Moss Tom. DNA-protein interactions: principles and protocols

Подождите немного. Документ загружается.

Assay of Restriction Endonucleases 483

each calculation. The experimental data entered into spreadsheet columns are

counts

complex

and the competitor DNA concentration [D

]

for each titration point.

3.4. Fluorescence Anisotropy

3.4.1. K

Determination. Example: K

≈

40 n

1. Set the fluorimeter to measure anisotropy in the “single-point polarization” mode.

Each anisotropy value should be measured 10 times (measurement time, 5 s) and

averaged (automatically carried out by the fluorimeter). Set the G-factor to “per

measurement” (see Note 14).

2. An excitation wavelength of 530 nm is used with all the slits on the excitation

monochromator set to 8.

3. Anisotropy is measured in the “L” format through the right photomultiplier with

a 570-nm long-pass filter between the sample and the photomultiplier. Using a

filter rather than a monochromator on the emission side greatly increases the

light intensity at the photomultiplier and so allows measurement of lower con-

centrations of fluorophore.

4. Place a 0.5-mL fluorescence cuvet in the fluorimeter. The cuvet should not be

removed from the fluorimeter or otherwise moved during the experiment. Add

fluorescence binding buffer such that the final volume after step 3 will be

0.5 mL. A measurement temperature of 25°C is used.

5. Add duplex DNA, one strand of which is labeled with hexachlorofluorescein

(Subheading 2.1., item 2) to give a final concentration of 10 nM (see Note 15).

Note the anisotropy.

6. Add EcoRV endonuclease in small aliquots (see Note 16), using a microliter

Hamilton syringe, to cover the range 0–400 nM. Mix thoroughly by gently with-

drawing the contents with a plastic pipet tip and readding to the cuvet. After each

addition, measure the anisotropy.

7. At the beginning (all oligonucleotide free in solution) and end (all oligonucle-

otide bound to endonuclease) of the titration, the fluorescence emission intensity

should be noted (see Note 14).

3.4.2. Data Analysis

1. The fluorescence anisotropy measurements were carried out at [D]

≈ K

(see

Note 15). Under these circumstances, [E]

≠ [E]

and so the simplified, hyper-

bolic, form of the binding equation cannot be used. Therefore, data must be fitted

using the full quadratic binding equation (see Note 4). A solution, in terms of

anisotropy, is

A = A

min

(

max –

–

min

)[

(

([D]

+ [E]

+ K

)–

√[

([D]

+ [E]

+ K

)

– (4 [D]

[E]

)

]

[D]

where A is the measured anisotropy; A

min

is the anisotropy of free DNA, and is

the A

max

anisotropy of DNA when fully bound to the endonuclease.

2. The data should be fitted to the above equation using GraFit (11). This requires

entry of A and [E]

into the spreadsheet provided by the software. [D]

is a con-

484 Connolly et al.

stant (i.e., the concentration of oligonucleotide present at the start of the experi-

ment). The titration consists of the addition of multiple aliquots of endonuclease

to the cuvet, resulting in a dilution of the components. Correction should be made

for the dilution of the endonuclease as aliquots are progressively added. Anisot-

ropy, a ratio of fluorescence intensities, does not depend on concentration and is,

therefore, unaffected by dilution. However, as [D]

is treated as a constant by this

software it is not possible to compensate for its dilution. Therefore, the final

dilution should not exceed about 10%. This may require endonuclease addition

from several stocks of different concentration. GraFit also requires an estimate of

min

, A

max

, and K

; suitable values for the first two can be obtained from the

measured anisotropy values at the start (before endonuclease addition) and (after

the final addition of endonuclease) at the end of the experiment. The software

uses nonlinear regression analysis to calculate best-fit values of these three

parameters. Note that the equation given above only holds if the fluorescence emis-

sion intensity of the free and protein-bound DNA are identical (see Note 14).

This is determined in Subheading 3.4.1., step 6.

3. A representative titration curve is shown in Fig. 7.

4. Notes

1. With restriction endonucleases, it is often found that the binding of DNA and

are rapid relative to subsequent events such as conformational changes,

hydrolysis, and product release. This means that rate constants measured under

single-turnover conditions are the same whether the reaction is initiated by add-

ing Mg

to an endonuclease–DNA solution or adding the enzyme to a solution

containing DNA and Mg

. This should, however, be checked by trying both

initiating methods.

2. The K

values for the EcoRV endonuclease and the oligonucleotides used in

Subheadings 3.1.1. and 3.1.2. vary between 2 and 40 nM. The DNA concentra-

tion used in Subheadings 3.1.1. and 3.1.2. (25 × K

for a K

of 40 nM) and an

endonuclease level 10 times higher ensures that ≥95.5% of the nucleic acid

is bound to the protein. Many other concentrations of the two macromolecules

formally meet this requirement. A useful combination, especially for proteins

that are unstable, poorly soluble, and prone to form inactive tetramers and

insoluble higher aggregates when free in solution is [D]

= 50 K

to100 K

and

[E]

= 2 × [D]

. This results in at least 98% of the DNA being bound and avoids a

large excess of the labile free protein. When bound to DNA, many endonucleases

are both stabilized and protected from aggregation. For substrates with poor K

and k

(especially so-called “star” sites, where one of the bases in the recognition

sequence is replaced by another natural base), dissociation rate constants are

much greater than the cleavage rate constants. Therefore, multiple protein

dissociations and associations occur prior to a productive cleavage event and so

very poor substrates are never really cleaved under true single-turnover condi-

tions. However, high concentrations of protein and nucleic acid ensure that the

rate (k

× [DNA][E]) of association is as fast as possible and minimizes the

Assay of Restriction Endonucleases 485

Fig. 7. Representative binding isotherm determined by fluorescence anisotropy

assay. An oligonucleotide composed of the complementary 14-mers 5'-Hex-

TCCGGATATCACCT and 5'-AGGTGATATCCGGA was titrated with the EcoRV

endonuclease. The binding buffer contained 50 mM Tris-HCl and 100 mM NaCl and

the experiment was carried out at 25°C. The solid circles are experimental points. The

value determined from the GraFit fit to the curve is 42 nM.

chances that it contributes to the rate-determining step. Whatever levels of pro-

tein and DNA are selected, the rate constants should be independent of protein

concentration if true single-turnover conditions apply. A useful empirical test is

to measure rates at several protein concentrations to confirm this assumption.

3. For the equilibrium,

Endonuclease + DNA

Endonuclease–DNA K

= k

off/

off

the stability of a protein–DNA complex depends on k

off

. Most proteins have k

values of between 10

and 10

–1

enabling an estimate of the half-life (0.693/

off

) expected for an endonuclease–DNA complex at different K

values:

= 1 µM (K

= 1 × 10

–1

); half-life between 0.01 and 1 s

= 1 nM (K

= 1 × 10

–1

); half-life between 10 s and 10 m

= 1 pM (K

= 1 × 10

–1

); half-life between 100 m and 150 h

= [E]

[D]

/[ED] = ([E]

– [ED]) × ([D]

– [ED])/[ED]

Therefore, the fractional saturation of the DNA with protein, [ED]/[D]

(the

parameter measured using filter binding, gel shift, or fluorescence anisotropy),

has a quadratic relationship to [E]

. However, under the condition [E]

= [E]

(which occurs if [E]

>> [ED]),

= [E]

× ([D]

– [ED])/[ED]

Here, [ED]/[D]

has a hyperbolic relationship to [E]

, simplifying data analysis.

In order to achieve the condition [E]

>> [ED] and to obtain a satisfactory titra-

486 Connolly et al.

tion it is also necessary to have [D]

< K

. DNA concentrations ≤15% K

are

sufficient to satisfy the assumption [D]

< K

and give insignificantly different

values whether the hyperbolic or quadratic binding equation is used. How-

ever, very low [D]

often gives binding curves of very poor quality. This may

result from insufficient ligand stabilization of the protein by the nucleic acid.

Therefore, a [D]

in the 5–15% K

range is recommended. Concentrations of pro-

tein are chosen so that the midpoint of the titration curve yields a protein con-

centration approx equal to the expected K

. Clearly, the K

will not be known

prior to experimentation, emphasizing the need for an initial estimation, followed

by a second, more accurate, determination.

5. Filter binding requires passage of the mixture through a nitrocellulose filter and a

brief washing of the filter with buffer to remove excess unbound DNA. The entire

process can be carried out in seconds. Gel retardation takes longer; about

1–2 min to load the samples and the same time to run them into the gel. Running

the gel takes 30–60 min. Therefore, gel retardation can suffer, to a greater extent

than filter binding, from dissociation of the protein–DNA complex. To some degree

the problem is self-diagnostic, because it results in a smeared protein–DNA band

rather than the tightly focused one which arises when no dissociation takes place.

In these cases, K

determination may be achieved by measurement of the free-

DNA band (16), but if this approach is used, an alternative method is best used to

check the validity of the result.

6. As an example, consider an unmodified DNA with a K

of 0.1 nM and four

modified DNAs with K

values that range from 1 nM to 1 mM. Direct equilib-

rium analyses are performed for the unmodified substrate and the modified site

with K

= 1 nM. Equilibrium competition analyses are also performed for the

four modified sites. If identical values of K

are obtained from both direct and

competition analyses for the modified site with K

= 1 nM, the modified site

serves as an internal control for the three modified sites which bind more weakly.

7. The relationship between fluorescence anisotropy, lifetime, and rotation is given

by the Perrin equation (7,8):

A = A

/(1 + τ/φ)

where A is the measured anisotropy, A

= intrinsic anisotropy (i.e., the anisot-

ropy measured under conditions where no depolarization takes place), τ is the

lifetime of the fluorescence excited state, and φ is the rotational correlation time,

a measure of rotational diffusion that depends on the molar volume and, hence,

the molecular mass attached to the fluorophore. The values of τ and φ must be

balanced: a very short lifetime, relative to the correlation time, would lead to

A = A

; a very long lifetime to A = 0. Use of this method requires 0 < A < A

. The

τ value of approx 3 ns for hexachlorofluorescein attached to oligonucleotides is

suitable for distinguishing between free and protein-bound oligonucleotides. The

measured anisotropy is given by:

A = (I

– I

⊥

)/(

⊥

+2I

⊥

)

Assay of Restriction Endonucleases 487

and I

⊥

are the intensities of the parallel and perpendicular components of the

emitted light when parallel excitation is used.

8. The buffer components used for k

and K

determination will obviously vary

somewhat depending on the enzyme under study and the particular experiment.

However, for most DNA-binding proteins, including EcoRI, EcoRV, and BamHI

restriction endonucleases there is pronounced nonlinearity in the dependence of

log K

versus salt concentration (17) at salt concentrations ≤0.1 M, presumably

because at these low salt concentrations, coupled protein–protein equilibria lead-

ing to aggregation (18) become significant. This phenomenon can produce mis-

leading quantitative comparisons and thus salt concentrations below 0.1 M should

be avoided. Bis-Tris propane is useful as it has two pK values (6.8 and 9), allow-

ing pH-dependence studies with a single buffer. However, other buffers, of

appropriate pK can equally well be used. Bovine serum albumin and DTT can

help to stabilize certain restriction endonucleases, especially at low concentra-

tions. With several DNA-binding proteins, including EcoRI endonuclease,

aggregation and ultimately precipitation, in the absence of the stabilizing DNA

substrate, is quite common at salt concentrations below 0.4 M. In these cases,

intermediate dilutions should be made with buffers containing 0.6 M NaCl. Five

percent glycerol, in the dilution buffers, also aids stability.

9. The “classic” gel-shift buffer, used to prepare and run nondenaturing polyacryla-

mide gel is 0.089 M Tris-borate, pH 8.0, containing 1 mM EDTA (19). Many

protein–DNA complexes, particularly those of high affinity, are stable in this

buffer. Therefore, it is often possible to incubate enzyme and DNA in a selected

binding buffer and then carry out gel retardation in Tris-borate electrophoresis

buffer without disturbing the equilibrium initially set up. However, the binding

of restriction endonucleases to DNA depends on pH, salt concentration (NaCl or

KCl), and divalent cations (Ca

, a Mg

mimic that does not allow hydrolysis,

often considerably strengthens binding) (18,20,21). Ideally it would be best to

use identical incubation and electrophoresis buffers. We have obtained satisfac-

tory gel retardation data using Tris-borate at pH 7.5. Replacing the EDTA with

up to 5 mM CaCl

is also not detrimental. However, many buffers are not suitable

for electrophoresis, giving poorly resolved or smeared bands, especially if NaCl

or KCl is added.

10. Gel retardation requires that the sample applied to the gel be denser than the

electrophoresis buffer and so sinks into the well. The presence of 3% glycerol in

the binding reactions ensures efficient sinking of the sample (18). If this method

is used, it should be checked that glycerol does not perturb K

values (e.g., by carry-

ing out filter binding ± glycerol. An alternative involves addition of half a volume

(i.e., in this case, 5 µL to the 10 µL binding reaction) of 25% (w/v) sucrose solution

following incubation and immediately prior to loading on the gel. This dilutes the

sample, and in cases of weak binding, the associated fast k

off

rates may lead to some

dissociation. However, if the binding is strong, the small dilution factor is unlikely

to be a problem. It is best not to add the bromophenol blue marker to the solution

of enzyme and DNA, as the dye can diminish (18) binding.

488 Connolly et al.

11. Phosphorimaging is the best way to quantitate the amount of radioactivity in

bands on gels. Phosphorimaging systems are now fairly standard in most labora-

tories. As an alternative autoradiography followed by scanning of the bands may

be used, but it is more cumbersome and less accurate.

12. Fitting to the full kinetic model is actually the least robust for the unmodified enzyme-

oligonucleotide pair because the rate constants are so similar numerically that the

program has difficulty partitioning among them. Thus, many closely spaced, and

accurate, time-points should be taken and this will usually require rapid quench.

13. The total concentration of endonuclease and radiolabeled reference DNA (D

)

are chosen to be at least fourfold over the predicted equilibrium dissociation con-

stant (K

) for their interaction. [E]

and [D

]

in the range K

to 30 K

have been

tested; the K

(equilibrium dissociation constant for competitor DNA) is not

affected significantly.

14. The change in anisotropy is only strictly related to the molecular mass of the

fluorophore and hence the amount of oligonucleotide complexed with endonu-

clease, if the following two criteria are met. First, the detector response to the

parallel and perpendicular polarized light must be identical. The G-factor mea-

sures this response, which is automatically corrected for, at each reading, with

the fluorimeter used here. Second, the quantum yield of the free oligonucleotide

and the enzyme-bound oligonucleotide must be the same (i.e., under identical

conditions, the free and bound oligonucleotide should emit the same amount of

light). If the quantum yield of the free and bound oligonucleotide varies the equa-

tion used to fit the data will not be valid (7,22,23). In the example shown in

Fig. 7, the emission intensity of the free and bound oligonucleotide varies less

than 10%. However, the use of other hexachlorofluorescein oligonucleotides and

different buffers with the EcoRV endonuclease often results in large intensity

changes (>10%). In these cases, fluorescence anisotropy cannot be used unless

corrections are made (7,22,23).

15. Good results have been obtained using a DNA concentration about equal to the

and varying protein concentrations from 0 to (20–30) K

. However, lower

concentrations of DNA can also be used, although 1 nM represents the detection

limit. DNA levels much higher than the K

should be avoided, as this promotes

stoichiometric titration (where each protein molecule added binds to the nucleic

acid), conditions under which accurate K

evaluation is not possible.

16. Purified endonucleases are often stored in buffers containing glycerol. Fluores-

cence anisotropy is very sensitive to viscosity and the addition of large amounts

of glycerol during the titration adversely effects the measurement. Therefore,

intermediate dilutions should be made with buffers lacking glycerol. This usually

ensures that the amount of glycerol added to the fluorescence cuvet is negligible

(≤2%). In the case of weak binding, when an enzyme may have to be added

directly from the storage buffer, dialysis is recommended to remove the glycerol.

Acknowledgments

The Newcastle group would like to thank G. Baldwin and S. Halford

(Bristol) for the use of their quench-flow apparatus and help with the data

Assay of Restriction Endonucleases 489

shown in Fig. 4. B.A.C. is supported by the UK BBSRC and MRC. L.J.J. is

supported by a grant (GM-29207) from the National Institutes of Health (USA).

B.A.C. and L.J.J. have a Wellcome Trust biomedical research collaboration grant.

References

1. Roberts, R. J. and Macelis, D. Rebase; the restriction enzyme database. http://

www. neb. com/rebase/rebase. html

2. Jen-Jacobson, L. (1995) Structural-perturbation approaches to thermodynamics

of site-specific protein-DNA interactions. Methods Enzymol. 259, 305–344.

3. Jen-Jacobson, L. (1997) Protein-DNA recognition complexes: conservation of

structure and binding energy in the transition state. Biopolymers 44, 153–180.

4. Lesser, D. R., Kurpiewski, M. R., Waters, T., Connolly, B. A., and Jen-Jacobson,

L. (1993) Facilitated distortion of the DNA site enhances EcoRI endonuclease

DNA interactions. Proc. Natl. Acad. Sci. USA 90, 7546–7552.

5. Kurpiewski, M. R., Koziolkiewicz, M., Wilk, A., Stec, W. J., and Jen-Jacobson,

L. (1996) Chiral phosphorothioates as probes of protein interactions with indi-

vidual DNA phosphoryl oxygens: essential interactions of the EcoRI endonuclease

with the phosphate at pGAATTC. Biochemistry 35, 8846–8854.

6. Lin, S. Y. and Riggs, A. D. (1972) lac Repressor binding to nonoperator DNA:

detailed studies and a comparison of equilibrium and rate competition methods. J.

Mol. Biol. 72, 671–690.

7. James, D. M and Sawyer, W. H. (1995) Fluorescence anisotropy applied to bimo-

lecular interactions. Methods Enzymol. 246, 283–300.

8. Hill, J. J. and Roger, C. A. (1997) Fluorescence approaches to the study of pro-

tein-nucleic acid complexes. Methods Enzymol. 278, 390–416.

9. Powell, L. M, Connolly, B. A., and Drained, D. T. F. (1998) The DNA binding

characteristics of trimeric EcoKI methyltransferase and its partially assembled

dimeric form detected by fluorescence polarization and DNA footprinting. J. Mol.

Biol. 283, 947–961.

10. Sambrook, J., Fritsch, E. F., and Maniatis, T. (1989) Molecular Cloning. A Labo-

ratory Manual, 2nd ed., Cold Spring Harbor Laboratory, Cold Spring Harbor,

NY, pp. 5.56–5.87 and 11.31–11.39.

11. GraFit (1998) Version 4.05, Erithacus Software, Staines, UK.

12. Scientist (1995) Version 2.0, MicroMath Software Inc., Salt Lake City, UT.

13. Sigma Plot (1996) Version 4.0 SPSS Inc, Chicago, IL.

14. Lesser, D. R., Kurpiewski, M. R., and Jen-Jacobson, L. (1990) The energetic basis of

specificity in the EcoRI endonuclease–DNA interaction. Science 250, 776–786.

15. Jen-Jacobson, L., Lesser, D. R., and Kurpiewski, M. R. (1991) DNA sequence

discrimination by the EcoRI endonuclease, in Nucleic Acids and Molecular

Biology, Vol. 5 (Eckstein, F. and Lilley, D. M. J., eds.), Springer-Verlag, NY, pp.

141–170.

16. Carey, J. (1991) Gel retardation. Methods Enzymol. 208, 103–118.

17. Record, M. T., Jr., Anderson, C. F., and Lohman, T. M. (1978) Thermodynamic

analysis of ion effects on the binding and conformational equilibria of proteins

490 Connolly et al.

and nucleic acids: the roles of ion association or release, screening, and ion effects

on water activity. Quart. Rev. Biophys. 11, 103–178.

18. Engler, L. E., Welch, K. K., and Jen-Jacobson, L. (1997) Specific binding by the

EcoRV restriction endonuclease to its recognition site GATATC. J. Mol. Biol.

269, 82–101.

19. Revzin, A. (1987) Gel electrophoresis assays for DNA–protein interactions

BioTechniques 7, 346–355.

20. Taylor, J. D. and Halford, S. E. (1989) Discrimination between DNA sequences

by the EcoRV restriction endonuclease. Biochemistry 28, 6198–6207.

21. Vipond, I. B. and Halford, S. E. (1995) Specific DNA recognition by EcoRV

restriction endonuclease induced by Ca

ions. Biochemistry 34, 1113–1119.

22. Eftink, M. R. (1997) Fluorescence methods for studying equilibrium macromol-

ecule-ligand interactions. Methods Enzymol. 278, 221–257.

23. Gutfreund, H. (1995) The role of light in kinetic investigations, in Kinetics for the

Life Sciences Receptors, Transmitters and Catalysts. Cambridge University Press,

Cambridge, pp. 291–295.

Intrinsic Fluorescence 491

491

From:

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

Analysis of DNA–Protein Interactions

by Intrinsic Fluorescence

Mark L. Carpenter,Anthony W. Oliver, and G. Geoff Kneale

1. Introduction

Changes in the fluorescence emission spectrum of a protein upon binding to

DNA can often be used to determine the stoichiometry of binding and equilib-

rium binding constants; in some cases the data can also give an indication of

the location of particular residues within the protein. The experiments are gen-

erally quick and easy to perform, requiring only small quantities of material

(1). Spectroscopic techniques allow one to measure binding at equilibrium

(unlike, for example, gel retardation assays and other separation techniques

that are strictly nonequilibrium methods). Fluorescence is one of the most sen-

sitive of spectroscopic techniques, allowing the low concentrations (typically

in the nanomolar to micromolar range) required for estimation of binding con-

stants for many protein–DNA interactions. Considerable care, however, needs

to be exercised in the experiment itself and in the interpretation of results. The

fundamental principles of fluorescence are discussed briefly in the remainder

of the Introduction.

A molecule that has been electronically excited with ultraviolet (UV)/vis-

ible light can lose some of the excess energy gained and return to its ground

state by a number of processes. In two of these, fluorescence and phosphores-

cence, this is achieved by emission of light. Phosphorescence is rarely observed

from molecules at room temperature and will not be considered further.

Although electrons can be excited to a number of higher-energy states, fluores-

cence emission in most cases only occurs from the first vibrational level of the

first excited state. This has two implications for the measurement of fluores-

cence emission spectra. First, some of the energy initially absorbed is lost prior

to emission, which means that the light emitted will be of longer wavelength

492 Carpenter, Oliver, and Kneale

(i.e., lower energy) than that absorbed. This is known as the Stoke’s shift. Sec-

ond, the emission spectrum and therefore the wavelength of maximum fluores-

cence will be independent of the precise wavelength used to excite the

molecule. Thus, for tyrosine, the wavelength of the fluorescence maximum is

observed around 305 nm, regardless of whether excitation is at the absorption

maximum (approx 278 nm) or elsewhere in the absorption band. Of course, the

fluorescence intensity will change as a consequence of the difference in the

amount of light absorbed at these two wavelengths.

The fraction of light emitted as fluorescence compared to that initially

absorbed is termed the quantum yield. The value of the quantum yield for a

particular fluorophore will depend on a number of environmental factors such

as temperature, solvent, and the presence of other molecules that may enhance

or diminish the probability of other processes deactivating the excited state.

The deactivation or quenching of fluorescence by another molecule, either

through collisional encounters or the formation of excited-state complexes,

forms the basis of many of the fluorescence studies on protein–DNA interactions.

The study of protein–nucleic acid interactions is greatly simplified by the

fact that all detectable fluorescence arises from the protein, all four of the natu-

rally occurring DNA bases being nonfluorescent by comparison. Tyrosine and

tryptophan residues account for almost all the fluorescence found in proteins.

As a general rule, when both residues are present, the emission spectrum will

be dominated by tryptophan, unless the ratio of tyrosines to tryptophans is very

high. The quantum yield of a tyrosine residue in a protein compared to that

observed in free solution is generally very low, illustrating the susceptibility of

tyrosine to quenching. Tryptophan residues are highly sensitive to the polarity

of the surrounding solvent, which affects the energy levels of the first excited

state with the result that the emission maximum for tryptophan can range from

330 nm in a hydrophobic environment to 355 nm in water. Thus, in proteins

containing only one tryptophan, the general environment surrounding the resi-

due can be ascertained. Tryptophan fluorescence, like that of tyrosine, can be

quenched by a number of molecules, including DNA. Unlike tyrosine, the emis-

sion maximum can also change if tryptophan is involved in the interaction and

this can also be used to monitor DNA binding (2).

The extent to which the fluorescence of a protein is quenched by DNA is

proportional to the concentration of quencher. As quenching is the result of the

formation of a complex between the protein and the DNA, the extent of quench-

ing is proportional to the amount of bound protein. Thus by determining the

extent to which the protein fluorescence is quenched when fully bound to DNA

(i.e., at saturation), the fraction of bound and free protein at any point in a

titration can be determined. From these data, the stoichiometry and binding

constant of the interaction can often be obtained. Note that to establish an