A streamlined protocol for performing an extensive biochemical and structural characterization of a carbohydrate substrate binding protein from Streptococcus pneumoniae is presented.
Development of new antimicrobials and vaccines for Streptococcus pneumoniae (pneumococcus) are necessary to halt the rapid rise in multiple resistant strains. Carbohydrate substrate binding proteins (SBPs) represent viable targets for the development of protein-based vaccines and new antimicrobials because of their extracellular localization and the centrality of carbohydrate import for pneumococcal metabolism, respectively. Described here is a rationalized integrated protocol to carry out a comprehensive characterization of SP0092, which can be extended to other carbohydrate SBPs from the pneumococcus and other bacteria. This procedure can aid the structure-based design of inhibitors for this class of proteins. Presented in the first part of this manuscript are protocols for biochemical analysis by thermal shift assay, multi angle light scattering (MALS), and size exclusion chromatography (SEC), which optimize the stability and homogeneity of the sample directed to crystallization trials and so enhance the probability of success. The second part of this procedure describes the characterization of the SBP crystals using a tunable wavelength anomalous diffraction synchrotron beamline, and data collection protocols for measuring data that can be used to resolve the crystallized protein structure.
S. pneumoniae (pneumococcus) is a gram-positive bacterium residing asymptomatically in the upper airways of the human respiratory tract with the ability to migrate to normally sterile niches causing otitis, pneumonia, sepsis, septicemia, and meningitis1,2. Moreover, pneumococcal infection is the leading cause of community-acquired pneumonia, which is contributing to a clinical and economic burden worldwide3,4. Antibiotic resistant strains of S. pneumoniae have spread across the globe and although a seven-valent and a thirteen-valent pneumococcal protein conjugate vaccine have helped reduce the rate of antimicrobial resistance, replacement strains from vaccine use have emerged and have led to increased demands for research into the development of new treatments for pneumococcal disease5,6,7,8.
The pneumococcus depends on sugars imported from the host as a carbon source9,10; indeed it devotes 30% of its import machinery to the transport of 32 different carbohydrates11,12,13. These importers include at least eight ABC-transporters13. In ABC transporters, the extracellular SBPs play a fundamental role in determining the specificity for the ligand and presenting it to the integral membrane transporter for uptake in the cell. SBPs represent valid targets for the design of new vaccines and antimicrobials because they are surface proteins and their vital role in cellular processes.
Target protein characterization and detailed description of structural features, like ligand pockets and interdomain flexibility, provide a useful tool for structure-based drug design14,15. X-ray crystallography is the method of choice for the structural characterization of proteins at near to atomic resolution, but the crystallization process is unpredictable, time-consuming, and not always successful. Systematic methods have improved the success rate and important factors are the sample quality and stability. The success rate of crystallization is influenced by the protein chemical properties and sample preparation methodology. The effect of these can be assessed and informed by biochemical characterization16,17.
A further complication for structure-based design is the crystallographic phase problem, which must be addressed. As more protein structures have become available, many structures can be resolved by the molecular replacement method, which requires a homologous structure18. As SBPs present a flexible domain structure, molecular replacement may also prove challenging19. If a structural model that is sufficiently similar to the target protein is not available, a number of techniques can be used to obtain the experimental phasing20. Among these, the Single-wavelength Anomalous Dispersion (SAD) method has emerged as the primary technique and has been extensively used to solve the phase problem21. The use of the SAD method has been further advanced with improvements in hardware and software, as well as data collection strategies to allow the detection and use of weak anomalous signals for phasing22,23,24. Furthermore, advances in direct methods for solving structures of macromolecules, which in the past required diffraction data to atomic resolution, can now be utilized by combining for example, stereochemical knowledge as implemented in the program ARCIMOLDO25. A useful review of methods for solving the phase problem in crystallography is given by Taylor26.
Here we present a rationalized protocol for the characterization of the carbohydrate transport SBP, SP0092 of S. pneumoniae, integrating biochemical and structural techniques (Figure 1). This step-by-step protocol provides a useful example test case of strategies to improve the success rate of structural studies on SBPs in general, which are found in all kingdoms of life. In particular, the protocol highlights the importance of characterizing the most stable oligomeric state of the protein in solution in a fast and effective method, and allows the identification of the best species to follow up for crystallization experiments. Although there are over 500 SBP structures reported in the Protein Data Bank27, molecular replacement can be challenging due to the inherent flexible nature of the two α/β domains, which are connected by a hinge region19. Thus, the second part of the protocol describes using the SAD method for phasing from bound metal ions, which is common in SBPs, as well as the incorporation of selenomethionine and use of selenium (Se) in SAD phasing.
NOTE: The coding sequence for which the signal peptide is deleted is cloned in the pOPINF vector following a standard in-fusion protocol; the native protein is expressed as a His-tag fusion in Escherichia coli BL21 Rosetta cells28,29. The selenomethionine labeled variant is expressed following standard methods according to the manufacturer30. The recombinant SBP is purified as previously described31,32.
1. Biochemical Characterization
2. Protein Preparation and Crystallization
3. Crystal Characterization and X-ray Data Collection
This integrated protocol has been proven to be successful with four (two published and two unpublished structures) of six carbohydrate binding protein targets from pneumococcus analyzed to date32,53. In this section, we present the biochemical and structural characterization of SP0092 as a representative result to guide structural studies of SBPs in general.
After the SBP SP0092 has been expressed and purified as defined previously32, the purified protein was analyzed for buffer stability using a thermal shift assay: SP0092 exhibits an increased Tm at pH 6.5 and in the presence of NaCl in the 0 – 0.2 M concentration range (Figure 3A). In light of this, the buffer solution for the following steps was defined as: 0.02 M MES pH 6.5, 0.2 M NaCl, 2.5% (v/v) glycerol, 0.5 mM TCEP. The absolute molar mass of the different oligomerization states of SP0092 was measured by MALS coupled to SEC measuring a molecular weight of 187.2, 140.8, 97.0, and 49.4 kDa, corresponding to tetrameric, trimeric, dimeric, and monomeric species, respectively (Figure 3B). The analysis of the SEC profile at different protein dilutions revealed that the oligomerization is triggered by increased protein concentration, suggesting that the larger oligomers are more stable at higher concentrations than typically used in crystallization. Indeed, the purified larger oligomeric species directed to crystallization trials, successfully produced protein crystals while the monomer species did not.
The optimized crystals obtained from the native and Se-methionine labeled forms of SP0092 were characterized by X-ray diffraction. Measurement of the X-ray fluorescence from these crystals revealed in both cases emission peaks for Zn being bound to the protein, while Se was detected only for the Se-methionine crystals as expected. Later, X-ray absorption scans at the Se and Zn edges were performed, which provided direct experimental data to tune the incident X-ray wavelength to the respective X-ray absorption edges of either the Zn or Se present in the crystals to maximize the anomalous signal obtainable from the resultant data (Figure 4A–B).
After measuring three diffraction patterns at low transmission, a complete anomalous data set was obtained using the data collection strategy suggested by EDNA (Figure 4C). The anomalous signal present triggers the automated phasing pipeline at the beamline to determine the sub-structure, and based on the initial experimental phases derived, produces initial maps and models, which can then be refined and validated (Figure 4D).
In summary, potential pitfalls of the technique are centered primarily on crystal availability and quality. Optimization of the buffer conditions to improve protein stability as well as identification of the most suitable oligomeric state of the protein to use, when more than one oligomer is identified in solution, can reduce the risk of failure at the crystallization stage. Exploiting the use of bound metal ions identified at an early stage can speed up structure solution and avoid unnecessary production of selenomethionine labeled protein when molecular replacement methods fail.
Figure 1. Diagram of workflow for biochemical and structural characterization of carbohydrate SBPs. Please click here to view a larger version of this figure.
Figure 2. Crystal characterization in GDA. (A) Screenshot of the X-ray fluorescence control tab of the GDA beamline control software. (B) Screenshot of the X-ray energy edge-scan control tab in GDA. (C) Screenshot of the data collection crystal screening tab in GDA. Please click here to view a larger version of this figure.
Figure 3. Biochemical characterization of SP009239-491. (A) 3D-surface graph plotting the melting temperature of SP009239-491 as a function of the pH and NaCl concentration of the buffer solution. (B) SEC and MALS results for SP009239-491. Absorption at 280 nm is shown in blue. The molar masses of the different oligomerization states are shown in red. Panel (B) has been modified from32. Please click here to view a larger version of this figure.
Figure 4. Structural characterization of SP009239-491. (A) and (B) X-ray absorption energy scan at the Zn and Se edge for SP009239-491 crystals. The measured fluorescence from the Zn and Se containing crystals is shown in blue and cyan, the calculated f' and f" anomalous scattering fractions are in green and red, respectively. (C) Examples of X-ray diffraction patterns collected from SP009239-491 crystals. (D) Cartoon representation of the SP009239-491 dimer structure. One protomer is colored white and the other magenta (residues 39 – 366) and violet (residues 367 – 491), and the anomalous scattering atoms are shown as blue and cyan balls for Zn and Se, respectively. Please click here to view a larger version of this figure.
0.1 M Citric Acid pH 4.0 | 0.1 M Citric Acid pH 4.5 | 0.1 M Phosphate pH 5.0 | 0.1 M Citrate pH 5.5 | 0.1 M Bis-tris pH 6 | 0.1 M ADA pH 6.5 | 0.1 M MOPS pH 7 | 0.1 M HEPES pH 7.5 | 0.1 M Imidazole pH 8.0 | 0.1 M Tris pH 8.5 | 0.1 M CHES pH 9 | 0.1 M CHES pH 9.5 | |
NaCl 0 M | 40 μL | 40 μL | 40 μL | 40 μL | 40 μL | 40 μL | 40 μL | 40 μL | 40 μL | 40 μL | 40 μL | 40 μL |
NaCl 0.1 M | 40 μL | 40 μL | 40 μL | 40 μL | 40 μL | 40 μL | 40 μL | 40 μL | 40 μL | 40 μL | 40 μL | 40 μL |
NaCl 0.2 M | 40 μL | 40 μL | 40 μL | 40 μL | 40 μL | 40 μL | 40 μL | 40 μL | 40 μL | 40 μL | 40 μL | 40 μL |
NaCl 0.5 M | 40 μL | 40 μL | 40 μL | 40 μL | 40 μL | 40 μL | 40 μL | 40 μL | 40 μL | 40 μL | 40 μL | 40 μL |
Table 1. Buffer composition for thermal shift assay.
In this paper, we describe and validate an integrated protocol for biochemical and structural characterization of carbohydrate SBPs with a specific emphasis on proteins from S. pneumoniae. Nevertheless, this can be used as a standard procedure for the analysis of other SBPs from different organisms and even other unrelated soluble proteins.
The first part of the protocol is focused on providing biochemical information on protein stability and quaternary structure, which can be exploited in the preparation of protein samples for crystallization. In the thermal shift assays section, we describe only the pH and NaCl concentration variations to maintain the general nature of this procedure. Despite this, many other buffer conditions can be tested in a similar way, for example, including any chemical compound used as a stabilizing additive: in particular the actual ligand(s) that bind to a specific SBP are remarkably effective in increasing the Tm by a few degrees Celsius31. In some cases, the denaturing curves can be poorly defined due to low signal or a high fluorescence background, which is caused by protein aggregation or partial unfolding. To avoid this, a protein:dye titration can be performed to optimize the unclear denaturing curves. If no improvement is obtained, screening various additives that can ameliorate the stability of the protein is advised, and suitable screens have been previously described54.
Typically, most SBPs are monomeric in their natural environment, but as shown here multimerization can occur at the higher concentrations used in crystallization experiments, thus the oligomerization behavior characterization provided by MALS and SEC is essential to assess the most favorable stable monodisperse oligomerization state for crystallization. Nevertheless, it is hard to predict the effect of different chemicals included in the crystallization condition on the oligomerization behavior of the proteins. If the SEC and MALS examination shows extensive aggregation of the protein sample, we would advise the following to reduce the likelihood of this occurring: use fresh protein sample (not freeze-thawed) and expand the stabilization analysis performed with thermal shift assays, testing possible additives and mild detergents as a last resource, to minimize aggregation. In this paper, we present basic guidelines for crystallization using high-throughput sparse matrix commercial crystallization screening to maintain the general nature of this protocol. However, obtaining high-resolution X-ray diffraction protein crystals might need iterative fine tuning to optimize the crystallization conditions with respect to precipitant concentration, pH, addition of chemical additives, different temperatures, and other factors changing equilibrium dynamics between the crystal drop and reservoir16,17.
The second part of the protocol describes the characterization of the protein crystals in order to define the optimal strategy for X-ray diffraction data collection with a specific focus on the acquisition of anomalous data for SAD phasing. Even if SBPs maintain a similar general architecture (and there are many deposited 3D structures potentially usable as starting models), phasing of these proteins by the molecular replacement method is not always straightforward because of the variability of the secondary structure elements and the intrinsic flexibility of these proteins. Hence, we propose the SAD method and highlight that these proteins may already have intrinsically bound metals or indeed non-specific binding of metals from the crystallization buffer conditions, which can provide a range of anomalous diffracting elements as a standard step in our general protocol.
In conclusion, this protocol defines a standard guided workflow of procedures enabling the detailed description of the biochemical and structural features of SBPs that can be exploited to increase the structure determination success rate, as well as accelerate the structural characterization of SBPs in general.
The authors have nothing to disclose.
We acknowledge OPPF-UK for assistance in cloning, Gemma Harris for SEC-MALLS and the scientists of beamlines I03 and I04 at Diamond Light Source.
SelenoMethionine Medium Complete | Molecular Dimensions | MD12-500 | Based on a synthetic M9 minimal media supplemented with glucose, vitamins and amino acids with the exception of L-methionine. Other equivalent products are commercially available by other companies. |
MicroAmp Optical 96-Well plate | Applied Biosystems | 4306737 | The Applied Biosystems MicroAmp Optical 96-Well Reaction Plate with Barcode is optimized to provide unmatched temperature accuracy and uniformity for fast, efficient PCR amplification. This plate, constructed from a single rigid piece of polypropylene in a 96-well format, is compatible with Applied Biosystems® 96-Well Real-Time PCR systems and thermal cyclers. |
SYPRO Orange | Molecular Probes | S6651 | SYPRO Orange Protein Gel Stain is a sensitive, ready-to-use fluorescent stain for proteins in 1D gels. Quite universal and well-established protein dye for hydrophobic regions. |
MicroAmp Optical Adhesive film | Applied Biosystems | 4311971 | The Applied Biosystems MicroAmp Optical Adhesive Film reduces the chance of well-to-well contamination and sample evaporation when applied to a microplate. It is ideal for optical measurement, because it gives low background. |
7500 Fast Real-time PCR System | Applied Biosystems | 4362143 | The Applied Biosystems 7500 Fast Real-Time PCR System enables standard 96-well format high speed thermal cycling, significantly reducing your run time for quantitative real-time PCR applications, delivering results in about 30 minutes. The Upgrade Kit is available to upgrade a standard 7500 Real-Time PCR System to the Fast Configuration via a field service installation. Other RT-PCR machine can be used. Data export not easy for the old data analysis software. |
Superdex 200 increase 10/300 GL column | GE Healthcare | 28990944 | Superdex 200 Increase 10/300 GL is a versatile, prepacked column for size exclusion chromatography in small-scale (mg) preparative purification as well as for characterization and analysis of proteins with molecular weights between 10 000 and 600 000, such as antibodies. Optimal separation ideal for high resolution biophysical techniques. |
DAWN HELEOS II | Wyatt | DAWN HELEOS II is the premier Multi-Angle static Light Scattering (MALS) detector for absolute characterization of the molar mass and size of macromolecules and nanoparticles in solution. The DAWN offers the highest sensitivity, the widest range of molecular weight, size and concentration, and the largest selection of configurations and optional modules for enhanced capabiliites. Other MALS detecting systems from other companies apart from Wyatt have not been tested, so no additional feedback can be provided. | |
Superdex 200 5/150 GL | GE Healthcare | 28906561 | Superdex 200 are prepacked size exclusion chromatography columns for high-resolution small-scale preparative and analytical separations of biomolecules. Superdex 200 has a separation range for molecules with molecular weights between 10 000 and 600 000. The peak separation is not as optimal as for the "increase" version but this model is ideal for the standard day by day use. |
HiLoad 16/600 Superdex 200 pg | GE Healthcare | 28989335 | HiLoad 16/600 Superdex 200 prep grade are prepacked XK columns designed for high-resolution preparative gel filtration chromatography. |
24 Well "Big" Sitting Drop Crystallization Plate | MiTeGen | XQ-P-24S-A | The 24 well "big" crystallization plate is used mainly for protein crystal screening by sitting drop vapor diffusion techniques, and for crystallization condition optimization. It has quite large reservoir well and sample container, which favor manual handling and big sized protein crystal growth. In addition, its flat surfaces are easily to be sealed by transparent tape or cover slips. |
PCT Pre-Crystallization Test | Hampton | HR2-140 | The PCT Pre-Crystallization Test is used to determine the appropriate protein concentration for crystallization screening. |
96 Well CrystalQuick | Greiner bio-one | 6098xx | Square-well plates have three crystallization wells per reservoir, making it possible to test 288 samples per plate. Generally used only for the inital screening because their squared edge make crystals fishing difficult. |
Uni-Puck | Molecular Dimensions | MD7-601 | The Universal V1-Puck (Uni-puck) is a sample pin storage and shipping container for use with the majority of automated sample mounting systems worldwide – includes the ACTOR, SAM and CATS systems amongst others (Diamond, Soleil, SPring-8, Photon Factory, CLSI and across the USA) |
Standard Foam Dewar | Molecular Dimensions | MD7-35 | 5.7" diameter by 2.8" deep. 800mL capacity. |
Mounted CryoLoop – 20 micron | Hampton | HR4-955 | Mounted CryoLoops with 20 micron diameter nylon. These nylon loops are bonded to hollow, stainless steel MicroTubes™ that are used to mount, freeze, and secure the crystal during cryocrystallographic procedures and X-ray data collection. Different sizes exist and they can be adapted in lenght. They are quite versatile tools. |
CryoWand | Molecular Dimensions | MD7-411 | |
Puck dewar loading tool | Molecular Dimensions | MD7-607 | This tool is used to separate uni-pucks to load them into the robot dewar. It consists of two pieces: a Teflon tube part and a metal rod part. |