Moss Tom. DNA-protein interactions: principles and protocols

Подождите немного. Документ загружается.

Calorimetry of Protein–DNA Complexes 533

19. Freire, E., Mayorga, O. L., and Straume, M. (1990) Isothermal titration. Anal.

Chem. 62, 950A–959A.

20. Breslauer, K. J., Freire, E., and Straume, M. (1992) Calorimetry: a tool for DNA

and ligand–DNA studies. Methods Enzymol. 211, 533–567.

21. Privalov, G., Kavina, V., Freire, E., and Privalov, P. L. (1995) Precise scanning

calorimeter for studying thermal properties of biological macromolecules in dilute

solution. Anal. Biochem. 232, 79–85.

22. Read, C. M., Cary, P. D., Preston, N. S., Lnenicek-Allen, M., and Crane-Robinson,

C. (1994) The DNA sequence specificity of HMG boxes lies in the minor wing of

the structure. EMBO J. 13, 5639–5646.

23. Wallace, R. B. and Miyada, C. G. (1987) Oligonucleotide probes for the screen-

ing of recombinant DNA libraries. Methods Enzymol. 152, 432–442.

Surface Plasmon Resonance 535

535

From:

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

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

Surface Plasmon Resonance Applied

to DNA–Protein Complexes

Malcolm Buckle

1. Introduction

1.1. The Relationship Between Refractive Index and Mass

Surface plasmon resonance (SPR) measures refractive index changes (∆n)

at or near a surface and relates these to changes in mass at the surface (Fig 1).

This relationship is given by the Clausius Mossotti form (Eq. 2) of the Debye

equation (Eq. 1):

ε – 1 N

(

α + µ

)

(1)

ε + 2

3ε

ε – 1 Nα

(2)

ε + 2

3ε

where ε is the real part of the dielectric constant or permittivity constant related

to the refractive index by ε = n

, N is the number density given by N

ρ/M

is Avogadro’s number, ρ is the density and M

is the molecular mass). It is

assumed that ∆n/∆C is a constant.

1.2. SPR Using the BIACORE Instrument

In physical terms, the detection system of this SPR machine consists of a

monochromatic, plane polarized light source, and a photodetector that are con-

nected optically through a glass prism (Fig. 1). A thin gold film (50 nm thick),

deposited on one side of the prism, is in contact with the sample solution. This

gold film is, in turn, covered with a long-chain hydroxyalkanethiol, which

forms a monolayer (approx 100 nm thick) at the surface. This layer essentially

serves as an attachment point for carboxymethylated dextran chains that create

a hydrophilic surface to which ligands can be covalently coupled. Light

536 Buckle

incident to the back side of the metal film is totally internally reflected onto the

diode-array detector. A property of this situation is that a nonpropagative eva-

nescent wave penetrates into the solution side of the prism away from the light

source. Free electrons in the gold layer enter into resonance with the evanes-

cent wave. In fact, such resonance implies that the amplitude vector character-

izing a transversal wave propagating along the gold surface (ks

→

p) is equal to

the component (kx

→

) of the evanescent wave. Because ε = n

, if ω is the fre-

quency of the wave and c the speed of light, then

ksp

−

ωε ε

εε

(3)

furthermore, given that for the evanescent wave,

θε

sin

(4)

Fig. 1. Schema showing the principle of surface plasmon resonance. Light

from a laser source arriving through a prism at a gold surface at the angle of

total internal reflection (θ) induces a nonpropagative evanescent wave that pen-

etrates into the flow cell opposite the prism. The intensity of the reflected light

is continuously monitored. At a given angle (λ) dependent on the refractive

index of the solution in the flow cell, resonance between the evanescent wave

and free electrons in the gold layer results in a reduction in the intensity of

reflected light. The change in angle of reduced intensity (∆λ) reflects changes

in the refractive index (n) of the solution in the flow cell immediately adjacent

to the gold layer. A dextran surface coupled to the gold layer allows immobili-

zation of ligands (e.g., DNA) within the evanescent field.

Surface Plasmon Resonance 537

when resonance occurs,

ksp

and the intensity of the reflected light

decreases at a sharply defined angle of incidence, the SPR angle, given by the

simple expression

sin·

()

g1 2

εε

εε ε

(5)

Thus, θ

, the SPR angle at which a decrease in the intensity of reflected light

occurs, measures the refractive index of the solution in contact with the gold

surface and is dependent on several instrumental parameters (e.g., the wave-

length of the light source and the metal of the film). When these parameters are

kept constant, the SPR angle shifts are dependent only on changes in refractive

index of a thin layer adjacent to the metal surface. Any increase of material at

the surface will cause a successive increase of the SPR angle, which is detected

as a shift of the position of the light intensity minimum on the diode array. This

change can be monitored over time, thus allowing changes in local concentra-

tion to be accurately followed. The SPR angle shifts obtained from different

proteins in solution have been correlated to surface concentrations determined

from radio-labeling techniques and found to be linear over a wide range of

surface concentration. The instrument output, the resonance signal, is indicated

in resonance units (RU); 1000 RU correspond to a 0.1° shift in the SPR

angle, and for an average protein, this corresponds to a surface concentra-

tion change of about 1 ng/mm

(for nucleic acids, see Note 1). It is remark-

able that the present instrument (Biacore 2000) can measure a deviation of

–3

°, in other words, a variation of 10

–5

in the refractive index .

1.3. Immobilization of DNA to a Surface

Although a variety of techniques exist for the immobilization of DNA on

the dextran surface, the most efficient for the majority of protein–DNA inter-

actions is the use of immobilized streptavidin that can then interact with a suit-

ably end-labeled DNA molecule. The streptavidin is immobilized via a

carbodiimide–N-hydroxyl succinimide coupling reaction to the carboxyl

groups of the dextran (Fig. 2). DNA is easily obtained either by direct purchase

of oligomers end-labeled with biotin, or, for larger fragments, direct poly-

merase chain reaction (PCR) from biotinylated oligomers. Unless a particu-

larly unusual configuration is required, biotin is generally present at one end of

the DNA molecule and on one strand if the DNA is double stranded. The end

biotinylated DNA is then flowed across the surface and allowed to bind to the

desired final concentration (Fig. 3).

1.4. Protein Binding to Immobilized DNA

The protein is flowed across the immobilized DNA in a buffer and at a tem-

perature suitable for the interaction being studied (Fig. 4). A range of concen-

538 Buckle

Fig. 2. Coupling of streptavidin to a flow cell activated by carbodiimide and

hydroxyl succinimide. The figure shows a sensorgram of a carboxymethylated dextran

surface (CM5) activated by carbodiimide and N-hydroxylsuccinimide prior to coupling

with streptavidin. Sharp changes in the resonance units (RU) reflect bulk refractive

index changes as a result of differences in the buffer. Ethanolamine is used to block all

unreacted activated carboxyl groups. The difference between the final RU value and

the initial value presents an accurate measure of the amount of streptavidin covalently

coupled to the surface.

trations should be investigated. A flow rate in excess of 10 µL/min is advised

and a sufficient contact time during the association phase to saturate the immo-

bilized DNA (as seen by a steady-state plateau for the RU values) at protein

concentrations in excess of the anticipated K

for the interaction. The dissocia-

tion phase should also be allowed to continue for at least sufficient time to

allow over a third of the complex to dissociate.

1.5. Binding Curve Analysis

1.5.1. Stoichiometry and Equilibrium Analysis

For an immobilized DNA fragment (D), the interaction with a mobile pro-

tein (P) can be written as

D + P

↔

DP (Scheme 1)

A classical Langmuir adsorption isotherm requires that the fraction of avail-

able sites on the DNA occupied by the protein (θ

) be given by

Surface Plasmon Resonance 539

Fig. 3. Immobilization of a biotinylated DNA fragment to a streptavidin surface. In

this example, a 200-bp fragment of DNA (10 µg/mL) containing a single biotin label

at one 5' end was flowed at 20 µL/min in HBS buffer over the streptavidin surface. The

initial bulk refractive index change was followed by a gradual increase reflecting DNA

binding to the streptavidin. At the end of the DNA injection phase, the bulk refractive

index change was recovered and the difference in absolute RU values compared to the

initial value reflects the number of molecules of DNA now bound to the surface.

(6)

Furthermore, in such a simple case, the equilibrium association constant K

given by the expression

KaP

(7)

1 + KaP

Thus, in an SPR experiment, the steady-state level of bound protein at a

given concentration of total protein should be calculated from the asymptote of

the sensorgram and the RU values converted into moles of bound protein.

Assuming that in the continuous-flow system typical of Biacore SPR machines,

[P]

= [P], a plot of θ

against [P]

should allow a direct fit by Eq. 7 to give an

estimation of K

, from which we obtain K

= 1/K

1.5.2. Kinetic Analysis

The protein that is injected across the surface should after an infinite time

arrive at an association equilibrium giving a signal R

, and the resonance sig-

nal R at time t during this process following injection at t = 0 when R = R

should, in simple instances, obey the expression

540 Buckle

= R

– (R

– R

) (1 – e

–k

obs

) (8)

Similarly, for the dissociation of the bound protein,

= R

+ (R

– R

) (e

–k

off

) (9)

assuming that the bound molecule completely dissociates from the immobi-

lized ligand. Consequently, the observed reaction rate k

obs

for the interaction is

given by

obs

+ k

[P] + k

off

(10)

There is thus a linear relationship between the value for k

obs

and the total

concentration of protein [P]. The value for k

obs

can be obtained from a direct fit

of the association phase using Eq. 8, or by linear regression of a semi-log plot.

It thus follows that linear regression analysis of the dependence of k

obs

on [P]

allows the calculation of k

and k

off

using Eq. 7. If we assume that the reaction

is in fact activation controlled (were it otherwise, then the association rate

Fig. 4. Protein binding to immobilized DNA. In this example, purified RNA poly-

merase (120 nM) from Eschericia coli was injected at 20 µL/min across an immobi-

lized 203-bp DNA fragment containing a promoter sequence (continuous line [a]).

The dotted line (b) shows the same protein flowing across a streptavidin surface with-

out immobilized DNA. Note the large bulk refractive index effects resulting princi-

pally from the presence of glycerol in the protein solution and nonspecific binding.

This is a complex phenomenon composed not only of electrostatic interactions with

the dextran but also necessary transient interactions with nonpromoter DNA. The dis-

sociation phase is characterized by a steady decrease in signal lending itself to the type

of analysis described in the text. The association phase is more complicated. In this

instance an example of how the association phase may be dealt with is given in ref. 7.

Surface Plasmon Resonance 541

would be of the order of 10

/M/s, which is well beyond the range of current

SPR devices), then

off

= k

(11)

Thus, the equilibrium dissociation constant (K

) can be obtained from the ratio

of the off and on rates.

There are many pitfalls to using SPR, which are covered in several fairly

recent reviews (2,3). In the case of the Biacore instrumentation, the new data

evaluation software deals with certain situations. What it cannot do is to deter-

mine the best strategy for setting up an experiment. In summary, certain

important points must be taken into account even in the simple analysis given

above. Several of these are covered in Notes 2–6.

2. Materials

1. An SPR device. In this chapter, a Biacore instrument is referred to either as the

classic Biacore or the Biacore 2000. It is recommended (but not essential) that

the machine be modified such that the two racks into which samples are placed in

the machine are separately thermostated. Rack 1 should be thermostated to 4°C;

rack 2 should be thermostated to the temperature at which the interaction is to be

measured. The protocols illustrated here require rack D in the first position and

rack A in the second position.

2. Streptavidin from Pierce resuspended in 0.22 µm filtered distilled water to a final

concentration of 5 mg/mL. This preparation may be stored at 4°C for up to 3 mo.

3. HBS buffer: 10 mM HEPES, pH 7.4, 150 mM NaCl, 3.4 mM EDTA, and 0.005%

Biacore surfactant.

4. N-Ethyl-N'-(diethylaminopropyl) carbodiimide (EDC) and N-hydroxyl

succinimide (NHS) purchased from Biacore as lyophilized powders are

resuspended in 0.22 µm filtered distilled water to a final concentration of

100 mM each.

5. 1 M ethanolamine hydrochloride (pH 8.5), purchased from Biacore, stored

at 4°C.

6. HBS buffer: 10 mM HEPES, pH 7.4, 150 mM NaCl, 3.4 mM EDTA, and 0.005%

surfactant P20.

7. Sensor chip surface CM5 research grade installed in the Biacore apparatus and

preprimed with HBS buffer.

8. Reaction vials for the Biacore (small, plastic = 7 mm; medium, glass = 16 mm;

large, glass = 2 mL) purchased from Biacore.

9. End biotinylated DNA suspended in HBS buffer to 10 µg/mL. This DNA can

either be purchased directly or constructed by polymerase chain reaction

(PCR) using templates and an oligomer primer carrying a biotin group (pur-

chased from Genset for example) as one of the primers. It is advisable to gel

purify or high-performance liquid chromatography (HPLC) purify the DNA prior

to immobilization.

542 Buckle

3. Methods

3.1. Immobilization of the Ligand on the Surface

3.1.1. Coupling of Streptavidin

1. Prime the apparatus with HBS buffer.

2. The thawed EDC solution, in an Eppendorf tube with the top removed, is placed

in rack 1 position a1 (r1a1).

3. The thawed NHS solution in an Eppendorf tube with the top removed, is placed in r1a2.

4. Streptavidin (5 mg/mL, 50 µL), in an Eppendorf tube with the top removed, is placed

in r1a3; 2 mL of filtered (0.2 µm) distilled water is placed in a large glass vial in r2f7.

5. 1 M, sodium acetate buffer (1 mL, pH 4.5) is placed in a large glass vial in r2f3.

6. 1 M ethanolamine (200 µL) is placed in a large tube in r2f4.

7. Two small clean plastic vials are placed in r2a1 and r2a2.

8. An empty large glass vial is placed in r2f5.

9. The following method is programmed into the Biacore or Biacore 2000, checked

for errors, and run.

DEFINE APROG mixing

FLOW 20

TRANSFER r1a1 r2a1 50 !rack1a1 = EDC

TRANSFER r1a2 r2a1 50 !rack1a2 = NHS

MIX r2a1 50 !rack2a1 = EDC/NHS mix

TRANSFER r2f7 r2a2 200 !rack2f7 = distilled water

TRANSFER r2f7 r2a2 290 !rack2f7 = distilled water

TRANSFER r2f3 r2a2 5 !rack2f3 = 1 M acetate pH 4.5

TRANSFER r1a3 r2a2 5 !rack1a3 = streptavidin (5 µg/mL)

MIX r2a2 50

END

DEFINE APROG bind

CAPTION activation

FLOW 20

* INJECT r2a1 50

-0:20 RPOINT EDC/NHS -b

* INJECT r2a2 30

-0:20 RPOINT streptavidin

* INJECT r2f4 35 !Ethanolamine (1 M)

-0:20 RPOINT ethanolamine

15:00 RPOINT bound

END

MAIN

FLOWCELL 1

APROG mixing

FLOWCELL 1

APROG bind

END

(See Note 3.)

Surface Plasmon Resonance 543

3.1.2. Immobilization of the DNA

1. Place the streptavidin-activated sensor chip surface CM5 research grade in the

Biacore apparatus and preprime with HBS buffer.

2. Select a surface pretreated with streptavidin.

3. Flow HBS buffer at 20 µL/min across the surface.

4. Inject the DNA solution across the surface, set the baseline to the point of injec-

tion, and monitor the change in RU during the injection phase. Ideally, between

20 and 100 RU of DNA should be immobilized (Fig. 3).

5. Wash the surface with a 50-µL injection of 1 M NaCl in filtered (0.2 µm) distilled

water.

6. Allow the surface to equilibrate in HBS buffer to a stable baseline, the difference

in RU between the beginning of the injection phase and the end of the wash

period reflects the amount of DNA bound. For stoichiometry, and availability of

sites, see Note 6.

7. Note that values in excess of 100 RU for DNA molecules of appro 100–1000 bp

are to be avoided for a number of reasons (see Notes 2–6).

3.2. Protein Binding to the Immobilized DNA

1. The protein should be prepared in the required buffer over a range of concentra-

tions, at least two orders of magnitude on either side of the suspected K

. The

detergent P20 should be present at concentrations of around 0.005% unless it has

been shown to have a deleterious effect on the interaction with the DNA.

2. Samples should be injected over the both the immobilized DNA surface and a

surface that has been treated with streptavidin and ethanolamine but no DNA as a

blank (Fig. 4).

3. The baseline should be stable with a slope inferior to 10 RU/min. If this is not the

case, then check the temperature of both the apparatus and the continuous-flow

buffer. If the problem persists, replace the sensor surface with an old used chip

and carry out a desorb and sanitize. Re-equilibrate the immobilized DNA chip in

new filtered (0.22-µm filters) buffer by running the prime command. If the prob-

lem persists, a potential reason may be degradation of the integrated fluid car-

tridge necessitating its replacement (see Note 8).

4. A typical sensorgram is shown in Fig. 4. Note that at low protein concentrations

it is very difficult to obtain steady-state saturation levels. Note also that there is a

nonspecific interaction with the control surface (curve b) and also a contribution

from the bulk refractive index effect and that both of these must be taken into

consideration when deducing kinetic or equilibrium values.

4. Notes

1. The relationship between mass and refractive index changes as measured by

changes in the angle at which resonance occurs in this system, although theoreti-

cally available through the additive properties of molar refractivity, has been

empirically established such that 10

–1

° is equivalent to 1000 resonance units (RU)