. Because the layers are not separable, the fraction of solid, which is degraded in each layer has to be averaged for the whole sample at any time (Eq. 6). The Beer-Lambert law describes the absorption of photons as given in Eq. 7, where σr is the molar Naperian absorption of the reactant, C0 is the initial concentration of the reactant, and 1-αn,t is the contribution to the fractional degradation in the nth layer at the time t. h

i la;n ¼ln;t 1‐exp ‐σ r bC 0 1‐α n;t

ð7Þ

Therefore, the introduction of the Beer-Lambert term (Eq. 7) in Eq. 6 permits transformation of Eq. 6 into Eq. 8. The Naperian absorption term is defined by the IUPAC as given in Eq. 9, where A is the attenuation of light, and l is the optical pathlength. η¼‐

h

i dα n Φ ¼ l 1‐exp ‐σ r bC 0 1‐α n;t LA bC 0 n;t dt

σr ¼

A loge 10 l

ð8Þ

ð9Þ

Considering infinitesimal thin layers in Eq. 8, the term b tends to zero, and then the limit of Eq. 8 is expressed by Eq. 10. ηb→0 ¼ −

 
 I dα n n;t ¼ σ r 1‐α n;t Φ dt b→0 LA

ð10Þ

The degradation kinetics of the whole solid are then given by the integration of Eq. 10 to Eq. 11. α total;t ¼

1 n

 Z n n dα n 1X ≡ α n;t n dt 1 b→0 1

ð11Þ

Based on Eq. 10, the depth B of the power bed of the solid affects the photo-degradation kinetics by simply increasing the amount of material in each of the n layers used to establish Eq. 10. An increase of either the quantum yield (ϕ), or the incident light intensity I0, increases the rate of degradation of the solid. For biotechnology products, the interaction of light with proteins in solids is controlled in various ways. We have, so far, described a photo-degradation process of a powder resulting from a drying process of protein initially solubilized in an aqueous solution and exposed to light. However, in solution, when proteins are placed in syringes, they can form aggregates, or colloids, which can adsorb on either the metal particles released by the needle or at the surface of the plunger of the syringe [65–67]. 2.1.3. Solid-liquid interfaces Therapies with monoclonal antibodies (mAbs), which have become powerful tools in the treatment of a wide range of diseases, can require the delivery of between 100 mg and 1 g of protein per dose, necessitating the preparation of formulations as crystalline suspensions [68] and high concentration solutions [69]. The latter present great challenges

for controlling the stability of the protein since high concentration and viscosity of the solutions can accelerate protein aggregation reactions [70,71]. For subcutaneous injection, a high concentration solution of mAb is filled into a syringe, and delivered by applying a pressure on the plunger of the syringe. Under such conditions, the formation of colloidal particles is likely to increase [72]. For example, Connolly et al. have observed that proteins starts to self-organize in colloidal structures, which adsorb on gold nanoparticles, resulting in the aggregation of the dispersion of the gold particles [73]. Such metal-protein adsorbate complexes are unstable under visible light [74], suggesting that the adsorption of the protein at the surface of the metal could catalyze photodegradation. Thus, the handling of proteins in solution using syringes obliges us to also consider the problem of interaction of light with proteins adsorbed on metals. 2.1.4. Solid protein on metal To understand how proteins could interact with metals, we will consider the protein solid as a homogenous solid disc film. In this ideal solid, we will also consider that only the protein adsorbed at the surface can absorb light. Considering these approximations, we can, therefore, simplify the mathematical treatment of a photochemical reaction, which occurs on the surface of the solid. In this approximation, the description of the kinetics of the photo-degradation of the solid is given by Eq. 10 [64]. Detailed reviews on photochemical processes on surfaces were published by Zhou et al [62], Ho [63], and Franchy [61]. In brief, photo-induced processes on surfaces involve three different mechanisms, a photo-induced desorption (PID), a photo-induced dissociation (PD), and a photo-induced reaction (PR) of the molecules absorbing light. A photo-induced process is the result of multiple reaction steps where PID, PD, and PR can occur independently or sequentially. The first step is the absorption of a photon by the adsorbed protein, or by the adsorbed complex of protein and excipient. Depending on the absorption cross sections and the rate constants, a photochemical reaction is a consequence of the absorption of photons from visible and UV light by suitable chromophores of the protein, i.e. the aromatic residues, the carbonyl groups, and/or the disulfide bonds, generating an excited electronic state. The following decay of the excitation occurs via an energy transfer, which can be either the emission of radiation, or a photochemical transformation of the excited protein. In case of the relaxation of the excited state through a radiative process, the emitted radiation can be either reabsorbed by another part of the same protein or by another molecule (e.g. a protein, or a molecule of the excipient). For the protein adsorbed on the surface, it is important to distinguish the photoinduced changes, which are thermally or non-thermally activated. A photo-induced electron transfer in a protein in solution is rather diffusion-controlled than thermally activated since most of the thermal energy of the protein can be dissipated through the solvent. However, in the solid, the absence of solvent and the rigidity of the matrix should make protein solids more prone to a thermally activated electron transfer upon photo-excitation. A photo-induced electron transfer is a two steps process proceeding via i) the photo-excitation of the thermal donor state to a near excited donor state, and ii) the electron transfer from the excited donor to the acceptor. Pollak et al. have proposed that the photo-induced electron transfer reaction rate in a polyatomic molecule is controlled by the frequency of the photo-excitation [75, 76]. Therefore, a thermally activated electron transfer depends on the energy distribution of the excited donor state. Cooling of the excited donor state upon photo-excitation causes an exponential decrease in the electron transfer rate. Inversely, heating of the excited donor will increase the electron transfer rate. Since cooling and heating of the excited donor depend on the photo-excitation frequency, it is, therefore, possible to modify the electron transfer rate by simply modulating the photo-excitation frequency [75]. These physical studies are crucial to understand the photo-degradation of proteins in solids. Proteins are often lyophilized in the presence of a sugar, such as trehalose [4,47,48, 77,78]. During the lyophilization process a trehalose glass is formed

Please cite this article as: O. Mozziconacci, C. Schöneich, Chemical degradation of proteins in the solid state with a focus on photochemical reactions, Adv. Drug Deliv. Rev. (2014), http://dx.doi.org/10.1016/j.addr.2014.11.016

O. Mozziconacci, C. Schöneich / Advanced Drug Delivery Reviews xxx (2014) xxx–xxx

[79,80]. Trehalose-derived glasses are known to support long range electron transfer reactions between donors and acceptors in proteins [81]. Navati et al. pointed out that thermal and photo-induced electron transfer reactions promoted by sugar-derived glasses could be beneficial for the development of solid state thermal fuel cells [81]. From the viewpoint of delivery of stable protein solid drugs, if sugar-derived glasses are able to promote long range photo-induced electron transfer reactions and, therefore, play a role in the activation of redox reactions in protein solids, these sugar-derived glasses could also be seen as a solid matrix, which could enhance the photo-degradation of protein in the solid. The theory of photo-induced thermal electron transfer reactions described so far does not take into consideration the isotropy of the solid. To our knowledge, there is no study on the variability of photodegradation mechanisms between amorphous and crystalline protein solids. In crystalline structures, the molecules are all aligned in parallel to one another. Therefore, theoretically, the crystal should absorb electromagnetic radiation, whose electric vector is oriented parallel to the chromophores of the molecules constituting the crystal. Then, the crystal is considered as dichroic and, if the molecules present in the crystals are photolabile, the reaction is initiated. The latter suggests that crystalline proteins should be degraded only by polarized light. The absence of experimental evidence in the literature on the photo-degradation of crystalline proteins does not permit to test this hypothesis. In our study on the photo-degradation of solid insulin [82], we could not find differences in the yields of photoproducts between amorphous and crystalline insulin. However, Murray et al. have studied dichroic crystals of several diazo compounds, and noticed that the irradiation of a polycrystalline sample with polarized light resulted in the decomposition of only the crystals suitably oriented to absorb the polarized light [83]. These studies suggest that protein crystals should probably be more accurately characterized than simply distinguishing them from an amorphous phase, to understand the effect, if any, of polarized light on protein crystals. 2.1.5. Colloids and nanoparticles Photochemical mechanisms in and on solids may become increasingly important because of the development of novel delivery platforms such as nanoparticles. For example, photochemical electron transfer reactions between acceptors and donors in proteins are enhanced when proteins are grafted to nanoparticles [84]. Studies have shown that nanoparticles can allow electron transfer reactions under visible light [84]. Therefore, photochemical processes at the interface proteinparticle could lead to photo-degradation of proteins, and these processes could also possibly be influenced by other interfaces [85]. The low toxicity of lipid structures such as liposomes, micelles, nanoemulsions, and solid lipid nanoparticles (SLN) makes lipid colloidal structures the most important carriers to administrate active molecules. Light exposure of SLN changes the crystallization form, which, during artificial light exposure, results in the formation of a gel network [86]. 2.1.6. Influence of packing material Packing materials are as much important for the stability of solid formulations as for liquid formulations. The oxygen permeability and the degree of light transmittance of the packing material are the two most significant extrinsic factors influencing the photostability of the embedded drugs. So far no study has specifically addressed the influence of packing material on the photostability of solid formulations. However, we can learn from the protection of food products. Food products are generally stored in the dark to minimize the loss of vitamins and, especially, riboflavin (which is a photosensitizer), β-carotene, and vitamin C [87]. 2.2. The mechanisms of photodegradation of proteins in the solid state Aromatic residues are the strongest chromophores in proteins and the exposure to UV light generates various excited states, which can

5

enter photochemical pathways and are likely to have harmful effects on protein structure. Disulfide bridges are excellent quenchers of the excited states of aromatic residues [23,25–28]. Flash photolysis studies of tryptophan (Trp) in solution have demonstrated that the excited state of Trp relaxes through two different channels. The first channel corresponds to the release of an electron from the singlet-excited state of Trp (1Trp⁎) to the solvent or an appropriate acceptor, generating a radical cation, Trp•+, which deprotonates immediately to generate the neutral radical, Trp•. In a second channel, 1Trp⁎ relaxes through intersystem crossing, generating a triplet-excited state (3Trp⁎). In solids, the low water content makes deprotonation of Trp•+ unlikely so that the back reaction between a photo-ejected electron and Trp•+ is highly probable. Thus, in the solid matrix, the predominant way for 1Trp⁎ to relax is to first yield a triplet-excited state, which ultimately may transfer energy and/or an electron to a nearby disulfide bond. The photoexcitation of tyrosine (Tyr) can also lead to the one-electron reduction of disulfide bonds. In solution, the absorption of near UV-light by Tyr results in the formation of a singlet-excited state (1TyrOH⁎). Once photoexcited, 1TyrOH⁎ can decay non-radiatively or via intersystem crossing, generating a triplet-excited state 3TyrOH⁎. At pH b pKa(TyrOH), 3 TyrOH⁎ can absorb a second photon, allowing for the ejection of a free electron, which is either solvated (e− aq) or captured by a suitable electron acceptor. In the solid, the occurrence of a biphotonic process to excite TyrOH is likely to occur since TyrOH is most likely to be protonated. A biphotonic reaction would unlikely result in the transfer of an electron from 3TyrOH⁎ to the solvent, because of the low water content. However, the electron could either be captured by a nearby Trp•+ or a disulfide bond, which would ultimately result in the formation of a ground-state Trp or a disulfide radical anion (SS•−), respectively. In the presence of molecular oxygen, the excited-state of Tyr could be quenched by O2, which would generate superoxide radical anion (O•− 2 ). Superoxide radical anion represents another potential oxidant, albeit a weak one, which may react with the solid protein. Superoxide radicals can also be generated by the absorption of energetic photons such as UV radiation by crystalline salts. Hosono et al. have demonstrated the occurrence of superoxide radical formation in crystalline calcium alumina salt [88]. Thus, during the lyophilization of protein in the presence of salts, the nature of the salts and their solid structures are crucial to determine whether or not they may act as photosensitizers. Disulfide bonds are efficient electron acceptors in proteins and are probably the most important electron acceptors in protein solids. The initial formation of a disulfide radical anion (SS•−) results in either the formation of perthiyl or thiyl radicals, respectively [89–93]. These radicals can involve in intramolecular hydrogen atom transfer reactions with C\H bonds [94–97], leading to the formation of carbon centered radicals, which can react with molecular oxygen to form peroxyl radicals [98]. In the presence of oxygen, the photo-degradation of recombinant bovine growth hormone (rbST) in the solid state is mediated through the formation of an excited state of Trp86, which transfers an electron to the nearest disulfide bond between Cys53 and Cys164 (the distance between Trp86 and Cys53–Cys164 is 6.42 Å) [99]. This electron transfer reaction results in the formation of a disulfide radical anion, which, in the presence of oxygen, ultimately leads to the formation of disulfoxide and/or sulfenyl sulfonate, and thiosulfonates [100]. In principle, Trp86 can also transfer an electron to the disulfide bond formed between Cys181 and Cys189. However, the distance between Trp86 and Cys181–Cys189 is 19 Å, suggesting that such electron transfer reaction would occur through hydrogen bonds and/or peptide bonds [99]. Miller et al have shown that during the photodegradation of lyophilized rbST, the fraction of lost rbST is inversely proportional to the thickness of the cake, suggesting that most of the light is absorbed at the surface of the solid. In this way, the results by Miller et al. are consistent with the model of Sande [64]. The degradation of solid insulin has been intensively studied [53,101–103], however until recently little was known about the nature of the products generated after the exposure of solid insulin to light. In the absence of oxygen, the most

Please cite this article as: O. Mozziconacci, C. Schöneich, Chemical degradation of proteins in the solid state with a focus on photochemical reactions, Adv. Drug Deliv. Rev. (2014), http://dx.doi.org/10.1016/j.addr.2014.11.016

6

O. Mozziconacci, C. Schöneich / Advanced Drug Delivery Reviews xxx (2014) xxx–xxx

significant damage observed by Mozziconacci et al, is the formation of a dithiohemiacetal at the C-terminal position of insulin, involving the cysteine residues Cys20 and Cys19 from chain A and B, respectively [82]. The authors have also observed the formation of a crosslink between Tyr19 and Cys19 from chain A and B, respectively [82]. Importantly, the formation of a dithiohemiacetal generates a photoproduct, which is isobaric to the original disulfide bond. This means that proteins that show apparently no variation of their mass after being stressed could be misinterpreted as non-damaged proteins. It is also important to consider that under further exposure to light dithiohemiacetals generate thioether and a thiolate [82]. Moreover, especially upon dissolution of the solid protein, the thiolate can induce disulfide scrambling. In this section, we have often referred to the “low water content” to rationalize the most probable photochemical reactions occurring in the solids. In solution, the solvent is an important parameter to understand photochemical reactions, since the molecules of the solvent are often involved in the formation and decay of the photo-excited states of the photosensitive molecule (e.g., in the photochemistry of benzophenone [104]). In solids, the low water content can, therefore, be a limiting factor for photochemical reactions. However, we need to define what “low water content” means in the solid. Sorensen was the first to measure the water content in crystallized proteins. They found that crystallized hemoglobin contains 0.35 g of water per gram of protein [105]. X-ray and neutron diffraction studies have demonstrated later that protein molecules can strongly associate with 0.20–0.70 g of water per gram of protein [106–108]. In human growth hormone formulations prepared in sucrose, trehalose, and stachyose, the water content represents between 0.7% and 2.8% of the mass of the lyophilisates [4]. The transfer of energy of a radiative process to internal molecules of water can generate hydroxyl radicals, which can ultimately degrade the protein solid [109]. The molecules of water associated with proteins are identified as internal water or peripheral water [110]. The internal water molecules, which diminish local charge-charge interactions and stabilize the protein towards unbound proton donors and acceptors [111] exchange with bulk water in a time scale of seconds to months [112,113]. In solution, Zhao and Han [114–117] have extensively studied hydrogen bonding in the ground and excited states of biological chromophores. Intermolecular hydrogen bonds formed between carbonyl chromophores and polar protic solvents (e.g. water) are significantly strengthened in electronically excited states of carbonyl chromophores [117]. Photo-induced electron transfer, excited state proton transfer (ESPT) and other photochemical processes are influenced by excited state intermolecular hydrogen bonding interactions [115,116]. In solids, where water relaxation is much slower compared with bulk water, the role of water and its structural organization with proteins is, therefore, fundamental for the mechanistic understanding for the photo-degradation of proteins in solids. For example, a study of 1-naphtol in water showed that excited state proton transfers (ESPT) are largely influenced by the nature and cluster size of proton acceptors [118]. To our knowledge, a similar study with proteins has not yet been published. 3. Non photochemical oxidation reactions Oxidation reactions of proteins frequently target methionine (Met), histidine (His), Trp, Cys, Phe, and Tyr residues. At neutral pH, the imidazole group of His can be oxidized to hydroxy-imidazole, the sulfhydryl group of Cys to sulfenic (RSOH), sulfinic (RSO2H), and sulfonic (RSO3H) acid, the indole group of Trp to several regio-isomers of hydroxy-indole, kynurenine, and N-formyl-kynurenine, the thioether group of Met to methionine sulfoxide (MetSO), Tyr to 3,4-dihydroxyphenylalanine (DOPA), and Phe to ortho-, meta- or para-hydroxyphenylalanine and DOPA [119]. The rates of formation of these oxidation products depend, to a large extent, on the accessibility of the target amino acid residues and their surrounding environments (e.g. solvent accessible surface area, presence of other oxidizable amino acids). In this section we will

briefly review the most important chemical oxidation reactions observed in solids and provide an introduction to potential radical reactions involved in the transformation of peptides and proteins in solids. 3.1. Oxidation of Met and deamination of amino-phenylalanine In solids, the mobility of peptides and proteins is limited. Thus oxidation rates in lyophilyzed protein formulations are largely controlled by the configuration of the solid matrix. For example, when rhodobacter sphaeroides is embedded in different trehalose/water systems, the relationship between the reaction center conformation dynamics and long range electron transfer varies [120]. The oxidation of Met to MetSO has been extensively studied since such modification can lead to the loss of biological activity, and can, for example, modify receptor binding [121,122] and biological half-life of therapeutic antibodies [123,124]. Solid formulations of human growth hormone (hGH) degrade essentially through the oxidation of Met14 to MetSO [125]. Pikal et al. have shown that even in an oxygen-depleted atmosphere (0.05% O2), the level of Met14 oxidation remains higher than the level of deamidation [51]. The oxidation of Met residues in solid human insulin-like growth factor I (hlGF-I) represents another well studied oxidation reaction [126], where the oxidation rate of Met is defined by a second order reaction with respect to the amount of protein and dissolved oxygen in the solid matrix. The transformation of Met to MetSO in hlGF-I is as fast in the solid as in solution. However, in the solid the amount of MetSO represents a greater fraction of the total oxidation products than in solution, suggesting that Met oxidation is the predominant pathway in solids. In addition, Fransson et al. observed that under light exposure the oxidation rate of Met is 30-fold faster than in the absence of light, suggesting the involvement of radical and/or excited species in the oxidation mechanisms of Met. However, the nature of the radical and excited species was never investigated [126]. A tentative hypothesis to rationalize the rapid oxidation of Met would be the formation of a two center-three electron (2c–3e) bond (symbolized as S∴O) between the sulfur atom of Met59 and the oxygen atom of the carbonyl function of arginine (Arg), Arg55, which are 3.98 Å apart. This distance is close to the sum of the van der Waals radii (ca. 3.32 Å), which defines a sulfur-oxygen collision within the framework of collision theory in chemical kinetics [127]. A neighboring group effect, involving the sulfur atom of Met35 and the oxygen atom of the carbonyl group of the C-terminal Ile31, was hypothesized to facilitate a one-electron oxidation of Met35 in β-amyloid peptide [128]. During the photo-degradation of hlGF-I [126], the hypothetical formation of an S∴O bond would require the formation of a sulfur centered radical cation on Met59, MetS•+. Photo-degradation of hlGF-I could also generate an excited state of one of the Tyr residues, which may initiate a succession of electron transfer reactions between the aromatic and sulfurcontaining aliphatic amino acids, which are both known to be relays in multi-step electron transfer reactions [129]. 3.2. Condensation reactions A condensation reaction joins two molecules through the elimination of a molecule of water. The low water content in solids renders the reverse of such condensation reaction rather unlikely. Therefore, condensation reactions such as the formation of Schiff bases and the formation of amide bonds should be favorable [130–132]. Thus, the condensation of the amino group of lysine (Lys) and the guaninine group of Arg with carbonyl groups of either an excipient (e.g. sugar, polyvinylpyrolididone) or the side chains of Asp and Glu, represents a major problem for the stability of solid protein formulations since these chemical reactions lead to the formation of covalent aggregates [133]. When proteins are formulated in the presence of sugars, the side chains of Lys and Arg can condense with the aldehyde moiety of the sugar to form an imine that is ultimately reduced to form a glycosylamino acid (Maillard reaction). For example, lyophilized human relaxin formulated

Please cite this article as: O. Mozziconacci, C. Schöneich, Chemical degradation of proteins in the solid state with a focus on photochemical reactions, Adv. Drug Deliv. Rev. (2014), http://dx.doi.org/10.1016/j.addr.2014.11.016

O. Mozziconacci, C. Schöneich / Advanced Drug Delivery Reviews xxx (2014) xxx–xxx

in the presence of glucose degrades through a Maillard reaction [132]. Proteins can also dimerize through condensation. The formation of dimers was observed during the storage of bovine somatotropin (rbSt) at 30 °C and 96% relative humidity [133]. Hageman et al. observed that the degradation of lyophilized rbSt resulted in the loss of protein monomer, and hypothesized that the degradation of the protein occurred essentially through the condensation of the amino group of a Lys residue and the carboxyl group of either an Asp or Glu residue [133]. Townsend et al. postulated a similar mechanism of aggregation to explain the degradation of ribonuclease A (RNase) during storage at 45 °C [134]. The involvement of Lys as a crosslinker was demonstrated by amino acid analysis that revealed the loss of Lys residues in the aggregates in comparison to the native ribonuclease A [134]. These examples show that the proximity of the residues in solid matrix permits the formation of crosslinks in proteins. In an attempt to generate zero-length crosslinks in proteins, Simons et al. have developed a facile method to dimerize RNase in the absence of a chemical reagent. In the lyophilized solid state, RNase dimerizes under vacuum at 85 °C through the condensation reactions between the amino group of Lys and carboxyl moieties [135]. 3.3. β-Elimination reactions The β-elimination from Cys, serine (Ser), and threonine (Thr) proceeds via the removal of HS− or HO−, respectively, from the βC carbon, generating a carbon-carbon double bond between the αC and β C carbons. Under alkaline conditions, the β-elimination is initiated by deprotonation at the αC-H bond. Through β-elimination, Cys, Ser, and Thr are transformed into dehydroalanine (Dha; 2-aminoacrylic acid) and 2-aminobut-2-enoic acid, respectively. In proteins, the presence of Dha creates a resonance system between the π-system at αC and βC of Dha, and the π-system of the amide bonds. This resonance system renders the βC carbon of Dha more electrophilic, and promotes reaction with nucleophiles such as thiols and amino groups. Lyophilized insulin degrades rapidly to form covalent and non-covalent aggregates [16]. After 24 hours of incubation at 50 °C and 96% relative humidity, Costantino et al. observed that the level of aggregates in solid insulin increased with the amount of free thiols. The authors postulated that the formation of free thiols was the result of β-elimination reactions at Cys residues. Thus, the presence of free thiols permits to explain the reshuffling of the disulfide bonds, which leads to the formation of reducible aggregates. However, the presence of free thiols also means that former Cys residues may be transformed into Dha, which can be involved in the formation of non-reducible aggregates proceeding via nucleophilic additions of either the amino group of Lys or the thiol group of Cys to Dha.

7

140]. These techniques should provide more accurate knowledge regarding the acid-base reactions in solids. Reducible aggregates are formed through disulfide bonds, and can, therefore, be reduced to thiols using dithiothreitol (DTT) or bis-(2mercaptoethyl) sulfone (BMS) as reducing agents. Since S-alkylated bovine serum albumin does not generate reducible aggregates, it has been rationalized that a free sulfhydryl group is required to start a thioldisulfide exchange reaction [141]. Therefore, when pH N pKa of the thiol groups, the thiolates present in the solid matrix (thiol group of Cys) can undergo intra- or inter-molecular thiol-disulfide exchange reaction, where one ionized thiol can attack another disulfide bond. In solution, such reaction is reversible and not regioselective [142–144], meaning that any disulfide can enter in a thiol-disulfide exchange reaction. In contrary, the degradation of tocinoic acid in the solid state showed that in the presence of reduced glutathione (GSH), the thioldisulfide exchange reaction is irreversible and regioselective [145]. The limited conformational flexibility in the solid state might prevent the reduced thiols of tocinoic acid to reform a disulfide bond, preventing the thiol-disulfide exchange reaction with GSH. At elevated temperature, human keratinocyte growth hormone (rhKGF) [146] and recombinant human albumin (rHA) [147] are prone to aggregation. Aggregation starts with the loss of tertiary structure. The unfolding of the protein leads to the formation of soluble aggregates, which ultimately recombine through the formation of new disulfide bonds. Non-reducible aggregates – Non-reducible aggregates correspond to oligomers bound through non-reducible covalent bonds. It was reported that Asn21 of chain A of lyophilized insulin can dimerize with either Phe1 of chain B, or Gly1 of chain A [101]. Strickley et al. have proposed that the deamidation of the side chain of Asn21 leads to the formation of a reactive cyclic anhydride, which reacts with the N-terminal Phe1 of chain B or Gly1 of chain A to form the desamidoAsn21-Phe1 or the desamido-Asn21-Gly1 α- and β-linked dimers, respectively [101]. Oligomers observed during the degradation of lyophilized tumor necrosis factor-alpha (TNFα) remain intact after treatment with β-mercaptoethanol, suggesting that the oligomers are non-reducible [148]. Obviously, the use of chemical reagents during manufacturing, which can induce polymerization, such as formaldehyde, are a suitable cause for the formation of covalent and nonreducible aggregates. For example, during the moisture-induced degradation of tetanus and diphtheria toxoids, which were produced in the presence of formaldehyde, Schwendeman et al. noticed that Lys and His residues had disappeared, and that the aggregates could not be reduced by DTT. The aggregates are likely the result of crosslinks induced by the reaction with formaldehyde [149]. 4. Hydrolytic reactions

3.4. Covalent aggregation of proteins Reducible aggregates – It is commonly noted in the literature that under “slightly alkaline conditions” and “exposure to heat”, most lyophilized proteins such as bovine serum albumin, ovalbumin, glucose oxidase, β-lactoglobulin, and monoclonal antibodies generate reducible aggregates [136–138]. Most of the time the term “alkaline condition” in the solid refers to a lyophilized sample obtained from alkaline solution. However, the definition of pH in the solid is not a trivial problem. In solids, the interaction of two condensed phases across an interphase depends on the functional groups capable of undergoing acid-base reactions. The pKa of an acid in solution is defined as the equilibrium between the concentration of protons, and the concentration of the conjugate base and the acid. At the interphase between solid and water, in which the proton-transfer reactions occur, it is impossible to measure the concentration or the thermodynamic activity of hydronium ion or hydroxide ion and, therefore, to define a pKa. However, it is possible to estimate the solution pH at which the functional group in the interphase is half-protonated. The development of new techniques such as contact angle titration are emerging to study interfacial acid-base reactions [139,

Specifically in peptides and proteins, a hydrolytic reaction breaks an amide bond with a molecule of water. This process results in either the transformation of a asparagine (Asn) into aspartic acid (Asp) (deamidation), the formation of protein fragments (peptide bond cleavage), or the cyclization of adjacent amino acid residues, such as Arg-Pro and Lys-Pro, which are particularly prone to the formation of diketopiperazine. 4.1. Deamidation Deamidation of Asn represents a major chemical degradation route of proteins in solution and in solids. Deamidation of Asn in peptides and proteins occurs via a cyclic imide formation in solution [150] and in amorphous solids [151,152], generating the L- and D-forms of aspartic acid (Asp) and iso-aspartic acid (isoAsp) [150–152]. The different factors controlling the deamidation of Asn in model peptides have been reviewed in several papers [103,153] and books [154]. The cyclic imide generated en route to deamidation of Asn in peptides (e.g. GFNG) in amorphous solids is a precursor for the formation of different types of degradants such as

Please cite this article as: O. Mozziconacci, C. Schöneich, Chemical degradation of proteins in the solid state with a focus on photochemical reactions, Adv. Drug Deliv. Rev. (2014), http://dx.doi.org/10.1016/j.addr.2014.11.016

8

O. Mozziconacci, C. Schöneich / Advanced Drug Delivery Reviews xxx (2014) xxx–xxx

hydrolysis and racemic products, amide-linked peptides, and aggregates [155]. The mechanisms of deamidation are pH- and water-dependent. At pH b 3, acid catalysis takes place, permitting a direct reaction of water with the protonated side chain amide of Asn (Fig. 1, A) At pH N 3, the cyclic imide is formed prior to the generation of Asp or iso-Asp. The cyclic imide results from an intramolecular attack of the carbonyl carbon of the side chain of Asn by the nitrogen of the peptide bond with the succeeding amino acid residue (Fig. 1, B). In the presence of H2O, the cyclic imide is hydrolyzed to generate either Asp or iso-Asp (Fig. 1, C). In solids, the low water content does not always permit the formation of Asp and iso-Asp [6,151,152]. Thus, the cyclic imide is often observed as a degradation product of Asn [150,156–162], and can have a significant impact on the safety or efficacy of a protein therapeutic [163–166]. As far as deamidation rates are concerned [159], a study with model peptides has shown that in the presence of polyvinylpyrrolidone (PVP), the degradation of VYPNGA, VYGNGA, VFGNGA, VIGNGA, VGGNGA, VGPNGA, and VGYNGA follows an apparent first order kinetics, and that the rate of deamidation of Asn is influenced by the nature of the amide acid residue preceding Asn [158,161,167]. The volume and the hydrophobicity of the neighboring amino acid residues influence the deamidation rate of Asn and the isomerization of Asp. The decrease of the overall rates of Asp formation in solids is attributed to a decrease of the conformational flexibility of the peptide backbone and the Asp side chain [151]. It has been shown that an increase of the volume and hydrophobicity result in a decrease of the deamidation rate of Asn [168]. Another model peptide study has demonstrated that the presence of Glu or Asp succeeding an Asn residue, enhances the deamidation rate by a factor 60 in the solid but not in solution [158]. The authors concluded that the presence of Glu or Asp provides local hydration sites for the hydrolysis of the succinimidyl group (Fig. 1, C), allowing for a deamidation process similar to that observed in solution (Fig. 1, A). Deamidation of Asn has been observed in solid insulin [101,102, 169]. However, the deamidation does not proceed via the formation of a cyclic imide as observed for peptides [170]. The C-terminal Asn21 in chain A of insulin deamidates through the formation of an intramolecular cyclic anhydride [101,102]. Because, the intramolecular cyclization of Asn21 requires modest flexibility, even in the glassy state, the intramolecular formation of the cyclic anhydride between the C-terminal carboxylic group and the side chain of Asn21 is moderately restricted. In the presence of H2O, the cyclic anhydride at Asn21 hydrolyzes to generate Asp21. Pikal et al. have documented that at high humidity, the deamidation of Asn3 of chain B is more favorable [53]. Because, protein mobility is dependent on the free volume within a solid matrix, an increase of the humidity with increasing plasticization also promotes the formation of covalent aggregates. Kossiakoff demonstrated that in

solution the primary factors influencing the deamidation of Asn in trypsin are not only the steric or electronic nature of the residues flanking Asn but also the tertiary structure of the protein. Computed models of trypsin suggest that the deamidation rate of Asn is related to the tertiary structure in controlling the geometry of the labile Asn [171]. Variability of the tertiary structure of proteins is linked to the presence of ions. Chen et al. investigated the effect of removal of exogenous calcium ions (Ca2+) on the deamidation of recombinant human deoxyribonuclease I (rhDNase) in solution and amorphous solids [172]. Their findings show that the presence or absence of Ca2+ had little effect on the secondary and tertiary structure of rhDNase. However, in solution, the removal of Ca2+ increased the rates of deamidation of Asn74 and the formation of soluble aggregates. The authors proposed that Ca2 + bound to the native protein protects the disulfide bonds against reduction. Ca2+ is bounded to the loops formed between residues Gly100 and Gly105 and between the residues Asp201 and Cys209. Both loops are close to the disulfides [Cys101 - Cys104], and [Cys173–Cys209]. The removal of Ca2 + would increase the flexibility of the protein to enhance the formation of the succinimidyl intermediate and ultimately the deamidation of Asn. Their findings also demonstrate that i) deamidation of rhDNase in lyophilized formulations is not Ca2+-dependent, and ii) the level of deamidation of rhDnase is lower in solids than in solution. These results appear consistent with the fact that deamidation is a hydrolytic process, i.e. requires water, and therefore the removal of H2O coincides with the reduction of the level of deamidation. 4.2. Peptide bond cleavage The low water content in solids does not favor the direct hydrolysis of peptide bonds. However, peptide bonds of Asp, and especially the sequences containing Asp-Pro, get hydrolyzed approximately 100-fold faster than any other peptide bond in solids. It is believed that under acidic conditions, the carboxylic acid group of Asp catalyzes the hydrolysis process (Fig. 2). Inglis has proposed that Asp would promote the formation of either a six-membered ring (Fig. 2, A) or a fivemembered ring (Fig. 2, B) to release the N-terminal peptide or the C-terminal sequences, respectively [173]. 4.3. Diketopiperazine formation In peptides and proteins, the nucleophilic addition of an N-terminal amine to the carbonyl group of the penultimate peptide bond results in the formation of 2,5-diketopiperazine (DKP). The sequence NH2-Gly-

Fig. 1. Deamidation reaction pathways of Asn (and Gln) at pH b3 in solution (A) and pH N5 in solution and solid (B and C).

Please cite this article as: O. Mozziconacci, C. Schöneich, Chemical degradation of proteins in the solid state with a focus on photochemical reactions, Adv. Drug Deliv. Rev. (2014), http://dx.doi.org/10.1016/j.addr.2014.11.016

O. Mozziconacci, C. Schöneich / Advanced Drug Delivery Reviews xxx (2014) xxx–xxx

9

Fig. 2. Hydrolysis of peptide bonds catalyzed by Asp.

Pro is particularly prone to DKP formation. Certain biologically relevant peptides such as the neuropeptide substance P (SP) and aspartame (R-aspartylphenylalanine methyl ester) degrade in solids through the formation of DKP. For example, in the solid, the degradation of SP leads to the sequential relase of N-terminal dipeptides, cyclo(Arg-Pro), and cyclo(Lys-Pro) [174]. The formation of DKP during the degradation of lyophilized aspartame corresponds essentially to the elimination of a molecule of methanol [13]. 5. Conclusion Non-photochemical and especially hydrolytic/solvolytic degradation mechanisms in solids are relatively well covered in the literature, including details about the relationships between protein and the solid matrix. However, comparatively little is known about oxidative and photochemical degradation mechanisms in solid protein formulations. In general, more mechanistic research effort is necessary to establish protein degradation reactions in the solid state and to design preventive measures. The formation of certain photoproducts can be immunogenic. The UV-irradiation of murine growth hormone in solution generates immunogenic photoproducts [175]. Although, these photoproducts are generated in solution, some of them can, in fact, be generated in solid proteins (e.g. dithiohemiacetal in solid insulin [82]). Acknowledgment We gratefully acknowledge support of our research by Genentech, Inc. and by Amgen, Inc.

[5]

[6]

[7]

[8]

[9] [10] [11]

[12] [13]

[14] [15]

[16] [17] [18] [19]

References [20] [1] F. de Mora, R. Torres, Biotechnology-Derived Medicines: What are They? A Pharmacological and a Historical Perspective, J. Gen. Med. 7 (2010) 145–157, http:// dx.doi.org/10.1057/jgm.2010.10. [2] R.S. Aggarwal, What's fueling the biotech engine, Nat. Biotechnol. 32 (2014) 32–39, http://dx.doi.org/10.1038/nbt.2794. [3] Kalorama Information, A division of Market Research Group, LLC, What's Next in Biopharmaceuticals? (mAbs, rDNA, Interferons, and Other Upcoming Biologics Drugs) 2011–2015, Kalorama, Information. (n.d.). [4] M.J. Pikal, D. Rigsbee, M.L. Roy, D. Galreath, K.J. Kovach, B.S. Wang, et al., Solid state chemistry of proteins: II. The correlation of storage stability of freeze-dried human

[21]

[22]

growth hormone (hGH) with structure and dynamics in the glassy solid, J. Pharm. Sci. 97 (2008) 5106–5121, http://dx.doi.org/10.1002/jps.21374. S.U. Sane, R. Wong, C.C. Hsu, Raman spectroscopic characterization of drying-induced structural changes in a therapeutic antibody: Correlating structural changes with long-term stability, J. Pharm. Sci. 93 (2004) 1005–1018, http://dx.doi.org/10. 1002/jps.20014. C. Oliyai, J.P. Patel, L. Carr, R.T. Borchardt, Solid state chemical instability of an asparaginyl residue in a model hexapeptide, PDA J. Pharm. Sci. Technol. 48 (1994) 167–173. M.C. Lai, R.L. Schowen, R.T. Borchardt, E.M. Topp, Deamidation of a model hexapeptide in poly(vinyl alcohol) hydrogels and xerogels, J. Pept. Res. 55 (2000) 93–101, http://dx.doi.org/10.1034/j.1399-3011.2000.00156.x. M.C. Lai, M.J. Hageman, R.L. Schowen, R.T. Borchardt, E.M. Topp, Chemical stability of peptides in polymers. 1. Effect of water on peptide deamidation in poly(vinyl alcohol) and poly(vinyl pyrrolidone) matrixes, J. Pharm. Sci. 88 (1999) 1073–1080, http://dx.doi.org/10.1021/js980227g. M.J. Hageman, The role of moisture in protein stability, Drug Dev. Ind. Pharm. 14 (1988) 2047–2070, http://dx.doi.org/10.3109/03639048809152002. M. Saltmarch, T.P. Labuza, Nonenzymatic browning via the Maillard reaction in foods, Diabetes 31 (1982) 29–36, http://dx.doi.org/10.2337/diab.31.3.S29. T.G. Park, W. Lu, G. Crotts, Importance of in vitro experimental conditions on protein release kinetics, stability and polymer degradation in protein encapsulated poly (d, l-lactic acid-co-glycolic acid) microspheres, J. Control. Release 33 (1995) 211–222, http://dx.doi.org/10.1016/0168-3659(94)00084-8. T.E. Creighton, Proteins, 2nd ed. Macmillan, New York, 1993. S.S. Leung, D.J.W. Grant, Solid state stability studies of model dipeptides: Aspartame and aspartylphenylalanine, J. Pharm. Sci. 86 (1997) 64–71, http://dx.doi. org/10.1021/js960228d. Aspartame degradation as a function of “water activity”, 302 (1991) 337–349, http://dx.doi.org/10.1007/978-1-4899-0664-9_19. J.L. Cleland, A. Mac, B. Boyd, J. Yang, E.T. Duenas, D. Yeung, et al., The stability of recombinant human growth hormone in poly(lactic-co-glycolic acid) (plga) microspheres, Pharm. Res. 14 (1997) 420–425, http://dx.doi.org/10.1023/A:1012031012367. H.R. Costantino, R. Langer, A.M. Klibanov, Moisture-induced aggregation of lyophilized insulin, Pharm. Res. 11 (1994) 21–29. Stability of solid pharmaceutical proteins, Institute of Technology, Massachusetts, 1995. Photostability of drugs and drug formulations, second edition, 2004. B.A. Kerwin, R.L.J. Remmele, Protect from light: Photodegradation and protein biologics, J. Pharm. Sci. 96 (2007) 1468–1479, http://dx.doi.org/10.1002/jps.20815. S.W. Baertschi, D. Clapham, C. Foti, P.J. Jansen, S. Kristensen, R.A. Reed, et al., Implications of in-use photostability: proposed guidance for photostability testing and labeling to support the administration of photosensitive pharmaceutical products, part 1: drug products administered by injection, J. Pharm. Sci. 102 (2013) 3888–3899, http://dx.doi.org/10.1002/jps.23717. S.W. Baertschi, K.M. Alsante, H.H. Tønnesen, A critical assessment of the ICH guideline on photostability testing of new drug substances and products (Q1B): Recommendation for revision, J. Pharm. Sci. 99 (2010), http://dx.doi.org/10.1002/jps.22076. D. Scheidegger, R.P. Pecora, P.M. Radici, S.C. Kivatinitz, Protein oxidative changes in whole and skim milk after ultraviolet or fluorescent light exposure, J. Dairy Sci. 93 (2010) 5101–5109, http://dx.doi.org/10.3168/jds. 2010-3513.

Please cite this article as: O. Mozziconacci, C. Schöneich, Chemical degradation of proteins in the solid state with a focus on photochemical reactions, Adv. Drug Deliv. Rev. (2014), http://dx.doi.org/10.1016/j.addr.2014.11.016

10

O. Mozziconacci, C. Schöneich / Advanced Drug Delivery Reviews xxx (2014) xxx–xxx

[23] D. Creed, The photophysics and photochemistry of the near-uv absorbing aminoacids.1. Tryptophan and its simple derivatives, Photochem. Photobiol. 39 (1984) 537–562. [24] M.J. Davies, R.J.W. Truscott, Photo-oxidation of proteins and its role in cataractogenesis, 63 (2001) 114–125, http://dx.doi.org/10.1016/S1011-1344(01) 00208-1. [25] D.V. Bent, E. Hayon, Excited state chemistry of aromatic amino acids and related peptides. I. Tyrosine, J. Am. Chem. Soc. 97 (1975) 2599–2606, http://dx.doi.org/ 10.1021/ja00843a002. [26] D. Creed, The photophysics and photochemistry of the near‐uv absorbing amino acids–II. Tyrosine and its simple derivatives, Photochem. Photobiol. 39 (2008) 563–575, http://dx.doi.org/10.1111/j.1751-1097.1984.tb03891.x. [27] D.V. Bent, E. Hayon, Excited state chemistry of aromatic amino acids and related peptides. II. Phenylalanine, J. Am. Chem. Soc. 97 (1975) 2606–2612, http://dx.doi. org/10.1021/ja00843a003. [28] D. Creed, The photophysics and photochemistry of the near‐uv absorbing amino acids‐III. Cystine and its simple derivatives, Photochem. Photobiol. 39 (2008) 577–583, http://dx.doi.org/10.1111/j.1751-1097.1984.tb03892.x. [29] L.I. Grossweiner, Photochemistry of proteins: A review, Curr. Eye Res. 3 (1983) 137–144. [30] L.I. Grossweiner, Y. Usui, Flash photolysis and inactivation of aqueous lysozyme, Photochem. Photobiol. 13 (1971) 195–214, http://dx.doi.org/10.1111/j.17511097.1971.tb06106.x. [31] Z. Li, W.E. Lee, W.C. Galley, Distance dependence of the tryptophan-disulfide interaction at the triplet level from pulsed phosphorescence studies on a model system, Biophys. J. (1989), http://dx.doi.org/10.1016/S0006-3495(89)82682-7. [32] G. Laustriat, C. Hasselmann, Photochemistry of Proteins, Photochem. Photobiol. 22 (1975) 295–298, http://dx.doi.org/10.1111/j.1751-1097.1975.tb06754.x. [33] A. Vanhooren, B. Devreese, K. Vanhee, J. Van Beeumen, I. Hanssens, Photoexcitation of tryptophan groups induces reduction of two disulfide bonds in goat alpha-lactalbumin, Biochemistry 41 (2002) 11035–11043, http://dx.doi.org/10.1021/ bi0258851. [34] C. Pigault, D. Gerard, Influence of the location of tryptophanyl residues in proteins on their photosensitivity, Photochem. Photobiol. 40 (1984) 291–296, http://dx.doi. org/10.1111/j.1751-1097.1984.tb04590.x. [35] S.C. Rao, C.M. Rao, D. Balasubramanian, The conformational status of a protein influences the aerobic photolysis of its tryptophan residues: melittin, β‐ lactoglobulin and the crystallins, Photochem. Photobiol. 51 (1990) 357–362, http://dx.doi. org/10.1111/j.1751-1097.1990.tb01722.x. [36] J.A. Silvester, G.S. Timmins, M.J. Davies, Photodynamically Generated Bovine Serum Albumin Radicals: Evidence for Damage Transfer and Oxidation at Cysteine and Tryptophan Residues, Free Radic. Biol. Med. 24 (1998) 754–766, http://dx.doi. org/10.1016/S0891-5849(97)00327-4. [37] N. Fujii, H. Uchida, T. Saito, The damaging effect of UV-C irradiation on lens alphacrystallin, Mol. Vis. 10 (2004) 814–820. [38] H.H. Tønnessen, Photostability of drugs and drug formulations, 2nd ed. CRC Press, Boca Raton, Fl, 2010. [39] Photostability of drugs and drug formulations, 1996. [40] D.E. Moore, Kinetic treatment of photochemical reactions, Int. J. Pharm. 63 (1990) R5. [41] D.D. Nolte, Diffraction and Light Scattering, Optical Interferometry for Biology and Medicine, Springer New York, New York, NY, 2011, pp. 49–94, http://dx.doi.org/10. 1007/978-1-4614-0890-1_2. [42] S.R. Logan, Does a photochemical reaction have a reaction order? J. Chem. Edu. 74 (1997) 1303, http://dx.doi.org/10.1021/ed074p1303. [43] A.M. Abdul-Fattah, V. Truong-Le, L. Yee, L. Nguyen, D.S. Kalonia, M.T. Cicerone, et al. , Drying-induced variations in physico-chemical properties of amorphous pharmaceuticals and their impact on stability (I): stability of a monoclonal antibody, J. Pharm. Sci. 96 (2007) 1983–2008, http://dx.doi.org/10.1002/jps.20859. [44] A.M. Abdul-Fattah, D.S. Kalonia, M.J. Pikal, The challenge of drying method selection for protein pharmaceuticals: Product quality implications, J. Pharm. Sci. 96 (2007) 1886–1916, http://dx.doi.org/10.1002/jps.20842. [45] H.R. Costantino, L. Firouzabadian, C. Wu, K.G. Carrasquillo, K. Griebenow, S.E. Zale, et al., Protein spray freeze drying. 2. Effect of formulation variables on particle size and stability, J. Pharm. Sci. 91 (2002) 388–395, http://dx.doi.org/10.1002/jps. 10059. [46] S.A. Desobry, F.M. Netto, T.P. Labuza, Comparison of Spray‐drying, Drum‐drying and Freeze‐drying for β‐Carotene Encapsulation and Preservation, J. Food Sci. 62 (1997) 1158–1162, http://dx.doi.org/10.1111/j.1365-2621.1997.tb12235.x. [47] L.L. Chang, D. Shepherd, J. Sun, X.C. Tang, M.J. Pikal, Effect of sorbitol and residual moisture on the stability of lyophilized antibodies: Implications for the mechanism of protein stabilization in the solid state, J. Pharm. Sci. 94 (2005) 1445–1455, http://dx.doi.org/10.1002/jps.20363. [48] L.L. Chang, D. Shepherd, J. Sun, D. Ouellette, K.L. Grant, X.C. Tang, et al., Mechanism of protein stabilization by sugars during freeze-drying and storage: Native structure preservation, specific interaction, and/or immobilization in a glassy matrix? J. Pharm. Sci. 94 (2005) 1427–1444, http://dx.doi.org/10. 1002/jps.20364. [49] L.L. Chang, M.J. Pikal, Mechanisms of protein stabilization in the solid state, J. Pharm. Sci. 98 (2009) 2886–2908, http://dx.doi.org/10.1002/jps.21825. [50] S. Yoshioka, K.M. Forney, Y. Aso, M.J. Pikal, Effect of sugars on the molecular motion of freeze-dried protein formulations reflected by nmr relaxation times, Pharm. Res. 28 (2011) 3237–3247, http://dx.doi.org/10.1007/s11095-011-0512-8. [51] M.J. Pikal, K.M. Dellerman, M.L. Roy, R.M. Riggin, The effects of formulation variables on the stability of freeze-dried human growth hormone, Pharm. Res. 8 (1991) 427–436.

[52] M.J. Pikal, K. Dellerman, M.L. Roy, Formulation and stability of freeze-dried proteins: effects of moisture and oxygen on the stability of freeze-dried formulations of human growth hormone, Dev. Biol. Stand. 74 (1992) 21–37 discussion 37–8. [53] M.J. Pikal, D.R. Rigsbee, The stability of insulin in crystalline and amorphous solids: observation of greater stability for the amorphous form, Pharm. Res. 14 (1997) 1379–1387. [54] G. Ragno, E. Cione, A. Garofalo, G. Genchi, G. Ioele, A. Risoli, et al., Design and monitoring of photostability systems for amlodipine dosage forms, Int. J. Pharm. 265 (2003) 125–132, http://dx.doi.org/10.1016/j.ijpharm.2003.07.001. [55] D.-J. Jang, E.J. Jeong, H.-M. Lee, B.-C. Kim, S.-J. Lim, C.-K. Kim, Improvement of bioavailability and photostability of amlodipine using redispersible dry emulsion, Eur. J. Pharm. Sci. 28 (2006) 405–411, http://dx.doi.org/10.1016/j.ejps.2006.04.013. [56] M. Adler, G. Lee, Stability and surface activity of lactate dehydrogenase in spraydried trehalose, J. Pharm. Sci. 88 (1999) 199–208, http://dx.doi.org/10.1021/ js980321x. [57] R.O. Fernández, R.A. Pizarro, Lethal effect induced in pseudomonas aeruglnosa exposed to ultraviolet-a radiation, Photochem. Photobiol. 64 (1996) 334–339, http:// dx.doi.org/10.1111/j.1751-1097.1996.tb02467.x. [58] F.M. Andreopoulos, E.J. Beckman, A.J. Russell, Light-induced tailoring of PEG-hydrogel properties, Biomaterials 19 (1998) 1343–1352, http://dx.doi.org/10.1016/ S0142-9612(97)00219-6. [59] K. Thoma, R. Kerker, Photoinstabilität von arzneimitteln. iii: Photoinstabilität und stabilisierung von nifedipin in arzneiformen, Pharm. Ind. 54 (1992) 359–365. [60] Y. Matsuda, R. Akazawa, R. Teraoka, M. Otsuka, Pharmaceutical evaluation of carbamazepine modifications: comparative study for photostability of carbamazepine polymorphs by using fourier‐transformed reflection‐absorption infrared spectroscopy and colorimetric measurement, J. Pharm. Pharmacol. 46 (1994) 162–167, http://dx.doi.org/10.1111/j.2042-7158.1994.tb03770.x. [61] R. Franchy, Surface and bulk photochemistry of solids, Rep. Prog. Phys. 61 (1999) 691–753, http://dx.doi.org/10.1088/0034-4885/61/6/003. [62] X.L. Zhou, X.Y. Zhu, J.M. White, Photochemistry at adsorbate/metal interfaces, Surf. Sci. Rep. 13 (1991) 73–220, http://dx.doi.org/10.1016/0167-5729(91)90009-M. [63] Surface photochemistry, Springer Series in Surface Sciences, Springer, Berlin Heidelberg, 1990, pp. 48–64, http://dx.doi.org/10.1007/978-3-642-84145-3_6. [64] Photostability Of Drugs And Drug Formulations, 1996. [65] W. Liu, R. Swift, G. Torraca, Y. Nashed-Samuel, Z.-Q. Wen, Y. Jiang, et al., Root cause analysis of tungsten-induced protein aggregation in pre-filled syringes, PDA J. Pharm. Sci. Technol. 64 (2010) 11–19. [66] L.S. Jones, A. Kaufmann, C.R. Middaugh, Silicone oil induced aggregation of proteins, J. Pharm. Sci. 94 (2005) 918–927, http://dx.doi.org/10.1002/jps.20321. [67] L. Liu, T.W. Randolph, J.F. Carpenter, Particles shed from syringe filters and their effects on agitation-induced protein aggregation, J. Pharm. Sci. 101 (2012) 2952–2959, http://dx.doi.org/10.1002/jps.23225. [68] M.X. Yang, B. Shenoy, M. Disttler, R. Patel, M. McGrath, S. Pechenov, et al., Crystalline monoclonal antibodies for subcutaneous delivery, Proc. Natl. Acad. Sci. U. S. A. 100 (2003) 6934–6939, http://dx.doi.org/10.1073/pnas.1131899100. [69] B. Dani, R. Platz, S.T. Tzannis, High concentration formulation feasibility of human immunoglubulin G for subcutaneous administration, J. Pharm. Sci. 96 (2007) 1504–1517, http://dx.doi.org/10.1002/jps.20508. [70] N. Harn, C. Allan, C. Oliver, C.R. Middaugh, Highly concentrated monoclonal antibody solutions: Direct analysis of physical structure and thermal stability, J. Pharm. Sci. 96 (2007) 532–546, http://dx.doi.org/10.1002/jps.20753. [71] A.W.P. Vermeer, W. Norde, The thermal stability of immunoglobulin: unfolding and aggregation of a multi-domain protein, Biophys. J. 78 (2000) 394–404, http://dx.doi.org/10.1016/S0006-3495(00)76602-1. [72] V.L. Vilker, C.K. Colton, K.A. Smith, The osmotic pressure of concentrated protein solutions: Effect of concentration and ph in saline solutions of bovine serum albumin, J. Colloid Interface Sci. 79 (1981) 548–566, http://dx.doi.org/10.1016/00219797(81)90106-5. [73] S. Connolly, S.N. Rao, D. Fitzmaurice, Characterization of protein aggregated gold nanocrystals, J. Phys. Chem. B 104 (2000) 4765–4776, http://dx.doi.org/10.1021/ jp992842u. [74] J.H. Seo, W.I. Jeon, U. Dembereldorj, S.Y. Lee, S.-W. Joo, Cytotoxicity of serum protein-adsorbed visible-light photocatalytic Ag/AgBr/TiO2 nanoparticles, J. Hazard. Mater. 198 (2011) 347–355, http://dx.doi.org/10.1016/j.jhazmat.2011.10.059. [75] E. Pollak, Y. He, Theory and control of thermal photoinduced electron transfer reactions in polyatomic molecules, J. Phys. Chem. B 105 (2001) 6500–6506, http:// dx.doi.org/10.1021/jp004264j. [76] E. Pollak, L. Plimak, Control of thermal photoinduced electron transfer reactions in the activated and activationless regimes, J. Chem. Phys. 115 (2001) 1867–1874, http://dx.doi.org/10.1063/1.1382815. [77] J. Obrzut, A. Anopchenko, J.F. Douglas, B.W. Rust, Relaxation and antiplasticization measurements in trehalose–glycerol mixtures – A model formulation for protein preservation, J. Non-Cryst. Solids 356 (2010) 777–781, http://dx.doi.org/10.1016/ j.jnoncrysol.2009.07.045. [78] Y. Han, B.-S. Jin, S.-B. Lee, Y. Sohn, J.-W. Joung, J.-H. Lee, Effects of sugar additives on protein stability of recombinant human serum albumin during lyophilization and storage, Arch. Pharm. Res. 30 (2007) 1124–1131. [79] L.M. Crowe, D.S. Reid, J.H. Crowe, Is trehalose special for preserving dry biomaterials? Biophys. J. 71 (1996) 2087–2093, http://dx.doi.org/10.1016/S00063495(96)79407-9. [80] D.S. Katayama, J.F. Carpenter, K.P. Menard, M.C. Manning, T.W. Randolph, Mixing properties of lyophilized protein systems: A spectroscopic and calorimetric study, J. Pharm. Sci. 98 (2009) 2954–2969, http://dx.doi.org/10.1002/ jps.21467.

Please cite this article as: O. Mozziconacci, C. Schöneich, Chemical degradation of proteins in the solid state with a focus on photochemical reactions, Adv. Drug Deliv. Rev. (2014), http://dx.doi.org/10.1016/j.addr.2014.11.016

O. Mozziconacci, C. Schöneich / Advanced Drug Delivery Reviews xxx (2014) xxx–xxx [81] M.S. Navati, J.M. Friedman, Sugar-derived glasses support thermal and photo-initiated electron transfer processes over macroscopic distances, J. Biol. Chem. 281 (2006) 36021–36028, http://dx.doi.org/10.1074/jbc.M606866200. [82] O. Mozziconacci, J. Haywood, E.M. Gorman, E. Munson, C. Schöneich, Photolysis of recombinant human insulin in the solid state: formation of a dithiohemiacetal product at the C-terminal disulfide bond, Pharm. Res. 29 (2012) 121–133, http:// dx.doi.org/10.1007/s11095-011-0519-1. [83] 29 (1964) 1268. [84] I. Delfino, S. Cannistraro, Optical investigation of the electron transfer protein azurin–gold nanoparticle system, Biophys. Chem. 139 (2009) 1–7, http://dx.doi. org/10.1016/j.bpc.2008.09.016. [85] D. Zhang, O. Neumann, H. Wang, V.M. Yuwono, A. Barhoumi, M. Perham, et al., Gold nanoparticles can induce the formation of protein-based aggregates at physiological pH, Nano Lett. 9 (2009) 666–671, http://dx.doi.org/10.1021/nl803054h. [86] C. Freitas, R.H. Müller, Effect of light and temperature on zeta potential and physical stability in solid lipid nanoparticle (SLN™) dispersions, Int. J. Pharm. 168 (1998) 221–229, http://dx.doi.org/10.1016/S0378-5173(98)00092-1. [87] J.O. Bosset, P.U. Gallmann, R. Sieber, Influence of light transmittance of packaging materials on the shelf-life of milk and dairy products — a review, in: M. Mathlouthi (Ed.), Food Packaging and Preservation, Springer US, Boston, MA, 1994, pp. 222–268, http://dx.doi.org/10.1007/978-1-4615-2173-0_13. [88] H. Hosono, Y. Abe, Occurrence of superoxide radical ion in crystalline calcium aluminate 12CaO.7Al2O3 prepared via solid-state reactions, Inorg. Chem. 26 (1987) 1192–1195, http://dx.doi.org/10.1021/ic00255a003. [89] C. Schöneich, Mechanisms of protein damage induced by cysteine thiyl radical formation, Chem. Res. Toxicol. 21 (2008) 1175–1179, http://dx.doi.org/10.1021/ tx800005u. [90] D. Becker, S. Swarts, M. Champagne, M.D.M. Sevilla, An esr investigation of the reactions of glutathione, cysteine and penicillamine thiyl radicals: competitive formation of RSO , R , RSSR -, and RSS , Int. J. Radiat. Biol. Relat. Stud. Phys. Chem. Med. 53 (1988) 767–786, http://dx.doi.org/10.1080/09553008814551121. [91] S.A. Everett, L.K. Folkes, P. Wardman, K.D. Asmus, Free-Radical Repair by a Novel Perthiol: Reversible Hydrogen Transfer and Perthiyl Radical Formation, 20 (2009) 387–400, http://dx.doi.org/10.3109/10715769409145638. [92] Perthiols as antioxidants: Radical-scavenging and prooxidative mechanisms, 251 (1995) 55–69, http://dx.doi.org/10.1016/0076-6879(95)51110-5. [93] C. Chatgilialoglu, C. Ferreri, A. Torreggiani, A.M. Salzano, G. Renzone, A. Scaloni, Radiation-induced reductive modifications of sulfur-containing amino acids within peptides and proteins, J. Proteome 74 (2011) 2264–2273. [94] T. Nauser, C. Schöneich, Thiyl radicals abstract hydrogen atoms from the αC−H bonds in model peptides: absolute rate constants and effect of amino acid structure, J. Am. Chem. Soc. 125 (2003) 2042–2043, http://dx.doi.org/10.1021/ ja0293599. [95] T. Nauser, W.H. Koppenol, C. Schöneich, Reversible hydrogen transfer reactions in thiyl radicals from cysteine and related molecules: absolute kinetics and equilibrium constants determined by pulse radiolysis, J. Phys. Chem. B 116 (2012) 5329–5341, http://dx.doi.org/10.1021/jp210954v. [96] O. Mozziconacci, V. Sharov, T.D. Williams, B.A. Kerwin, C. Schöneich, Peptide cysteine thiyl radicals abstract hydrogen atoms from surrounding amino acids: the photolysis of a cystine containing model peptide, J. Phys. Chem. B 112 (2008) 9250–9257, http://dx.doi.org/10.1021/jp801753d. [97] O. Mozziconacci, T.D. Williams, C. Schöneich, Intramolecular hydrogen transfer reactions of thiyl radicals from glutathione: formation of carbon-centered radical at Glu, Cys, and Gly, Chem. Res. Toxicol. 25 (2012) 1842–1861, http://dx.doi.org/10. 1021/tx3000494. [98] H. Terryn, V. Deridder, C. Sicard-Roselli, B. Tilquin, C. Houée-Levin, Radiolysis of proteins in the solid state: an approach by EPR and product analysis, J. Synchrotron Radiat. 12 (2005) 292–298, http://dx.doi.org/10.1107/S0909049505003304. [99] B.L. Miller, M.J. Hageman, T.J. Thamann, L.B. Barròn, C. Schöneich, Solid-state photodegradation of bovine somatotropin (bovine growth hormone): evidence for tryptophan-mediated photooxidation of disulfide bonds, J. Pharm. Sci. 92 (2003) 1698–1709, http://dx.doi.org/10.1002/jps.10418. [100] F. Freeman, C.N. Angeletakis, α-Disulfoxide formation during the mchloroperoxybenzoic acid oxidation of S-(2, 2-dimethylpropyl) 2, 2dimethylpropanethiosulfinate, J. Am. Chem. Soc. 104 (1982) 5766–5774. [101] R.G. Strickley, B.D. Anderson, Solid-state stability of human insulin i. mechanism and the effect of water on the kinetics of degradation in lyophiles from pH 2–5 solutions, Pharm. Res. 13 (1996) 1142–1153, http://dx.doi.org/10.1023/A:1016043715791. [102] R.G. Strickley, B.D. Anderson, Solid-state stability of human insulin II. Effect of water on reactive intermediate partitioning in lyophiles from pH 2-5 solutions: Stabilization against covalent dimer formation, J. Pharm. Sci. 86 (1997) 645–653, http://dx.doi.org/10.1021/js9700311. [103] M.C. Lai, E.M. Topp, Solid-state chemical stability of proteins and peptides, J. Pharm. Sci. 88 (1999) 489–500, http://dx.doi.org/10.1021/js980374e. [104] B.M. Baughman, E. Stennett, R.E. Lipner, A.C. Rudawsky, S.J. Schmidtke, Structural and spectroscopic studies of the photophysical properties of benzophenone derivatives, J. Phys. Chem. A 113 (2009) 8011–8019, http://dx.doi.org/10.1021/ jp810256x. [105] S.P.L. Sorenson, M. Sorensen, Compt. Rend. Trav. Lab. Carlsberg. 19 (n.d.) 1–34. [106] W. Saenger, Structure and dynamics of water surrounding biomolecules, Annu. Rev. Biophys. Biophys. Chem. 16 (1987) 93–114, http://dx.doi.org/10.1146/ annurev.bb.16.060187.000521. [107] N. Thanki, J.M. Thornton, J.M. Goodfellow, Distributions of water around aminoacid residues in proteins, J. Mol. Biol. 202 (1988) 637–657. [108] E.N. Baker, R.E. Hubbard, Hydrogen bonding in globular proteins, Prog. Biophys. Mol. Biol. 44 (1984) 97–179.

11

[109] C. von Sonntag, Advanced oxidation processes: mechanistic aspects, Water Sci. Technol. 58 (2008) 1015–1021, http://dx.doi.org/10.2166/wst.2008.467. [110] R. Pethig, Protein-water interactions determined by dielectric methods, Annu. Rev. Phys. Chem. 43 (1992) 177–205, http://dx.doi.org/10.1146/annurev.pc.43.100192. 001141. [111] J.T. Edsall, H.A. McKenzie, Water and proteins. II. The location and dynamics of water in protein systems and its relation to their stability and properties, Adv. Biophys. 16 (1983) 53–183. [112] A.A. Kossiakoff, Protein dynamics investigated by the neutron diffraction–hydrogen exchange technique, Nature 296 (1982) 713–721, http://dx.doi.org/10.1038/ 296713a0. [113] E. Tuchsen, C. Woodward, Hydrogen exchange of primary amide protons in basic pancreatic trypsin inhibitor: evidence for amino group rotation in buried asparagine side chains, Biochemistry 26 (1987) 8073–8078, http://dx.doi.org/10.1021/ bi00399a008. [114] G.-J. Zhao, K.-L. Han, Site-specific solvation of the photoexcited protochlorophyllide a in methanol: formation of the hydrogen-bonded intermediate state induced by hydrogen-bond strengthening, Biophys. J. 94 (2008) 38–46, http://dx.doi.org/10. 1529/biophysj.107.113738. [115] G.-J. Zhao, K.-L. Han, Hydrogen bonding in the electronic excited state, Acc. Chem. Res. 45 (2012) 404–413, http://dx.doi.org/10.1021/ar200135h. [116] G.-J. Zhao, J.-Y. Liu, L.-C. Zhou, K.-L. Han, Site-selective photoinduced electron transfer from alcoholic solvents to the chromophore facilitated by hydrogen bonding: a new fluorescence quenching mechanism, J. Phys. Chem. B 111 (2007) 8940–8945, http://dx.doi.org/10.1021/jp0734530. [117] G.-J. Zhao, K.-L. Han, Effects of hydrogen bonding on tuning photochemistry: concerted hydrogen-bond strengthening and weakening, ChemPhysChem 9 (2008) 1842–1846, http://dx.doi.org/10.1002/cphc.200800371. [118] R. Knochenmuss, S. Leutwyler, Proton transfer from 1-naphthol to water: Small clusters to the bulk, J. Chem. Phys. 91 (1989) 1268, http://dx.doi.org/10.1063/1. 457202. [119] E.R. Stadtman, R.L. Levine, Free radical-mediated oxidation of free amino acids and amino acid residues in proteins, Amino Acids 25 (2003) 207–218, http://dx.doi. org/10.1007/s00726-003-0011-2. [120] G. Palazzo, A. Mallardi, A. Hochkoeppler, L. Cordone, G. Venturoli, Electron transfer kinetics in photosynthetic reaction centers embedded in trehalose glasses: Trapping of conformational substates at room temperature, Biophys. J. 82 (2002) 558–568, http://dx.doi.org/10.1016/S0006-3495(02)75421-0. [121] L.C. Teh, L.J. Murphy, N.L. Huq, A.S. Surus, H.G. Friesen, L. Lazarus, et al., Methionine oxidation in human growth hormone and human chorionic somatomammotropin. Effects on receptor binding and biological activities, J. Biol. Chem. 262 (1987) 6472–6477. [122] S.R. Labrenz, M.A. Calmann, G.A. Heavner, G. Tolman, The oxidation of methionine54 of epoetinum alfa does not affect molecular structure or stability, but does decrease biological activity, PDA J. Pharm. Sci. Technol. 62 (2008) 211–223. [123] H. Pan, K. Chen, L. Chu, F. Kinderman, I. Apostol, G. Huang, Methionine oxidation in human IgG2 Fc decreases binding affinities to protein A and FcRn, Protein Sci. 18 (2009) 424–433, http://dx.doi.org/10.1002/pro.45. [124] I. Correia, Stability of IgG isotypes in serum, mAbs 2 (2014) 221–232, http://dx.doi. org/10.4161/mabs.2.3.11788. [125] G.W. Becker, P.M. Tackitt, W.W. Bromer, D.S. Lefeber, R.M. Riggin, Isolation and characterization of a sulfoxide and a desamido derivative of biosynthetic human growth hormone, Biotechnol. Appl. Biochem. 10 (1988) 326–337, http://dx.doi. org/10.1111/j.1470-8744.1988.tb00023.x. [126] J. Fransson, E. Florin-Robertsson, K. Axelsson, C. Nyhlén, Oxidation of human insulin-like growth factor I in formulation studies: kinetics of methionine oxidation in aqueous solution and in solid state, Pharm. Res. 13 (1996) 1252–1257. [127] A. Bondi, van der Waals Volumes and Radii, J. Phys. Chem. 68 (1964) 441–451, http://dx.doi.org/10.1021/j100785a001. [128] D. Pogocki, C. Schöneich, Redox properties of Met(35) in neurotoxic beta-amyloid peptide. A molecular modeling study, Chem. Res. Toxicol. 15 (2002) 408–418. [129] B. Giese, M. Wang, J. Gao, M. Stoltz, P. Müller, M. Graber, Electron relay race in peptides, J. Org. Chem. 74 (2009) 3621–3625, http://dx.doi.org/10.1021/ jo900375f. [130] D.C. Dubost, M.J. Kaufman, J.A. Zimmerman, M.J. Bogusky, A.B. Coddington, S.M. Pitzenberger, Characterization of a solid state reaction product from a lyophilized formulation of a cyclic heptapeptide. A novel example of an excipient-induced oxidation, Pharm. Res. 13 (1996) 1811–1814. [131] K. Eichner, M. Ciner-Doruk, Formation and decomposition of browning intermediates and visible sugar-amine browning reactions, Elsevier, 1981., http://dx.doi.org/ 10.1016/b978-0-12-591350-8.50028-0. [132] S. Li, T.W. Patapoff, D. Overcashier, C. Hsu, T.H. Nguyen, R.T. Borchardt, Effects of reducing sugars on the chemical stability of human relaxin in the lyophilized state, J. Pharm. Sci. 85 (1996) 873–877, http://dx.doi.org/10.1021/js950456s. [133] M.J. Hageman, J.M. Bauer, P.L. Possert, R.T. Darrington, Preformulation studies oriented toward sustained delivery of recombinant somatotropins, J. Agric. Food Chem. 40 (1992) 348–355, http://dx.doi.org/10.1021/jf00014a036. [134] M.W. Townsend, P.P. Deluca, Nature of aggregates formed during storage of freeze‐ dried ribonuclease A, J. Pharm. Sci. 80 (1991) 63–66, http://dx.doi.org/10.1002/jps. 2600800116. [135] B.L. Simons, M.C. King, T. Cyr, M.A. Hefford, H. Kaplan, Covalent cross‐linking of proteins without chemical reagents, Protein Sci. 11 (2002) 1558–1564, http://dx. doi.org/10.1110/ps.4390102. [136] W.R. Liu, R. Langer, A.M. Klibanov, Moisture‐induced aggregation of lyophilized proteins in the solid state, Biotechnol. Bioeng. 37 (1991) 177–184, http://dx.doi. org/10.1002/bit.260370210.

Please cite this article as: O. Mozziconacci, C. Schöneich, Chemical degradation of proteins in the solid state with a focus on photochemical reactions, Adv. Drug Deliv. Rev. (2014), http://dx.doi.org/10.1016/j.addr.2014.11.016

12

O. Mozziconacci, C. Schöneich / Advanced Drug Delivery Reviews xxx (2014) xxx–xxx

[137] G.M. Jordan, S. Yoshioka, T. Terao, The aggregation of bovine serum albumin in solution and in the solid state, J. Pharm. Pharmacol. 46 (1994) 182–185. [138] J.D. Andya, C.C. Hsu, S.J. Shire, Mechanisms of aggregate formation and carbohydrate excipient stabilization of lyophilized humanized monoclonal antibody formulations, AAPS J. 5 (2003) 21–31, http://dx.doi.org/10.1208/ps050210. [139] E. McCafferty, J.P. Wightman, Determination of the surface isoelectric point of oxide films on metals by contact angle titration, J. Colloid Interf. Sci. 194 (1997) 344–355, http://dx.doi.org/10.1006/jcis.1997.5138. [140] L.-K. Chau, M.D. Porter, Surface isoelectric point of evaporated silver films: Determination by contact angle titration, J. Colloid Interf. Sci. 145 (1991) 283–286, http://dx.doi.org/10.1016/0021-9797(91)90121-N. [141] H.R. Costantino, R. Langer, A.M. Klibanov, Solid-phase aggregation of proteins under pharmaceutically relevant conditions, J. Pharm. Sci. 83 (1994) 1662–1669. [142] P.L. Yeo, D.L. Rabenstein, Characterization of the thiol/disulfide chemistry of neurohypophyseal peptide hormones by high-performance liquid chromatography, Anal. Chem. 65 (1993) 3061–3066, http://dx.doi.org/10.1021/ac00069a019. [143] Thiol/disulfide exchange equilibria and disulfidebond stability, Methods in Enzymology, Elsevier, 1995, pp. 8–28, http://dx.doi.org/10.1016/0076-6879(95)51107-5. [144] P.A. Fernandes, M.J. Ramos, Theoretical insights into the mechanism for thiol/disulfide exchange, Chem. Eur. J. 10 (2004) 257–266, http://dx.doi.org/10.1002/chem. 200305343. [145] L. Zhang, T.D. Williams, E.M. Topp, Reversibility and regioselectivity in thiol/disulfide interchange of tocinoic acid with glutathione in lyophilized solids, J. Pharm. Sci. 98 (2009) 3312–3318, http://dx.doi.org/10.1002/jps.21516. [146] B.L.B. Chen, T.T. Arakawa, E.E. Hsu, L.O.L. Narhi, T.J.T. Tressel, S.L.S. Chien, Strategies to suppress aggregation of recombinant keratinocyte growth factor during liquid formulation development, J. Pharm. Sci. 83 (1994) 1657–1661, http://dx.doi.org/ 10.1002/jps.2600831204. [147] H.R. Costantino, R. Langer, A.M. Klibanov, Aggregation of a lyophilized pharmaceutical protein, recombinant human albumin: effect of moisture and stabilization by excipients, Biotechnology (N.Y.) 13 (1995) 493–496. [148] M.S. Hora, R.K. Rana, F.W. Smith, Lyophilized formulations of recombinant tumor necrosis factor, Pharm. Res. 9 (1992) 33–36. [149] S.P. Schwendeman, H.R. Costantino, R.K. Gupta, G.R. Siber, A.M. Klibanov, R. Langer, Stabilization of tetanus and diphtheria toxoids against moisture-induced aggregation, Proc. Natl. Acad. Sci. U. S. A. 92 (1995) 11234–11238. [150] T. Geiger, S. Clarke, Deamidation, isomerization, and racemization at asparaginyl and aspartyl residues in peptides. Succinimide-linked reactions that contribute to protein degradation, J. Biol. Chem. 262 (1987) 785–794. [151] C. Oliyai, J.P. Patel, L. Carr, R.T. Borchardt, Chemical pathways of peptide degradation. VII. Solid state chemical instability of an aspartyl residue in a model hexapeptide, Pharm. Res. 11 (1994) 901–908, http://dx.doi.org/10.1023/A:1018998312503. [152] Y. Song, R.L. Schowen, R.T. Borchardt, E.M. Topp, Effect of ‘pH’ on the rate of asparagine deamidation in polymeric formulations: ‘pH’–rate profile, J. Pharm. Sci. 90 (2001) 141–156, http://dx.doi.org/10.1002/1520-6017(200102)90:2b141::AIDJPS5N3.0.CO;2-Y. [153] A.A. Wakankar, R.T. Borchardt, Formulation considerations for proteins susceptible to asparagine deamidation and aspartate isomerization, J. Pharm. Sci. 95 (2006) 2321–2336, http://dx.doi.org/10.1002/jps.20740. [154] D.W. Aswad, Deamidation and isoaspartate formation in peptides and proteins, CRC Press Inc, Boca raton, Florida, 1994. [155] M.P. Dehart, B.D. Anderson, A mechanism‐based kinetic analysis of succinimide‐ mediated deamidation, racemization, and covalent adduct formation in a model peptide in amorphous lyophiles, J. Pharm. Sci. 101 (2012) 3096–3109, http://dx. doi.org/10.1002/jps.23061. [156] N.E. Robinson, Robinson A B, Molecular clocks, 1st ed. Althouse Press, Cave Junction, 2004. [157] S. Clarke, Propensity for spontaneous succinimide formation from aspartyl and asparaginyl residues in cellular proteins, J. Pept. Res. 30 (1987) 808–821, http:// dx.doi.org/10.1111/j.1399-3011.1987.tb03390.x.

[158] B. Li, E.M. Gorman, K.D. Moore, T. Williams, R.L. Schowen, E.M. Topp, et al., Effects of acidic N + 1 residues on asparagine deamidation rates in solution and in the solid state, J. Pharm. Sci. 94 (2005) 666–675, http://dx.doi.org/10.1002/jps.20263. [159] S. Capasso, Estimation of the deamidation rate of asparagine side chains, J. Pept. Res. 55 (2000) 224–229, http://dx.doi.org/10.1034/j.1399-3011.2000.00172.x. [160] C. Oliyai, R.T. Borchardt, Chemical pathways of peptide degradation. IV. Pathways, kinetics, and mechanism of degradation of an aspartyl residue in a model hexapeptide, Pharm. Res. 10 (1993) 95–102, http://dx.doi.org/10.1023/A:1018981231468. [161] K. Patel, R.T. Borchardt, Chemical pathways of peptide degradation. III. Effect of primary sequence on the pathways of deamidation of asparaginyl residues in hexapeptides, Pharm. Res. 7 (1990) 787–793. [162] A.B. Joshi, L.E. Kirsch, The relative rates of glutamine and asparagine deamidation in glucagon fragment 22-29 under acidic conditions, J. Pharm. Sci. 91 (2002) 2331–2345, http://dx.doi.org/10.1002/jps.10213. [163] J. Cacia, R. Keck, L.G. Presta, J. Frenz, Isomerization of an aspartic acid residue in the complementarity-determining regions of a recombinant antibody to human IgE: identification and effect on binding affinity, Biochemistry 35 (1996) 1897–1903, http://dx.doi.org/10.1021/bi951526c. [164] M. Xie, Z. Shahrokh, M. Kadkhodayan, W.J. Henzel, M.F. Powell, R.T. Borchardt, et al., Asparagine deamidation in recombinant human lymphotoxin: Hindrance by three‐dimensional structures, J. Pharm. Sci. 92 (2003) 869–880, http://dx.doi. org/10.1002/jps.10342. [165] C.L. Stevenson, A.R. Friedman, T.M. Kubiak, M.E. Donlan, R.T. Borchardt, Effect of secondary structure on the rate of deamidation of several growth hormone releasing factor analogs, J. Pept. Res. 42 (1993) 497–503, http://dx.doi.org/10.1111/j. 1399-3011.1993.tb00356.x. [166] M.L. di Salvo, S.D. Fratte, D. De Biase, F. Bossa, V. Schirch, Purification and characterization of recombinant rabbit cytosolic serine hydroxymethyltransferase, Protein Expr. Purif. 13 (1998) 177–183, http://dx.doi.org/10.1006/prep.1998.0890. [167] T.V. Brennan, S. Clarke, Spontaneous degradation of polypeptides at aspartyl and asparaginyl residues: Effects of the solvent dielectric, Protein Sci. 2 (1993) 331–338, http://dx.doi.org/10.1002/pro.5560020305. [168] B. Li, R.L. Schowen, E.M. Topp, R.T. Borchardt, Effect of N-1 and N-2 residues on peptide deamidation rate in solution and solid state, Aaps J. 8 (2006) E166–E173, http:// dx.doi.org/10.1208/aapsj080120. [169] H.T. Wright, Sequence and structure determinants of the nonenzymatic deamidation of asparagine and glutamine residues in proteins, Protein Eng. 4 (1991) 283–294, http://dx.doi.org/10.1093/protein/4.3.283. [170] R. Tyler-Cross, V. Schirch, Effects of amino acid sequence, buffers, and ionic strength on the rate and mechanism of deamidation of asparagine residues in small peptides, J. Biol. Chem. 266 (1991) 22549–22556. [171] A. Kossiakoff, Tertiary structure is a principal determinant to protein deamidation, Science 240 (1988) 191–194, http://dx.doi.org/10.1126/science.3353715. [172] B. Chen, H.R. Costantino, J. Liu, C.C. Hsu, S.J. Shire, Influence of calcium ions on the structure and stability of recombinant human deoxyribonuclease I in the aqueous and lyophilized states, J. Pharm. Sci. 88 (1999) 477–482, http://dx.doi.org/10.1021/ js980273g. [173] A.S. Inglis, Cleavage at aspartic acid, Meth. Enzymol. 91 (1983) 324–332, http://dx. doi.org/10.1016/S0076-6879(83)91030-3. [174] U. Kertscher, M. Bienert, E. Krause, N.F. Sepetov, B. Mehlis, Spontaneous chemical degradation of substance P in the solid phase and in solution, Int. J. Pept. Protein Res. 41 (1993) 207–211. [175] A.H. Fradkin, O. Mozziconacci, C. Schoeneich, J.F. Carpenter, T.W. Randolph, UV photodegradation of murine growth hormone: Chemical analysis and immunogenicity consequences, Eur. J. Pharm. Biopharm. 87 (2014) 395–402, http://dx.doi. org/10.1016/j.ejpb.2014.04.005.

Please cite this article as: O. Mozziconacci, C. Schöneich, Chemical degradation of proteins in the solid state with a focus on photochemical reactions, Adv. Drug Deliv. Rev. (2014), http://dx.doi.org/10.1016/j.addr.2014.11.016