Moss Tom. DNA-protein interactions: principles and protocols

Подождите немного. Документ загружается.

544 Buckle

and relates to a change in mass of 1 ng of a globular protein at the surface. Using

these values, a similar relationship has been demonstrated where 0.78 ng of a

nucleic acid gives the same response (1000 RU) (4,5).

2. Mass transport: In instances where the association rate is particularly elevated

and the diffusion rate of the nonimmobilized molecules is not especially fast,

then the interaction of the free molecules with the immobilized ligand may deplete

a layer of solvent immediately surrounding the immobilized ligand such that the

rate-limiting step for association now becomes the rate of repletion of this layer

from the bulk solvent. Two practical solutions to this problem are first to use a

low immobilization density and, second, to use relatively elevated flow rates

(>20 µL/min).

3. Estimation of the amplitude of the signal allowing calculation of the final

equilibrium or steady-state level is often hampered by bulk refractive index

changes that mask the initial and final phases of the injection period. Blank runs

over free surfaces may be used to correct for this provided that the free surface

reflects as close as possible the ligand surface. For example, if streptavidin is

being used to immobilize the DNA, then a surface containing a comparable quan-

tity of streptavidin to the sample surface should be used as a blank (Fig. 4). Inci-

dentally, this is again an argument in favor of low levels of DNA immobilization

so that the control surface is very similar in refractive properties to the surface

under study. Alternatively, and indeed if possible, preferably samples being

injected across the surface should be desalted into the injection buffer so as to

minimize bulk refractive index effects. An ideal method is to use fast desalting

columns on HPLC/FPLC systems such as the SMART (Amersham Pharmacia

Biotech) which produce little dilution, are rapid and allow automatic quantifica-

tion of the protein.

4. The asymptote of the binding curve should provide the stoichiometry of the reac-

tion assuming that all the DNA molecules are available for binding. If this is

uncertain, one way of ensuring that a double-stranded target DNA is accessible

would be to hybridize the protein binding site in situ on the surface by immobi-

lizing a single strand and then hybridizing a second homologous strand to consti-

tute the double stranded site. The mass increase in the second hybridization step

would give the number of accessible DNA molecules because it results from a

successful hybridization at the surface. It goes without saying however that this

requires great care in eliminating nonspecific adsorption of DNA onto the surface.

5. Steric hindrance: Let us imagine for the sake of argument that we have immobi-

lized 1000 RU of a 100-bp double-stranded DNA molecule via biotin/streptavidin

to a surface. This constitutes 1.2 × 10

–14

mole of DNA or 2 × 10

molecules. Let

us further assume that these molecules are evenly and randomly distributed across

the 1-mm

surface so a simple calculation shows that each molecule is separated

from its neighbor by 70 nm. The DNA molecules are around 30 nm long, so if the

dextran is itself flexible, then these DNA molecules are going to be in contact.

This may produce problems such as occlusion of sites, creation of new potential

binding sites, or unusual DNA structures possessing aberrant binding modes for

Surface Plasmon Resonance 545

target proteins. Even if a relatively large molecule such as a prokaryotic RNA

polymerase (500 kDa, approximate diameter-10 nm) were to bind to a single site

on the DNA, then there would be every possibility that adjacent bound molecules

are going to interact with complicated consequences upon binding kinetics.

6. The amplitude of each signal should be estimated manually from the steady-state

signal at the end of the injection period or, if this is difficult because of a slow

overall relaxation time or because of a perturbing bulk refractive index effect,

from the projected R

value calculated from direct fits. Note that the projected

values should be identical for fits of both the association and dissociation

phases. It is essential that in the equilibrium analysis of these curves, the stoichi-

ometry reflects the biology of the situation. An immobilized DNA ligand con-

taining a single site for a protein should bind only one protein per site at

saturation. If this is not the case, then no single- or double-binding site analysis is

going to provide meaningful affinity values. It will be clear from careful inspec-

tion of the dissociation phase, for example, that several interactions are involved

and a simple analysis cannot be made.

7. This procedure will activate surface 1. Other surfaces can be activated simply by

changing the designated flow cell.

8. The Biacore machines are robust but susceptible to poorly prepared solutions.

All solutions entering the fluid system must be filtered through 0.2-µm mem-

branes and preferentially be sterile. The machine should be cleaned periodically

with the desorb and sanitize protocols, and when not in use, it is recommended to

maintain a continuous flow of distilled water containing 0.005% P20 at 20°C.

9. As pointed out in ref. 6, there are at least three consistency tests that should be

applied to a given analysis: First, the equilibrium or steady-state analysis

provides a dissociation constant K

from the Langmuir isotherm derived from

Eq. 6:

[P] = R

+ (R

sat

– R

)

[P]

(12)

+ [P]

where R

sat

corresponds to the asymptote of binding curves of the type shown in

Fig. 4. The value obtained for K

here should be equal to that obtained from the

thermodynamic relationship described in Eq. 11. In cases where this is not the

case neither treatment may reflect the correct situation. Second, the use of Eq. 10

in a linear least squares analysis provides a value for k

off

. This value should agree

with that obtained from direct analysis of the dissociation phase of all sensor-

grams at all concentrations of P (Eq. 9). Third, it should be obvious from Eq. 10

that the value for k

off

should be inferior to that of k

obs

over all values of [P] because

off

must always be in excess of zero. Any values that do not comply to these

simple tests are thus derived from an erroneous treatment of the interaction.

10. Recapture: At high levels of immobilization, when a captured ligand dissociates

from the immobilized surface, it may subsequently be recruited to an adjacent

molecule. The effect of this will be to decrease the numerical values ascribed to

derived dissociation constants. Thus, in practice, the density of immobilization

546 Buckle

must be adjusted so that the dissociation rate is independent of immobilized

ligand density. If this proves difficult, then one can, with considerable error,

extrapolate to infinite dilution. Finally, a free ligand may be included during the

dissociation phase in order to calculate an affinity for the competitor and thus

allow an estimation of a true dissociation constant.

References

1. Buc, H. (1998) L’utilisation de la résonance de plasmon de surface pour la déter-

mination des constantes d’equilibre). Regard Biochim. 2, 21–26.

2. Schuck, P. (1997) Reliable determination of binding affinity and kinetics using

surface plasmon resonance biosensors. Curr. Opin. Biotechnol. 8(4), 498–502.

3. Schuck, P. and Minton, A. P. (1996) Analysis of mass transport-limited binding

kinetics in evanescent wave biosensors. Analy. Biochem. 240(2), 262–272.

4. Buckle, M., et al. (1996) Real time measurements of elongation by a reverse

transcriptase using surface plasmon resonance. Proc. Natl. Acad. Sci. USA 93(2),

889–894.

5. Fisher, R. J., et al. (1994) Real-time DNA binding measurements of the ETS1

recombinant oncoproteins reveal significant kinetic differences between the p42

and p51 isoforms. Protein Sci. 3(2), 257–266.

6. Schuck, P. and Minton, A. P. (1996) Kinetic analysis of biosensor data: elementary

tests for self-consistency [see comments]. Trends Biochem. Sci. 21(12), 458–460.

7. Adelman, K., et al. (1998) Stimulation of bacteriophage T4 middle transcription

by the T4 proteins MotA and AsiA occurs at two distinct steps in the transcription

cycle. Proc. Natl. Acad. Sci. USA 95, 15,247–15,252.

Reconstitution of Complexes 547

547

From:

Methods in Molecular Biology, vol. 148: DNA–Protein Interactions: Principles and Protocols, 2nd ed.

Edited by: T. Moss © Humana Press Inc., Totowa, NJ

Reconstitution of Protein–DNA Complexes

for Crystallization

Rachel M. Conlin and Raymond S. Brown

1. Introduction

An increasing number of structural studies are aimed at identifying the prin-

ciples that govern protein-DNA recognition in gene regulation (1). This work

depends on the successful reconstitution of protein–DNA complexes from their

purified components. X-ray crystallography and two-dimensional nuclear mag-

netic resonance (NMR) techniques require large amounts of pure protein and

DNA. These can be supplied through expression in bacteria of the cDNA cod-

ing for intact proteins or their smaller DNA-binding domains and the auto-

mated chemical synthesis of DNA in the laboratory.

Expression of the protein of interest is usually achieved at high levels in bacte-

ria. An increasingly popular system is the combination of Escherichia coli strain

BL21(DE3) transformed with a pRSET expression vector containing a strong

phage promoter adjacent to the cloning site (2). This particular bacterial strain

has an integrated T7 RNA polymerase gene that can be induced with IPTG (3).

Target DNA sequences are easily identified by DNase I footprinting of a

radioactively labeled DNA restriction fragment to which the protein is bound

(4). Little technical difficulty is experienced in the chemical synthesis of these

DNA sequences, up to about 40 base pairs in length, in amounts necessary to

perform structural studies.

Considerable progress has been made in solving three-dimensional struc-

tures of protein–DNA complexes (1), largely because proteins and their iso-

lated DNA-binding domains are able to recognize and form stable complexes

with quite short duplexes containing the binding sequence. Indeed, protein–

DNA complexes have been crystallized that contain duplexes that are as short

as 8 base pairs in length (5).

548 Colin and Brown

In this chapter, we describe the preparation of a protein–DNA complex com-

posed of a six-zinc-finger fragment (22 kDa) of Xenopus laevis transcription

factor TFIIIA and a synthetic DNA duplex. The experimental strategy, recon-

stitution conditions, and technical problems to be discussed are probably quite

similar to those encountered with most other protein–DNA complexes. In this

example, the solution properties of the protein are known and the limits of the

target DNA sequence are precisely defined. It has generally been assumed that

association of the components takes place spontaneously to produce protein–

DNA complexes of the desired molar composition. We have chosen to perform

biochemical analysis in order to optimize authentic binding and ensure the

efficient scaling up of reconstitution conditions.

2. Materials

1. Expression plasmid pRSET B (Invitrogen, Carlsbad, CA).

2. Buffer A: 500 mM NaCl, 2 mM benzamidine–HCl, 1 mM dithiothreitol (DTT),

1 mM NaN

50% (v/v) glycerol, and 50 mM Tris-HCl, pH 7.5.

3. Standard protein solution: 1 mg/mL bovine serum albumin (BSA).

4. Protein assay concentrate (Bio-Rad, Richmond, CA). Dilute 1:5 with water and

filter through Whatman 2

paper.

5. 0.01% (w/v) Coomassie brilliant blue R250 in 20% (v/v) ethanol and 10% (v/v)

acetic acid.

6. 1 M Tris-HCl, pH 8.0.

7. DEAE–Sephacel (Pharmacia, Piscataway, NJ).

8. 2 M sodium acetate.

9. Repelcote solution (Hopkin and Williams, Chadwell Heath, Essex, UK).

10. Millex HA and Millex HV4 (0.45 µm) filter units (Millipore, Bedford, MA).

11. Buffer B: 100 mM NaCl and 10 mM Tris-HCl, pH 8.0.

12. Amberlite MB-150 resin (Sigma, St. Louis, MO).

13. 150-mL (0.2-µm) NYL/50 filter unit (Sybron Corp., Rochester, NY).

14. Thermal cycler for polymerase chain reaction (PCR).

15. 5X binding buffer: 250 mM NaCl, 5 µM MgCl

, 5 mM DTT, 50 µM zinc acetate,

50% (v/v) glycerol and 100 mM Tris-HCl, pH 7.5.

16. 1 M HEPES–NaOH, pH 7.5.

17. 5 M NaCl.

18. Buffer C: 100 mM NaCl, 1 mM DTT, 1 mM NaN

, and 20 mM Tris-HCl, pH 7.5.

19. Collodion bags and 300-mL glass vacuum dialysis flask (Sartorius AG,

Göttingen, Germany).

20. Magnetic micro flea 5-mm × 2-mm spinbar (Bel-Art products, Pequannock, NJ).

21. Microcon-10 microconcentrators (Amicon, Beverly, MA).

22. Natrix nucleic acid sparse matrix kit (Hampton Research, Lagnua Niguel, CA).

23. Linbro tissue culture multiwell plate with cover. Twenty-four flat-bottomed wells

(1.7 cm × 1.6 cm) (Flow Laboratories, McLean, VA).

24. AquaSil water-soluble siliconizing fluid (Pierce, Rockford, IL).

Reconstitution of Complexes 549

25. DiSPo plastic cover slips M6100 (American Scientific Products, McGaw Park, IL).

26. High-vacuum grease (Dow Corning, Midland, MI).

27. A 10cc plastic B-D syringe (Becton-Dickinson, Rutherford, NJ).

3. Methods

3.1. Purification of a Recombinant DNA-Binding Protein

Standard laboratory techniques for protein purification (6) will not be

described in detail in this chapter. Advantage is usually taken of the rather

basic nature of DNA-binding proteins to isolate them from the bacterial cell

extract. Few of the bacterial proteins bind so strongly to chromatography

matrices, such as CM–Sepharose, Bio-Rex 70, S-Sepharose, phosphocellulose,

and hydroxyapatite–Ultrogel. The protein of interest can usually be eluted with

a NaCl gradient. Affinity chromatography with heparin–Sepharose, an immo-

bilized dye–Sepharose, DNA–agarose or phenyl–Sepharose often provides an

adequate final purification step.

3.1.1. Preparation of Protein

The cDNA for the 22-kDa fragment of TFIIIA (amino acids 1–190) was

cloned into the vector pRSET B and expressed in strain BL21 (DE3) (see Note 1).

After sonication, the cell pellet is stirred in 0.5 M NaCl and 7 M urea for 48 h at

4°C. Purification of the protein (Fig. 1) is carried out in 7 M urea on columns

of Bio-Rex 70 and heparin–Sepharose (see Notes 2–4). Workup of the protein

for use in DNA binding is described in detail as follows:

1. Pool those fractions that contain protein after heparin–Sepharose column

chromatography.

2. Concentrate the protein to 5 mg/mL by vacuum dialysis at 4°C in a collodion bag

(see Note 5).

3. Dialyze against ice-cold buffer A.

4. Store the protein at –20°C in a 1.5-mL Eppendorf tube.

3.1.2. Measurement of Protein Concentration

1. Construct a calibration curve of protein concentration versus absorbance from

samples (in triplicate) containing 15, 25, 35, and 45 µL of the standard protein

solution and water added to 100 µL in glass tubes (see Note 6).

2. Add 5 mL of protein assay reagent and read the absorbance at 595 nm with a

spectrophotometer after the blue color has developed for 10 min.

3. Calculate the relative concentration from comparison of the values for the pro-

tein sample with the calibration curve (see Notes 7 and 8).

4. Examine the purity of the protein sample by standard sodium dodecyl sulfate

(SDS)/polyacrylamide gel electrophoresis. A 5% stacking/13.5% resolving slab

gel (30:0.8 acrylamide:bis) is suitable for this purpose (see Note 9). At least 0.1 µg

550 Colin and Brown

of protein is detected in a band with Coomassie blue stain. If necessary, adjust

the protein concentration by an estimate of its gel purity.

3.2. Preparation of Synthetic Oligomers

Typically a 1-µmol scale synthesis with a commercially available machine

provides 1–2 mg of a purified 31-mer. The standard laboratory methods for

isolation of the 5'-dimethoxytrityl full-length oligomer and subsequent chemi-

cal removal of base protecting groups will not be described here (see Note 10).

Oligomers pure enough for this work can be obtained with a fast protein liquid

chromatography system (FPLC) and suitable columns (7).

3.2.1. Recovery and Concentration of Oligomers

1. Adjust the pH to 8.0 and apply the pooled Mono Q column fractions containing

the oligomer onto a 0.5-mL DEAE-Sephacel column equilibrated with 50 mM

Tris-HCl (pH 8.0) at room temperature.

Fig. 1. (A) 13.5% SDS-PAGE gel showing the expression of TFIIIA (residues

1–190) in E. coli BL21(DE3). Dalton markers are in lane 1. Whole-cell extracts are

run in lanes 2 and 3 before and after induction with 1 mM IPTG. The TFIIIA protein

after purification is shown in lane 4. (B) 6.5% PAGE mobility shift gel for analysis of

protein-DNA binding. One microgram of the synthetic 31-mer DNA duplex was mixed

with 0.5 µg (lane 1) or 1 µg (lane 2) of the 22.1-kDa TFIIIA fragment. Bands that

contain DNA are visualized by ethidium bromide stain and UV illumination.

Reconstitution of Complexes 551

2. Elute the bound oligomer with 5X 1 mL of 2 M sodium acetate into a silanized

30-mL Corex tube.

3. Add 20 mL of 95% ethanol and precipitate the oligomer overnight at –20°C.

4. Centrifuge (45 min at 7500g, 4°C) to recover the oligomer.

5. Wash the pellet with 25 mL of 75% ethanol to remove excess sodium acetate and

centrifuge as in step 4. Pour off the supernatant and dry pellet under vacuum.

6. Redissolve the oligomer in 1 mL of water (see Note 11) and then filter through a

Millex HV4 unit (0.45 µm).

7. Measure the absorbance at 260 nm with a spectrophotometer (dilute the oligomer

as necessary with 100 mM NaCl). The concentration is calculated with a conver-

sion factor of 25 A

260

units = 1 mg. Store the oligomer at 1 mg/mL at –20°C.

3.2.2. Annealing the Duplex

1. Mix equal amounts of complementary DNA strands in 100 µL of buffer B at

1 mg/mL in 0.25-mL PCR tubes.

2. Heat at 95°C for 5 min and then slowly cool to 4°C (see Note 12).

3. Examine duplex formation by electrophoresis at 50 V in a nondenaturing 6.5%

polyacrylamide slab gel at room temperature. One-tenth microgram of DNA in a

gel band is easily visible after staining in ethidium bromide (1 mg/L) and illumi-

nation with ultraviolet (UV) light.

3.3. Testing the Reconstitution Conditions

The duplex is titrated with increasing amounts of protein in order to mea-

sure DNA-binding activity (Fig. 1). The resulting complexes are monitored by

mobility shift on a nondenaturing 6.5% polyacrylamide gel (see Notes 13–15).

This method is used to systematically optimize 1:1 molar complex formation

with respect to incubation time, temperature, concentration of monovalent and

divalent ions, pH, and the presence of glycerol and nonionic detergents like

Nonidet P-40.

1. Mix 1 µg of DNA, 2 µL of 5X binding buffer, and 5 M NaCl (to make 0.5 M NaCl

after step 2) in a 1.5-mL Eppendorf tube.

2. Add 0.5, 1, 1.5, and 2X molar excess of protein and stir in gently with the

Pipetman tip (see Note 16).

3. Dilute with 10 µL of 1X binding buffer and incubate at room temperature for 15 min.

4. Apply the 20-µL samples, without tracking dyes, to a nondenaturing 6.5% poly-

acrylamide gel (see Note 17).

5. Load dyes (bromophenol blue and xylene cyanol FF) in adjacent tracks to indicate

progress of the electrophoresis. The protein–DNA complex migrates between the dyes.

6. Stain the gel first with ethidium bromide (1 mg/L) for DNA and then with 0.01%

Coomassie blue for protein.

3.3.1. Scaling-Up Reconstitution of the Complex

In low-salt conditions, the protein shows a strong tendency to aggregate.

This is detected as material trapped at the top of the nondenaturing polyacryla-

552 Colin and Brown

mide gel. An excess of protein results in formation of some protein–DNA

complexes with higher molar stoichiometries. We have discovered that these

unwanted effects can be substantially eliminated by dilution of NaCl in the

scaled-up reconstitution mix in steps; from 0.75 M to 0.225 M NaCl (see Note 18).

1. Mix together 500 µg of DNA, 200 µL of 5X binding buffer, and 5 M NaCl water

in a 50-mL polypropylene Falcon tube (the final reconstitution mix, including

the protein, is 0.75 M NaCl).

2. Add 625 µg of protein and mix together by gentle swirling.

3. Dilute with 1 mL of 1X binding buffer and incubate for 5 min at room temperature.

4. Add 1 mL of 1X binding buffer and incubate for 5 min.

5. Add 1 mL of 1X binding buffer and incubate for 5 min.

6. Concentrate the complex to 10 mg/mL by centrifugation 4°C with Microcon-10

microconcentrator units (Amicon).

3.3.2. Crystallization Trials

Crystallization conditions are often screened according to an incomplete

factorial design (8). Supplies and kits are available from Hampton Research.

Some typical crystallization results, mainly employing the hanging drop/vapor

diffusion method, are shown in Table 1. Conditions are screened at 4°C and at

room temperature with small droplets (1–2 µL) of the protein–DNA complex

in which the concentrations of salts, buffers, and precipitants are systemati-

cally varied by small amounts. Crystallization often occurs at or close to the

point of precipitation of the protein–DNA complex.

1. Pass all solutions of salts, buffers and precipitants through 0.45 µ filters (Millex

HA units) before use.

2. Blow dust from plastic pipet tips and then place 1-µL aliquots of the protein–

DNA complex onto silanized plastic cover slips (see Note 19).

3. Mix with 1 µL of appropriate salts, buffer and precipitant or the well solution

(see Note 20).

4. Apply vacuum grease to the upper rim of the wells of the tissue culture plate (see

Note 21).

5. Into each well put 1 mL of suitable precipitant.

6. Carefully invert a cover slip without disturbing the droplet and place onto the

greased rim so as to seal each well.

7. Store the tissue culture plates in closed polystyrene boxes to avoid excessive

vibration and changes in temperature.

8. After a week inspect the droplets for signs of crystallization with a stereo micro-

scope at ×50 magnification.

4. Notes

1. Bacteria grown in LB broth are induced by addition of 1 mM IPTG and 100 µM

zinc acetate for 2 h at 37°C.

2. Losses of protein during column chromatography may be decreased by addition

of 10% (v/v) glycerol.

Reconstitution of Complexes 553

Table 1

Crystallization Conditions for Protein–DNA Complexes

Protein–DNA Complex Method Well solution

Sso7d: 8-mer, DNA and protein in 2 mM Tris-HCl, pH 6.5, 15% PEG 400

Gao et al. (5) and 2.6% PEG 400 at room temperature

NF-κB p50/p65: 12-mer, Final conditions are 50 mM sodium

Chen et al. (9) acetate (pH 5.5) 100 mM CaCl

, 0.125%

β-octyl-glucoside, 1 mM spermine,

10 mM DTT, and 8% PEG 3350 at 18°C

Antp homeodomain: Mix protein and DNA in 5 mM bis-tris– 1–5% MPD

15-mer, propane (pH 7.0) with 1 vol of well 20 mM bis-tris–

Fraenkel and Pabo (10) solution at room temperature propane (pH 7.0),

and10 mM NiCl

Stat3beta: 18-mer DNA and protein in 20 mM HEPES 10% v/v glycerol,

glycerol, (pH 7.0) 200 mM NaCl, 10 mM MgCl

, 0.1–0.4 M NaCl.

Becker et al. (11) 5 mM DTT, and 0.5 mM PMSF. Mix 5 mM MgSO

with 1 volume of well solution at 50 mM MES,

room temperature pH 5.6–6.0, 0.1 M

ammonium aceta te

NFAT/AP–1: 20-mer, Final conditions are 10 mM HEPES 300 mM ammonium

Chen et al. (12) (pH 7.5), 1 mM DTT, 100 mM NaCl, 20% acetate

(v/v) glycerol, 500 mM ammonium acetate

GABPalpha/beta: 21-mer, Mix protein and DNA in 1 mM 100 mM bis-tris–

Batchelor et al. (13) EDTA, 1 mM DTT, 20 mM propane, pH9.0

Tris-HCl, pH 8.0, and 0.001% NaN

5 mM cobaltic

with 1 vol of the well solution at 20°C hexammine

chloride, and 9%

PEG 1000

Topo I: 22-mer Mix 1µL of DNA in 6 mM NaCl with 2 µL 100 mM MgCl

, 10 mM

Redinbo et al. (14) of protein in 10 mM Tris-HCl (pH 7.5), DTT, 100 mM

1 mM EDTA, 5 mM DTT, and 3 µL H

O. Tris-HCl (pH 7.7),

Add to 3 µL of well solution at 22°C and 24% PEG 400

T7 DNA polymerase: Microseeding 100 mM ACES,

21-mer + 26-mer pH 7.5, 30 mM

Doublie et al. (15) MgCl

, 120 mM

ammonium sulfate,

5 mM DTT, and

12% PEG 8000

TFIIIA (residues 1–190): Final conditions are 165 mM NaCl,

31-mer, 35 mM sodium acetate, 3.2 mM DTT,

Nolte et al. (16) 9.2% (v/v) glycerol, 1.8 mM NaN

, 1.8 mM

cadaverine–2HCl, 5.5 mM Tris-HCl

(pH 8.0), and 22.5% PEG 4000 at 18°C

DtxR repressor: 33-mer, Mix 2µL of protein and DNA in 1 mM 6–12% PEG 4000,

10 mM

White et al. (17) NiCl

, and 100 mM Tris-HCl, pH 7.5 with MgCl

, and 100 mM

1 µL well solution at room temperature MES (pH 6.0)

Abbreviations: ACES: 2-[(2-Amino–2-oxoethly)amino]ethanesulfonic acid; DTT: dithiothreitol;

EDTA: disodium ethylenediaminetetraacetic acid; HEPES: (N-[2-hydroxyethyl]piperazine-N'-

[2-ethanesulfonic acid]); MES: 2-[N-morpholino]ethane-sulfonic acid; MPD: 2-methyl-2,

4-pentanediol; PEG: polyethylene glycol; PMSF: phenylmethylsulfonyl flluoride.