New comprehensive biochemistry vol 11 modern physical methods in biochemistry part b

MODERN PHYSICAL METHODS
IN BIOCHEMISTRY, PART B


New Comprehensive Biochemistry

Volume 11B

General Editors

A. NEUBERGER
London

L.L.M. van DEENEN
Utrecht

ELSEVIER
AMSTERDAM. NEW YORK . OXFORD


Modern Physical Methods in
Biochemistry
Part B

Editors

A. NEUBERGER and L.L.M. VAN DEENEN
London and Utrecht

1988

ELSEVIER
AMSTERDAM * NEW YORK . OXFORD


0 1988, Elsevier Science Publishers B.V. (Biomedical Division)

All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or
transmitted in any form or by any means, electronic, mechanical, photocopying, recording or otherwise,
without the prior written permission of the Publisher, Elsevier Science Publishers B.V. (Biomedical
Division), P.O. Box 1527, loo0 BM Amsterdam, The Netherlands.

No responsibility is assumed by the Publisher for any injury and/or damage to persons or property as a
matter of products liability, neghgence or otherwise, or from any use or operation of any methods;
products, instructions or ideas contained in the material herein. Because of the rapid advances in the
medical sciences, the Publisher recommends that independent verification of diagnoses and drug dosages
should be made.
Special regulations for readers in the USA. This publication has been registered with the Copyright
Clearance Center, Inc. (CCC), Salem, Massachusetts. Information can be obtained from the CCC about
conditions under which the photocopying of parts of this publication may be made in the USA. All other
copyright questions, including photocopying outside of the USA, should be referred to the Publisher.

ISBN 0-444-80968-6(volume)
ISBN 0-444-80303-3 (series)
Published by:
Elsevier Science Publishers B.V. (Biomedical Division)
P.O. Box 211
loo0 AE Amsterdam
The Netherlands
Sole distributors for the USA and Canada:
Elsevier Science Publishing Company, Inc.

52 Vanderbilt Avenue
New York, NY 10017
USA

Library of Congress Cataloging-in-PublicationData
(Revised for volume 11 B)
Modem physical methods in biochemistry.
(New comprehensive biochemistry; v. 11 A, B)
Includes bibliographies and index.
1. Spectrum analysis. 2. Biochemistry-Technique.
I. Neuberger, Albert. 11. Deenen, Laurens L.M. van.
QD415.N48 vol. 11 A, etc. 574.19’2 s [574.19’283] 85-4402
[QP519.9S6]
ISBN 0-444-80649-0 (v. 11 A) 0-444-80968-6 (v. 11 B)
Acknowledgment
Many illustrations and diagrams in this volume have been obtained from other publications. In all cases
reference is made to the original publication. ThejuN source can be found in the reference list. Permission for
the reproduction of this material is gratefully acknowledged.

Printed in The Netherlands


Preface

In the former series of Comprehensive Biochemistry the contributions of physical
methods to biochemistry were considered in volumes 1-4, a section which was
devoted to the physicochemical and organic aspects of biochemistry. In 1962 the
series editors M. Florkin and E.H. Stotz emphasized the importance of these basic
sciences for the future progress in the life sciences. Since that time, the application
of physical methods to biological problems has solved many questions and opened

new avenues of research.
Volume 11,part A, of the present series contained chapters on protein crystallography, nuclear magnetic resonance spectroscopy, electron spin resonance, mass
spectroscopy, circular dichroism and optical rotatory dispersion. In this volume the
range of spectroscopic techniques is extended to chapters on fluorescence and
Raman spectroscopy. One chapter deals extensively with neutron and X-ray solution scattering techniques, and a choice of rapid reaction methods is discussed in a
further chapter. The use of electron microscopy has been another very important
development in the biological sciences and the results are illustrated by a chapter
with emphasis on biomembranes. The New Comprehensive Biochemistry series
contains a volume (8) devoted to separation methods. This area is now supplemented by a chapter in the present volume on high performance liquid chromatography of nucleic acids and a chapter on reversed phase HPLC of peptides and
proteins. The editors hope that the publication of this volume may serve the needs
of many biochemists and thus contribute to further research in the biological
sciences.

A. Neuberger
L.L.M. van Deenen


This Page Intentionally Left Blank


Contents

Preface

V

Chapter I
Fluorescence spectroscopy; principles and application to biological macromolecules
J.R. Lakowicz (Baltimore, MD, USA)

1

1. The phenomenon of fluorescence
2. Factors affecting the fluorescence emission
2.1. Solvent polarity and viscosity
2.2. Emission spectra of melittin
2.3. Quenching of fluorescence
2.4. Fluorescence energy transfer
2.5. Fluorescence anisotropy
3. Time-resolved fluorescence spectroscopy
3.1. Resolution of the emission spectrum of liver alcohol dehydrogenase
3.2. Pulsed lasers for time-resolved fluorescence
3.3. Frequency-domain resolution of protein fluorescence
3.4. Anisotropy decays of protein fluorescence
4. Harmonic-content frequency-domain fluorometry
5. summary

1
4
4
6
7
9
11
13
15
18
19
21
23

25

Acknowledgements
References

25
26

Chapter 2
Raman and resonance Raman spectroscopy
P.R. Carey (Ottawa, Ont., Canada)

27

1.
2.
3.
4.
5.
6.
7.

Introduction
The units used in Raman spectroscopy
A model for Raman scattering based on classical physics
Raman and resonance Raman scattering: a quantum mechanical interpretation
Polarisation properties of Raman scattering
Basic experimental aspects
Raman studies on biological materials
7.1. Proteins

7.1.1. Amide I and amide I11 features
7.1.2. Side chain contributions to the Raman spectrum
7.1.3. Applications
7.1.4. UV excited resonance Raman spectra of proteins

27
29
31
34
37
38
40
40
40
42
43
43


...

Vlll

7.2. Proteins containing a natural, visible chromophore
7.3. Resonance Raman labels
7.4. Nucleic acids
7.4.1. The purine and pyrimidine bases
7.4.2. Conformation of the (deoxy)ribose-phosphate backbone
7.4.3. Resonance Raman studies of nucleic acids
7.5. Viruses

7.6. Lipids and membranes
7.6.1. The C-C stretching region between 1050 and 1150 cm-'
7.6.2. The C-H stretching region between 2800 and 3000 cm-'
7.6.3. Deuterated lipids as selective probes
7.6.4. Lipid protein interactions and natural membranes

44
48
50
50
52
53
54
56
57
57
58
59

References

61

Chapter 3
Rapid reaction methods in biochemistry
Quentin H. Gibson (Ithaca, NY, USA)

65

1. Introduction

2. Continuous flow
3. Stopped flow
3.1. Miscellaneous stopped-flow devices
3.2. Relaxation methods
3.2.1. Flash sources
3.2.2. Observation light sources
3.2.3. Light detectors
4. Combinations of flash photolysis with other techniques
5. Temperature jump
6. Miscellaneous methods
6.1. Time-resolved resonance Raman spectroscopy
6.2. Competitive methods
7. Data reduction

65
65
69
71
72
73
75
75
76
76
77
77
78
78

References

83

Chapter 4
High performance liquid chromatography of nucleic acids
M. Colpan and D. Riesner (Dusseldorf, FRG)

85

1. Introduction
2. Techniques
2.1. Size exclusion chromatography
2.2. Anion-exchange chromatography
2.3. Reversed phase and hydrophobic interaction chromatography
2.4. RPC-5 and other mixed mode chromatography
2.5. Sample preparation and recovery
3. Applications
3.1. Oligonucleotides
3.2. Natural RNA
3.3. DNA fragments
3.4. Plasmids
4. Concluding remarks

85
86
86
88
91
94
95

96
96
98
99
101
102


Acknowledgements
References

Chapter 5
Reversed phase high per,,rmance liquid chromatography of peptides and proteins
M.T.W. Hearn and M.I. Aguilar (Clayton, Vic., Australia)

103
103

107

1. Introduction
2. Retention relationships of peptides in RP-HPLC
3. The relationship between peptide retention behaviour and hydrophobicity coefficients
4. Bandwidtb relationships of peptides in RP-HPLC
5. Dynamic models for interconverting systems
6. Conclusion

107
111
120

126
131
139

Acknowledgements
References

139
140

Chapter 6
X-ray and neutron solution scattering
S.J. Perkins (London, UK)

143

1. Introduction
Part A: Theoretical and Practical Aspects
2. Theory of X-ray and neutron scattering
2.1. Scattering phenomena and their angular ranges
2.1.1. X-ray scattering
2.1.2. Neutron scattering
2.1.3. Scattering angles, vectors and resolution
2.2. The scattering event and the Debye equation
2.3. Scattering densities and allowance for solvent
2.3.1. Concept of scattering densities
2.3.2. Scattering densities and volumes
2.3.3. The contrast difference A p
2.3.4. Mean macromolecular scattering densities p
2.3.5. Scattering density fluctuations pF(r)

2.4. The Guinier plot: Z ( 0 ) and R ,
2.4.1. The innermost scattering curve
2.4.2. Cross-sectional and thickness Guinier analyses
2.5. Analyses of I ( 0 ) values
2.6. Analyses of R, values
2.7. Non-uniform scattering densities and contrast variation
2.7.1. The Stuhrmann plot
2.7.2. Solvent penetration and exchange effects
2.7.3. Isomorphous replacement
2.7.4. Matchpoints of multicomponent systems
2.8. Label triangulation
2.9. Wide-angle scattering and modelling strategies
2.9.1. Spheres and ellipsoids
2.9.2. Scattering curves at large Q
2.9.3. Independent parameters from scattering
2.9.4. Debye curve simulations
2.9.5. Interparticle interference
2.10. Distance distribution functions

143
144
144
144
144
145
146
147
149
149
150

152
154
154
160
160
162
163
165
167
167
170
172
173
173
175
175
177
178
178
180
180


X

3. Experimental practice and instrumentation
3.1. Sample preparation and measurement
3.1.1. Sample monodispersity and concentrations
3.1.2. Sample assays
3.1.3. Sample backgrounds

3.1.4. Sample holders
3.1.5. Instrumental calibration
3.2. Labelling techniques and deuteration
3.3. Sources of X-rays and neutrons
3.3.1. Anode sources
3.3.2. Synchrotron radiation
3.3.3. Reactor neutron sources
3.3.4. Spallation neutron sources
3.4. Scattering instrumentation
3.4.1. X-ray cameras
3.4.2. Neutron cameras
3.5. Data reduction
Part B: Biochemical Applications to Proteins, Carbohydrates, Lipids and Nucleic Acids
4. Applications of X-ray and neutron scattering
4.1. Introduction
4.2. X-ray studies on globular proteins
4.2.1. Relationship between R , and M ,
4.2.2. Comparison of crystal and solution structures
4.2.3. Conformational changes and ligand binding
4.2.4. AUostericism
4.2.5. Molecular modelling of proteins
4.2.6. Associative systems and time-resolved synchrotron radiation studies
4.2.7. Interparticle interference
4.2.8. X-ray contrast variation and anomalous scattering
4.2.9. Label triangulation of heavy metal probes
4.3. Neutron studies on globular proteins
4.3.1. Contrast variation studies
4.3.2. Label triangulation and deuteration
4.4. X-ray and neutron studies on glycoproteins
4.4.1. Plasma glycoproteins, proteoglycans and polysaccharides

4.4.2. Immunoglobulins
4.4.3. Components of complement
4.5. Lipids, detergents, membrane proteins and lipoproteins
4.5.1. Lipid vesicles and complexes with proteins
4.5.2. Detergent micelles and complexes with proteins
4.5.3. Lipoproteins
4.6. Nucleic acids and nucleoproteins
4.6.1. DNA studies by X-ray scattering
4.6.2. X-ray and neutron studies on transfer RNA
4.6.3. Protein-nucleic acid interactions by neutron scattering
4.6.4. Chromatin and chromosomes by X-rays and neutrons
4.6.5. Ribosomes and their constituents
4.6.6. Viruses
5. Conclusions
Acknowledgements
References

182
182
183
183
184
184
185
186
187
187
187
189
190

190
190
191
193
194
194
194
194
194
196
196
198
199
201
203
204
207
208
208
21 1
213
21 3
218
219
221
221
224
226
230
230

231
234
236
239
244
249
25 1
251


Chapter 7
Electron microscopy
W.F. Voorhout and A.J. Verkleij (Utrecht, The Netherlands)

267

1. Introduction
2. Negative staining and metal shadowing
2.1. Negative staining
2.2. Metal shadowing
3. Thin sectioning
4. Low-temperature techniques
4.1. Cryofmation
4.2. Freeze-fracturing
4.2.1. The freeze-fracture technique
4.2.2. Biological membranes
4.2.3. Lipid phase transitions and lipid polymorphism as visualized by freeze-fracturing
4.3. Localization studies
4.3.1. Introduction
4.3.2. Immunocytochemistry

4.3.3. Marker system
4.3.4. Cryo-ultramicrotomy
4.3.5. Cryo-fractures
4.3.6. Label-efficiency
5. Conclusions

261
268
268
269
210
212
212
214
214
215
219
286
286
281
288
289
290
293
295

Acknowledgements
References

295

295

Subject index

301


This Page Intentionally Left Blank


A. Neuberger and L.L.M. Van Deenen (Eds.) Modern Physical Methoak in Biochemimy. Part B
0 1988 Elsevier Science Publishers B.V. (Biomedical Division)

1

CHAPTER 1

Fluorescence spectroscopy; principles
and application to biological macromolecules *
JOSEPH R.LAKOWICZ
University of Maryland at Baltimore School of Medicine, Department of Biological Chemistry,
660 West Redwood Street, Baltimore, M D 21201, USA

1. The phenomenon of fluorescence
Luminescence is the emission of photons from electronically excited states.
Luminescence is divided into two types, fluorescence and phosphorescence. In
phosphorescence, the emission is from an excited triplet state to a ground state
singlet. Since this transition is forbidden the rate of return to the ground state is
slow, which means the decay times are long (msec to sec). Fluorescence is the
emission from excited singlet states, also yielding a ground state singlet. These

allowed transitions to occur rapidly, with rates near lo8 sec-'. Consequently, the
decay times for fluorescence are typically near lo-' sec or 10 nsec. In this chapter
we will discuss primarily fluorescence, but the concepts are also applicable to events
on a slower timescale if the phosphorescence is observed. The nanosecond timescale
of fluorescence provides much of its usefulness in biophysical chemistry. In solutions near room temperature, a variety of molecular events can occur within 10 nsec
and alter the emission. These events include rotational diffusion, collisions with
quenchers, solvent reorientation, and energy transfer. These events alter one or more
of the spectral observables, and can thus be detected by analysis of the emission.
Substances which display fluorescence are generally delocalized aromatic systems
with or without polar substituents (Fig. 1).It is difficult to predict which molecules
will be fluorescent or non-fluorescent because exceptions can usually be found.
However, several general rules are generally true. Rigid molecules are usually more
fluorescent, or at least their fluorescence more predictable, than molecules with the
possibility of internal rotation. Hence, perylene and anthracene fluoresce with high
efficiencies, whereas stilbene can be much less efficient. In viscous solvents, in which
rotational reorientation to cis-stilbene cannot occur, trans-stilbene is highly fluorescent. In non-viscous solution stilbene is only weakly fluorescent. This illustrates
an important aspect of fluorescence, which is that the excited states are involved,
* Dedicated to Professor Gregorio Weber on the occasion of his seventieth birthday.


2

Perylene

Anfhracene

I ndole

P PO

trans- Stilbene

2-Naphthol

Fig. 1. Typical fluorescent molecules.

and these states have a different electronic distribution which may alter their
chemical properties. In the excited state trans-stilbene (Fig. 1) isomerizes to the
non-fluorescent cis-stilbene. The altered electronic distribution can also alter chemical reactivity. For instance, the pKa of the hydroxyl group on naphthol decreases
from 9 to 2 upon excitation, presumably as the result of transfer of electron density
from the oxygen into the aromatic ring. The emission from an aqueous solution of
naphthol can be due to unionized naphthol, naphtholate, or both, depending upon
pH and the concentration of basic species available to accept the dissociated proton.
The presence of substitutes for carbon in the aromatic system generally alters the
emission from the aromatic nucleus. Insertion of oxygen or nitrogen into the ring
system often results in good fluorescence. Hence, indole, fluorescein, PPO, the
rhodamines and similar substances are fluorescent (Fig. 1). The presence of sulfur,
nitro groups or heavy atoms like iodide generally result in quenching of fluorescence.
Biological systems contain a variety of intrinsic (natural) fluorophores (Fig. 2). In
proteins, tryptophan is the most highly fluorescent amino acid, accounting for 90%
of the emission from most proteins. Emission from tyrosine residues is also
observed, especially in proteins lacking tryptophan, in denatured proteins, or in
those with a high ratio of tyrosine to tryptophan. Tyrosine is highly fluorescent in
solution, but its emission is often quenched in native proteins, due either to the
quenching effects of hydrogen bonding to the hydroxyl group or because of energy
transfer from tyrosine to tryptophan. The emission of phenylalanine from proteins
is less studied.
The nucleotides and nucleic acids are generally non-fluorescent. However, some
notable exceptions are known. Phenylalanine transfer RNA from yeast (tRNAPhe)
contains a single highly fluorescent base, called the Y-base, whch has an emission

maximum near 470 nm. The presence of this intrinsic fluorophore has resulted in
numerous studies of tRNAPheby fluorescence spectroscopy. Regarding the “nonfluorescent” nucleic acids, it should be noted that they do fluoresce, but with very
low yields and with short decay times.


3

CH

/I

CH

I

CH

II

CH

I

CH

II

ANS

DNS-CI

DPH

-ATP

Ethidium Bromide

Acridine Orange

Fig. 2. Intrinsic biological fluorophores. For NADH and FAD we only showed the fluorescent part of the
molecule.

Other natural fluorophores include NADH and FAD, whose fluorescent moieties
are shown in Fig. 2. In both cases the amount of fluorescence depends upon their
local environments. For instance, the emission of NADH is usually increased about
three-fold upon binding to proteins, whereas the emission of FAD is usually
quenched.
C02H

C02H
I

a;H2
6
H2N-CHI

H

Tryptophan

Y,- b a s e

H2N-CH

I

C02H

I

H 2N-CH

6
I

\

OH

Tyrosine

NADH

Phenylalanine

FAD

Fig. 3. Typical extrinsic fluorophores used to label macromolecules.

4

In instances where nature has not provided an appropriate fluorophore, one can
often add an extrinsic label. The earliest probes include dansyl chloride [l]and ANS
(Fig. 3). Dansyl chloride can be covalently attached to macromolecules by reaction
with amino groups. ANS often binds spontaneously but non-covalently to proteins
and membranes, probably by hydrophobic and electrostatic interactions. The emission of both molecules is sensitive to the polarity of the surrounding environment.
ANS is nearly non-fluorescent in water, but fluoresces strongly upon association
with serum albumin, immunoglobulins and other proteins. A wide variety of
covalent and non-covalent probes are available [2,3].
Studies of cell membranes by fluorescence depends almost exclusively upon the
use of extrinsic probes. This is because most lipids are not fluorescent, and the
emission from membrane-bound proteins is too heterogeneous for interpretation of
the data. The probe DPH (Fig. 3) is typical of membrane probes, as is perylene (Fig.
1).These non-polar molecules partition spontaneously into membranes. And finally,
Fig. 3 shows several extrinsic probes for nucleic acids. Addition of the etheno bridge
to ATP results in a highly fluorescent residue. Unfortunately, this modification also
disrupts the base pairing of the nucleotide. Alternatively, nucleic acids can be
labeled by spontaneous binding of planar cations such as ethidium bromide and
acridine orange (Fig. 3). Depending upon the structure, the fluorescence of the
probe may be quenched or enhanced upon intercalation into DNA, and the
emission may depend upon whether the intercalation site is adjacent to A-T or
G-C pairs. For instance, the fluorescence yield of ethidium bromide is enhanced
about 30-fold upon intercalation into DNA. Other intercalating dyes such as
proflavin and 9-aminoacridine are quenched by their interactions with DNA [4].

2. Factors affecting the fluorescence emission
2. I . Solvent polarity and viscosity
The variety of factors which can affect the fluorescence emission are illustrated by
the modified Jablonski diagram [5] shown in Fig. 4. In this diagram we emphasize

fluorescence emission and quenching, and hence we have not included the higher
electronic states or the triplet states. Upon absorption of light the fluorophore
arrives instantaneously in the first singlet state (Sl), usually with some excess
vibrational energy. This excess energy is usually dissipated quickly in lo-'' sec by
interaction with the solvent, resulting in a molecule in the lowest vibrational level of
S,. The fluorophores remain at this level for the mean duration of the excited state,
which is typically 10 nsec. Any process or interaction which occurs during this
interval can alter the fluorescence emission. These processes and interactions are the
origin of much of the information available from fluorescence spectroscopy.
Fluorophores with polar groups are often sensitive to solvent polarity. Interaction
between the excited fluorophore and surrounding polar groups lowers the energy of
the excited state, which shifts the emission to longer wavelengths. The relative
amounts of emission from the relaxed and the unrelaxed states depend upon the


5

\j

S L a R e l a Y
Donor Emission

k,

Energy Transfer
Acceptor
absorption

Xkil

h

a

so

Fig. 4. Jablonski diagram for fluorescence emission and quenching.

relative rates of depopulation of the excited state ( r+ Cki) and that of solvent
relaxation ( k R ) .The rate of emission is r and the rate of return to the ground state
exclusive of emission is Cki. In fluid solvents near room temperature the rate of
solvent reorientation is near 10-1'-10-12 sec. Hence, this process is mostly complete prior to emission, so the observed emission is that of the relaxed state. If the
solution is cold or viscous, or if the probe is bound to a rigid site on a macromolecule, then the rates of relaxation and emission can be comparable, so emission is
seen from both the relaxed and the unrelaxed states. If the solution is very viscous
then solvent relaxation does not occur during the lifetime of the excited state, and
the observed emission is from the higher energy (shorter wavelength) unrelaxed
state.
The effects of solvent polarity are best understood by specific examples. To
model the fluorescence emission of proteins we examine spectra for N-acetyl-Ltryptophanamide (NATA). This molecule is analogous to tryptophan in proteins. It
is a neutral molecule, and its emission is more homogeneous than that of tryptophan
itself. In solvents of increasing polarity the emission spectra shift towards longer
wavelengths (Fig. 5). The emission maxima of NATA in dioxane, ethanol and water
are 333, 344 and 357, respectively. These solvents are non-viscous, so the emission is
dominantly from the relaxed state (Fig. 4). The spectral shifts can be used to
calculate the change in dipole moment which occurs upon excitation [6]. More
typically, the emission spectrum for a sample is compared with that found for the
same fluorophore in various solvents, and the environment judged as polar or
non-polar. While this approach is qualitative, it is simple and reliable, and does not
involve the use of theoretical models or complex calculations.
The timescale of the relaxation process also affects the emission. This effect is

illustrated for NATA in propylene glycol (Fig. 6). At room temperature the
relaxation is mostly complete, and a red shifted spectrum is observed (348 nm).
Lowering the temperature results in a progressive shift of the emission to shorter
wavelengths, with an emission maximum of 329 nm at - 60 O C . As the temperature
is decreasing the relaxation rate ( k R )becomes slow relative to that of the decay rate
( r+ Cki). Hence, an increasing proportion of the emission is from the unrelaxed
and the intermediate states (Fig. 4), which have higher energies and shorter emission


6

WAVELENGTH ,(nanometers)

Fig. 5 . Emission spectra of N-acetyl-L-tryptophanamidein various solvents of different polarities.

wavelengths (Fig. 6). In proteins it is probable that the emission maxima are
affected by both the average environment of the tryptophan residues [7] and by the
relaxation rates [8]. Rather detailed data and analysis are needed for an unambiguous separation of these effects, but the average environment of the tryptophan
residues seems to be the dominant determinant of the emission maxima.
2.2. Emission spectra of melittin
Melittin is an amphipathic peptide component of bee venom which associates with
cell membranes, enhances the phospholipase activity of venom and participates in
the disruption of cell membranes. This protein has been studied extensively by

WAVELENGTH [nanometers)

Fig. 6. Emission spectra of N-acetyl-L-tryptophanamidein propylene glycol at 25 and

- 60

C


WAVELENGTH (nanometers)
Fig. 7. Emission spectra of melittin in the absence (data are courtesy of N. Joshi.

) and presence (-

- -)

of 2 M NaCl. The

fluorescence and other physical methods [9,10], and its X-ray structure is known
[ll]. In solution melittin exists as either a monomer or a self-associated tetramer.
The self-association is driven by high salt concentration, which apparently shields
the positive charges on the monomer from each other and allows the hydrophobic
interactions to cause association. The monomeric form of melittin is thought to be
largely random coil with a high degree of segmental mobility. In the tetrameric state
the monomeric units are mostly a-helical.
Melittin is an ideal protein to illustrate the effects of structure on the fluorescence spectral properties. Each monomer contains a singly tryptophan residue and
no tyrosine residues. The X-ray structure of the tetrameric form shows that the
tryptophan residues are buried in a non-polar pocket and are not directly exposed
to the aqueous phase.
The emission spectra of melittin illustrate the effects of solvent exposure on the
tryptophan emission (Fig. 7). In the absence of salt, the emission maximum of 360
nm is comparable to that found for NATA in water. In the presence of 2 M NaCl
the emission maximum is blue shifted by 12 nm to 348 nm. This shift is a result of
shielding of the indole ring from the aqueous phase. Hence, solvent relaxation
proceeds to a lesser extent because there is less solvent available for interaction with
the fluorophore.

2.3. Quenching of fluorescence

Collisional quenching of fluorescence requires contact between the fluorophore and
the quencher. For quenching to occur the quencher must diffuse to and collide with
the fluorophore in the excited state. If this occurs the fluorophore returns to the
ground state without emission of a photon (Fig. 4). Many small molecules act as
collisional quenchers of fluorescence [6,12]. These include iodide, acrylamide,
halogenated hydrocarbons and occasionally amines and metal ions. The excited
state lifetimes provide ample opportunity for quenching. For instance, acrylamide is
known to be an efficient quencher of tryptophan fluorescence [12,13]. Su.ppose its


8

I
P

Melittin
7.01 2 5 O C , p H - 7

Monomer
K =

7,2 M-'

k = 2.1x I 0 ' M-'s-'

5.0-

-FF.

0.6

0.4

0.2

0

0.8

[Acrylamidel ,M

Fig. 8. Stern-Volmer plot for acrylamide quenching of melittin monomer and tetramer. From M. Eftink,
University of Mississippi, Chemistry Department, unpublished observations. The lifetimes are from [35].
The broken lines are the initial slopes, corresponding to the values on the figure.

diffusion coefficient is lop5cm2/sec. In 10 nsec an acrylamide molecule can diffuse
a distance of 44 A, as calculated using A x 2 = 2 0 7 , where A x is the distance, D is
the diffusion coefficient and 7 is the fluorescence lifetime. This distance is comparable to the diameter of many proteins. Hence, we expect quenching to occur to a
measurable degree and the extent of quenching to be sensitive to the average degree
of exposure of the tryptophan to the aqueous phase.
Once again melittin illustrates the effect of protein structure on the fluorescence
emission. Acrylamide quenching data for melittin monomer and tetramer are shown
in Fig. 8. Stern-Volmer plots are often used to present quenchmg data. The
Stern-Volmer equation is
FO
= 1+ k.,[Q]

F

=

1

+K [ Q ]

where Fo and F are the fluorescence intensities in the absence and presence of
quenching, respectively, T~ is the fluorescence lifetime in the absence of quenchmg,
[ Q ]is the concentration of quencher, k is the bimolecular quenching constant, and
K is the Stern-Volmer quenching constant. The lifetime T,, is the reciprocal of the
rates which depopulate the excited state. From Fig. 4,

If every collisional event results in quenching, the bimolecular rate constant can be
estimated using the diffusion constants of the fluorophore ( D F )and quencher ( DQ)
and the radius expected for contact ( R ) ,
k

(3)

= 4~rRDN/1000

where N is Avogadro's number and D

= D,

+ DQ. If the fluorophore is exposed to


9

the solvent we expect k to be near 0.5 X lo1' M-' sec-'. If the residue is shielded
from collisional encounters this rate will be smaller. This comparison is the basis for
estimating the extent of exposure from quenching data. The quenching constant
measured for a protein is compared with that expected for a completely exposed
fluorophore. Typically, model compounds with no possibilities for shielding are
studied to account for lack of precise knowledge of diffusion coefficients, and the
possibility that the quenching encounter is not 100% efficient.
The quenching data for both the monomeric and tetrameric forms of melittin
indicate the tryptophan residues are accessible to acrylamide with the accessibility
being greater in the monomeric state. This conclusion is reached by comparison
with acrylamide quenching data for NATA. At 25°C in water the acrylamide
quenching constant for NATA is 0.58 X 10" M-' sec-' [47]. For the monomer the
quenching constant is about one-third of this value, which is indicative of a rather
fully exposed residue [13]. The value of k for the tetramer is less, indicating
shielding of the tryptophan residue from the aqueous phase. It should be noted that
the relative shielding is only 40%, which probably indicates considerable penetration
of the tetramer by acrylamide. In other more extensive studies Eftink and Ghiron
showed that acrylamide quenching reflects the average degree of tryptophan exposure to the aqueous phase [13]. The penetration of proteins by quenchers has been
known for some time [14,15]. For melittin tetramer the penetration by acrylamide is
not unexpected since acrylamide is neutral and the tryptophans are located in a
loosely packed non-polar region of the protein [ll].
2.4. Fluorescence energy transfer

Another process which can occur during the excited state is fluorescence energy
transfer, which is the transfer of the excited state energy from a donor (D) to an
acceptor (A) (Fig. 4). The transfer is called radiation-less because it occurs without
the appearance of a photon. This process is strongly dependent upon distance
because it is the result of dipole-dipole coupling between the donor and the
acceptor [16]. A requirement for energy transfer is that the emission spectrum of the

donor overlaps with the absorption spectrum of the acceptor. The rate of transfer
( kT) is gven by
(4)
where R , is the distance at which 50% of the energy is transferred, and r is the
donor-to-acceptor distance. The value of R o can be calcuated from the spectral
properties of donor and acceptor [6,16]. The efficiency of energy transfer is given by
the ratio of the rates of transfer to the total rate of depopulation of the donor.
Hence,


10

Usually, both the transfer efficiency ( E ) and R , are determined experimentally.
Then, the donor-to-acceptor distance is calculated using

This method is widely used to measure the distance between sites on a macromolecule, and has been the subject of considerable experimentation and discussion
[17- 191.
Energy transfer has been used to measure the self-association of melittin. The
melittin was labeled with a N-methylanthraniloyl (NMA) residue on one of the
lysine residues. This fluorophore serves as the energy acceptor for the single
tryptophan residue. Only a small fraction (5%) of the melittin monomers was
labeled with NMA. In the monomer there is only one tryptophan residue near the
acceptor, whereas four such residues are present in the tetramer. Hence, the extent
of tryptophan to NMA energy transfer should be sensitive to and increased by
melittin self-association. In this experiment the intention is not to determine a
distance, but rather to use the association-dependent energy transfer to determine
the extent of self-association [20].

300

350

400

450

500

WAVELENGTH (nanometers)

Fig. 9. Emission spectra of N-methylanthraniloyl-labeled melittin. Spectra are shown for the monomer (0
M NaC1) and for the tetramer (2 M NaCI). From [20]. The broken lines are the emission spectra of the
unlabeled melittin.


11

250

300

350

400

WAVELENGTH (nanometers 1

Fig. 10. Excitation spectra of NMA-labeled melittin. From [20].

Emission spectra of the labeled melittin are shown in Fig. 9. Recall that only a

small fraction of the melittin contains a NMA label. Hence, the emission spectra are
mostly characteristic of tryptophan, with shoulders at 430 nm due to the NMA
emission. In the presence of 2 M NaCl the NMA emission is enhanced, reflecting
increased energy transfer from the additional tryptophan residues.
Excitation spectra are often used to study energy transfer. This is because energy
transfer can be detected by enhanced emission from the acceptor when the excitation is centered at the donor absorption. The effects of melittin self-association are
evident from the excitation spectra (Fig. 10). For these spectra the emission
monochromator is centered on the NMA emission (430 nm) and the intensity
recorded as the excitation monochromator is scanned through the absorption bands
of the NMA label (350 nm) and the tryptophan absorption (280 nm). Increasing salt
concentrations result in increased intensity of the tryptophan excitation band (280
nm). This increase in energy transfer is due to the close proximity of the three
additional donors to the NMA acceptor.

2.5. Fluorescence anisotropy
The timescale of fluorescence emission is comparable to that of rotational diffusion
of proteins and the timescale of segmental motions of protein domains or individual
amino acid residues. The polarization or anisotropy of the emission provides a
measure of these processes. Suppose a sample is excited with vertically polarized
light (Fig. ll),and that the sample is viscous so that the fluorophores do not rotate
during the lifetime of the excited state. Then the emission is polarized, usually also
in the vertical direction. This polarization occurs because the polarized excitation
selectively excites those fluorophores in the isotropic solution whose absorption


12

LIGHT

SOURCE-

m+
f/

DETEC~OR
Fig. 11. Measurement of fluorescence anisotropies.

moments are aligned vertically. The extent of polarization is most conveniently
defined by the anisotropy [6,40].

where I refers to the intensities, and the subscripts indicate the parallel (11) or
component.
perpendicular (I)
A number of processes can result in the loss of anisotropy, the most common
being rotational diffusion. Melittin is expected to have rotational correlation times
near 2 and 8 nsec in the monomeric and tetrameric states, respectively. The effect of
rotational diffusion on the anisotropy is described by the Perrin equation,

where r, is the anisotropy in the absence of rotational diffusion, r is the anisotropy,
is the lifetime and 6 is the correlation time. The value of r, is usually measured in
a separate low-temperature experiment. Its value depends upon the excitation
wavelength, and is typically in the range of 0.1 to 0.4. The r, value is a measure of
the angle between the absorption and emission transition moments of the fluorophore. For tryptophan the value of ro on the long wavelength side of the absorption
is near 0.32. Values of r, which are less than 0.1 are usually not useful because the
difference between 11,and I* will be small, and the precision of the measurements
will be decreased.
When 7, and 8 are of similar magnitude then the measured anisotropy is
dependent upon the correlation time. Self-association of melittin is expected to
increase its correlation time about four-fold. Since the lifetime of melittin fluoresence is near 3 nsec we expect self-association to have a dramatic effect on the
7,