Jacob Berger
The Effect of Precipitant Chirality on the Crystallization of Lysozyme Thesis Submitted in Partial Fulfillment of the Requirements of the Jay and Jeanie Schottenstein Honors Program
Yeshiva College Yeshiva University May 2011
Jacob Berger Mentor: Professor Neer Asherie, Departments of Physics and Biology
Abstract Protein crystallization has been studied for over 150 years but is still not well understood. Protein crystals are extremely difficult to grow since protein crystallization is sensitive to many parameters. In this thesis I investigate if the chirality of a precipitant, 2methyl-2,4-pentanediol (MPD), influences the crystal growth of the protein lysozyme. It appears that R-MPD has a single conformation within the crystal, while S-MPD may take on a few different conformations. Although there appears to be a chiral difference within the structure, this does not seem to dramatically affect the kinetics of crystal growth as both crystal structures form in the same period of time.
1. Introduction Proteins control many of the processes needed for organisms to function. Knowing the structure of proteins is essential to understand how they help regulate life, and also how mutations of proteins can harm life. Only by knowing the structure can the function of the protein be fully determined. While determining the function of the protein may seem like the more difficult task, determining the structure of proteins is, in fact, a much greater problem. The structure of a protein can be determined by x-ray diffraction of a crystallized protein. The difficulty is that conditions needed to crystallize proteins are still not well understood. There is no general phase diagram for proteins (like those that exists for homogeneous structures such as water or iron), the only way to actually crystallize proteins is by an arduous and lengthy guess and check method of trying many different experimental conditions. Proteins are long, complex polymers of amino acids. This makes it exceedingly difficult to understand how to crystallize them. In principle, each protein could crystallize under different conditions and have a unique phase diagram. Currently, no systematic procedure has been found that will guarantee protein crystals for a protein that has never been crystallized before. Fewer than three percent of proteins from cloned genes have had their structure determined. 1 Growing diffraction quality crystals represent the largest challenge facing crystallographers. This inability to effectively crystallize proteins limits the biomedical and industrial applications. Some diseases, such as genetic cataracts, are caused by the crystallization of
1
proteins in the body.2 If the phase diagram were known for all proteins, a method could be employed to solubilize the proteins and reverse the effects of these diseases. Industrially, crystallization is one of the most effective forms of purification. Knowing each protein’s phase behavior would revolutionize the industrial purification process. At the moment, there appears to be no general phase diagram for all proteins. Crystallization of proteins is not a recent problem; scientists have attempted to crystallize proteins for over 150 years. 3 The first documented crystallized protein was hemoglobin from an earthworm, published in 1840 by Hünefeld. He demonstrated that protein crystals could be obtained from the slow and controlled dehydration of concentrated protein – a technique still in use today. For a decade no additional method was developed to reproducibly grow crystals of hemoglobin until Fünke in 1851. He employed methods of dissolving protein in purified water or organic solvents, also a commonly used method today. Thirty years later, the next advances were made. Gruber, Ritthausen, and Osborne, developed new methods of crystallization. They were the first to use temperature as a controlled variable and dialysis against low ionic strength solutions. They also further studied the use of organic solvents as precipitating agents. Slowly, crystallographers began to realize the sensitivity of crystallization to many different conditions. Crystallizing proteins is still extremely difficult, and the methods developed 150 years ago remain in use today. The phase diagram for proteins is a multi-dimensional graph that depends of many variables including protein concentration, pH, temperature, ionic strength, and the precipitant used. Crystallization is extremely sensitive to small changes in any of these conditions. Changing the pH or concentration slightly can be the difference between highly ordered crystals and amorphous precipitate. Combining the perfect combination of conditions to get crystals is often mostly luck. The general theory behind protein crystallization is to allow a supersaturated solution of protein to precipitate as ordered crystals in order to reach equilibrium with the solvent (figure 1). In the under-saturated zone, crystals will never form because the most stable phase is the solution phase. If the protein is only slightly super-saturated, it may remain supersaturated in the metastable zone due to the free energy barrier that prevents nucleation, although crystals can grow.4 In the nucleation zone a critical nucleus can form, a prerequisite for crystal growth. If the solution is overly super-saturated, then uncontrolled precipitation
2
may occur, resulting in an amorphous solid. The boundaries shown schematically in figure 1 represent regions to investigate for possible crystal growth. The actual lines on a protein phase diagram may be completely different depending on the crystallization conditions. Often there is a large region of no growth and uncontrolled amorphous precipitation, but a very small, narrow region of optimal crystal growth. There may never even be a region where ordered crystals can grow, as we have seen with some of our experiments.
Figure 1. Schematic illustration of a protein crystallization phase diagram. Adjustable parameters include precipitant or additive concentration, pH and temperature. The arrows indicate different paths crystal growth will take depending on the method. The two methods used in this thesis are microbatch (i) and hanging drop (ii). The filled black circles represent the starting conditions. The solubility is defined as the concentration of protein in the solute that is in equilibrium with crystals. The supersolubility curve is defined as the line separating conditions under which spontaneous nucleation (or phase separation or precipitation) occurs from those under which the crystallization solution remains clear if left undisturbed.5
Our group has recently shown that even the chirality of the precipitant used can greatly affect the crystallization process. Chiral molecules have the same chemical groups, but are mirror images of each other and non-superposable (figure 2). Chirality is derived from the Greek word for hand and refers to the “handedness” of the molecule. The most commonly used examples for chirality are left and right hands since they are mirror images of each other but cannot be superposed onto one another.
3
Figure 2. A chiral molecule and its mirror image.
Our group initially established the important role of chirality in protein crystallization when we investigated the solubility of thaumatin and sodium tartrate. We noticed an apparent contradiction in the literature: some groups claimed the solubility of the protein thaumatin was direct (increases with temperature), while other groups said it was retrograde (decreases with temperature). This difference occurs because L- and D- tartrate have different effects on the solubility of thaumatin. We measured the solubility of thaumatin with each enantiomer of tartrate and resolved the apparent contradiction. We showed that thaumatin crystallized with L-tartrate
has direct solubility while thaumatin crystallized with D-tartrate displays retrograde
solubility. The different enantiomers not only altered the thermodynamics of crystallization, but the kinetics of crystallization as well. For example,
L-
and
D-tartrate
both lead to the
formation of bipyramidal crystals of thaumatin; with L-tartrate these crystals form in hours, while in D-tartrate they form in days. Using the separate crystals grown from each of the enantiomers, the high-resolution x-ray diffraction data determined that the different enantiomers of tartrate take different conformations within the unit cell, therefore leading to completely different crystallization kinetics (figure 3). The L-tartrate (figure 3A) only takes on one conformation within the crystal, while conformations within the structure of thaumatin.
4
D-tartrate
(figure 3B) can take on two
Figure 3. L-tartrate (A) only takes on one conformation while D-tartrate (B) takes on two conformations. The dashed red and black lines denote the bonds for the two conformers of D-tartrate.6
While these findings are extraordinarily interesting, they may be specific to thaumatin and tartrate. If our results were generalized to other protein-precipitant systems, it would further elucidate the conditions and considerations needed for effective crystallization of proteins. Lysozyme is one of the most studied proteins, which makes it a good candidate to investigate the affects of precipitant chirality. Lysozyme is an ideal model protein because it is relatively inexpensive and easy to crystallize. Hen egg white lysozyme (muramidase or Nacetlymuramide glycanohydrolase)8 was the second protein and the first enzyme to have its structure determined by x-ray diffraction.7 Lysozyme also has had its structure determined to the highest resolution (0.65 Å) of any protein in the Protein Data Bank (PDB ID 2VB1); the tetragonal structure we study here has been determined to 0.94 Å (PDB ID 1IEE). Lysozyme is a relatively small secretory enzyme that catalyzes the hydrolysis of specific polysaccharides contained in cell walls of bacteria. Modified lysozyme molecule is used in many areas, such as to prevent infections and simultaneously to act as a natural antibiotic or stimulant of the immune system. Lysozyme from hen’s egg white is a polypeptide of 129 amino acid residues having a molecular weight of 14.3 kDa. Lysozyme is a strongly basic protein and with an isoelectric point (pI) of 10–11. Because of the high isoelectric point, lysozyme binds to negatively charged residues. Four disulfide bonds of the single polypeptide chain of the enzyme make this small protein molecule unusually compact and highly stable.8
5
MPD is the most commonly used additive in crystallization. 9 MPD promotes stabilization of the protein by preferential hydration, which is facilitated by attachment of MPD molecules to the hydrophobic surface. In crystallization experiments, MPD can act as precipitant through a combination of activities, including competition for water, hydrophobic exclusion of protein solutes, and lowering of the solution dielectric. The dielectric constant of pure MPD is 25. MPD is a chiral molecule (figure 4). The only high purity form of MPD that is commercially available is the racemate (RS-MPD, an equimolar mixture of R- and SMPD).
Figure 4. R and S enantiomers of MPD. The hydrogen atoms are not shown.8
Lysozyme has previously been crystallized with MPD, but as was the case with thaumatin, the previously published results are inconsistent. Weiss et al. reports that they only find the R enantiomer of MPD in their crystals (PDB ID 1DPW).10 Michaux et al. on the other hand, publish that they only find the S enantiomer of MPD in their crystals (PDB ID 3B72).11 In both cases the investigators use the racemic mixture of MPD. In this thesis I crystallize lysozyme with the separate enantiomers of MPD as wells as the racemic mixture. I investigate whether the chirality of the precipitant affects the crystallization of the protein.
2. Experimental Section 2.1. Materials. Lysozyme (Batch 36P9210) was purchased from Worthington Biochemical Cooperation. Purity was checked using High Performance Liquid Chromatography (HPLC) and quasielastic light scattering (QLS). R-MPD and S-MPD were synthesized by RCA separations (Freiburg, Germany). Enantiomeric purity was greater than 99%. RS-MPD (CAS no. 107-41-5), Sodium azide (cat. no. S227), monobasic sodium phosphate (cat. no. BP330), dibasic sodium phosphate (cat. no. BP332), sodium acetate (cat. no. BP334), Tris base (cat.
6
no. BP512), sodium chloride (cat. no. BP512), and sodium hydroxide (cat. no. BP359) were purchased from Fisher Scientific, Pittsburgh, PA. All materials were used without further purification. Deionized water was obtained from an Integral 3 deionization system (Millipore, Billerica, MA). Solutions were filtered through a Nalgene disposable 0.22 µm filter unit (Nalge Nunc International, Rochester, NY) prior to use. 2.2. Size-Exclusion and Cation-Exchange Chromatography. Both size-exclusion and cation-exchange high-performance liquid chromatography (SE-HPLC and CE-HPLC, respectively) were carried out on a Beckman System Gold apparatus at a flow rate of 1 mL/min (Beckman Coulter, Fullerton, CA). SE-HPLC measurements were carried out by isocratic elution on a Superdex 75 10/300 GL column (GE Healthcare, Piscataway, NJ). The buffer used was 100 mM sodium phosphate buffer (pH 7.1, σ=11.4 mS/cm) with 0.02% sodium azide. CE-HPLC measurements were carried out on a Bakerbond Wide-Pore (5 µm particle size) CBx column (Mallinckrodt Baker, Phillipsburg, NJ). The proteins were eluted in 20 mM Tris acetate containing 0.02% sodium azide (pH 6.5, σ=1.5 mS/cm) with a salt gradient of 0-100% 0.5 M sodium acetate (pH 6.5, σ=30.5 mS/cm) for 41 min. 2.3. Quasielastic Light Scattering. Quasielastic light scattering (QLS) was performed on a home-built optical system using a 35 mW He-Ne laser (Coherent, Santa Clara, CA), a custom-made scattering cell (Precision Detectors, Bellingham, MA), and a PD2000DLSPLUS 256 channel correlator (Precision Detectors, Bellingham, MA). The scattering angle was 90° and all measurements were carried out at 20±0.3 °C. The measured correlation functions were analyzed by a constrained regularization method as implemented in the PrecisionDeconvolve software (version 5.4) provided by Precision Detectors. This software computes the distribution of scattered intensity as a function of the diffusion coefficient. To convert the diffusion coefficient to a hydrodynamic radius,12 we took the viscosity of the solution to be that of water (1.002 mPa s). The QLS measurements were done in 275 mM acetate buffer (pH 4.5, σ=8.08 mS/cm) with 0.02% sodium azide, and all samples were filtered through an Anotop 10 0.02 µm filter (Whatman, Maidstone, UK).
7
2.4. Electrospray Ionization Mass Spectroscopy (ESIMS). Samples for mass spectrometry were dialyzed into deionized water using Ultrafree-MC centrifugal concentrators (3 kDa membrane; Millipore, Billerica, MA) and then filtered through an Ultra-MC centrifugal filter (0.22 µm; Millipore, Billerica, MA). Mass spectrometry was performed at the Laboratory for Macromolecular Analysis and Proteomics at the Albert Einstein College of Medicine of Yeshiva University. 2.5. Data Collection, Refinement and Structural Analysis. All X-ray diffraction data were recorded at beamline X6A (Brookhaven National Laboratory, National Synchrotron Light Source, Upton, NY, USA) between 13.5 and 15.1 keV. All data were recorded at 100 K using an ADSC Q210 CCD detector (Poway, CA, USA) at X6A and a MAR345 image plate (Mar Research, Norderstedt, Germany) at ID15A. Data were indexed, integrated and scaled in HKL 2000. 13 Lysozyme crystal structures were solved by molecular replacement using MOLREP28 and the model from the PDB entry 1IEE. Each model was refined by restrained maximum-likelihood refinement with REFMAC14,15 with individual anisotropic temperature factors and manual building performed in Coot.16 After the final refinement, stereochemistry of the structures were assessed with PROCHECK.17 2.6. Production of Crystals. Lysozyme protein was crystallized using a few different methods. Originally the lysozyme protein was dialyzed in an Amicon ultrafiltration cell with a 10 kDa membrane into 50 mM sodium phosphate buffer (pH 8.0; σ=7.56 mS/cm) with 0.002% sodium azide. It was then filtered through a 0.22 µm filter and concentrated up to approximately 100 mg/mL in a 3 kDa centrifugal filter unit (Millipore, Billerica, MA). Since the protein for Worthington is already > 99% pure and our HPLC results confirmed this purity, we eventually skipped the Amicon step and washed the protein directly in a 15 ml or 3 ml, 3 kDa centrifugal filter unit. We concentrated the protein in a centrifuge to about 200 µl, then added additional buffer to wash any small molecules out. This step was repeated three times to assure that any small molecules would be washed out of the protein. Lysozyme is very stable at high concentrations of protein, but only at low concentrations of salt (less than 300mM). If any turbidity or solid phase could be seen by eye, an aliquot was removed and inspected for crystals by bright field and polarized microscopy with an AxioImager A1m
8
microscope (Carl Zeiss, Gottingen, Germany). The first method of crystallization was using microbatch trays. Two different microbatch methods were used to produce crystals for X-ray diffraction studies. The first method was where 1 µL of precipitant was added to 1 µL of protein solution and then covered with approximately 15 µL of mineral oil in a microbatch 72-well plate (Hampton Research, Aliso Viejo, CA). We checked that the mineral oil is not miscible with MPD. The second method was “millibatch.” The only difference between the two methods is that an equal volume of protein and precipitant was added to an eppendorf vial and mixed, then 2µL of solution was pipetted into the 72 well microbatch tray. Only hydrophobic plates (cat. no. HR3-086) were used since we found that the crystals tended to stick to the hydrophilic plates. The plates were left at a fixed temperature (typically, 4 C or room temperature, i.e., 20±2 C) and inspected periodically by bright field microscopy with an AxioImager A1m microscope. Observations were carried out for up to 2 months. Crystals were harvested with mounted cryoloops (Hampton Research, Aliso Viejo, CA) and dipped in a cryoprotectant (usually, 10-30% glycerol solution with 0.5 M of the precipitant) immediately before the diffraction measurements. The other method of crystallization was adapted from Weiss et al.10 and employed a crystallization method of vapor diffusion using hanging drop trays. We used EasyXtal 15Well Tool trays (cat. no. 132006) purchased from Qiagen Sciences (Germantown, Maryland). This method is kinetically different from microbatch and allows for the hanging drop to slowly concentrate due to the diffusion of the water in the drop into the well. The protein was prepared the same way as described above. We filled the wells of the hanging drop trays with varying percentages of precipitant solution from 60 to 80% (v/v) MPD and 100 to 200 mM Tris; the volume on the reservoir was around 0.3 mL. The precipitant solution was prepared, filtered and allowed to cool in a cold room at 4 °C. In the cold room, the wells of the hanging drop trays were filled to prevent any condensation. During crystallization we combined equal volumes of well solution and protein that was then added to a small eppendorf vials and mixed. The final concentration of the drops was between 30 and 35 mg/mL of protein and 30 to 40% (v/v) MPD. 5 µL drops were added to the caps and then the wells and were secured tightly. All the conditions studied for both methods are tabulated in the Appendix.
9
3. Results and Discussion 3.1 Sample Purity. Purity is an important factor for protein crystallization. We rigorously checked the purity of our protein lysozyme. Figure 5 shows a SE-HPLC chromatogram of our protein. The single peak shows the homogeneity of our protein. There is a little tail at the end of the elution, but that is due to protein sticking to the column.
DAD-280nm Lys_Purity
500
500
0 0
5
10
mAU
1000
Retention Time
13.550
mAU
1000
0 15
20
25
30
Minutes
Figure 5. Size exclusion high-performance liquid chromatography of pure lysozyme. The single peak demonstrated the high level of purity. The slight tail is due to the fact that some protein sticks to the column and needs more time to elute.
We also examine the charge distribution of the protein using CE-HPLC (figure 6). Almost 92 percent of the protein is a single peak, also confirming the overall purity. The other small peaks may be due to slightly different charged states of the protein. ESIMS shows that the size of the large peak is 14306.0±1.4 Da in agreement with the theoretical molecular weight of lysozyme, which is 14305.1 Da.
DAD-280nm Lys_Check
Retention Time
500
500
0
10
20
30
40
45.767
36.650 37.117
0
38.883 40.133 40.850
mAU
1000
mAU
1000
0 50
Minutes
Figure 6. Cation-exchance high-performance liquid chromatography of pure lysozyme. The vast majority of the protein elutes at one time indicating its high level of purity.
10
QLS is a powerful technique to measure the RH (hydrodynamic radius) of a protein and is very sensitive to large molecules that scatter light. Figure 7 shows the result for our QLS measurements on lysozyme. The proteins were measured at concentration 1.79 mg/mL lysozyme and solution conditions 20 °C, 275 mM acetate buffer at pH 4.5 with 0.002% sodium azide. The hydrodynamic radius was measured to be 1.77nm. This is consistent with other groups. Parmar et al. measured the hydrodynamic radius to be 1.90 nm but their conditions are a little different. They concentrate lysozyme to 20 mg/mL,18 their larger size is probably due to the association between the protein molecules at high concentration. We can also compare our data for lysozyme with our previous data for thaumatin20 to show that, as expected with globular proteins, RH scales approximately as M1/3 (where M is the molecular weight of the protein).
3
M Lys 14.3 3 0.864 MTha 22.2
R H _ Lys R H _ Tha
1.77 0.808 2.19
The relatively good agreement (within 10%) of the two ratios demonstrates that the hydrodynamic radius we measure is a reasonable value.
Figure 7. The QLS data also confirms the purity of lysozyme in a sodium acetate solution. The X-axis is the hydrodynamic radius and the Y-axis is the normalized scattering intensity. The small hydrodynamic radius demonstrates the absence of large particles and the purity of our lysozyme.
3.2 Lysozyme and Sodium Phosphate. We initially decided to try crystallizing lysozyme
with MPD using a pH 8.0 sodium phosphate buffer using a similar procedure to the Weiss et
11
al.10 The motivation for using phosphate was that it is a very good buffer and its pH changes little with temperature. We chose the relatively high pH since the pI value (isoelectric point) for lysozyme is 10-11. As the pH approaches the pI value the charge of the protein is reduced, thereby decreasing the electrostatic repulsion between two protein molecules, which should favor crystallization. We tried a wide range of conditions but very few diffraction quality crystals grew (figure 8).
Figure 8. Figure 8A shows spherulites forming in regions of too much MPD and lysozyme in a hanging drop (53.2 mg/mL with 50% (v/v) RS-MPD). Figure 8B is the brown precipitate that forms if too much MPD is used (29.2 mg/mL with 50% (v/v) RS-MPD). Figure 8C is a single hexagonal crystal (40 mg/mL with 16.7% (v/v) RMPD). Figure 8D is the meta-stable LLPS (liquid-liquid phase separation) phase (43.2 mg/mL with 25% (v/v) S-MPD).
Figure 9 graphically displays the results of one set of conditions that was tried. (The complete set of results and conditions can be found in the Appendix.) In general, few crystals grew using the phosphate buffer. The crystals that did grow usually did not contain any MPD
12
in their structure. We investigated the phase diagram altering the concentration of protein, the percent of MPD in the solution and the molarity of the buffer used. Crystallizing with sodium phosphate at pH 8.0 produced a few different phases (figure 8). When we exceeded around 30% (v/v) MPD, an amorphous phase of rapid brown precipitation occurred (figure 8B). When we slightly reduced either protein concentration or the percentage of MPD, there was liquid-liquid phase separation (LLPS) in the wells (figure 8D). In principle, this is a metastable state, although, only rarely did we see crystals grow after the LLPS occurred. Next was a narrow region where we managed to grow a few crystals after a long period of time (figure 8C). Only some of the crystals that formed in that region were diffraction quality, but they too did not have any MPD in the crystals. Finally, there was also a large region of no growth at all.
Figure 9. The blue points indicate that no crystal grew. The brown points indicate brown precipitate. The red points are liquid-liquid phase separation (LLPS). The green points indicate where we did see crystal growth. The yellow point indicates an amorphous crystal.
3.3 Lysozyme and Sodium Chloride. We were not successful in growing crystals in sodium
phosphate and were concerned about the possibility that the protein itself was the problem. Unlike Weiss et al.10 and Michaux et al.11, we did not use lysozyme purchased from SigmaAlrich, but the purer protein from Worthington.19 Lysozyme is easily crystallized with NaCl in a pH 4.5 sodium acetate buffer. Under these conditions, we were able to grow crystals
13
with and without MPD (figure 10). Clearly our protein was not the problem.
A
B
Figure 10. Crystals grown with NaCl in sodium acetate at pH 4.5. Figure 10A is lysozyme in 1 M NaCl (37.5 mg/mL with 0% (v/v) MPD) and figure 10B is lysozyme crystallized with 1 M NaCl (43.3 mg/mL with 15% (v/v) MPD).
3.4 Literature Replication. After we failed to produce crystals using a phosphate buffer, we
switched to a pH 8.0 Tris Buffer, which Weiss et al.10 and Michaux et al.11 used to grow their crystals. Both papers showed that two chloride ions were in the structure of crystallized lysozyme. I thought that these chloride ions might be reducing repulsion between the positively charged lysozyme molecules in specific areas of the crystal that phosphate could not fit. This would allow for nucleation by increasing the screening of the protein molecules, which the phosphate buffer could not provide. Since the Tris buffer is titrated using HCl, this buffer would provide the chloride ions needed for screening.i It is also possible that the Tris interacts with lysozyme specifically in a way that phosphate cannot. Weiss et al.10 actually finds a Tris molecule in the structure of the crystallized lysozyme. 3.5 Lysozyme and Tris. We decided to follow the procedure of Weiss et al.9 more strictly.
They were able to grow crystals with drops initially at 10 mg/mL lysozyme and 35% (v/v) i We hadn’t quite exhausted our conditions using a phosphate buffer, for example, the pH. Although we tried a sodium phosphate buffer at pH 7 and 8, raising the pH to around 8.5 may have provided the necessary charge reduction of lysozyme to further reduce electrostatic repulsion and allow for crystal nucleation. This possibility remains to be tested in the future.
14
MPD in 50mM Tris. Ultimately, their approach was unsuccessful for us. Only after a month did we see a few small, poorly formed crystals, but most of the wells yielded nothing. The impurities in the Sigma-Aldrich lysozyme could be the source of heterogeneous nucleation and the reason that they were able to grow crystals, and we were unable to reproduce Weiss’s results. We have previously shown with thaumatin that protein purity is an important factor in crystallization.20 It was immediately apparent that the Tris buffer was affecting the phase behavior of lysozyme differently than phosphate. At room temperature (20 °C) the lysozyme in Tris was not nearly as soluble as it was in phosphate. High concentrations of protein in around 300mM Tris and above would cause irreversible precipitation since there was too much electrostatic screening. We also noticed that at 4 °C the rate of precipitation increased. We began by trying an array of static growing conditions using the microbatch method of crystallization. We had more success with this method over our previous efforts with phosphate, specifically at 200mM Tris. We did see regions of crystal growth, although we were not sure how many of those crystals actually contained MPD. We noticed an interesting competition in the microbatch wells. In the control well or wells with low percentages of MPD (between 0 to 25% [v/v]) lots of small crystals grew. Using a little more MPD there would be no, or very little crystal growth. But at higher levels of MPD, around 40% (v/v) we saw crystal growth again (figure 11).
15
Figure 11. The blue points indicate no crystal growth; the green points indicate crystals, the orange point indicates sticks, and the grey points indicate precipitate. There may be a kinetic competition only allowing crystals to grow at low or high concentrations of MPD, but not in between.
There seemed to be competing pathways of crystallization. At low percentages of MPD, the protein can diffuse in solution uninhibited, and probably grew into crystals without MPD. Since MPD is extremely viscous, it is possible that there is a region where the diffusion of the protein in solution is too slow to allow for any crystal growth. But at higher percentages of MPD, the MPD is abetting the crystallization and actually allowing the crystals to grow despite the high viscosity. There was no significant difference between the sizes of the crystals between these two conditions, although the crystals with more MPD were less defined (figure 12). This effect was also seen at room temperature (20 °C) and in the cold (4 °C) microbatch trays with both enantiomers. There is no visible difference in the crystals habits of these crystals. The tetragonal unit cell of lysozyme is the same whether there is MPD in the crystal or not. That makes it difficult to tell by eye whether one enantiomer of MPD affects the crystal formation. Although microbatch was successful in producing crystals, most were too small for diffraction or not well formed.
16
Figure 12. Figure 12A shows lysozyme crystallized in 200mM Tris (39.9 mg/mL with 25% (v/v) R-MPD) using the microbatch method. Figure 12B shows lysozyme crystallized at the same conditions with 40% (v/v) R-MPD.
It is know among crystallographers that different methods of crystallization can affect the growth of the crystal.5 This is logical because the crystallization kinetics of the hanging drop method is completely different than the microbatch method. The hanging drop method allows for the slow concentration of the protein by the diffusion of the drop into the well. Using hanging drop crystallization we were able to reproducibly grow large diffraction quality crystals (figure 13).
17
A
B
C
D
Figure 13. Crystal grown in hanging drop wells. Figure 13A shows lysozyme crystallized with S-MPD (31 mg/mL 30% (v/v) S-MPD). Figure 13B shows lysozyme crystallized with R-MPD (31 mg/mL with 30% (v/v) R-MPD). Figure 13C shows crystals growing out of the precipitate (35.4 mg/mL with 32.5% (v/v) R-MPD) (200 pixels are equal to 271.6 µm). Figure 13D shows lysozyme grown without any MPD (36 mg/mL with 0% (v/v) MPD).
We consistently noticed that our crystals grew better at 4 °C, so most of the hangingdrop wells were crystallized at that temperature. This is likely related to the effect of the temperature change on the Tris buffer. The pKa of Tris is sensitive to temperature. A pH 8.0 Tris buffer at room temperature (20 °C) becomes pH 8.62 at 0 °C. This further moves the pH closer to the pI value and reduces the charge of the positive lysozyme molecules. The weaker electrostatic repulsion may favor nucleation. The phase diagram for pH 8.62, at 200mM Tris, has not been completely investigated
18
but we gained an important insight into the sensitive kinetics of the crystallization method. Figure 14 shows our investigated areas on the phase diagram for lysozyme at these conditions. The black points represent areas of stable precipitation of tiny crystals. The red points indicate areas in which there was an initial metastable state of precipitation, but eventually larger crystals would begin to form out of the precipitation. This final state was clearly the true equilibrium in this region. The green points represent the region where crystals could grow without any precipitation. The blue points represent regions of no growth at all. The growth in the controls is also not always uniform. Eventually after time, both crystals and precipitate will grow at the same time indiscriminately, which does not occur in our hanging drops with MPD. We closely watched the crystal growth for over 40 days. The conditions in figure 14 are just the initial conditions of the drops on the hanging drop caps. The power of the hanging drop method is the fact that these conditions will slowly change over time. Theoretically, the conditions of the drops will eventually be twice the initial conditions (e.g., the protein concentration in the drop will double). The crystal may actually grow anywhere between the initial and final conditions. This makes it difficult to determine exactly the optimal conditions for growth, but once the conditions are determined, they can be tested using the more static microbatch method.
19
Figure 14. The sharp phase boundaries can easily be seen and again demonstrates the sensitivity of crystallization. The black points indicate precipitation. The red points indicate initial precipitation that eventually forms crystal. The green points indicate crystal growth without precipitation. The blue points indicate no crystal growth.
Unlike thaumatin crystals, it still remains unclear if there is a difference in the crystal habit between different enantiomers of MPD used during crystallization. This makes it hard to determine whether the enantiomers of MPD are having any structural influence on the growth of the crystal. During the growth of crystals with R-, S-, and RS-MPD, there was an initial structural difference. The crystals with R-MPD formed rounded crystals that would develop sharp edges after time. The crystals grown with S- and RS-MPD on the other hand did not have this initially rounded structure. These results need to be repeated to confirm the trend, but it may indicate that the MPD molecule has some difficulty finding the best conformation within the crystal. The X-ray data below helps corroborate this hypothesis. 3.6 X-ray Diffraction. The preliminary data seem to indicate that the chiral MPD displays a
slight difference within the structure of the crystal, although, it is unclear what role this may play in the solubility and habit of the crystallized lysozyme. Both Weiss et al.10 and Michaux et al.11 claim that they can clearly see R and S respectively in their crystals. It seems unlikely that they see the MPD molecule as clearly as they believe because the resolution on their
20
structures is 1.64 Å and 1.50 Å respectively. All the structures that we are using in this study have better resolution. The crystal structure using just R MPD has a very defined conformation in that position (figure 15A). The same is not the case for the crystals using S and RS MPD in the wells (figure 15B).
Figure 15. Figure 15A shows the electron density around R-MPD in crystallized lysozyme (resolution 1.45 Å). Figure 15B shows the density of RS-MPD (assigned S-MPD) in the structure of crystallized lysozyme (resolution 1.20 Å). Red areas indicate regions of too little density while green areas indicate too much density. The red lines represent oxygen atoms and the yellow or green lines show the carbon backbone of the protein. The hydrogen atoms are not shown. Figures 15C and 15D show the same structure as 15A and 15B respectively, without the electron density to clearly see the chiral MPD molecule within the crystal. The gray colored backbone indicates that it is part of a symmetry-related protein in the unit cell.
While the resolution of the crystal with RS-MPD (assigned S-MPD) is 1.20 Å, in the specific location of the MPD molecule the electron density is not well defined. In the case of the
21
lysozyme crystallized with RS-MPD, this lack of density could be an averaging of the two enantiomers throughout the crystal. But oddly we notice a similar uncertainty in the electron density when the lysozyme is just crystallized with S-MPD. This too could be an averaging of the possible conformations that the S-MPD takes within the crystal. MPD can rotate freely around its axis and it may be able to configure itself stably inside the crystal structure using multiple conformations. Although MPD is a chiral molecule, it is almost perfectly symmetric. In its most stable conformation the C4-O4 and C2-O2 bonds point in the same direction so that an intramolecular hydrogen bond can form between the two oxygen atoms (cf. R-MPD in figure 4). 9 It appears that MPD occupies this conformation in the protein crystal as well. In addition to the intramolecular hydrogen bonds, both R- and S-MPD form two intermolecular hydrogen bonds directly with neighboring lysozyme molecules, establishing a crystal contact between them. In R-MPD, the hydrogen bonds are between the hydrogen of the O4 atom on MPD and the carbonyl oxygen of Phe34 on the same lysozyme molecule. The other hydrogen bond is between the hydrogen on the O2 atom of MPD with the carbonyl oxygen of the Gly22 on the symmetric lysozyme molecule. In RS-MPD the same bonds are formed except that the O2 takes the place of O4 and vice-versa. The molecules of MDP in the structures of Weiss et al.10 and Michaux et al.11 make the same hydrogen bonds are we do here. The similarity between the bonds formed by the two enantiomers may explain why they report different enantiomers in their crystal structures. Under our specific growing conditions, the R-MPD enantiomer seems to naturally fit into the crystal lattice, based on the clear electron density surrounding the MPD molecule. The electron density for S- and RS-MPD is not nearly as defined. There are large errors in the location of MPD even though the protein backbone itself has a very high resolution. The errors could indicate that S- or RS-MPD has difficultly selecting the best conformation. The errors may be due to an averaging of different possible configurations within the crystal. Although the density of R-MPD seems to indicate that it takes a single conformation within the structure of the crystal, it does not clearly indicate that one enantiomer is kinetically favored; otherwise, we would expect to only see R-MPD in lysozyme crystallized with RSMPD. It is still unclear how lysozyme selects which enantiomer to use as the crystal contact. Nevertheless, this chiral effect may not be significant enough to influence the crystallization
22
as much as we had seen in the tartrate-thaumatin system.
4. Conclusions MPD is clearly an important precipitant for the crystallization of lysozyme. Crystals grown with MPD form much faster then crystals without MPD. This is because MPD makes a crystal contact between the unit cells that allows the protein to form order crystals. Although the enantiomers of MPD may cause an initial difference in crystal habit, it does not appear that the chirality MPD plays a significant role in the crystallization of lysozyme. The X-ray data show that each enantiomer can structure itself within the crystal, even though the exact conformation within the structure is not the same. The rotational freedom of the MPD molecule allows its OH groups to point in the same direction for either enantiomer so this may allow each to make the necessary crystal contacts and insert itself in the unit cell of the crystal. The R-MPD enantiomer has a more uniform conformation within the crystal, while S- and RS-MPD have more areas of uncertainty that could indicate averaging of different possible conformations. While the S-MPD enantiomer may be displacing some water molecules in the unit cell, the OH groups can make the same crystal contacts that allow both enantiomers to fit in the crystal. Even though there is little difference in the crystal habits with R- and S-MPD that does not mean there is not a difference in the kinetics of crystal growth. I would also measure the solubility of crystals crystallized with R- and S-MPD to see if there is a difference similar to thaumatin and tartrate. Our previous work on thaumatin showed that small changes in chirality can affect the conformation of the enantiomer in the crystal, and possibly the entire crystal. It is possible that if lysozyme were crystallized with a less-symmetric chiral molecule there would be a greater chiral affect. Molecules with two chiral centers (like tartrate) will be less symmetric and may play a larger role in the crystallization. Molecules that are more rigid than MPD may also be able to show a more significant difference in crystallization. In the future I would like to try and crystallize lysozyme with tartrate and test this theory. Another possibility is that other model proteins may be more sensitive to small changes in its structure. Therefore, I would try using MPD as a precipitant for others proteins such as ribonuclease A.
23
5. Acknowledgments I thank Professor Neer Asherie for his guidance throughout this entire project and his many useful conversations. I also thank the other members of the Asherie lab including Mark Stauber, Ari Greenbaum, Sam Blass and Charles Ginsberg. I thank the National Science Foundation (DMR # 0901260) for providing a summer stipend and the Jay and Jeanie Schottenstein Honors Program for additional funding. 1 http://targetdb.sbkb.org/statistics/TargetStatistics.html#general 2 Pande, A.; Pande, J.; Asherie, N.; Lomakin, A.; Ogun, O.; King, J.; Benedek, G. B.; Proc. Natl. Acad. Sci., USA 2001, 98, 6116–20. 3 A. McPherson, Crystallization of Biological Macromolecules, CSHL Press, Cold Spring Harbor, 1999. 4 Asherie, N. Methods 2004, 34, 266-272. 5 Chayen, N. E.; Curr. Opin. Struct. Biol. 2004, 14, 577–583. 6 Asherie N.; Jakoncic, J.; Ginsberg, C.; Greenbaum, A.; Stojanoff, V.; Hrnjez, B.; Blass S.; and Berger, J.; Cryst. Growth Des. 2009, 9, 4189-4198. 7 Johnson, L. N.; Nature 1998, 5 number 11, 1998 8 R. Huopalahti, R. López-Fandiño, M. Anton, Rüdiger Schade Bioactive Egg Compounds Chapter 6 2007. 9 Anand, K.; Pal, D.; and Hilgenfeld, R.; Acta Cryst. D, 2002, 58, 1722-1728 10 Weiss, M. S.; Palm, G. J.; and Hilgenfeld, R.; Acta Cryst. D, 2000 56, 952-958. 11 Michaux, C.; Pouyez, J.; Wouters, J.; Privé, G. G.; BMC Struct. Bio. 2008, 8, 29 12 Lomakin, A.; Teplow, D. B.; Benedek, G. B.; Methods Mol. Biol. 2005, 299, 153–174. 13 Otwinowski, Z.; Minor,W.; Methods Enzymol., 1997, 276, 307–326. 14 Murshudov, G. N.; Vagin, A. A.; Dodson, E. J.; Acta Crystallogr. D, 1997, 53, 240–255. 15 CCP4; Collaborative Computational Project, Number 4; Acta Crystallogr. D, 1994, 50, 760-763 16 Emsley, P.; Cowtan, K.; Acta Crystallogr. D, 2004, 60, 2126–2132. 17 Laskowski, R. A.; MacArthur, M. W.; Moss, D. S.; Thornton, J. M. J. Appl.; Crystallogr. 1993, 26, 283–291. 18 Parmar, A. S.; Muschol, M.; Biophys. 2009, 97, 590-598. 19 Parmar, A. S.; Gorrschall, P. E.; Muschol, M.; Biophys. Chem. 2007, 129, 224-234. 20 Asherie, N.; Ginsberg, C.; Greenbaum, A.; Blass, S.; Knafo, S.; Cryst. Growth Des. 2008, 8, 4200-4207.
24
Appendix All the conditions we attempted to crystallize lysozyme are presented bellow. The results describe what we saw in the wells after giving them more than 3 weeks to reach equilibrium. A backslash in the results section indicates that multiple results were seen. Lysozyme Crystallized with R-, S-, and RS-MPD with pH 8.0 10mM sodium phosphate using the microbatch method at 20 °C. Concentration of Protein (mg/mL) 73.1 73.1 48.8 48.8 24.4 24.4 43.2 28.8 14.4 43.6 43.6 42.6 42.6 38.3 38.3 0 43.5 0 43.5 0 43.5 43.5 39 39 39 39 0 39
Percent MPD (v/v) 0 25 0 50 0 75 25 50 75 0 33 0 16.7 0 25 10 10 20 20 30 30 0 40 30 20 10 40 0
Enantiomer Used
Results Nothing Brown Precipitate (BP) Nothing Brown Precipitate Nothing Brown Precipitate Nothing Brown Precipitate Brown Precipitate Nothing Brown Precipitate Nothing Brown Precipitate Nothing Nothing Nothing Nothing Nothing Nothing Nothing Small Sticks Nothing BP/Spherulites Small Sticks Nothing Nothing Nothing Nothing
R, S R, S R, S R, S R, S R, S R, S R, S R, S R, S R, S R, S R, S R, S R, S R, S, RS R, S, RS R, S, RS R, S, RS R, S, RS R, S, RS R, S, RS R, S, RS R, S, RS R, S, RS R, S, RS R, S, RS R, S, RS
25
Lysozyme Crystallized with R-, S-, and RS-MPD with pH 8.0 10mM sodium phosphate using the microbatch method at 4 °C. Concentration of Protein (mg/mL) 73.1 73.1 48.8 48.8 24.4 24.4 43.6 43.6 42.6 42.6 38.3 38.3 0 38.3 0 42.6 0 44.5 0 43.5 0 43.5 0 43.5 43.5 39 39 39 39 0 39
Percent MPD (v/v) 0 25 0 50 0 75 0 33 0 16.7 0 25 25 25 16.6 16.6 12.5 12.5 10 10 20 20 30 30 0 40 30 20 10 40 0
Enantiomer Used
Results Crystals Brown Precipitate Crystals Brown Precipitate Nothing Brown Precipitate Nothing Brown Precipitate Nothing LLPS Nothing Crystal/LLPS Nothing LLPS Nothing BP/LLPS/Crystal Nothing Crystal/LLPS Nothing Nothing/LLPS Nothing LLPS Nothing LLPS Nothing BP/LLPS LLPS Nothing/LLPS Nothing Nothing Nothing
R, S R, S R, S R, S R, S R, S R, S R, S R, S R, S R, S R, S R, S R, S R, S R, S R, S R, S R, S, RS R, S, RS R, S, RS R, S, RS R, S, RS R, S, RS R, S, RS R, S, RS R, S, RS R, S, RS R, S, RS R, S, RS R, S, RS
26
Lysozyme Crystallized with RS-MPD with sodium phosphate using the hanging drop method at 20 °C. Concentration of Protein (mg/mL) 25.6 25.6 25.6 25.6 25.6 26.6 26.6 26.6 26.6 26.6 30.9 30.9 30.9 30.9 30.9 29.2 29.2 29.2 29.2 29.2
Percent MPD (v/v) 0 10 15 20 25 0 10 15 20 25 0 10 15 20 25 0 10 15 20 25
Molatiry of Buffer (mM) 10 10 10 10 10 10 10 10 10 10 5 5 5 5 5 5 5 5 5 5
pH 7 7 7 7 7 8 8 8 8 8 7 7 7 7 7 7 8 8 8 8
27
Enantiomer
RS RS RS RS RS RS RS RS RS RS RS RS RS RS RS RS RS RS RS RS
Result Nothing Nothing Nothing Small sticks Precipitate Nothing Nothing Nothing Sticks Spherulites Nothing Nothing Small Sticks Nothing Precipitate Nothing Nothing Nothing Small Sticks Spherulites/Precipitate
Lysozyme Crystallized with RS-MPD with sodium phosphate using the hanging drop method at 4 °C. Concentration of Protein (mg/mL) 25.6 25.6 25.6 25.6 25.6 26.6 26.6 26.6 26.6 26.6 30.9 30.9 30.9 30.9 30.9 29.2 29.2 29.2 29.2 29.2
Percent MPD (v/v) 0 10 15 20 25 0 10 15 20 25 0 10 15 20 25 0 10 15 20 25
Molatiry of Buffer (mM) 10 10 10 10 10 10 10 10 10 10 5 5 5 5 5 5 5 5 5 5
pH
7 7 7 7 7 8 8 8 8 8 7 7 7 7 7 7 8 8 8 8
Enantiomer
RS RS RS RS RS RS RS RS RS RS RS RS RS RS RS RS RS RS RS RS
28
Result
Nothing Nothing Precipitate Precipitate Brown Precipitate Nothing Nothing Precipitate Precipitate/BP Precipitate/BP Nothing Nothing Precipitate/BP Precipitate/BP Brown Precipitate Nothing Nothing Precipitate Precipitate/BP Brown Precipitate
Lysozyme crystallized with R-, S-, and RS-MPD in H2O using the microbatch method at 20 °C. Some wells contain 0.002% sodium azide. Concentration of Protein (mg/mL) 38.5 38.5 29.2 29.2 19.2 19.2 71.8 50.7 36.3 29.3 71.8 0
Percent MPD (v/v) 0 33 0 50 0 66.7 50 40 30 20 0
Enantiomer Used R, S R, S R, S R, S R, S R, S R, S, RS R, S, RS R, S, RS R, S, RS R, S, RS
50 R, S, RS
Result Precipitate Brown Precipitate Precipitate Brown Precipitate Precipitate Amorphous Crystals Amorphous Crystals Amorphous Crystals Amorphous Crystals Nothing Nothing/Amorphous Crystal Nothing
Azide No No No No No No Yes Yes Yes Yes Yes Yes
Lysozyme crystallized with R-, S-, and RS-MPD in H2O using the microbatch method at 4 °C. Some wells contain 0.002% sodium azide. Concentration of Protein (mg/mL) 38.5 38.5 29.2 29.2 19.2 19.2 71.8 50.7 36.3 29.3 71.8 0
Percent MPD (v/v) 0 33 0 50 0 66.7 50 40 30 20 0 50
Enantiomer Used R, S R, S R, S R, S R, S R, S R, S, RS R, S, RS R, S, RS R, S, RS R, S, RS R, S, RS
Result Nothing Amorphous Crystals Nothing Crystals/nothing Nothing Nothing/Amorphous Crystals Nothing Nothing Nothing Nothing Nothing Nothing
29
Azide No No No No No No Yes Yes Yes Yes Yes Yes
Lysozyme crystallized with RS-MPD in H2O (with 0.002% sodium azide) using the hanging drop method at 20 °C. Concentration Percent of Protein MPD (v/v) (mg/mL) 52.2 0 52.2 30 52.2 35 52.2 40 52.2 45 26.1 0 26.1 30 26.1 35 26.1 40 26.1 45
Enantiomer Used
Result Nothing Nothing Nothing Brown Precipitate Brown Precipitate Nothing Nothing Nothing Brown Precipitate Brown Precipitate
RS RS RS RS RS RS RS RS
Lysozyme crystallized with RS-MPD in H2O (with 0.002% sodium azide) using the hanging drop method at 4 °C. Concentration of Protein (mg/mL) 52.2 52.2 52.2 52.2 52.2 26.1 26.1 26.1 26.1 26.1
Percent MPD (v/v)
Enantiomer Used
Result
0 30 RS 35 RS 40 45 0 30 35 40 45
Nothing Nothing Nothing Nothing Nothing Nothing Nothing Nothing Nothing Nothing
RS RS RS RS RS RS
30
Lysozyme crystallized with R-, S-, and RS-MPD in pH 8.0 Tris using the microbatch method at 20 °C. Concentration Percent Molatiry of of Protein MPD Buffer (mM) (mg/mL) (v/v)
39.5 39.5 39.5 39.5 39.5 0 39.5 39.5 39.5 39.5
0 30 35 45 50 50 0 30 35 40
39.5 39.5 0 39.5, 39.9 39.5, 39.9 39.5, 39.9 39.5, 39.9 39.5, 39.9 0 38.3 38.3 38.3 38.3 38.3 0 61.7 61.7 61.7 61.7 61.7 0
45 50 50 0 25 30 35 40 40 0 25 30 35 37.5 37.5 0 25 30 35 37.5 37.5
Enantiomer Used
Results
400 400 400 400 400 400 300 300 300
RS RS RS RS RS RS RS RS RS
Nothing LLPS LLPS/Crystal Crystals Crystals Crystals/Spherulites Nothing LLPS LLPS/Crystal
300 300 300 300 200 200 200 200 200 200 50 50 50 50 50 50 50 50 50 50 50 50
RS RS RS RS R, S R, S R, S R, S R, S R, S R, S R, S R, S R, S R, S R, S R, S R, S R, S R, S R, S R, S
LLPS/Crystal Crystals LLPS/Crystal Nothing Nothing Nothing/Precipitate LLPS/Crystal LLPS LLPS Nothing Nothing Nothing Nothing Nothing Nothing Nothing Nothing Nothing Nothing Nothing Nothing Nothing
31
Lysozyme crystallized with R-, S-, and RS-MPD in pH 8.6 Tris using the microbatch method at 4 °C. Concentration of Protein (mg/mL) 39.5 39.5 39.5 39.5 39.5 39.5 0 39.5 39.5 39.5 39.5 39.5 39.5 0 40.9, 39.5, 39.9 40.9, 39.5, 39.9 40.9, 39.5, 39.9 40.9, 39.5, 39.9 40.9, 39.5, 39.9 0 38.3 38.3 38.3 38.3 38.3 0 61.7 61.7 61.7 61.7 61.7 0
Percent Molatiry of MPD Buffer (mM) (v/v) 0 400 30 400 35 400 40 400 45 400 50 400 50 400 0 300 30 300 35 300 40 300 45 300 50 300 50 300 0 200 25 200 30 200 35 200 40 200 40 200 0 50 25 50 30 50 35 50 37.5 50 37.5 50 0 50 25 50 30 50 35 50 37.5 50 37.5 50
32
Enantiomer Used RS RS RS RS RS RS RS RS RS RS RS RS RS RS R, S R, S R, S R, S R, S R, S R, S R, S R, S R, S R, S R, S R, S R, S R, S R, S R, S R, S
Results Crystals/Precipitate Crystals Crystals Crystals Precipitate Crystals/Precipitate Nothing Nothing Nothing Precipitate Sticks Sticks/Crystal Crystal/Precipitate Nothing Crystals/Precipitate Crystals/Precipitate Nothing/Crystals Sticks/Crystals Crystals Nothing Crystals Nothing/Crystals Nothing Nothing/Crystals Sticks Nothing Crystals Nothing Nothing Nothing/Crystals Nothing Nothing
Lysozyme crystallized with RS-MPD in pH 8.0 Tris using the millibatch method at 20 °C. Concentration of Protein (mg/mL)
Percent MPD (v/v)
38.8 38.8 38.8 38.8 38.8 0
0 25 30 35 40 40
Molatiry of Buffer (mM) 200 200 200 200 200 200
Enantiomer Used RS RS RS RS RS RS
Results
Nothing/Crystals Nothing Nothing Nothing Nothing Nothing
Lysozyme crystallized with RS-MPD in pH 8.6 Tris using the millibatch method at 4 °C.
Concentration of Protein (mg/mL)
Percent MPD (v/v)
38.8 38.8 38.8 38.8 38.8
0 25 30 35 40
Molatiry of Buffer (mM)
200 200 200 200 200
33
Enantiomer Used
RS RS RS RS RS
Results
Crystals Nothing Nothing Sticks/Nothing Sticks/Crystals
Lysozyme crystallized with RS-MPD in pH 8.0 Tris using the hanging drop method at 20 °C. Concentration of Protein (mg/mL)
Percent MPD (v/v)
10 10 10 10 10 36.6 36.6 36.6 37.1 37.1
27.5 30 32.5 35 37.5 0 35 40 0 35
37.1
40
Molatiry of Buffer (mM) 50 50 50 50 50 100 100 100 200 200
Enantiomer Used RS RS RS RS RS RS RS RS RS RS
200 RS
34
Results
Nothing Nothing Nothing Light Precipitate Light Precipitate Nothing Precipitate Precipitate Nothing Crystals Growing from Precipitate (CGP) Precipitate
Lysozyme crystallized with R-, S-, and RS-MPD in pH 8.6 Tris using the hanging drop method at 4 °C.
Concentration of Protein (mg/mL) 10 10 10 10 10 36.6 36.6
Percent MPD (v/v)
36.6 37.1 37.1
40 0 35
100 RS 200 RS 200 RS
37.1 0 35.2 35.2 35.2 35.2 28.4
40 35 0 35 30 25 35
200 200 200 200 200 200 200
28.4 28.4 23
30 25 35
200 RS 200 RS 200 RS
23 23 36 36 36 31.5 31.5 31.5 35.4 35.4
30 20 0 35 30 0 35 30 0 32.5
200 200 200 200 200 200 200 200 200 200
35.4 31
30 32.5
200 R, S 200 R, S
31
30
200 R, S
27.5 30 32.5 35 37.5 0 35
Molatiry of Buffer (mM) 50 50 50 50 50 100 100
Enantiomer Used RS RS RS RS RS RS RS
RS RS RS RS RS RS RS
RS RS R, S R, S R, S R, S R, S R, S R, S R, S
35
Results
Nothing Nothing Light Precipitate Light Precipitate Light Precipitate CGP/Precipitate Crystals Growing from Precipitate Precipitate Crystals Crystals Growing from Precipitate Precipitate Nothing Nothing CGP/Precipitate Crystals Nothing Crystals Growing from Precipitate Crystals Nothing Crystals Growing from Precipitate Crystals Sticks Crystals/Precipitate Precipitate Crystals Crystals/Precipitate Crystals Crystals Nothing Crystals Growing from Precipitate Crystals Crystals Growing from Precipitate Crystals
Yeshiva College Yeshiva University May 2011
Jacob Berger Mentor: Professor Neer Asherie, Departments of Physics and Biology
Abstract Protein crystallization has been studied for over 150 years but is still not well understood. Protein crystals are extremely difficult to grow since protein crystallization is sensitive to many parameters. In this thesis I investigate if the chirality of a precipitant, 2methyl-2,4-pentanediol (MPD), influences the crystal growth of the protein lysozyme. It appears that R-MPD has a single conformation within the crystal, while S-MPD may take on a few different conformations. Although there appears to be a chiral difference within the structure, this does not seem to dramatically affect the kinetics of crystal growth as both crystal structures form in the same period of time.
1. Introduction Proteins control many of the processes needed for organisms to function. Knowing the structure of proteins is essential to understand how they help regulate life, and also how mutations of proteins can harm life. Only by knowing the structure can the function of the protein be fully determined. While determining the function of the protein may seem like the more difficult task, determining the structure of proteins is, in fact, a much greater problem. The structure of a protein can be determined by x-ray diffraction of a crystallized protein. The difficulty is that conditions needed to crystallize proteins are still not well understood. There is no general phase diagram for proteins (like those that exists for homogeneous structures such as water or iron), the only way to actually crystallize proteins is by an arduous and lengthy guess and check method of trying many different experimental conditions. Proteins are long, complex polymers of amino acids. This makes it exceedingly difficult to understand how to crystallize them. In principle, each protein could crystallize under different conditions and have a unique phase diagram. Currently, no systematic procedure has been found that will guarantee protein crystals for a protein that has never been crystallized before. Fewer than three percent of proteins from cloned genes have had their structure determined. 1 Growing diffraction quality crystals represent the largest challenge facing crystallographers. This inability to effectively crystallize proteins limits the biomedical and industrial applications. Some diseases, such as genetic cataracts, are caused by the crystallization of
1
proteins in the body.2 If the phase diagram were known for all proteins, a method could be employed to solubilize the proteins and reverse the effects of these diseases. Industrially, crystallization is one of the most effective forms of purification. Knowing each protein’s phase behavior would revolutionize the industrial purification process. At the moment, there appears to be no general phase diagram for all proteins. Crystallization of proteins is not a recent problem; scientists have attempted to crystallize proteins for over 150 years. 3 The first documented crystallized protein was hemoglobin from an earthworm, published in 1840 by Hünefeld. He demonstrated that protein crystals could be obtained from the slow and controlled dehydration of concentrated protein – a technique still in use today. For a decade no additional method was developed to reproducibly grow crystals of hemoglobin until Fünke in 1851. He employed methods of dissolving protein in purified water or organic solvents, also a commonly used method today. Thirty years later, the next advances were made. Gruber, Ritthausen, and Osborne, developed new methods of crystallization. They were the first to use temperature as a controlled variable and dialysis against low ionic strength solutions. They also further studied the use of organic solvents as precipitating agents. Slowly, crystallographers began to realize the sensitivity of crystallization to many different conditions. Crystallizing proteins is still extremely difficult, and the methods developed 150 years ago remain in use today. The phase diagram for proteins is a multi-dimensional graph that depends of many variables including protein concentration, pH, temperature, ionic strength, and the precipitant used. Crystallization is extremely sensitive to small changes in any of these conditions. Changing the pH or concentration slightly can be the difference between highly ordered crystals and amorphous precipitate. Combining the perfect combination of conditions to get crystals is often mostly luck. The general theory behind protein crystallization is to allow a supersaturated solution of protein to precipitate as ordered crystals in order to reach equilibrium with the solvent (figure 1). In the under-saturated zone, crystals will never form because the most stable phase is the solution phase. If the protein is only slightly super-saturated, it may remain supersaturated in the metastable zone due to the free energy barrier that prevents nucleation, although crystals can grow.4 In the nucleation zone a critical nucleus can form, a prerequisite for crystal growth. If the solution is overly super-saturated, then uncontrolled precipitation
2
may occur, resulting in an amorphous solid. The boundaries shown schematically in figure 1 represent regions to investigate for possible crystal growth. The actual lines on a protein phase diagram may be completely different depending on the crystallization conditions. Often there is a large region of no growth and uncontrolled amorphous precipitation, but a very small, narrow region of optimal crystal growth. There may never even be a region where ordered crystals can grow, as we have seen with some of our experiments.
Figure 1. Schematic illustration of a protein crystallization phase diagram. Adjustable parameters include precipitant or additive concentration, pH and temperature. The arrows indicate different paths crystal growth will take depending on the method. The two methods used in this thesis are microbatch (i) and hanging drop (ii). The filled black circles represent the starting conditions. The solubility is defined as the concentration of protein in the solute that is in equilibrium with crystals. The supersolubility curve is defined as the line separating conditions under which spontaneous nucleation (or phase separation or precipitation) occurs from those under which the crystallization solution remains clear if left undisturbed.5
Our group has recently shown that even the chirality of the precipitant used can greatly affect the crystallization process. Chiral molecules have the same chemical groups, but are mirror images of each other and non-superposable (figure 2). Chirality is derived from the Greek word for hand and refers to the “handedness” of the molecule. The most commonly used examples for chirality are left and right hands since they are mirror images of each other but cannot be superposed onto one another.
3
Figure 2. A chiral molecule and its mirror image.
Our group initially established the important role of chirality in protein crystallization when we investigated the solubility of thaumatin and sodium tartrate. We noticed an apparent contradiction in the literature: some groups claimed the solubility of the protein thaumatin was direct (increases with temperature), while other groups said it was retrograde (decreases with temperature). This difference occurs because L- and D- tartrate have different effects on the solubility of thaumatin. We measured the solubility of thaumatin with each enantiomer of tartrate and resolved the apparent contradiction. We showed that thaumatin crystallized with L-tartrate
has direct solubility while thaumatin crystallized with D-tartrate displays retrograde
solubility. The different enantiomers not only altered the thermodynamics of crystallization, but the kinetics of crystallization as well. For example,
L-
and
D-tartrate
both lead to the
formation of bipyramidal crystals of thaumatin; with L-tartrate these crystals form in hours, while in D-tartrate they form in days. Using the separate crystals grown from each of the enantiomers, the high-resolution x-ray diffraction data determined that the different enantiomers of tartrate take different conformations within the unit cell, therefore leading to completely different crystallization kinetics (figure 3). The L-tartrate (figure 3A) only takes on one conformation within the crystal, while conformations within the structure of thaumatin.
4
D-tartrate
(figure 3B) can take on two
Figure 3. L-tartrate (A) only takes on one conformation while D-tartrate (B) takes on two conformations. The dashed red and black lines denote the bonds for the two conformers of D-tartrate.6
While these findings are extraordinarily interesting, they may be specific to thaumatin and tartrate. If our results were generalized to other protein-precipitant systems, it would further elucidate the conditions and considerations needed for effective crystallization of proteins. Lysozyme is one of the most studied proteins, which makes it a good candidate to investigate the affects of precipitant chirality. Lysozyme is an ideal model protein because it is relatively inexpensive and easy to crystallize. Hen egg white lysozyme (muramidase or Nacetlymuramide glycanohydrolase)8 was the second protein and the first enzyme to have its structure determined by x-ray diffraction.7 Lysozyme also has had its structure determined to the highest resolution (0.65 Å) of any protein in the Protein Data Bank (PDB ID 2VB1); the tetragonal structure we study here has been determined to 0.94 Å (PDB ID 1IEE). Lysozyme is a relatively small secretory enzyme that catalyzes the hydrolysis of specific polysaccharides contained in cell walls of bacteria. Modified lysozyme molecule is used in many areas, such as to prevent infections and simultaneously to act as a natural antibiotic or stimulant of the immune system. Lysozyme from hen’s egg white is a polypeptide of 129 amino acid residues having a molecular weight of 14.3 kDa. Lysozyme is a strongly basic protein and with an isoelectric point (pI) of 10–11. Because of the high isoelectric point, lysozyme binds to negatively charged residues. Four disulfide bonds of the single polypeptide chain of the enzyme make this small protein molecule unusually compact and highly stable.8
5
MPD is the most commonly used additive in crystallization. 9 MPD promotes stabilization of the protein by preferential hydration, which is facilitated by attachment of MPD molecules to the hydrophobic surface. In crystallization experiments, MPD can act as precipitant through a combination of activities, including competition for water, hydrophobic exclusion of protein solutes, and lowering of the solution dielectric. The dielectric constant of pure MPD is 25. MPD is a chiral molecule (figure 4). The only high purity form of MPD that is commercially available is the racemate (RS-MPD, an equimolar mixture of R- and SMPD).
Figure 4. R and S enantiomers of MPD. The hydrogen atoms are not shown.8
Lysozyme has previously been crystallized with MPD, but as was the case with thaumatin, the previously published results are inconsistent. Weiss et al. reports that they only find the R enantiomer of MPD in their crystals (PDB ID 1DPW).10 Michaux et al. on the other hand, publish that they only find the S enantiomer of MPD in their crystals (PDB ID 3B72).11 In both cases the investigators use the racemic mixture of MPD. In this thesis I crystallize lysozyme with the separate enantiomers of MPD as wells as the racemic mixture. I investigate whether the chirality of the precipitant affects the crystallization of the protein.
2. Experimental Section 2.1. Materials. Lysozyme (Batch 36P9210) was purchased from Worthington Biochemical Cooperation. Purity was checked using High Performance Liquid Chromatography (HPLC) and quasielastic light scattering (QLS). R-MPD and S-MPD were synthesized by RCA separations (Freiburg, Germany). Enantiomeric purity was greater than 99%. RS-MPD (CAS no. 107-41-5), Sodium azide (cat. no. S227), monobasic sodium phosphate (cat. no. BP330), dibasic sodium phosphate (cat. no. BP332), sodium acetate (cat. no. BP334), Tris base (cat.
6
no. BP512), sodium chloride (cat. no. BP512), and sodium hydroxide (cat. no. BP359) were purchased from Fisher Scientific, Pittsburgh, PA. All materials were used without further purification. Deionized water was obtained from an Integral 3 deionization system (Millipore, Billerica, MA). Solutions were filtered through a Nalgene disposable 0.22 µm filter unit (Nalge Nunc International, Rochester, NY) prior to use. 2.2. Size-Exclusion and Cation-Exchange Chromatography. Both size-exclusion and cation-exchange high-performance liquid chromatography (SE-HPLC and CE-HPLC, respectively) were carried out on a Beckman System Gold apparatus at a flow rate of 1 mL/min (Beckman Coulter, Fullerton, CA). SE-HPLC measurements were carried out by isocratic elution on a Superdex 75 10/300 GL column (GE Healthcare, Piscataway, NJ). The buffer used was 100 mM sodium phosphate buffer (pH 7.1, σ=11.4 mS/cm) with 0.02% sodium azide. CE-HPLC measurements were carried out on a Bakerbond Wide-Pore (5 µm particle size) CBx column (Mallinckrodt Baker, Phillipsburg, NJ). The proteins were eluted in 20 mM Tris acetate containing 0.02% sodium azide (pH 6.5, σ=1.5 mS/cm) with a salt gradient of 0-100% 0.5 M sodium acetate (pH 6.5, σ=30.5 mS/cm) for 41 min. 2.3. Quasielastic Light Scattering. Quasielastic light scattering (QLS) was performed on a home-built optical system using a 35 mW He-Ne laser (Coherent, Santa Clara, CA), a custom-made scattering cell (Precision Detectors, Bellingham, MA), and a PD2000DLSPLUS 256 channel correlator (Precision Detectors, Bellingham, MA). The scattering angle was 90° and all measurements were carried out at 20±0.3 °C. The measured correlation functions were analyzed by a constrained regularization method as implemented in the PrecisionDeconvolve software (version 5.4) provided by Precision Detectors. This software computes the distribution of scattered intensity as a function of the diffusion coefficient. To convert the diffusion coefficient to a hydrodynamic radius,12 we took the viscosity of the solution to be that of water (1.002 mPa s). The QLS measurements were done in 275 mM acetate buffer (pH 4.5, σ=8.08 mS/cm) with 0.02% sodium azide, and all samples were filtered through an Anotop 10 0.02 µm filter (Whatman, Maidstone, UK).
7
2.4. Electrospray Ionization Mass Spectroscopy (ESIMS). Samples for mass spectrometry were dialyzed into deionized water using Ultrafree-MC centrifugal concentrators (3 kDa membrane; Millipore, Billerica, MA) and then filtered through an Ultra-MC centrifugal filter (0.22 µm; Millipore, Billerica, MA). Mass spectrometry was performed at the Laboratory for Macromolecular Analysis and Proteomics at the Albert Einstein College of Medicine of Yeshiva University. 2.5. Data Collection, Refinement and Structural Analysis. All X-ray diffraction data were recorded at beamline X6A (Brookhaven National Laboratory, National Synchrotron Light Source, Upton, NY, USA) between 13.5 and 15.1 keV. All data were recorded at 100 K using an ADSC Q210 CCD detector (Poway, CA, USA) at X6A and a MAR345 image plate (Mar Research, Norderstedt, Germany) at ID15A. Data were indexed, integrated and scaled in HKL 2000. 13 Lysozyme crystal structures were solved by molecular replacement using MOLREP28 and the model from the PDB entry 1IEE. Each model was refined by restrained maximum-likelihood refinement with REFMAC14,15 with individual anisotropic temperature factors and manual building performed in Coot.16 After the final refinement, stereochemistry of the structures were assessed with PROCHECK.17 2.6. Production of Crystals. Lysozyme protein was crystallized using a few different methods. Originally the lysozyme protein was dialyzed in an Amicon ultrafiltration cell with a 10 kDa membrane into 50 mM sodium phosphate buffer (pH 8.0; σ=7.56 mS/cm) with 0.002% sodium azide. It was then filtered through a 0.22 µm filter and concentrated up to approximately 100 mg/mL in a 3 kDa centrifugal filter unit (Millipore, Billerica, MA). Since the protein for Worthington is already > 99% pure and our HPLC results confirmed this purity, we eventually skipped the Amicon step and washed the protein directly in a 15 ml or 3 ml, 3 kDa centrifugal filter unit. We concentrated the protein in a centrifuge to about 200 µl, then added additional buffer to wash any small molecules out. This step was repeated three times to assure that any small molecules would be washed out of the protein. Lysozyme is very stable at high concentrations of protein, but only at low concentrations of salt (less than 300mM). If any turbidity or solid phase could be seen by eye, an aliquot was removed and inspected for crystals by bright field and polarized microscopy with an AxioImager A1m
8
microscope (Carl Zeiss, Gottingen, Germany). The first method of crystallization was using microbatch trays. Two different microbatch methods were used to produce crystals for X-ray diffraction studies. The first method was where 1 µL of precipitant was added to 1 µL of protein solution and then covered with approximately 15 µL of mineral oil in a microbatch 72-well plate (Hampton Research, Aliso Viejo, CA). We checked that the mineral oil is not miscible with MPD. The second method was “millibatch.” The only difference between the two methods is that an equal volume of protein and precipitant was added to an eppendorf vial and mixed, then 2µL of solution was pipetted into the 72 well microbatch tray. Only hydrophobic plates (cat. no. HR3-086) were used since we found that the crystals tended to stick to the hydrophilic plates. The plates were left at a fixed temperature (typically, 4 C or room temperature, i.e., 20±2 C) and inspected periodically by bright field microscopy with an AxioImager A1m microscope. Observations were carried out for up to 2 months. Crystals were harvested with mounted cryoloops (Hampton Research, Aliso Viejo, CA) and dipped in a cryoprotectant (usually, 10-30% glycerol solution with 0.5 M of the precipitant) immediately before the diffraction measurements. The other method of crystallization was adapted from Weiss et al.10 and employed a crystallization method of vapor diffusion using hanging drop trays. We used EasyXtal 15Well Tool trays (cat. no. 132006) purchased from Qiagen Sciences (Germantown, Maryland). This method is kinetically different from microbatch and allows for the hanging drop to slowly concentrate due to the diffusion of the water in the drop into the well. The protein was prepared the same way as described above. We filled the wells of the hanging drop trays with varying percentages of precipitant solution from 60 to 80% (v/v) MPD and 100 to 200 mM Tris; the volume on the reservoir was around 0.3 mL. The precipitant solution was prepared, filtered and allowed to cool in a cold room at 4 °C. In the cold room, the wells of the hanging drop trays were filled to prevent any condensation. During crystallization we combined equal volumes of well solution and protein that was then added to a small eppendorf vials and mixed. The final concentration of the drops was between 30 and 35 mg/mL of protein and 30 to 40% (v/v) MPD. 5 µL drops were added to the caps and then the wells and were secured tightly. All the conditions studied for both methods are tabulated in the Appendix.
9
3. Results and Discussion 3.1 Sample Purity. Purity is an important factor for protein crystallization. We rigorously checked the purity of our protein lysozyme. Figure 5 shows a SE-HPLC chromatogram of our protein. The single peak shows the homogeneity of our protein. There is a little tail at the end of the elution, but that is due to protein sticking to the column.
DAD-280nm Lys_Purity
500
500
0 0
5
10
mAU
1000
Retention Time
13.550
mAU
1000
0 15
20
25
30
Minutes
Figure 5. Size exclusion high-performance liquid chromatography of pure lysozyme. The single peak demonstrated the high level of purity. The slight tail is due to the fact that some protein sticks to the column and needs more time to elute.
We also examine the charge distribution of the protein using CE-HPLC (figure 6). Almost 92 percent of the protein is a single peak, also confirming the overall purity. The other small peaks may be due to slightly different charged states of the protein. ESIMS shows that the size of the large peak is 14306.0±1.4 Da in agreement with the theoretical molecular weight of lysozyme, which is 14305.1 Da.
DAD-280nm Lys_Check
Retention Time
500
500
0
10
20
30
40
45.767
36.650 37.117
0
38.883 40.133 40.850
mAU
1000
mAU
1000
0 50
Minutes
Figure 6. Cation-exchance high-performance liquid chromatography of pure lysozyme. The vast majority of the protein elutes at one time indicating its high level of purity.
10
QLS is a powerful technique to measure the RH (hydrodynamic radius) of a protein and is very sensitive to large molecules that scatter light. Figure 7 shows the result for our QLS measurements on lysozyme. The proteins were measured at concentration 1.79 mg/mL lysozyme and solution conditions 20 °C, 275 mM acetate buffer at pH 4.5 with 0.002% sodium azide. The hydrodynamic radius was measured to be 1.77nm. This is consistent with other groups. Parmar et al. measured the hydrodynamic radius to be 1.90 nm but their conditions are a little different. They concentrate lysozyme to 20 mg/mL,18 their larger size is probably due to the association between the protein molecules at high concentration. We can also compare our data for lysozyme with our previous data for thaumatin20 to show that, as expected with globular proteins, RH scales approximately as M1/3 (where M is the molecular weight of the protein).
3
M Lys 14.3 3 0.864 MTha 22.2
R H _ Lys R H _ Tha
1.77 0.808 2.19
The relatively good agreement (within 10%) of the two ratios demonstrates that the hydrodynamic radius we measure is a reasonable value.
Figure 7. The QLS data also confirms the purity of lysozyme in a sodium acetate solution. The X-axis is the hydrodynamic radius and the Y-axis is the normalized scattering intensity. The small hydrodynamic radius demonstrates the absence of large particles and the purity of our lysozyme.
3.2 Lysozyme and Sodium Phosphate. We initially decided to try crystallizing lysozyme
with MPD using a pH 8.0 sodium phosphate buffer using a similar procedure to the Weiss et
11
al.10 The motivation for using phosphate was that it is a very good buffer and its pH changes little with temperature. We chose the relatively high pH since the pI value (isoelectric point) for lysozyme is 10-11. As the pH approaches the pI value the charge of the protein is reduced, thereby decreasing the electrostatic repulsion between two protein molecules, which should favor crystallization. We tried a wide range of conditions but very few diffraction quality crystals grew (figure 8).
Figure 8. Figure 8A shows spherulites forming in regions of too much MPD and lysozyme in a hanging drop (53.2 mg/mL with 50% (v/v) RS-MPD). Figure 8B is the brown precipitate that forms if too much MPD is used (29.2 mg/mL with 50% (v/v) RS-MPD). Figure 8C is a single hexagonal crystal (40 mg/mL with 16.7% (v/v) RMPD). Figure 8D is the meta-stable LLPS (liquid-liquid phase separation) phase (43.2 mg/mL with 25% (v/v) S-MPD).
Figure 9 graphically displays the results of one set of conditions that was tried. (The complete set of results and conditions can be found in the Appendix.) In general, few crystals grew using the phosphate buffer. The crystals that did grow usually did not contain any MPD
12
in their structure. We investigated the phase diagram altering the concentration of protein, the percent of MPD in the solution and the molarity of the buffer used. Crystallizing with sodium phosphate at pH 8.0 produced a few different phases (figure 8). When we exceeded around 30% (v/v) MPD, an amorphous phase of rapid brown precipitation occurred (figure 8B). When we slightly reduced either protein concentration or the percentage of MPD, there was liquid-liquid phase separation (LLPS) in the wells (figure 8D). In principle, this is a metastable state, although, only rarely did we see crystals grow after the LLPS occurred. Next was a narrow region where we managed to grow a few crystals after a long period of time (figure 8C). Only some of the crystals that formed in that region were diffraction quality, but they too did not have any MPD in the crystals. Finally, there was also a large region of no growth at all.
Figure 9. The blue points indicate that no crystal grew. The brown points indicate brown precipitate. The red points are liquid-liquid phase separation (LLPS). The green points indicate where we did see crystal growth. The yellow point indicates an amorphous crystal.
3.3 Lysozyme and Sodium Chloride. We were not successful in growing crystals in sodium
phosphate and were concerned about the possibility that the protein itself was the problem. Unlike Weiss et al.10 and Michaux et al.11, we did not use lysozyme purchased from SigmaAlrich, but the purer protein from Worthington.19 Lysozyme is easily crystallized with NaCl in a pH 4.5 sodium acetate buffer. Under these conditions, we were able to grow crystals
13
with and without MPD (figure 10). Clearly our protein was not the problem.
A
B
Figure 10. Crystals grown with NaCl in sodium acetate at pH 4.5. Figure 10A is lysozyme in 1 M NaCl (37.5 mg/mL with 0% (v/v) MPD) and figure 10B is lysozyme crystallized with 1 M NaCl (43.3 mg/mL with 15% (v/v) MPD).
3.4 Literature Replication. After we failed to produce crystals using a phosphate buffer, we
switched to a pH 8.0 Tris Buffer, which Weiss et al.10 and Michaux et al.11 used to grow their crystals. Both papers showed that two chloride ions were in the structure of crystallized lysozyme. I thought that these chloride ions might be reducing repulsion between the positively charged lysozyme molecules in specific areas of the crystal that phosphate could not fit. This would allow for nucleation by increasing the screening of the protein molecules, which the phosphate buffer could not provide. Since the Tris buffer is titrated using HCl, this buffer would provide the chloride ions needed for screening.i It is also possible that the Tris interacts with lysozyme specifically in a way that phosphate cannot. Weiss et al.10 actually finds a Tris molecule in the structure of the crystallized lysozyme. 3.5 Lysozyme and Tris. We decided to follow the procedure of Weiss et al.9 more strictly.
They were able to grow crystals with drops initially at 10 mg/mL lysozyme and 35% (v/v) i We hadn’t quite exhausted our conditions using a phosphate buffer, for example, the pH. Although we tried a sodium phosphate buffer at pH 7 and 8, raising the pH to around 8.5 may have provided the necessary charge reduction of lysozyme to further reduce electrostatic repulsion and allow for crystal nucleation. This possibility remains to be tested in the future.
14
MPD in 50mM Tris. Ultimately, their approach was unsuccessful for us. Only after a month did we see a few small, poorly formed crystals, but most of the wells yielded nothing. The impurities in the Sigma-Aldrich lysozyme could be the source of heterogeneous nucleation and the reason that they were able to grow crystals, and we were unable to reproduce Weiss’s results. We have previously shown with thaumatin that protein purity is an important factor in crystallization.20 It was immediately apparent that the Tris buffer was affecting the phase behavior of lysozyme differently than phosphate. At room temperature (20 °C) the lysozyme in Tris was not nearly as soluble as it was in phosphate. High concentrations of protein in around 300mM Tris and above would cause irreversible precipitation since there was too much electrostatic screening. We also noticed that at 4 °C the rate of precipitation increased. We began by trying an array of static growing conditions using the microbatch method of crystallization. We had more success with this method over our previous efforts with phosphate, specifically at 200mM Tris. We did see regions of crystal growth, although we were not sure how many of those crystals actually contained MPD. We noticed an interesting competition in the microbatch wells. In the control well or wells with low percentages of MPD (between 0 to 25% [v/v]) lots of small crystals grew. Using a little more MPD there would be no, or very little crystal growth. But at higher levels of MPD, around 40% (v/v) we saw crystal growth again (figure 11).
15
Figure 11. The blue points indicate no crystal growth; the green points indicate crystals, the orange point indicates sticks, and the grey points indicate precipitate. There may be a kinetic competition only allowing crystals to grow at low or high concentrations of MPD, but not in between.
There seemed to be competing pathways of crystallization. At low percentages of MPD, the protein can diffuse in solution uninhibited, and probably grew into crystals without MPD. Since MPD is extremely viscous, it is possible that there is a region where the diffusion of the protein in solution is too slow to allow for any crystal growth. But at higher percentages of MPD, the MPD is abetting the crystallization and actually allowing the crystals to grow despite the high viscosity. There was no significant difference between the sizes of the crystals between these two conditions, although the crystals with more MPD were less defined (figure 12). This effect was also seen at room temperature (20 °C) and in the cold (4 °C) microbatch trays with both enantiomers. There is no visible difference in the crystals habits of these crystals. The tetragonal unit cell of lysozyme is the same whether there is MPD in the crystal or not. That makes it difficult to tell by eye whether one enantiomer of MPD affects the crystal formation. Although microbatch was successful in producing crystals, most were too small for diffraction or not well formed.
16
Figure 12. Figure 12A shows lysozyme crystallized in 200mM Tris (39.9 mg/mL with 25% (v/v) R-MPD) using the microbatch method. Figure 12B shows lysozyme crystallized at the same conditions with 40% (v/v) R-MPD.
It is know among crystallographers that different methods of crystallization can affect the growth of the crystal.5 This is logical because the crystallization kinetics of the hanging drop method is completely different than the microbatch method. The hanging drop method allows for the slow concentration of the protein by the diffusion of the drop into the well. Using hanging drop crystallization we were able to reproducibly grow large diffraction quality crystals (figure 13).
17
A
B
C
D
Figure 13. Crystal grown in hanging drop wells. Figure 13A shows lysozyme crystallized with S-MPD (31 mg/mL 30% (v/v) S-MPD). Figure 13B shows lysozyme crystallized with R-MPD (31 mg/mL with 30% (v/v) R-MPD). Figure 13C shows crystals growing out of the precipitate (35.4 mg/mL with 32.5% (v/v) R-MPD) (200 pixels are equal to 271.6 µm). Figure 13D shows lysozyme grown without any MPD (36 mg/mL with 0% (v/v) MPD).
We consistently noticed that our crystals grew better at 4 °C, so most of the hangingdrop wells were crystallized at that temperature. This is likely related to the effect of the temperature change on the Tris buffer. The pKa of Tris is sensitive to temperature. A pH 8.0 Tris buffer at room temperature (20 °C) becomes pH 8.62 at 0 °C. This further moves the pH closer to the pI value and reduces the charge of the positive lysozyme molecules. The weaker electrostatic repulsion may favor nucleation. The phase diagram for pH 8.62, at 200mM Tris, has not been completely investigated
18
but we gained an important insight into the sensitive kinetics of the crystallization method. Figure 14 shows our investigated areas on the phase diagram for lysozyme at these conditions. The black points represent areas of stable precipitation of tiny crystals. The red points indicate areas in which there was an initial metastable state of precipitation, but eventually larger crystals would begin to form out of the precipitation. This final state was clearly the true equilibrium in this region. The green points represent the region where crystals could grow without any precipitation. The blue points represent regions of no growth at all. The growth in the controls is also not always uniform. Eventually after time, both crystals and precipitate will grow at the same time indiscriminately, which does not occur in our hanging drops with MPD. We closely watched the crystal growth for over 40 days. The conditions in figure 14 are just the initial conditions of the drops on the hanging drop caps. The power of the hanging drop method is the fact that these conditions will slowly change over time. Theoretically, the conditions of the drops will eventually be twice the initial conditions (e.g., the protein concentration in the drop will double). The crystal may actually grow anywhere between the initial and final conditions. This makes it difficult to determine exactly the optimal conditions for growth, but once the conditions are determined, they can be tested using the more static microbatch method.
19
Figure 14. The sharp phase boundaries can easily be seen and again demonstrates the sensitivity of crystallization. The black points indicate precipitation. The red points indicate initial precipitation that eventually forms crystal. The green points indicate crystal growth without precipitation. The blue points indicate no crystal growth.
Unlike thaumatin crystals, it still remains unclear if there is a difference in the crystal habit between different enantiomers of MPD used during crystallization. This makes it hard to determine whether the enantiomers of MPD are having any structural influence on the growth of the crystal. During the growth of crystals with R-, S-, and RS-MPD, there was an initial structural difference. The crystals with R-MPD formed rounded crystals that would develop sharp edges after time. The crystals grown with S- and RS-MPD on the other hand did not have this initially rounded structure. These results need to be repeated to confirm the trend, but it may indicate that the MPD molecule has some difficulty finding the best conformation within the crystal. The X-ray data below helps corroborate this hypothesis. 3.6 X-ray Diffraction. The preliminary data seem to indicate that the chiral MPD displays a
slight difference within the structure of the crystal, although, it is unclear what role this may play in the solubility and habit of the crystallized lysozyme. Both Weiss et al.10 and Michaux et al.11 claim that they can clearly see R and S respectively in their crystals. It seems unlikely that they see the MPD molecule as clearly as they believe because the resolution on their
20
structures is 1.64 Å and 1.50 Å respectively. All the structures that we are using in this study have better resolution. The crystal structure using just R MPD has a very defined conformation in that position (figure 15A). The same is not the case for the crystals using S and RS MPD in the wells (figure 15B).
Figure 15. Figure 15A shows the electron density around R-MPD in crystallized lysozyme (resolution 1.45 Å). Figure 15B shows the density of RS-MPD (assigned S-MPD) in the structure of crystallized lysozyme (resolution 1.20 Å). Red areas indicate regions of too little density while green areas indicate too much density. The red lines represent oxygen atoms and the yellow or green lines show the carbon backbone of the protein. The hydrogen atoms are not shown. Figures 15C and 15D show the same structure as 15A and 15B respectively, without the electron density to clearly see the chiral MPD molecule within the crystal. The gray colored backbone indicates that it is part of a symmetry-related protein in the unit cell.
While the resolution of the crystal with RS-MPD (assigned S-MPD) is 1.20 Å, in the specific location of the MPD molecule the electron density is not well defined. In the case of the
21
lysozyme crystallized with RS-MPD, this lack of density could be an averaging of the two enantiomers throughout the crystal. But oddly we notice a similar uncertainty in the electron density when the lysozyme is just crystallized with S-MPD. This too could be an averaging of the possible conformations that the S-MPD takes within the crystal. MPD can rotate freely around its axis and it may be able to configure itself stably inside the crystal structure using multiple conformations. Although MPD is a chiral molecule, it is almost perfectly symmetric. In its most stable conformation the C4-O4 and C2-O2 bonds point in the same direction so that an intramolecular hydrogen bond can form between the two oxygen atoms (cf. R-MPD in figure 4). 9 It appears that MPD occupies this conformation in the protein crystal as well. In addition to the intramolecular hydrogen bonds, both R- and S-MPD form two intermolecular hydrogen bonds directly with neighboring lysozyme molecules, establishing a crystal contact between them. In R-MPD, the hydrogen bonds are between the hydrogen of the O4 atom on MPD and the carbonyl oxygen of Phe34 on the same lysozyme molecule. The other hydrogen bond is between the hydrogen on the O2 atom of MPD with the carbonyl oxygen of the Gly22 on the symmetric lysozyme molecule. In RS-MPD the same bonds are formed except that the O2 takes the place of O4 and vice-versa. The molecules of MDP in the structures of Weiss et al.10 and Michaux et al.11 make the same hydrogen bonds are we do here. The similarity between the bonds formed by the two enantiomers may explain why they report different enantiomers in their crystal structures. Under our specific growing conditions, the R-MPD enantiomer seems to naturally fit into the crystal lattice, based on the clear electron density surrounding the MPD molecule. The electron density for S- and RS-MPD is not nearly as defined. There are large errors in the location of MPD even though the protein backbone itself has a very high resolution. The errors could indicate that S- or RS-MPD has difficultly selecting the best conformation. The errors may be due to an averaging of different possible configurations within the crystal. Although the density of R-MPD seems to indicate that it takes a single conformation within the structure of the crystal, it does not clearly indicate that one enantiomer is kinetically favored; otherwise, we would expect to only see R-MPD in lysozyme crystallized with RSMPD. It is still unclear how lysozyme selects which enantiomer to use as the crystal contact. Nevertheless, this chiral effect may not be significant enough to influence the crystallization
22
as much as we had seen in the tartrate-thaumatin system.
4. Conclusions MPD is clearly an important precipitant for the crystallization of lysozyme. Crystals grown with MPD form much faster then crystals without MPD. This is because MPD makes a crystal contact between the unit cells that allows the protein to form order crystals. Although the enantiomers of MPD may cause an initial difference in crystal habit, it does not appear that the chirality MPD plays a significant role in the crystallization of lysozyme. The X-ray data show that each enantiomer can structure itself within the crystal, even though the exact conformation within the structure is not the same. The rotational freedom of the MPD molecule allows its OH groups to point in the same direction for either enantiomer so this may allow each to make the necessary crystal contacts and insert itself in the unit cell of the crystal. The R-MPD enantiomer has a more uniform conformation within the crystal, while S- and RS-MPD have more areas of uncertainty that could indicate averaging of different possible conformations. While the S-MPD enantiomer may be displacing some water molecules in the unit cell, the OH groups can make the same crystal contacts that allow both enantiomers to fit in the crystal. Even though there is little difference in the crystal habits with R- and S-MPD that does not mean there is not a difference in the kinetics of crystal growth. I would also measure the solubility of crystals crystallized with R- and S-MPD to see if there is a difference similar to thaumatin and tartrate. Our previous work on thaumatin showed that small changes in chirality can affect the conformation of the enantiomer in the crystal, and possibly the entire crystal. It is possible that if lysozyme were crystallized with a less-symmetric chiral molecule there would be a greater chiral affect. Molecules with two chiral centers (like tartrate) will be less symmetric and may play a larger role in the crystallization. Molecules that are more rigid than MPD may also be able to show a more significant difference in crystallization. In the future I would like to try and crystallize lysozyme with tartrate and test this theory. Another possibility is that other model proteins may be more sensitive to small changes in its structure. Therefore, I would try using MPD as a precipitant for others proteins such as ribonuclease A.
23
5. Acknowledgments I thank Professor Neer Asherie for his guidance throughout this entire project and his many useful conversations. I also thank the other members of the Asherie lab including Mark Stauber, Ari Greenbaum, Sam Blass and Charles Ginsberg. I thank the National Science Foundation (DMR # 0901260) for providing a summer stipend and the Jay and Jeanie Schottenstein Honors Program for additional funding. 1 http://targetdb.sbkb.org/statistics/TargetStatistics.html#general 2 Pande, A.; Pande, J.; Asherie, N.; Lomakin, A.; Ogun, O.; King, J.; Benedek, G. B.; Proc. Natl. Acad. Sci., USA 2001, 98, 6116–20. 3 A. McPherson, Crystallization of Biological Macromolecules, CSHL Press, Cold Spring Harbor, 1999. 4 Asherie, N. Methods 2004, 34, 266-272. 5 Chayen, N. E.; Curr. Opin. Struct. Biol. 2004, 14, 577–583. 6 Asherie N.; Jakoncic, J.; Ginsberg, C.; Greenbaum, A.; Stojanoff, V.; Hrnjez, B.; Blass S.; and Berger, J.; Cryst. Growth Des. 2009, 9, 4189-4198. 7 Johnson, L. N.; Nature 1998, 5 number 11, 1998 8 R. Huopalahti, R. López-Fandiño, M. Anton, Rüdiger Schade Bioactive Egg Compounds Chapter 6 2007. 9 Anand, K.; Pal, D.; and Hilgenfeld, R.; Acta Cryst. D, 2002, 58, 1722-1728 10 Weiss, M. S.; Palm, G. J.; and Hilgenfeld, R.; Acta Cryst. D, 2000 56, 952-958. 11 Michaux, C.; Pouyez, J.; Wouters, J.; Privé, G. G.; BMC Struct. Bio. 2008, 8, 29 12 Lomakin, A.; Teplow, D. B.; Benedek, G. B.; Methods Mol. Biol. 2005, 299, 153–174. 13 Otwinowski, Z.; Minor,W.; Methods Enzymol., 1997, 276, 307–326. 14 Murshudov, G. N.; Vagin, A. A.; Dodson, E. J.; Acta Crystallogr. D, 1997, 53, 240–255. 15 CCP4; Collaborative Computational Project, Number 4; Acta Crystallogr. D, 1994, 50, 760-763 16 Emsley, P.; Cowtan, K.; Acta Crystallogr. D, 2004, 60, 2126–2132. 17 Laskowski, R. A.; MacArthur, M. W.; Moss, D. S.; Thornton, J. M. J. Appl.; Crystallogr. 1993, 26, 283–291. 18 Parmar, A. S.; Muschol, M.; Biophys. 2009, 97, 590-598. 19 Parmar, A. S.; Gorrschall, P. E.; Muschol, M.; Biophys. Chem. 2007, 129, 224-234. 20 Asherie, N.; Ginsberg, C.; Greenbaum, A.; Blass, S.; Knafo, S.; Cryst. Growth Des. 2008, 8, 4200-4207.
24
Appendix All the conditions we attempted to crystallize lysozyme are presented bellow. The results describe what we saw in the wells after giving them more than 3 weeks to reach equilibrium. A backslash in the results section indicates that multiple results were seen. Lysozyme Crystallized with R-, S-, and RS-MPD with pH 8.0 10mM sodium phosphate using the microbatch method at 20 °C. Concentration of Protein (mg/mL) 73.1 73.1 48.8 48.8 24.4 24.4 43.2 28.8 14.4 43.6 43.6 42.6 42.6 38.3 38.3 0 43.5 0 43.5 0 43.5 43.5 39 39 39 39 0 39
Percent MPD (v/v) 0 25 0 50 0 75 25 50 75 0 33 0 16.7 0 25 10 10 20 20 30 30 0 40 30 20 10 40 0
Enantiomer Used
Results Nothing Brown Precipitate (BP) Nothing Brown Precipitate Nothing Brown Precipitate Nothing Brown Precipitate Brown Precipitate Nothing Brown Precipitate Nothing Brown Precipitate Nothing Nothing Nothing Nothing Nothing Nothing Nothing Small Sticks Nothing BP/Spherulites Small Sticks Nothing Nothing Nothing Nothing
R, S R, S R, S R, S R, S R, S R, S R, S R, S R, S R, S R, S R, S R, S R, S R, S, RS R, S, RS R, S, RS R, S, RS R, S, RS R, S, RS R, S, RS R, S, RS R, S, RS R, S, RS R, S, RS R, S, RS R, S, RS
25
Lysozyme Crystallized with R-, S-, and RS-MPD with pH 8.0 10mM sodium phosphate using the microbatch method at 4 °C. Concentration of Protein (mg/mL) 73.1 73.1 48.8 48.8 24.4 24.4 43.6 43.6 42.6 42.6 38.3 38.3 0 38.3 0 42.6 0 44.5 0 43.5 0 43.5 0 43.5 43.5 39 39 39 39 0 39
Percent MPD (v/v) 0 25 0 50 0 75 0 33 0 16.7 0 25 25 25 16.6 16.6 12.5 12.5 10 10 20 20 30 30 0 40 30 20 10 40 0
Enantiomer Used
Results Crystals Brown Precipitate Crystals Brown Precipitate Nothing Brown Precipitate Nothing Brown Precipitate Nothing LLPS Nothing Crystal/LLPS Nothing LLPS Nothing BP/LLPS/Crystal Nothing Crystal/LLPS Nothing Nothing/LLPS Nothing LLPS Nothing LLPS Nothing BP/LLPS LLPS Nothing/LLPS Nothing Nothing Nothing
R, S R, S R, S R, S R, S R, S R, S R, S R, S R, S R, S R, S R, S R, S R, S R, S R, S R, S R, S, RS R, S, RS R, S, RS R, S, RS R, S, RS R, S, RS R, S, RS R, S, RS R, S, RS R, S, RS R, S, RS R, S, RS R, S, RS
26
Lysozyme Crystallized with RS-MPD with sodium phosphate using the hanging drop method at 20 °C. Concentration of Protein (mg/mL) 25.6 25.6 25.6 25.6 25.6 26.6 26.6 26.6 26.6 26.6 30.9 30.9 30.9 30.9 30.9 29.2 29.2 29.2 29.2 29.2
Percent MPD (v/v) 0 10 15 20 25 0 10 15 20 25 0 10 15 20 25 0 10 15 20 25
Molatiry of Buffer (mM) 10 10 10 10 10 10 10 10 10 10 5 5 5 5 5 5 5 5 5 5
pH 7 7 7 7 7 8 8 8 8 8 7 7 7 7 7 7 8 8 8 8
27
Enantiomer
RS RS RS RS RS RS RS RS RS RS RS RS RS RS RS RS RS RS RS RS
Result Nothing Nothing Nothing Small sticks Precipitate Nothing Nothing Nothing Sticks Spherulites Nothing Nothing Small Sticks Nothing Precipitate Nothing Nothing Nothing Small Sticks Spherulites/Precipitate
Lysozyme Crystallized with RS-MPD with sodium phosphate using the hanging drop method at 4 °C. Concentration of Protein (mg/mL) 25.6 25.6 25.6 25.6 25.6 26.6 26.6 26.6 26.6 26.6 30.9 30.9 30.9 30.9 30.9 29.2 29.2 29.2 29.2 29.2
Percent MPD (v/v) 0 10 15 20 25 0 10 15 20 25 0 10 15 20 25 0 10 15 20 25
Molatiry of Buffer (mM) 10 10 10 10 10 10 10 10 10 10 5 5 5 5 5 5 5 5 5 5
pH
7 7 7 7 7 8 8 8 8 8 7 7 7 7 7 7 8 8 8 8
Enantiomer
RS RS RS RS RS RS RS RS RS RS RS RS RS RS RS RS RS RS RS RS
28
Result
Nothing Nothing Precipitate Precipitate Brown Precipitate Nothing Nothing Precipitate Precipitate/BP Precipitate/BP Nothing Nothing Precipitate/BP Precipitate/BP Brown Precipitate Nothing Nothing Precipitate Precipitate/BP Brown Precipitate
Lysozyme crystallized with R-, S-, and RS-MPD in H2O using the microbatch method at 20 °C. Some wells contain 0.002% sodium azide. Concentration of Protein (mg/mL) 38.5 38.5 29.2 29.2 19.2 19.2 71.8 50.7 36.3 29.3 71.8 0
Percent MPD (v/v) 0 33 0 50 0 66.7 50 40 30 20 0
Enantiomer Used R, S R, S R, S R, S R, S R, S R, S, RS R, S, RS R, S, RS R, S, RS R, S, RS
50 R, S, RS
Result Precipitate Brown Precipitate Precipitate Brown Precipitate Precipitate Amorphous Crystals Amorphous Crystals Amorphous Crystals Amorphous Crystals Nothing Nothing/Amorphous Crystal Nothing
Azide No No No No No No Yes Yes Yes Yes Yes Yes
Lysozyme crystallized with R-, S-, and RS-MPD in H2O using the microbatch method at 4 °C. Some wells contain 0.002% sodium azide. Concentration of Protein (mg/mL) 38.5 38.5 29.2 29.2 19.2 19.2 71.8 50.7 36.3 29.3 71.8 0
Percent MPD (v/v) 0 33 0 50 0 66.7 50 40 30 20 0 50
Enantiomer Used R, S R, S R, S R, S R, S R, S R, S, RS R, S, RS R, S, RS R, S, RS R, S, RS R, S, RS
Result Nothing Amorphous Crystals Nothing Crystals/nothing Nothing Nothing/Amorphous Crystals Nothing Nothing Nothing Nothing Nothing Nothing
29
Azide No No No No No No Yes Yes Yes Yes Yes Yes
Lysozyme crystallized with RS-MPD in H2O (with 0.002% sodium azide) using the hanging drop method at 20 °C. Concentration Percent of Protein MPD (v/v) (mg/mL) 52.2 0 52.2 30 52.2 35 52.2 40 52.2 45 26.1 0 26.1 30 26.1 35 26.1 40 26.1 45
Enantiomer Used
Result Nothing Nothing Nothing Brown Precipitate Brown Precipitate Nothing Nothing Nothing Brown Precipitate Brown Precipitate
RS RS RS RS RS RS RS RS
Lysozyme crystallized with RS-MPD in H2O (with 0.002% sodium azide) using the hanging drop method at 4 °C. Concentration of Protein (mg/mL) 52.2 52.2 52.2 52.2 52.2 26.1 26.1 26.1 26.1 26.1
Percent MPD (v/v)
Enantiomer Used
Result
0 30 RS 35 RS 40 45 0 30 35 40 45
Nothing Nothing Nothing Nothing Nothing Nothing Nothing Nothing Nothing Nothing
RS RS RS RS RS RS
30
Lysozyme crystallized with R-, S-, and RS-MPD in pH 8.0 Tris using the microbatch method at 20 °C. Concentration Percent Molatiry of of Protein MPD Buffer (mM) (mg/mL) (v/v)
39.5 39.5 39.5 39.5 39.5 0 39.5 39.5 39.5 39.5
0 30 35 45 50 50 0 30 35 40
39.5 39.5 0 39.5, 39.9 39.5, 39.9 39.5, 39.9 39.5, 39.9 39.5, 39.9 0 38.3 38.3 38.3 38.3 38.3 0 61.7 61.7 61.7 61.7 61.7 0
45 50 50 0 25 30 35 40 40 0 25 30 35 37.5 37.5 0 25 30 35 37.5 37.5
Enantiomer Used
Results
400 400 400 400 400 400 300 300 300
RS RS RS RS RS RS RS RS RS
Nothing LLPS LLPS/Crystal Crystals Crystals Crystals/Spherulites Nothing LLPS LLPS/Crystal
300 300 300 300 200 200 200 200 200 200 50 50 50 50 50 50 50 50 50 50 50 50
RS RS RS RS R, S R, S R, S R, S R, S R, S R, S R, S R, S R, S R, S R, S R, S R, S R, S R, S R, S R, S
LLPS/Crystal Crystals LLPS/Crystal Nothing Nothing Nothing/Precipitate LLPS/Crystal LLPS LLPS Nothing Nothing Nothing Nothing Nothing Nothing Nothing Nothing Nothing Nothing Nothing Nothing Nothing
31
Lysozyme crystallized with R-, S-, and RS-MPD in pH 8.6 Tris using the microbatch method at 4 °C. Concentration of Protein (mg/mL) 39.5 39.5 39.5 39.5 39.5 39.5 0 39.5 39.5 39.5 39.5 39.5 39.5 0 40.9, 39.5, 39.9 40.9, 39.5, 39.9 40.9, 39.5, 39.9 40.9, 39.5, 39.9 40.9, 39.5, 39.9 0 38.3 38.3 38.3 38.3 38.3 0 61.7 61.7 61.7 61.7 61.7 0
Percent Molatiry of MPD Buffer (mM) (v/v) 0 400 30 400 35 400 40 400 45 400 50 400 50 400 0 300 30 300 35 300 40 300 45 300 50 300 50 300 0 200 25 200 30 200 35 200 40 200 40 200 0 50 25 50 30 50 35 50 37.5 50 37.5 50 0 50 25 50 30 50 35 50 37.5 50 37.5 50
32
Enantiomer Used RS RS RS RS RS RS RS RS RS RS RS RS RS RS R, S R, S R, S R, S R, S R, S R, S R, S R, S R, S R, S R, S R, S R, S R, S R, S R, S R, S
Results Crystals/Precipitate Crystals Crystals Crystals Precipitate Crystals/Precipitate Nothing Nothing Nothing Precipitate Sticks Sticks/Crystal Crystal/Precipitate Nothing Crystals/Precipitate Crystals/Precipitate Nothing/Crystals Sticks/Crystals Crystals Nothing Crystals Nothing/Crystals Nothing Nothing/Crystals Sticks Nothing Crystals Nothing Nothing Nothing/Crystals Nothing Nothing
Lysozyme crystallized with RS-MPD in pH 8.0 Tris using the millibatch method at 20 °C. Concentration of Protein (mg/mL)
Percent MPD (v/v)
38.8 38.8 38.8 38.8 38.8 0
0 25 30 35 40 40
Molatiry of Buffer (mM) 200 200 200 200 200 200
Enantiomer Used RS RS RS RS RS RS
Results
Nothing/Crystals Nothing Nothing Nothing Nothing Nothing
Lysozyme crystallized with RS-MPD in pH 8.6 Tris using the millibatch method at 4 °C.
Concentration of Protein (mg/mL)
Percent MPD (v/v)
38.8 38.8 38.8 38.8 38.8
0 25 30 35 40
Molatiry of Buffer (mM)
200 200 200 200 200
33
Enantiomer Used
RS RS RS RS RS
Results
Crystals Nothing Nothing Sticks/Nothing Sticks/Crystals
Lysozyme crystallized with RS-MPD in pH 8.0 Tris using the hanging drop method at 20 °C. Concentration of Protein (mg/mL)
Percent MPD (v/v)
10 10 10 10 10 36.6 36.6 36.6 37.1 37.1
27.5 30 32.5 35 37.5 0 35 40 0 35
37.1
40
Molatiry of Buffer (mM) 50 50 50 50 50 100 100 100 200 200
Enantiomer Used RS RS RS RS RS RS RS RS RS RS
200 RS
34
Results
Nothing Nothing Nothing Light Precipitate Light Precipitate Nothing Precipitate Precipitate Nothing Crystals Growing from Precipitate (CGP) Precipitate
Lysozyme crystallized with R-, S-, and RS-MPD in pH 8.6 Tris using the hanging drop method at 4 °C.
Concentration of Protein (mg/mL) 10 10 10 10 10 36.6 36.6
Percent MPD (v/v)
36.6 37.1 37.1
40 0 35
100 RS 200 RS 200 RS
37.1 0 35.2 35.2 35.2 35.2 28.4
40 35 0 35 30 25 35
200 200 200 200 200 200 200
28.4 28.4 23
30 25 35
200 RS 200 RS 200 RS
23 23 36 36 36 31.5 31.5 31.5 35.4 35.4
30 20 0 35 30 0 35 30 0 32.5
200 200 200 200 200 200 200 200 200 200
35.4 31
30 32.5
200 R, S 200 R, S
31
30
200 R, S
27.5 30 32.5 35 37.5 0 35
Molatiry of Buffer (mM) 50 50 50 50 50 100 100
Enantiomer Used RS RS RS RS RS RS RS
RS RS RS RS RS RS RS
RS RS R, S R, S R, S R, S R, S R, S R, S R, S
35
Results
Nothing Nothing Light Precipitate Light Precipitate Light Precipitate CGP/Precipitate Crystals Growing from Precipitate Precipitate Crystals Crystals Growing from Precipitate Precipitate Nothing Nothing CGP/Precipitate Crystals Nothing Crystals Growing from Precipitate Crystals Nothing Crystals Growing from Precipitate Crystals Sticks Crystals/Precipitate Precipitate Crystals Crystals/Precipitate Crystals Crystals Nothing Crystals Growing from Precipitate Crystals Crystals Growing from Precipitate Crystals
