ARTICLE pubs.acs.org/Langmuir
Tuning in-meso-Crystallized Lysozyme Polymorphism by Lyotropic Liquid Crystal Symmetry Alexandru Zabara, Idit Amar-Yuli, and Raffaele Mezzenga* ETH Zurich, Food & Soft Materials Science, Institute of Food, Nutrition & Health Schmelzbergstrasse 9, LFO E23, 8092 Z€urich, Switzerland
bS Supporting Information ABSTRACT: Lipid-based lyotropic liquid crystals (LLCs) show great potential for applications in fields as diverse as food technology, cosmetics, pharmaceutics, or structural biology. Recently, these systems have provided a viable alternative to the difficult process of membrane protein crystallization, owing to their similarities with cell membranes. Nonetheless, the process of in-meso crystallization of proteins still remains poorly understood. In this study, we demonstrate that in-meso crystal morphologies of lysozyme (LSZ), a model hydrophilic protein, can be controlled by both the composition and symmetry of the mesophase, inferring a possible general influence of the LLC space group on the protein crystal polymorphism. Lysozyme was crystallized in-meso from three common LLC phases (lamellar, inverse hexagonal, and inverse bicontinuous cubic) composed of monolinolein and water. Different mixing ratios of mesophase to crystallization buffer were used in order to tune crystallization both in the bulk mesophase and in excess water conditions. Two distinct mechanisms of crystallization were shown to take place depending on available water in the mesophases. In the bulk mesophases, protein nuclei form and grow within structural defects of the mesophase and partially dehydrate the system inducing order-to-order transitions of the liquid crystalline phase toward stable symmetries in conditions of lower hydration. The formed protein crystals eventually macrophase separate from the mesophase allowing the system to reach its final symmetry. On the other hand, when excess water is available, protein molecules diffuse from the water channels into the excess water, where the crystallization process can take place freely, and with little to no effect on the structure and symmetry of the lyotropic liquid crystals.
’ INTRODUCTION Lyotropic liquid crystalline (LLC) phases based on the selfassembly of lipids and water are characterized by a high internal order and have been the subject of extensive research due to their complex structural features and their potential for practical applications in a variety of fields such as food technology, cosmetic, biomedical (pharmaceutical), and structural biology.111 In the past years significant progress has been made in the encapsulation and controlled release of various biomolecules from inverted cubic lyotropic liquid crystals.1214 The diversity of molecules that has been studied ranges from lipophilic and hydrophilic drugs13 to amino acids,12,14 peptides,3,5,12 or nucleic acids,12 and emphasizes the potential applicability of liquid crystals as delivery systems. Additionally, given similarities with cellular lipid bilayers, inverted cubic phases have been also used as protein hosting reservoirs from which membrane proteins crystals grow, a crystallization procedure referred as in-meso.6,7,15 Ever since the development of the in-meso approach by Landau and Rosenbusch in 1996,15 this alternative method for membrane protein crystallization has been successfully used for high-resolution structure determination studies of several membrane proteins.6,7,16 r 2011 American Chemical Society
Solubilization of hydrophilic proteins such as R-chymotrypsin, lysozyme, cytochrome c, hemoglobin or insulin, or membrane proteins such as bacteriorhodopsin or dopamine D2L receptor, their effect on the liquid crystalline structure and their interactions with the lipidic component, have been extensively studied.1726 There have also been several reports regarding the in-meso crystallization of hydrophilic proteins, in most cases lysozyme (LSZ) being used as a model protein.2731 A crystallization mechanism for lysozyme inside the cubic phase, based on crystals nucleating from concentrated solutions of lysozyme found within mesoscopic sized defects of the liquid crystalline structure was proposed by Tanaka et al.30 Because, however, this mechanism was limited solely to cubic phases of Im3m symmetry, a broader and more comprehensive study is still required to provide a general mechanism for the in-meso crystallization process of hydrophilic proteins in lyotropic liquid crystals of varying and diverse group symmetries. In this work we use the in-meso approach to understand the crystallization process of lysozyme, selected as a model water-soluble protein, in three different structures of LLC: lamellar, hexagonal, Received: February 23, 2011 Revised: March 28, 2011 Published: April 20, 2011 6418
dx.doi.org/10.1021/la200710p | Langmuir 2011, 27, 6418–6425
Langmuir and Pn3m cubic and we trigger the crystallization by using the hanging drop vapor diffusion method; we then correlate, for the very first time, the crystallization behavior and crystal polymorphism to the symmetry of the hosting mesophase. Lysozyme is an enzyme that catalyzes the hydrolysis of polysaccharides contained within the cellular wall.32 It has been well characterized and used as a model for a number of studies including protein stability, enzymatic activity and protein crystallization.3335 Additionally it is the most commonly used hydrophilic protein for the in-meso crystallization studies of water-soluble proteins.27,28,30,31 None of these studies, however, has clearly established the connection existing between the crystal structure and the symmetry of the mesophase, which constitutes one of the main objectives of the present work. To date, four polymorphic crystal forms of hen egg white lysozyme (HEWL) have been well established, using only conventional methods of crystallization, and provided high resolution structures: tetragonal, orthorhombic, monoclinic, and triclinic.35,36 In what follows we discuss the effect of the symmetry of the hosting mesophase on the lysozyme crystallization process as a function of time. We use the hanging-drop vapor diffusion method to induce crystallization, and a combination of crosspolarized optical microscopy and wide- and small-angle X-ray scattering (WAXS and SAXS) as main characterization tools.
’ MATERIALS AND METHODS Materials. Dimodan U/J was a gift of Danisco (Denmark) and was used as received. This commercial-grade form of monolinolein contains more than 98 wt % monoglyceride. The same batch of Dimodan was used throughout the whole work. Linoleic acid (LA) was purchased from sigma-Aldrich (Buchs, Switzerland). Chicken egg-white lysozyme was purchased form Sigma-Aldrich (Schnelldorf, Germany). All of the necessary salts for the crystallization procedures were purchased from Sigma Aldrich-Chemie (Steinheim, Germany). Limbro 24-well hanging drop crystallization plates as well as all other crystallization accessories were purchased from Jena Bioscience (Jena, Germany). Samples Preparation. The composition of each LLC system (empty and loaded) was as follows: in case of LLC loaded with protein, 5 wt % LSZ was dispersed in a 0.1 M Na acetate (pH 4.5) buffer prior to its incorporation into the LC mesophases. The first LLC system consisted of a lamellar phase (LR) and was composed of 90 wt % GML and 10 wt % water. The second was a cubic phase (Pn3m) containing 65.3 wt % GML and 34.7 wt % water. The third LLC phase was a columnar hexagonal phase (HII) and consisted of 63.75 wt % GML, 11.25 wt % LA, and 25.00 wt % water. The liquid crystalline mesophases were prepared by mixing weighed quantities of GML, protein solution (protein buffer in the case of the empty phases) and/or LA followed by heating to 45 °C and vortexing until a homogeneous mixture was obtained. The prepared mesophase was then allowed to cool down to room temperature. All of the buffers and protein solution were prepared using ultrapure water and the pH was adjusted using a 1 M solution of HCl. Crystallization of Chicken Egg-White Lysozyme. LSZ crystals were grown at 20 °C using the hanging drop vapor diffusion method. The reservoir solution contained 500 μL of 0.1 M Na acetate buffer (pH 4.5) and 0.8 M NaCl. The drop for the control experiment contained 4 μL of 50 mg/mL LSZ solution in sodium acetate buffer and 1, 2, or 4 μL of the reservoir solution (for the
ARTICLE
Figure 1. Control experiment for crystallization of LSZ. (a) Optical microscope image of tetragonal LSZ crystal obtained by hanging drop vapor diffusion crystallization method. (b) 2D diffraction pattern of LSZ tetragonal crystals.
different mixing ratios). The drops for the in-meso crystallization were prepared by adding the proportional amount of reservoir buffer (according to the desired mixing ratio) on top of the weighted amount of LSZ-containing mesophase. Hanging drop vapor diffusion blank experiments without the protein were also performed in order to rule-out unexpected morphological changes in the liquid crystalline phases due to the higher ionic strength of the crystallization buffer (see Supporting Information). Cross Polarized Optical Microscopy. Microscopy observations regarding protein crystal growth and morphology were achieved under cross-polarized light using a Zeiss Axioskop 2 MOT optical microscope and a magnification of 10. All of the pictures were taken with a Hamamatsu C5810 CCD camera. Small Angle and Wide Angle X-ray Scattering. Small angle and Wide angle X-ray scattering experiments were performed on a MicroMax-002þ microfocused beam, the applied voltage and filament current being 45 kV and 0.88 mA, respectively. The Nifiltered Cu KR radiation (λCuKR = 1.5418 Å) was collimated by three pinhole (0.4, 0.3, and 0.8 mm) collimators. The data was collected either by a two-dimensional argon-filled detector (for SAXS) or with the help of a Fuji Film BAS-MS 2025 imaging plate system: 15.2 15.2 cm, 50 μm resolution (for WAXS). An effective scattering-vector range of 0.03 Å1 < q < 2.5 Å1 was obtained, where q is the scattering wave-vector defined as q = 4π sin(θ)/λCuKR, with a scattering angle of 2θ. For all measurements the samples were placed inside a Linkam HFS91 stage for X-ray spectroscopy measurements.
’ RESULTS A. Free Crystallization of Lysozyme from Buffer. We first assessed lysozyme crystals growth and morphology in solutions without external influences. A control experiment using a 50 mg/mL LSZ solution and different ratios (4:1, 2:1, and 1:1) of protein solution to crystallization solution (0.1 M NaAc, 0.8 M NaCl, and pH 4.5) has been carried out. At all protein solution/ buffer ratios LSZ crystals of tetragonal space group were observed as it could be easily recognized by cross-polarized optical microscopy on samples incubated 2 days at 20 °C (Figure 1a). These findings were confirmed by SAXS on single LSZ crystals, in which 2D diffraction patterns (Figure 1b) confirmed a tetragonal (P43212) symmetry of the LSZ crystals. 6419
dx.doi.org/10.1021/la200710p |Langmuir 2011, 27, 6418–6425
Langmuir
ARTICLE
Figure 2. 1D SAXS patterns of the lamellar phase with 5 wt % LSZ (a) in the absence of crystallization agent and during in-meso crystallization process for different weight ratios of mesophase to crystallization buffer 4:1 (b), 2:1 (c), and 1:1 ratio (d).
B. Crystallization of Lysozyme Starting from the Lamellar (Lr) Phase. The crystallization process requires a large number
of water molecules in order to occur; for example, the tetragonal LSZ crystal was found to contain 33.5 wt % water.37 Therefore it is reasonable to assume that the liquid crystalline phase with the least amount of water, i.e., LR containing 10 wt % water and 90 wt % glycerol monolinoleate (GML), would provide the harshest conditions for crystallization to occur. Positively charged lysozyme molecules (pH 4.5) with a hydrodynamic diameter of 36 Å, as measured by dynamic light scattering, were encapsulated within the bulk lamellar phase characterized by a lattice parameter of 36 Å and a thickness of water layers of 3.7 Å (at a concentration of 50 mg/mL which is corresponding to 5 wt % of the water, in the absence of crystallization agent) and the effect on the structure of the mesophase was monitored for a period of 30 days (Figure 2a). Due to the mismatch between the size of the protein and the water layers, a high penetration of the protein within the lipid polar heads interfacial region is expected. Structural changes, possibly corresponding to the process of protein aggregation within the mesophase were already noticeable after 1 day of incubation, by the appearance of a new peak at a relatively low Q value of approximately 0.14 Å1, corresponding in real space to 4.5 nm. After 2 days and beyond, two new peaks were observed at lower Q values of 0.11 and 0.05 Å1, respectively, corresponding to real space features of 5.7 and 12.6 nm (all peaks are indicated by arrows in Figure 2a). It is well-known that increasing the amount of water in the system may lead to order-to-order transitions in the lyotropic liquid crystal mesophases.9 Therefore, it was opted to gradually increase the amount of water in the system by starting with a fixed amount of the water inside the mesophase and adding larger quantities of crystallization buffer (4:1, 2:1, and 1:1 weight ratios of mesophase to crystallization buffer, as explained in the experimental part). Addition of crystallization buffer to the lamellar LLC mesophase in 4:1 weight ratio (mesophase:crystallization buffer) increased the total amount of water to 28 wt % and led to a rapid (after 1 day incubation) order-to-order transition from the lamellar phase to the gyroid Ia3d cubic phase. The structural symmetry of the gyroid Ia3d cubic √ phase√is clearly √ depicted by the diffraction peaks in the ratio of 6, 8, 14,
√
√ √ 16, 20, and 22 and in agreement with the binary phase diagram (ref 9, Figure 2b). Decreasing the mesophase:crystallization buffer ratio to 2:1 (that is increasing the total amount of water to 40 wt %), led to a gradual transformation from lamellar into a mixture of gyroid Ia3d and diamond Pn3m cubic, after 1 day of incubation, based on the coexistence of peaks having ratios of spacing specific to the two phases (Figure 2c). Only after approximately two weeks of incubation the system reached the equilibrium state and transformed to the double diamond the diffraction √ √ Pn3m √ cubic √ √phase exhibiting √ peak ratios of 2, 3, 4, 6, 8, and 9, as expected.9 A further decrease in the ratio of mesophase to crystallization buffer to 1:1 (increasing the total water amount to 55 wt %), allows a transformation after only 1 day of incubation, from lamellar into double diamond Pn3m cubic phase coexisting with excess water (Figure 2d). Cross polarized optical microscopy was used to monitor crystal growth in the different conditions explored. At low total water content (28 wt %) in the system with 4:1 ratio no protein crystals could be observed even after two months, inferring that either the water content or the diffusion of protein in these conditions are insufficient for crystallization to occur (Figure 3a). However, once the total amount of water was raised to 40 wt % (2:1 ratio) where a mixture of Ia3d and Pn3m LLC mesophases were formed at equilibrium, very thin needle-like crystal clusters were observed after 34 days (Figure 3b). Moreover, when Pn3m coexists with excess water (55 wt % water, 1:1 ratio) thick cylindrical crystal clusters were observed after 1012 days, coexisting with monoclinic crystals (P21 space group), at the interface between the liquid crystalline media and the aqueous phase (crystallization buffer solution) (Figure 3c). The different location, crystal morphology and time-span of the crystallization process suggest that there are two distinct mechanisms for the crystallization process, according to the total amount of water available in the system and the corresponding symmetry of the LLC mesophase. Because isolation of single crystals from the liquid crystalline phase turned out to be a challenging process, WAXS measurements were performed directly on the mesophases containing the protein crystals (Figure 3d). These measurements reinforce the findings based on optical microscopy of the absence of 6420
dx.doi.org/10.1021/la200710p |Langmuir 2011, 27, 6418–6425
Langmuir
ARTICLE
Figure 3. Cross polarized optical microscopy images of LSZ crystals obtained for the different weight ratios of mesophase to crystallization buffer (a) 4:1 ratio, no crystals were observed, (b) 2:1 ratio, thin needle-like crystals clustering together, (c) 1:1 ratio, thick cylindrical rods crystal clusters in coexistence with monoclinic crystals. (d) WAXS spectra for the three tested conditions.
Figure 4. 1D SAXS spectra of scattered intensities versus scattering vector of the hexagonal phase with 5 wt % LSZ: (a) in the absence of crystallization agent and during in-meso crystallization for different weight ratios of mesophase to crystallization buffer: 4:1 (b), 2:1 (c), and 1:1 ratio (d).
ordered structures in the case of low water content (4:1) and the presence of highly ordered structures for the high water content (1:1), as revealed by a sharp peak at approximately 1.4 Å1. We note here that the sharp peak in the WAXS region corresponds to a real space feature of 4.5 Å, and as such it only reflects order inside individual LSZ proteins, as a result of the new state of the protein within the crystals. The thin needles that appeared in the 2:1 weight ratio were probably too brittle to withstand the transfer from the crystallization plate to the measuring unit cell. C. Crystallization of Lysozyme within the Inverted Hexagonal (HII) Phase. A 5 wt % LSZ in aqueous solution was used to form HII phase composed of the glycerol monolinoleate/ linoleic Acid/aqueous solution, containing a total amount of 25 wt % of water and characterized by a lattice parameter of
61.6 Å and a diameter of the water channel of 32 Å. Once again, protein confinement in LLC mesophase was tested over 30 days of incubation using SAXS. A decrease of peak intensities and a broadening of the peaks was observed in the first three days of incubation, followed after seven days, by the appearance of a peak at 0.05 Å1 (compare the lamellar phase case, Figure 2bd), corresponding to a spatial correlation of 12.6 nm and associated with the process of protein aggregation (Figure 4a). The crystallization process under different mixing ratios of LLC mesophase to crystallization buffer solution (4:1, 2:1, and 1:1, as in the lamellar case) had a similar effect on the mesophase structure, regardless of the available water content, as at all the weight ratios considered, the inverted hexagonal phase can either be found as a bulk phase or coexisting with excess of water, 6421
dx.doi.org/10.1021/la200710p |Langmuir 2011, 27, 6418–6425
Langmuir
ARTICLE
Figure 5. Cross polarized optical microscopy images of the different types of protein crystals obtained for the tested mixtures of mesophase and crystallization buffer. The case of (a) 4:1 ratio, no crystals can be observed due to the high birefringence of the bulk phase; (b) 2:1 ratio, “sea-urchin” clusters of very thin needles; (c) 1:1 ratio, thick “needle-like” cylindrical crystals. (d) The corresponding 1D WAXS spectra of the liquid crystalline phases containing the different crystal morphologies.
differently from the lamellar case, where addition of buffer caused phase transformations. A broadening of peak intensities and the rapid (after 1 day) appearance of the new peak in the low Q region summarizes the structural changes that occur within the hexagonal phase (Figure 4bd). No crystals could be observed by polarized optical microscopy, during the crystallization process in the presence of the lowest amount of water (total 40 wt %, Figure 5a) although the strong peak in the WAXS profile suggests presence of crystalline species (Figure 5d). Increasing the excess water content in the reverse hexagonal LLC system promoted the appearance of different crystal morphologies that can be imaged at the interface between the crystalline mesophase and the aqueous phase (crystallization buffer solution). It the case of 2:1 weight ratio of hexagonal phase to crystallization buffer “sea-urchin” type of very thin needle clusters appeared after 34 days (Figure 5b), whereas at higher water content (62.5 wt %, 1:1 weight ratio) single, thick, cylindrical needle-like crystals were observed after 1012 days of incubation (Figure 5c). Wide-angle X-ray scattering patterns were recorded for all the three ratios studied. The sharp peak observed for all three ratios provides further evidence for the crystallization of LSZ within the bulk hexagonal phase (Figure 5d), whereas variations in the WAXS profile suggest a rich polymorphism of these crystals depending on the ratio used. D. Crystallization of Lysozyme within the Double Diamond (Pn3m) Cubic Phase. A 5 wt % LSZ solution was used to form the double diamond, Pn3m cubic mesophase (with a lattice parameter of 95 Å and a radius of the water channel of 39.3 Å) and the crystallization process was studied under different mixing ratios of LLC mesophase to crystallization buffer solution (4:1 and 2:1, corresponding to 40 wt % and 56.7 wt % water solution, that is bulk Pn3m and Pn3m þ excess water, respectively), with respect to its influence on the cubic symmetry (Figure 6ac). Addition of crystallization buffer to the cubic LLC mesophase to reach 2:1 ratio increased the total amount of water to 56.7 wt %, ensuring a
sufficient excess of water in the system, making larger buffer amounts unnecessary for the Pn3m study. Following the protein solubilization in the bulk Pn3m prior to buffer addition, two additional peaks appeared at Q values of approximately 0.8 Å1, after 2 days, and at 0.05 Å1 after 8 days (Figure 6a). Furthermore, the specific ratios of spacing between the recorded peaks indicated the coexistence of two types of structures, respectively the Ia3d and Pn3m cubic mesophases. When 4:1 LLC:crystallization buffer solution weight ratio was used, an order-to-order transition from the Pn3m to the Ia3d cubic symmetry occurred during the first three days of crystallization (Figure 6b). Additionally, from the third day to the seventh an increase in lattice parameter was detected and followed by slow and gradual order-to-order transition back to its initial Pn3m symmetry (reached after 30 days). These findings provide an insight for the in-meso crystallization mechanism (which will be elaborated in the discussion section). Once the total amount of water further increased to approximately 56.7 wt % (2:1 weight ratio), no major structural changes were noted throughout the crystallization process (Figure 6c). The Pn3m cubic phase in excess with water, maintained the initial double diamond symmetry throughout the entire process and the only noticeable change was the appearance of the ubiquitous peak at a Q value of 0.05 Å1. Optical microscopy revealed two distinct crystal morphologies emerging from the two different tested conditions. At 4:1 ratio (bulk Pn3m), in-meso crystallization, although a lengthy process (first crystals being noticed after approximately four weeks) led to the formation of orthorhombic LSZ crystals belonging to the P212121 space group (Figure 7a), whereas the use of excess water led to the more rapid formation (after 34 days) of tetragonal LSZ crystals (Figure 7b), belonging to the P43212 space group, thus identical to those found in the control experiment in free buffer solution. Wide angle X-ray data were also collected for the two crystal morphologies (Figure 7c), providing further evidence 6422
dx.doi.org/10.1021/la200710p |Langmuir 2011, 27, 6418–6425
Langmuir
ARTICLE
Figure 6. SAXS patterns showing the overall effect in time on the mesophase structure and symmetry in the case of: (a) solubilization of 5 wt % of LSZ, (b) crystallization of LSZ in a 4:1 ratio of mesophase to crystallization buffer, (c) crystallization of LSZ in a 2:1 ratio of mesophase to crystallization buffer.
Figure 7. Optical microscopy images of LSZ crystals, obtained in-meso from Pn3m in bulk and water excess conditions. (a) Orthorhombic crystals obtained from the bulk phase crystallization process (4:1 ratio). (b) Tetragonal crystals obtained from crystallization process (2:1 ratio) in the Pn3m with excess water. (c) WAXS spectra of liquid crystalline cubic phases containing the two morphological types of LSZ crystals.
for differences in the intramolecular contacts/internal structure of LSZ in the two different crystal space groups.
’ DISCUSSION The initial and necessary step in the formation of nuclei is consumption of water molecules by the protein. Therefore water molecules will dehydrate the polar head groups of the lipidic mesophase in order to hydrate the protein. This phenomenon is obviously more easily detectable when limited quantities of water are present (at 4:1 for LR, HII, and Pn3m and at 2:1 for LR and HII systems) compared to the case of excess water. As a result, the time period of crystallization and the structural transitions within the
LLC symmetry are different between these two cases (bulk versus excess water conditions). The outcomes of this study give evidence for two distinct mechanisms of protein crystallization taking place in accordance with the total amount of water available in the mixture. A. Crystallization Mechanism within the Bulk Liquid Crystalline Phases. In case of crystallization within the bulk liquid crystalline phase SAXS data depict a clear trend of the transformations that occur during protein crystallization. As a result of the limited amount of water, a dehydration of all three initial LLC mesophases can be seen at the onset of protein crystallization. The initial lamellar phase, which upon addition of the crystallization buffer alone undergoes a fast transition toward the Pn3m symmetry, under these conditions exhibits a much slower and 6423
dx.doi.org/10.1021/la200710p |Langmuir 2011, 27, 6418–6425
Langmuir gradual order-to-order transition toward the double diamond Pn3m cubic phase, preceded by seven days of a coexistence of Ia3d and Pn3m cubic phases (Figure 2c). The process of migration of water molecules to the protein, thereby overall dehydrating the system, is responsible for the long coexistence of Ia3d (lower hydration) and Pn3m (higher hydration). As the crystallization process continues, the protein crystals reach their final morphological state and macro-phase separate from the liquid crystalline phase as revealed by optical microscopy. This in turn allows the remaining water to equilibrate the system to its final structural conformation and then (∼two weeks) a complete transformation to Pn3m can finalize. The dehydrating effect was detected also in the case of reverse hexagonal phase by the decrease in the lattice parameter up to the 14th day: only when protein crystallization was completed, the lattice parameter relaxed to its initial value (day 30), suggesting a full hydration of the polar heads. Finally, when the initial LLC was the Pn3m cubic mesophase, due to water dehydration from the system, again an order-to-order transition toward Ia3d cubic phase was detected during the first seven days. The coexistence of Pn3m and Ia3d cubic phases vanished to yield solely Pn3m only when the protein crystallization process was completed. We shall emphasize here that this systematic dehydration effect is not observed in hanging drop vapor diffusion blank experiments analogous at those reported here but in the absence of the protein (see Supporting Information): thus, this can unambiguously be attributed to the in-meso crystallization of LSZ. Previous research proposed a mechanism for in-cubo crystallization of LSZ based on structural “defects” within the structure of the cubic phase where the protein nuclei might reside.30 This theory is compatible with our findings, especially with respect to the growth of the crystals to macroscopic sizes; nonetheless, the SAXS study presented here, completes the picture with the assessment of the role of water associated with the in-meso protein crystallization in bulk and offers a comprehensive rationale for the transient mesophase order-to-order transitions (or lattice shrinkage) observed during the crystallization process. B. Crystallization Mechanism in the Presence of Excess Water. Addition of a larger amount of water (at 2:1 for Pn3m and at 1:1 for LR and HII systems) led to a different dynamics and mechanism of the protein crystallization. This can be immediately realized from the lack of structural changes occurring within the mesophase, the stability of the lattice parameter, and the shorter time required for protein crystals to be formed. In the case of the lamellar phase addition of an equal amount of water to that of the liquid crystalline phase (1:1 weight ratio) led to a fast transition to the cubic Pn3m phase (after 1 day), after which no noticeable changes could be observed. At the same time protein crystals were detected at the interface of the aqueous and lipid phase rather than within the bulk mesophase, which clearly indicates that a different mechanism took place. The hexagonal and cubic phases showed a similar behavior with respect to LLC structural changes and location of protein crystals. Hence, it can reasonably be assumed that after the initial change in structural symmetry caused by the addition of excess water the protein molecules are allowed to diffuse outside the water channels into the interface bulk excess water, and crystal formation can take place at the mesophase-excess water interface. The role of mesophase structure may then play a role in the crystallization process essentially in controlling different diffusion rates for the protein to escape the LLC mesophases. Our results are consistent with this picture, protein crystals being noticed after 34 days in the cubic phase (2:1 ratio), 1012 days
ARTICLE
from the initial lamellar phase (1:1 ratio), and 1415 days in the inverse hexagonal phase (1:1 ratio). C. Possible Effect of Mesophase Symmetry on Protein Crystal Morphology. Although identical conditions were used in all three LLC mesophases, with respect to protein concentration, choice of crystallization buffer and mixing ratios, different crystal morphologies were obtained. Crystal morphology appears to change according not only to the structural symmetry of the mesophase but also with the total amount of water available in the system. Within the bulk mesophases, clusters of very thin needle-like protein crystals emerged out of the initial lamellar phase, “seaurchin” from the hexagonal phase and large orthorhombic crystals were obtained from the double diamond cubic phase. This is possibly a result of the fact that only the double diamond cubic phase can provide a pseudoisotropic environment for protein crystallization, whereas the other phase are characterized by 1D and 2D anisotropic symmetries, with substantial lower diffusion rates.38 Similarly, different types of crystals were formed in excess water, according to the initial mesophase structure that incorporated the protein molecules. If the coexistence of monoclinic crystals and clusters of thick, cylindrical rods was observed for the 1:1 ratio of lamellar phase to crystallization buffer, the crystal morphology changed dramatically in the case of the hexagonal and cubic phase. Single thick rod-like crystals were seen at the hexagonal phase-excess water interface, whereas the cubic phase led to the formation of very large tetragonal crystals. This might again be a very direct consequence of the diffusion rates following the trend DPn3m . DLAM g DHEX.38 From the point of view of the total amount of water added to the same structural type of mesophase, obvious differences in crystal morphology were noted in all the systems tested, the two distinct types of large crystals obtained in the cubic phase (orthorhombic in bulk and tetragonal in excess water) being the most relevant examples. The orthorhombic f tetragonal polymorphic change observed in this case is understood to be the result of confined versus free protein crystallization. Furthermore, in the initial lamellar and hexagonal phase, even when enough excess crystallization buffer is added, the diffusion process is too slow to allow the crystallization of LSZ into the tetragonal form expected for free buffer crystallization (Figure 1), leading to the rich polymorphic crystal examples observed in this work. With all this in mind, it seems realistic to conclude that the topology of the mesophase, the location of the mesophase in the phase diagram (bulk vs excess water) and the inherent diffusion processes controlled by different mesophase symmetries all concur to determine the protein crystal space group and their macroscopic growth. The present work brings further insights into the process of in-meso protein crystal formation and while it settles for the first time a general frame for engineering the protein crystals at the molecular and macroscopic level, it also highlights once more the subtle balance of the many parameters controlling this delicate, yet fascinating crystallization pathway.
’ ASSOCIATED CONTENT
bS
Supporting Information. Additional supporting figures. This material is available free of charge via the Internet at http:// pubs.acs.org.
6424
dx.doi.org/10.1021/la200710p |Langmuir 2011, 27, 6418–6425
Langmuir
’ AUTHOR INFORMATION Corresponding Author
*E-mail: raff
[email protected]. Tel (office): þ41 446329140. Tel (secretary): þ41 446323284. Web site: http://www.ilw.agrl.ethz.ch/lwm/.
’ REFERENCES (1) Mezzenga, R.; Schurtenberger, P.; Burbidge, P.; Michel, M. Understanding foods as soft materials. Nat. Mater. 2005, 4, 729–740. (2) Mezzenga, R.; Lee, W. B.; Fredrickson, G. H. Design of liquidcrystalline foods via field theoretic computer simulations. Trends Food Sci. Technol. 2006, 17, 220–226. (3) Libster, D.; Aserin, A.; Wachtel, E.; Shoham, G.; Garti, N. An HII liquid crystal-based delivery system for cyclosporine A: Physical characterization. J. Colloid Interface Sci. 2007, 308, 514–524. (4) Feher, A.; Urban, E.; Eros, I.; Szabo-Revesz, P.; Csanyi, E. Lyotropic liquid crystal preconcentrates for the treatment of periodontal disease. Int. J. Pharm. 2008, 358, 23–26. (5) Cohen-Avrahami, M.; Aserin, A.; Garti, N. HII mesophase and peptide cell penetrating enhancers for improved transdermal delivery of sodium diclofenac. Colloids Surf. B 2010, 77, 131–138. (6) Caffrey, M.; Cherezov, V. Crystallizing membrane proteins using lipidic mesophases. Nat. Protocols 2009, 4, 706–731. (7) Misquitta, L. V.; Misquitta, Y.; Cherezov, V.; Slattery, O.; Mohan, J. M.; Hart, D.; Zhalnina, M.; Cramer, W. A.; Caffrey, M. Membrane protein crystallization in lipidic mesophases with tailored bilayers. Structure 2004, 12, 2113–2124. (8) Larsson, K. Aqueous dispersions of cubic lipid-water phases. Curr. Opin. Colloid Interface Sci. 2000, 5, 64–69. (9) Mezzenga, R.; Grigorov, M.; Zhang, Z.; Servais, C.; Sagalowicz, L.; Romoscanu, A. I.; Khanna, V.; Meyer, C. Polysaccharide-induced order-to-order transitions in lyotropic liquid crystals. Langmuir 2005, 21, 6165–6169. (10) Mezzenga, R.; Meyer, C.; Servais, C.; Romoscanu, A. I.; Sagalowicz, L.; Hayward, R. C. Shear rheology of lyotropic liquid crystals. Langmuir 2005, 21, 3322–3333. (11) Qiu, H.; Caffrey, M. The phase diagram of the monoolein/ water system: metastability and equilibrium aspects. Biomaterials 2000, 21, 223–234. (12) Clogston, J.; Caffrey, M. Controlling release from the lipidic cubic phase. Amino acids, proteins and nucleic acids. J. Controlled Release 2005, 107, 97–111. (13) Shah, J. C.; Sadhale, Y.; Chilukuri, D. M. Cubic phase gels as drug delivery systems. Adv. Drug Delivery Rev. 2001, 47, 229–250. (14) Mohammady, S. Z.; Pouzot, M.; Mezzenga, R. Oleoylethanolamide-based Lyotropic Liquid Crystals as vehicles for delivery of amino acids in aqueous environment. Biophys. J. 2009, 96, 1537–1546. (15) Landau, E. M.; Rosenbusch, J. P. Lipidic cubic phases: A novel concept for the crystallization of membrane proteins. Proc. Natl. Acad. Sci. U.S.A. 1996, 93, 14532–14535. (16) Cherezov, V.; Yamashita, E.; Liu, W.; Zhalnina, M.; Cramer, W. A.; Caffrey, M. In Meso Structure of the Cobalamin Transporter BtuB, at 1.95 Å resolution. J. Mol. Biol. 2006, 364, 716–734. (17) Portmann, M.; Landau, E. M.; Luisi, P. L. Spectroscopic and Rheological studies of Enzymes in rigid Lipidic matrices: The case of R-Chymotrypsin in a Lysolecithin/water Cubic phase. J. Phys. Chem. 1991, 95, 8437–8440. (18) Kraineva, J.; Nicolini, C.; Thiyagarajan, P.; Kondrashkina, E.; Winter, R. Incorporation of R-chymotrypsin into the 3D channels of bicontinous cubic lipid mesophases. Biochim. Biophys. Acta 2006, 1764, 424–433. (19) Razumas, V.; Talaikyte, Z.; Barauskas, J.; Larsson, K.; Miezis, Y.; Nylander, T. Effects of distearoylphosphatidyglycerol and lysozyme on the structure of the monoolein-water cubic phase: X-ray diffraction and Raman studies. Chem. Phys. Lipids 1996, 84, 123–138.
ARTICLE
(20) Tanaka, S. Structural transitions of the mono-olein bicontinous cubic phase induced by inclusion of protein lysozyme solutions. Phys. Rev. E 2006, 73, 061510. (21) Mishraki, T.; Libster, D.; Aserin, A.; Garti, N. Temperature dependent behaviour of lysozyme within the reverse hexagonal mesophases (HII). Colloids Surf. B 2010, 75, 391–397. (22) Razumas, V.; Larsson, K.; Miezis, Y.; Nylander, T. A cubic Monoolein-Cytochrome c-Water phase: X-ray diffraction, FT-IR, Differential Scanning Calorimetric and Electrochemical studies. J. Phys. Chem. 1996, 100, 11766–11774. (23) Leslie, S. B.; Puvvada, S.; Ratna, B. R.; Rudolph, A. S. Encapsulation of hemoglobin in a bicontinuous cubic phase lipid. Biochim. Biophys. Acta 1996, 1285, 246–254. (24) Sadhale, Y.; Shah, J. C. Stabilization of insulin against agitationinduced aggregation by the GMO cubic phase gel. Int. J. Pharm. 1999, 191, 51–64. (25) Conn, C. E.; Darmanin, C.; Sagnella, S. M.; Mulet, X.; Greaves, T. L.; Varghese, J. N.; Drummond, C. J. Incorporation of the dopamine D2L receptor amd bacteriorhodopsin within bicontinous cubic lipid phases. 1. Relevance to in meso crystallization of integral membrane proteins in monoolein systems. Soft Matter 2010, 6, 4828–4837. (26) Conn, C. E.; Darmanin, C.; Sagnella, S. M.; Mulet, X.; Greaves, T. L.; Varghese, J. N.; Drummond, C. J. Incorporation of the dopamine D2L receptor amd bacteriorhodopsin within bicontinous cubic lipid phases. 2. Relevance to in meso crystallization of integral membrane proteins in novel lipid systems. Soft Matter 2010, 6, 4838–4846. (27) Rummel, G.; Hardmeyer, A.; Widmer, C.; Chiu, M. L.; Nollert, P.; Locher, K. P.; Pedruzzi, I.; Landau, E. M.; Rosenbusch, J. P. Lipidic Cubic Phases: New matrices for the three-dimensional crystallization of membrane proteins. J. Struct. Biol. 1997, 121, 82–91. (28) Landau, E. M.; Rummel, G.; Cowan-Jacob, S. W.; Rosenbusch, J. P. Crystallization of a polar protein and small molecules from the aqueous compartment of lipidic cubic phases. J. Phys. Chem. B 1997, 101, 1935–1937. (29) Caffrey, M. Membrane protein crystallization. J. Struct. Biol. 2003, 142, 108–132. (30) Tanaka, S.; Egelhaaf, S. U.; Poon, W. C. K. Crystallization of a globular protein in lipid cubic phase. Phys. Rev. Lett. 2004, 92, 128102. (31) Mishraki, T.; Libster, D.; Aserin, A.; Garti, N. Lysozyme entrapped within reverse hexagonal mesophases: Physical properties and structural behaviour. Colloids Surf., B 2010, 75, 47–56. (32) Jolles, J.; Jolles, P. Lysozymes esterase activity. FEBS 1983, 162, 120–122. (33) Velicelebi, G.; Sturtevant, J. M. Thermodynamics of the denaturation of Lysozyme in alcohol-water mixtures. Biochemistry 1979, 18, 1180–1186. (34) Mine, S.; Tate, S. I.; Ueda, T.; Kainosho, M.; Imoto, T. Analysis of the relationship between enzyme activity and its internal motion using Nuclear Magnetic Resonance: 15N Relaxation studies of wild-type and mutant lysozyme. J. Mol. Biol. 1999, 286, 1547–1565. (35) Hogle, J.; Rao, S. T.; Mallikarjunan, M.; Beddell, C.; McMullan, R. K.; Sundaralingam, M. Studies of monoclinic Hen Egg White Lysozyme. I. Structure solution at 4 Å resolution and molecular-packing comparisons with tetragonal and triclinic lysozymes. Acta Cryst. B 1981, 37, 591–597. (36) Artymiuk, P. J.; Blake, C. C. F.; Rice, D. W.; Wilson, K. S. The structures of Monoclinic and Orthorhombic forms of Hen Egg-White Lysozyme at 6Å resolution. Acta Cryst. B 1981, 38, 778–783. (37) Phillips, D. C. The hen egg-white lysozyme molecule. Proc. Natl. Acad. Sci. U.S.A. 1967, 57, 483–495. (38) Sagalowicz, L.; Mezzenga, R.; Leser, M. E. Investigating reversed liquid crystalline mesophases. Curr. Opin. Colloid Interface Sci. 2006, 11, 224–229.
6425
dx.doi.org/10.1021/la200710p |Langmuir 2011, 27, 6418–6425
