Surface treatment by oxidizing the plates can alter the response of protein crystallization
This report describes the modification of crystallization plates by simply oxidizing the surface of the protein wells. The oxidized crystallization plates were tested in standard protein crystallization screening and reproducibility studies. The results showed that the protein wells of the treated plates were smoother and more optically transparent than those of the untreated plates, and more importantly, protein crystallization was significantly promoted after the oxidation treatment. Because there is no change to the routine screening protocol, this method is simple and easy to apply in protein crystallization.
Protein crystallization is important because it provides protein crystals for the structural determination of protein molecules using X-ray crystallography. However, it is not always easy to obtain suitable protein crystals for X-ray diffraction because of the difficulties in crystallizing the proteins. Over the past century, many methods have been developed for promoting protein crystallization. The progress in the field of protein crystallogenesis has been especially rapid in recent decades. New technologies and the use of automated facilities in the field of protein crystallization have greatly increased the likelihood of obtaining protein crystals. For examples, utilization of sparse matrix screens (Jancarik & Kim, 1991), new screening kits (Cudney et al., 1994; Brzozowski & Walton, 2001; Newman et al., 2005; McPherson & Cudney, 2006; Gorrec, 2009), microfluidics (Hansen et al., 2002; Zheng et al., 2004), automation (Mueller et al., 2001; Santarsiero et al., 2002; Sulzenbacher et al., 2002; Bard et al., 2004; Stock et al., 2005) etc. have contributed a lot in this field. However, the overall success rate remains at approximately 10%, without obvious improvements in the process from purified soluble proteins to structure determination (Kim et al., 2008). Therefore, the rapidly increasing number of solved structures in the Protein Data Bank (PDB; Berman et al., 2000) should be attributed to the much larger number of trials that are possible owing to factors such as new (automated) technologies, more efficient software, and many more researchers and targets, among other factors.
Why does the overall success rate remain low despite the progress in protein crystallogenesis? One reason might be that many methods that are useful for enhancing protein crystallization have not been fully adopted in routine protein crystallization screening and optimization experiments. A typical example is the utilization of heterogeneous nucleants. In the most widely used method, i.e., the vapor diffusion method, the crystallization of proteins occurs primarily through heterogeneous nucleation because a solid surface can provide a heterogeneous nucleation site for crystallization. However, the solid surface, for example, the surface of the sitting pits in sitting-drop crystallization plates, has never been optimized for heterogeneous nucleation. Instead, other types of materials have been proposed and tested as heterogeneous nucleants. Now it is well known, and also widely accepted, that heterogeneous nucleants can successfully promote protein crystallization using various substrates and/or nucleants. For example, substrates that act as nucleation sites, such as mineral substrates (McPherson & Shlichta, 1988; Paxton et al., 1999), modified substrates (Fermani et al., 2001; Liu et al., 2007), lipid layers (Hemming et al., 1995) and molecularly imprinted polymers (MIPs; Saridakis et al., 2011; Reddy et al., 2012; Saridakis & Chayen, 2013), have been investigated and effectively improved crystallization. A variety of nucleants like porous materials (Chayen et al., 2001, 2006; Kertis et al., 2012), polystyrene nanospheres (Kallio et al., 2009), dried seaweed (Thakur et al., 2007) and hair (Bergfors, 2003; D'Arcy et al., 2003; Georgieva et al., 2007; Thakur et al., 2007) have also been shown to successfully induce nucleation. When heterogeneous nucleants are used, it is usually necessary to add the nucleants into each of the crystallization droplets. However, because it is troublesome to do so, many laboratories prefer to use the standard crystallization plates directly without adding any nucleants in their routine crystallization experiments. Thus, to avoid an extra step of adding nucleants, directly treating the surface of crystallization plates may help to promote crystallization.
On the basis of the above idea, we attempted to modify the surface of crystallization plates. Commercially available crystallization plates were used for the modification test. These plates are made of polystyrene (Jacquamet et al., 2004), which can be oxidized using a strong oxidizing acid solution or a corrosive organic solution. The oxidation treatment may change the physical and chemical characteristics of the surface so that the hydrophilicity and the adsorption capacity of the surface will be modified (Martins et al., 2003), and thus, crystallization on such a surface may be affected. Therefore, we applied the oxidation treatment to the crystallization plates. After the treatment, the plates were utilized in standard sitting-drop vapor diffusion protein crystallization experiments. Both crystallization screening experiments and reproducibility experiments were used to verify the effects of the oxidation treatment on the final crystallization results. On the basis of the experimental results, we discuss the applicability of this treatment method to routine protein crystallization experiments.
Potassium dichromate (K2Cr2O7) was purchased from Tianjin Chemical Reagent Factory (China) and concentrated sulfuric acid (98%) from Xi'an Sanpu Chemical Reagent Limited Company (China). The crystallization plates (96-2-well sitting-drop Intelli-Plates, HR3-143) were from Hampton Research Company (USA).
The oxidation process was as follows: (1) Preparation of the oxidation solution: potassium dichromate, concentrated sulfuric acid and distilled water were mixed at a mass ratio of K2Cr2O7:H2SO4:H2O = 7:8:100. (2) Oxidation treatment: the crystallization plates were immersed in the oxidation solution in a beaker and kept at 330 K for 1.5 h in an oven (DHG-9146A, Shanghai Jingwang Laboratory Equipment Company Ltd, China). (3) Cleaning of the plates: the crystallization plates were put into an ultrasonic cleaner (5510, Branson, USA) with distilled water and treated for 30 min. (4) Drying of the plates: the plates were dried in an electric oven at 333 K for 4 h. After drying, the crystallization plates (oxidized crystallization plates, OCPs) were ready for use.
Atomic force microscopy (AFM) and infrared spectrum analysis were used to characterize the surface properties of the crystallization plates before and after the surface oxidation.
The atomic force microscope (BioScope, Bruker, USA) scanning conditions were as follows: probe: NSC11/AIBS from NT-MDT (Russia); scanning area: 10 × 10 µm; and scanning mode: tapping mode. The surface roughness was calculated from the scanning data.
2.2.2. Infrared spectrum analysis and X-ray photoelectron spectroscopy
The properties of the plates before and after oxidation treatment were characterized using Fourier transform infrared (FTIR) spectroscopy (TENSOR27 FTIR spectrometer, Bruker, Germany).
The ionic species on the surface of the crystallization plates were analyzed using X-ray photoelectron spectroscopy (Thermo Fisher Scientific, USA) with monochromatic Al Kα radiation (150 W, 15 kV, 1486.6 eV). The vacuum in the spectrometer was set to 10−7 Torr.
The contact angle between the crystallization plate and water can provide information about the wettability between the crystallization surface and the crystallization (aqueous) solution. We measured the contact angle between water and the crystallization plates before and after the oxidation treatment. The measurement apparatus, which was designed and manufactured by our group, was composed of a temperature controller [a copper water jacket with flowing water from a circulator (Polyscience 9712, Polyscience, USA)] and a CCD camera (Ultra Micro-CCD UN43H, ELMO Company, Japan). The sample (1 × 1 cm) was a flat piece cut from a crystallization plate. The temperature was 293 K. During the measurement, water was slowly ejected from an accurately controlled syringe onto the horizontal surface of the sample until the volume of the droplet reached 100 µl. The volume was kept constant for approximately 1 min and then slowly decreased using the same syringe. The time course of the side view of the droplet was recorded using the CCD camera. A sequence of images was then captured from the video and analyzed using a program written by our group to obtain the contact angle.
In this study, 12 commercially available proteins were used without further purification. Proteinase K (P6556), thaumatin (T7638), catalase (C40), concanavalin A (L7647), α-chymotrypsinogen A I (C4879), ribonuclease A III (R5125), ribonuclease A XII (R5500), subtilisin A VIII (P5380), papain (P3152) and lactalbumin (K7015) were purchased from Sigma–Aldrich Company. Hen egg-white lysozyme (HEWL, Lot No. 100940) was purchased from Seikagaku Kogyo Company. Glucose isomerase (HR7-100) was purchased from Hampton Research Company.
The concentrations of the proteins in the crystallization solutions (after mixing) were 3.5 mg ml−1 for glucose isomerase and 10 mg ml−1 for the remaining proteins. The buffers used for dissolving the proteins were 0.2 M sodium acetate (pH 4.6) for lysozyme and 0.25 M HEPES sodium buffer (pH 7.00) for the remaining proteins. The Index screening kit was utilized (HR2-144, Hampton Research Company, USA).
Following the standard protocol for the sitting-drop vapor diffusion method (Snook et al., 2000), we set up the crystallization trials by mixing the protein solution with the crystallization reagents from the crystallization screening kit at a volume ratio of 1:1 using a crystallization robot (Screenmaker 96+8, Innovadyne Technologies Inc., USA). Then, the crystallization droplets with their corresponding crystallization reagents (80 µl each) were sealed using Crystal Clear Sealing Tape (HR4-506, Hampton Research). The crystallization samples were placed into a sealed chamber whose temperature was controlled by flowing water from a circulator (PolyScience 9712, PolyScience, USA). The incubation temperature was controlled at 293 K.
The crystallization droplets were examined after 2 and 6 d of incubation using an automated image reader (Xtal Finder, XtalQuest Inc., China) equipped with a UV light source (X-talight 100, Molecular Dimensions, Germany). A polarization microscope (Olympus PLAPO, Japan) equipped with a digital camera (Cannon EOS 550D, Japan) was used to aid the visualization of the crystals at a higher resolution. Crystallization hits (here a `hit' is defined as a crystallization condition that yielded protein crystals observable by the automated image reader) were identified using the captured images.
To verify the crystallization results from the screening tests, we conducted reproducibility tests (Yin et al., 2008). Using the 96-well crystallization plates, we set up 96 repeated conditions. To ensure that the concentration conditions in each droplet were identical, the protein solution and the reservoir solution were premixed at a volume ratio of 1:1 and then dispensed into each protein well of the 96-well sitting-drop crystallization plates (before and after oxidation treatment) using a crystallization robot. The final concentrations in the droplets were as follows: lysozyme: 15 and 20 mg ml−1; buffer: 0.2 M sodium acetate pH 4.6; and NaCl: 30 mg ml−1. The volume of the droplets was 2 µl, and the volume of the reservoir was 80 µl. All other procedures were the same as described for the screening study.
The crystallization plates were kept in a temperature-controlled chamber at 293 K. After incubation for 2 and 6 d, images of the droplets were captured to examine the crystallization reproducibility for the crystallization plates before and after oxidation.
To validate the applicability of the OCP in protein crystallization, X-ray diffraction of crystals grown in the OCP and the control was carried out. Two commercially available proteins (HEWL and proteinase K) were used as the model proteins. The crystallization conditions were as follows: protein concentration: 20 mg ml−1 in 0.25 M HEPES sodium buffer (pH 7.00); the reservoir solutions were Index No. 94 and No. 5 for lysozyme and proteinase K, respectively. The volume of the crystallization droplets was 2 µl (1 µl protein solution + 1 µl reservoir solution), and the volume of the reservoir was 90 µl. The crystallization plates were kept in the temperature-controlled chamber at 293 K for 4 d. After the crystals had grown, they were harvested and gradually equilibrated in the reservoir solution containing cryoprotectant (20% glycerol), and then flash-frozen in liquid nitrogen for the subsequent diffraction experiment. Collections of diffraction data were carried out using an X-ray diffractometer (Marresearch, Marux, Germany). The diffraction data were then indexed and reduced with HKL-2000 (Otwinowski & Minor, 1997).
3.1. Effect of oxidation treatment of the crystallization plates on the number of crystallization screening hits
In the screening experiments, 12 proteins were tested using the standard sitting-drop vapor diffusion method. Table 1 shows the results of the number of crystallization hits for the control plates (without oxidation treatment of the crystallization plates) and the OCPs. `+' and `−' represent the number of new crystallization conditions that were not found in the control group and the number of crystallization conditions that were found in the control group but were not present in the OCP group, respectively. From the table, we can observe that there was a clear increase in the number of crystallization screening hits for the plates after oxidation treatment. Overall, there were 26 more crystallization hits when using the OCPs than when using untreated plates after 2 d of incubation, and there were 46 more after 6 d of incubation. The difference was especially obvious for ribonuclease A XII, ribonuclease A III, lactalbumin and subtilisin A VIII, for which no crystals were found in the control plates but crystals were found in the treated plates. The difference was also obvious in the cases of lysozyme and α-chymotrypsinogen, for which 12 and ten more crystallization hits, respectively, were found in the OCP group than in the control group.
To show the effect of the oxidation treatment more directly, we performed a statistical comparison using the data in Table 1. Fig. 1 shows the averaged normalized hits based on the hits in the OCP group at 6 d. From this figure, we can see that at 2 d the number of normalized hits was 1 in the OCP group and 0.68 in the control group. At 6 d, the number of normalized hits was 1 in the OCP group and 0.50 in the control group. These results indicate that, compared with the crystallization in the control group, the crystallization in the OCP group was 46% greater at 2 d and 98% greater at 6 d. An independent-samples t-test was used to evaluate the significance of the improvement. The data were significant both at 2 d and at 6 d.
3.2. Effect of oxidation treatment of the crystallization plates on the crystallization success rate in the reproducibility experiment
In the reproducibility experiment, the crystallization of lysozyme at two different concentrations was performed. Each condition was repeated 96 times (i.e. 96 crystallization droplets were dispensed using the crystallization robot from the same premixed crystallization solution into the protein wells of the sitting-drop crystallization plates). We utilized the `crystallization success rate' to examine the effect of the oxidation treatment of the crystallization plates. The crystallization success rate is defined as the ratio of the number of crystallized droplets to the total number of crystallization droplets. The numbers of crystals obtained with and without the oxidation treatment of the plates were also counted for comparison.
Fig. 2 shows the results obtained from the reproducibility experiments. The effect of oxidation treatment on the crystallization success rate was obvious for both concentration levels (Figs. 2a and 2c). In addition, the number of crystals was higher in all cases (Figs. 2b and 2d) for the oxidation treatment compared with the control. Furthermore, the differences in the crystallization success rate and number of crystals were larger at the earlier time (2 d) than at the later time (6 d), indicating that the nucleation time was shorter when the crystallization plates were oxidized. This phenomenon is very important in understanding the effect of oxidation on protein crystallization. These results also show that, at the lower concentration, the differences in both the crystallization success rate and the number of crystals were larger than at the higher concentration. [To understand this effect, imagine the case when the concentration is very high: the crystallization success rate will be the same (= 1) in both cases, with and without oxidation treatment, because lysozyme can always crystallize at high concentrations. At a very low concentration (for example, close to 0 mg ml−1), the crystallization success rate will also be the same (= 0) in both cases.]
We also examined and compared the crystal morphologies after the experiments. The crystals grown in the OCPs were more often larger with more well defined faces than the crystals grown in the control plates. Fig. 3 shows some typical examples [(a)–(d) crystals grown in OCPs and (a′)–(d′) crystals grown in control plates]. The crystals in the OCPs looked better than those in the control plates, and the number of crystals was usually smaller in the OCPs.
Crystal size was used to compare the morphologies quantitatively. There were 207 crystallization conditions that yielded crystals in both the OCP and the control group. For these crystallization conditions, we counted the numbers of conditions in the OCP group in which the crystals were larger, smaller than and comparable to those grown in the control plates. Fig. 4 shows the percentages for the three cases. It can be observed that 45% of the crystals grown in the OCPs were larger than those grown in the control plates, whereas only 21% of the crystals grown in the control plates were larger than those grown in the OCPs.
To check if the crystals grown in the OCP diffract or not, we conducted diffraction analysis of the crystals grown in both the treated and the untreated crystallization plates. Judging from the handling experience, we did not find any clear difference in crystal harvesting, i.e. it is not more difficult to harvest crystals from the OCP than from the control. The diffraction experiment can also be done smoothly. Table 2 summarizes the resulting diffraction data for lysozyme and proteinase K. The results showed that the crystals grown in the OCP could diffract well (the tested crystals showed better diffraction quality than the control, as can be seen from the data of mosaicity, Rmerge and resolution), validating that the oxidized crystallization plates are applicable in protein crystallization.
The effect of oxidation of the crystallization plates on crystallization is clearly closely related to the surface modification due to the treatment. Therefore, collecting information on what occurred after the oxidation treatment is useful to understand the phenomena observed in the crystallization experiments. The surface characterization methods used in this study include morphology and roughness analysis, wettability (contact angle) analysis, and FTIR spectroscopy and X-ray photoelectron spectroscopy (XPS).
We scanned the treated and untreated surfaces of the crystallization plates using AFM. Fig. 5 shows a pair of surfaces with and without oxidation treatment. Figs. 5(a) and 5(b) are micrographs of the surfaces, respectively. The surface roughness became smaller after the surface treatment. To present the results in a clearer manner, we list the data for surface roughness for both cases in Table 3.
The data in the table show that there is a clear difference in the surface roughness due to the oxidation treatment. The average roughness was reduced from 17.67 nm in the control group to 4.52 nm in the OCP group. These data indicate that the surface of the treated crystallization plates was smoother than the surface of the untreated plates.
We measured the contact angle between the crystallization plate and a water droplet to examine the wettability of the plate's surface before and after the oxidation treatment. Fig. 6 shows a set of images used to measure the contact angles for plates with (Fig. 6a) and without (Fig. 6b) oxidation treatment. We observed an obvious decrease in the contact angle after the oxidation treatment: the average contact angle before the treatment was 63°, and the average contact angle after the treatment was 46°.
To analyze the functional groups on the surface of the crystallization plates before and after the oxidation treatment, we examined the infrared spectra of the surfaces. Fig. 7 shows the infrared spectra of the surfaces. From the figure, it can be observed that, on the surfaces of both the OCP and the control, there exist phenyl groups (716 and 778 cm−l) and the main component –CH2 of the polystyrene band (1450 cm−1). The stretching modes for –CH2 found in polyethylene (2864 and 2941 cm−1) and –CH, which appeared with a reasonable intensity (at approximately 1250 and 1380 cm−1) (Liang & Krimm, 2003), could also be detected in the spectra. The most evident difference between the two was that there was no peak at 1730 cm−1 for the OCPs, indicating that the –CH group was altered on the OCP surface.
The oxidation treatment might have left some residues which may affect the crystallization. To find out if there are any residues on the plate, analytical tools of high sensitivity must be used. XPS is very sensitive to trace elements on surfaces. If there are any different residues left on the OCP, we could find them using this technique. Thus, to obtain more information about the chemical composition of the surfaces of the plates before and after treatment, XPS was employed. It was found that only the oxidation treatment result can be verified. Fig. 8 shows the XPS spectra in both wide and narrow scan. The high-resolution spectra showed that there were a hydrocarbon peak at 285.0 eV, an aromatic carbon peak at 284.7 eV and a hydroxyl peak at 286.1 eV for the surface of the OCPs. For the control plates, the peak at 285.0 eV showed the existence of polystyrene hydrocarbons on the surface (Browne et al., 2004). Thus, the results showed that hydroxyl groups were present after the oxidation treatment.
According to the above results, protein crystallization was promoted by the oxidized crystallization plates. Studying the effects of the surface treatment may help to explain the phenomena observed in the experiment. Below, we will discuss the possible mechanisms from the viewpoints of protein adsorption, contact angle and surface roughness and their effects on protein crystallization.
According to the reactants used, the mixing process can be written as shown in equation (1); oxygen, which could oxidize the crystallization plate, was produced when potassium dichromate and concentrated sulfuric acid were mixed.
Functional groups containing oxygen, such as hydroxyl groups, can increase the wettability (Lee et al., 1998) of a surface. Therefore, the presence of hydroxyl groups will result in an increase in wettability, and consequently, the contact angle will decrease. The experimental measurement of the contact angle on the OCPs and control plates (see Fig. 6) confirmed this effect. Thus, the OCP surface was more hydrophilic than that of the control plates. Owing to the decrease in the contact angle, the adsorption of protein molecules will be enhanced (Arima & Iwata, 2007). A direct result of the enhancement of protein adsorption is that a localized highly supersaturated region on the surface will be created, and nucleation will be easier under these conditions. From Fig. 2, it can be observed that the crystallization success rate and the number of crystals were larger at the earlier time (2 d) than at the later time (6 d), which indicated that the induction time for nucleation was indeed shorter for the OCPs than for the control plates. This observation is in good agreement with the previous results that the free-energy barrier for nucleation is increased at higher contact angles and that the induction time is increased with increasing contact angle (Liu et al., 2007).
4.1.2. The new functional group enhanced the adsorption of proteins
Sigal et al. (1998) suggested that the hydroxyl group is able to form hydrogen bonds with side-chain groups of polar proteins, which can facilitate the adsorption of the proteins. Furthermore, hydroxyl groups can help stabilize the globular conformation of proteins (Martins et al., 2003). These factors are all beneficial in creating a localized region of high supersaturation at the surface of the treated plate, and these factors also stabilize the protein. Therefore, the nucleation of protein crystals is easier on the OCPs than on the untreated plates.
From Fig. 6, we can roughly estimate the different contact areas between the liquid and the solid surface in both cases (the OCP and the control groups) according to the diameter of the contact region. Using the photographs in Fig. 6 as examples, we obtained a ratio of the solid–liquid contact area of the OCPs to that of the control plates of SOCP/Scontrol = 28.32/21.32 = 1.76. Using Fig. 5, we could compare the contact areas between the OCP and control groups. The contact area was larger for the OCP than the control plates, and the larger contact area provided more opportunities for nucleation (Guo et al., 2010). Thus, there would be a greater likelihood of obtaining crystals in the OCP group.
4.2.2. Larger surface area could promote the evaporation of the solvent in the crystallization droplet
Similar to the above calculation, the ratio of the liquid–air contact area of the OCPs to that of the control plates could also be obtained from Fig. 5: SOCP/Scontrol = 205.89/176.53 = 1.17. This ratio shows that the liquid–air contact area was larger for the OCPs. Therefore, the evaporation of the solvent in the crystallization droplet should be faster for the OCPs than for the control plates. Supersaturation at early time points is controlled by the dynamics of solvent evaporation (Baird, 1999), and a faster increase in supersaturation could induce nucleation because the localized supersaturated region can reach a higher concentration, leading to more crystallization hits.
Porous materials were found to be beneficial for improving the crystallization success rate (Chayen et al., 2001; Rong et al., 2004). Compared with a smooth surface, the surfaces of the oxidized plates and the control plates both had much higher surface areas and many more small nanopores. Theoretically, the free-energy barrier for nucleation is decreased at a higher roughness (Guo et al., 2010; Liu et al., 2007). As shown in Table 3, the surface of the oxidized plates was smoother than that of the untreated plates. This result suggests that nucleation might be enhanced on the surface of the untreated plates. However, the average roughness was not sufficient to judge the effect of the structural characteristics of the surface on nucleation because roughness represents only the height in the vertical direction, which is not sensitive to the horizontal structures (Fang et al., 1997). The horizontal structures should also be considered important for nucleation.
According to our results, there were several factors (increased wettability, decreased contact angle, increased contact area for nucleation and increased evaporation surface) that might have exerted positive effects on nucleation, and there was one factor (roughness) that was not an obvious positive or negative factor affecting nucleation. The experimental results showed that the combined effects of all the above factors (and others not mentioned) facilitated nucleation and thus promoted crystallization.
In this study, we compared the crystallization with and without oxidation treatment of the crystallization plates. The comparison clearly shows that oxidation treatment was beneficial to facilitate protein crystallization. The potential power of this approach (i.e. treatment of crystallization plates) can be further improved in future studies by systematically comparing it with many well documented methods like doping with horse hair (D'Arcy et al., 2003), MIPs (Saridakis et al., 2011) and porous particles (Chayen et al., 2006), and on the basis of the comparison results, better plate treatment methods may be developed.
In this study, we modified protein crystallization plates using oxidation in an attempt to facilitate crystallization. When the oxidized plates were used, the number of protein crystallization screening hits was increased and the reproducibility was improved. We anticipate that this modification can be used in protein crystallization and can help facilitate the structural analysis of proteins. The testing of treatment methods other than oxidation to modify the surfaces of crystallization plates may yield even better modification methods.
‡These authors contributed equally to this work.
This work was supported by the National Natural Science Foundation of China (grant No. 31170816 and grant No. 11202167), the National Basic Research Program of China (973 Program, grant No. 2011CB710905), the Doctorate Foundation of Northwestern Polytechnical University (grant No. CX201121), the Ministry of Education Fund for Doctoral Students Newcomer Awards of China, the Natural Science Basic Research Plan in Shaanxi Province of China (grant No. 2012JQ3009), the Fundamental Research Foundation of Northwestern Polytechnical University in China (grant No. JC201164 and grant No. JC201163), the Hong Kong Scholar Program (No. XJ2011028) and a special financial grant from the China Postdoctoral Science Foundation (grant No. 201104682).
Arima, Y. & Iwata, H. (2007). Biomaterials, 28, 3074–3082. Web of Science CrossRef PubMed CAS
Baird, J. K. (1999). J. Cryst. Growth, 204, 553–562. Web of Science CrossRef CAS
Bard, J., Ercolani, K., Svenson, K., Olland, A. & Somers, W. (2004). Methods, 34, 329–347. Web of Science CrossRef PubMed CAS
Bergfors, T. (2003). J. Struct. Biol. 142, 66–76. Web of Science CrossRef PubMed CAS
Berman, H. M., Westbrook, J., Feng, Z., Gilliland, G., Bhat, T. N., Weissig, H., Shindyalov, I. N. & Bourne, P. E. (2000). Nucleic Acids Res. 28, 235–242. Web of Science CrossRef PubMed CAS
Browne, M., Lubarsky, G., Davidson, M. & Bradley, R. (2004). Surf. Sci. 553, 155–167. Web of Science CrossRef CAS
Brzozowski, A. M. & Walton, J. (2001). J. Appl. Cryst. 34, 97–101. Web of Science CrossRef CAS IUCr Journals
Chayen, N. E., Saridakis, E., El-Bahar, R. & Nemirovsky, Y. (2001). J. Mol. Biol. 312, 591–595. Web of Science CrossRef PubMed CAS
Chayen, N. E., Saridakis, E. & Sear, R. P. (2006). Proc. Natl Acad. Sci. USA, 103, 597–601. Web of Science CrossRef PubMed CAS
Cudney, R., Patel, S., Weisgraber, K., Newhouse, Y. & McPherson, A. (1994). Acta Cryst. D50, 414–423. CrossRef CAS IUCr Journals
D'Arcy, A., Mac Sweeney, A. & Haber, A. (2003). Acta Cryst. D59, 1343–1346. Web of Science CrossRef CAS IUCr Journals
Fang, S. J., Haplepete, S., Chen, W., Helms, C. R. & Edwards, H. (1997). J. Appl. Phys. 82, 5891–5898. Web of Science CrossRef CAS
Fermani, S., Falini, G., Minnucci, M. & Ripamonti, A. (2001). J. Cryst. Growth, 224, 327–334. Web of Science CrossRef CAS
Georgieva, D. G., Kuil, M. E., Oosterkamp, T. H., Zandbergen, H. W. & Abrahams, J. P. (2007). Acta Cryst. D63, 564–570. Web of Science CrossRef CAS IUCr Journals
Gorrec, F. (2009). J. Appl. Cryst. 42, 1035–1042. Web of Science CrossRef CAS IUCr Journals
Guo, Y., Yin, D., Lu, Q., Wang, X. & Liu, J. (2010). Cryst. Res. Technol. 45, 158–166. Web of Science CrossRef CAS
Hansen, C. L., Skordalakes, E., Berger, J. M. & Quake, S. R. (2002). Proc. Natl Acad. Sci. USA, 99, 16531–16536. Web of Science CrossRef PubMed CAS
Hemming, S. A., Bochkarev, A., Darst, S. A., Kornberg, R. D., Ala, P., Yang, D. S. & Edwards, A. M. (1995). J. Mol. Biol. 246, 308–316. CrossRef CAS PubMed Web of Science
Jacquamet, L., Ohana, J., Joly, J., Borel, F., Pirocchi, M., Charrault, P., Bertoni, A., Israel-Gouy, P., Carpentier, P., Kozielski, F., Blot, D. & Ferrer, J. L. (2004). Structure, 12, 1219–1225. Web of Science CrossRef PubMed CAS
Jancarik, J. & Kim, S.-H. (1991). J. Appl. Cryst. 24, 409–411. CrossRef CAS Web of Science IUCr Journals
Kallio, J. M., Hakulinen, N., Kallio, J. P., Niemi, M. H., Kärkkäinen, S. & Rouvinen, J. (2009). PLoS One, 4, e4198. Web of Science CrossRef PubMed
Kertis, F., Khurshid, S., Okman, O., Kysar, J. W., Govada, L., Chayen, N. & Erlebacher, J. (2012). J. Mater. Chem. 22, 21928. Web of Science CrossRef
Kim, Y. et al. (2008). Nat. Methods, 5, 853–854. Web of Science CrossRef PubMed CAS
Lee, J. H., Khang, G., Lee, J. W. & Lee, H. B. (1998). J. Colloid Interface Sci. 205, 323–330. Web of Science CrossRef CAS PubMed
Liang, C. & Krimm, S. (2003). J. Polym. Sci. 27, 241–254. CrossRef Web of Science
Liu, Y. X., Wang, X. J., Lu, J. & Ching, C. B. (2007). J. Phys. Chem. B, 111, 13971–13978. Web of Science CrossRef PubMed CAS
Martins, M. C., Ratner, B. D. & Barbosa, M. A. (2003). J. Biomed. Mater. Res. A, 67, 158–171. Web of Science CrossRef PubMed
McPherson, A. & Cudney, B. (2006). J. Struct. Biol. 156, 387–406. Web of Science CrossRef PubMed CAS
McPherson, A. & Shlichta, P. (1988). Science, 239, 385–387. CrossRef PubMed CAS Web of Science
Mueller, U., Nyarsik, L., Horn, M., Rauth, H., Przewieslik, T., Saenger, W., Lehrach, H. & Eickhoff, H. (2001). J. Biotechnol. 85, 7–14. Web of Science CrossRef PubMed CAS
Newman, J., Egan, D., Walter, T. S., Meged, R., Berry, I., Ben Jelloul, M., Sussman, J. L., Stuart, D. I. & Perrakis, A. (2005). Acta Cryst. D61, 1426–1431. Web of Science CrossRef CAS IUCr Journals
Otwinowski, Z. & Minor, W. (1997). Methods Enzymol. 276, 307–326. CrossRef CAS Web of Science
Paxton, T., Sambanis, A. & Rousseau, R. (1999). J. Cryst. Growth, 198–199, 656–660. Web of Science CrossRef
Reddy, S. M., Phan, Q. T., El-Sharif, H., Govada, L., Stevenson, D. & Chayen, N. E. (2012). Biomacromolecules, 13, 3959–3965. Web of Science CrossRef CAS PubMed
Rong, L., Komatsu, H., Yoshizaki, I., Kadowaki, A. & Yoda, S. (2004). J. Synchrotron Rad. 11, 27–29. Web of Science CrossRef CAS IUCr Journals
Santarsiero, B. D., Yegian, D. T., Lee, C. C., Spraggon, G., Gu, J., Scheibe, D., Uber, D. C., Cornell, E. W., Nordmeyer, R. A., Kolbe, W. F., Jin, J., Jones, A. L., Jaklevic, J. M., Schultz, P. G. & Stevens, R. C. (2002). J. Appl. Cryst. 35, 278–281. Web of Science CrossRef CAS IUCr Journals
Saridakis, E. & Chayen, N. E. (2013). Trends Biotechnol. 31, 515–520. Web of Science CrossRef CAS PubMed
Saridakis, E., Khurshid, S., Govada, L., Phan, Q., Hawkins, D., Crichlow, G. V., Lolis, E., Reddy, S. M. & Chayen, N. E. (2011). Proc. Natl Acad. Sci. USA, 108, 11081–11086. Web of Science CrossRef CAS PubMed
Sigal, G. B., Mrksich, M. & Whitesides, G. M. (1998). J. Am. Chem. Soc. 120, 3464–3473. Web of Science CrossRef CAS
Snook, C. F., Purdy, M. D. & Wiener, M. C. (2000). J. Appl. Cryst. 33, 344–349. Web of Science CrossRef CAS IUCr Journals
Stock, D., Perisic, O. & Löwe, J. (2005). Prog. Biophys. Mol. Biol. 88, 311–327. Web of Science CrossRef PubMed CAS
Sulzenbacher, G. et al. (2002). Acta Cryst. D58, 2109–2115. Web of Science CrossRef CAS IUCr Journals
Thakur, A. S., Robin, G., Guncar, G., Saunders, N. F., Newman, J., Martin, J. L. & Kobe, B. (2007). PLoS One, 2, e1091. Web of Science CrossRef PubMed
Yin, D., Wakayama, N. I., Lu, H., Ye, Y., Li, H., Luo, H. & Inatomi, Y. (2008). Cryst. Res. Technol. 43, 447–454. Web of Science CrossRef CAS
Zheng, B., Tice, J. D., Roach, L. S. & Ismagilov, R. F. (2004). Angew. Chem. Int. Ed. 43, 2508–2511. Web of Science CrossRef CAS
© International Union of Crystallography. Prior permission is not required to reproduce short quotations, tables and figures from this article, provided the original authors and source are cited. For more information, click here.