IIHIHHIIIHIll!HIHHllllllIIIIHIIIHIllllHlllHlllllJiHl
3 1293 00561 3678

 

LIERARY
Michigan State
University

 

 

 

This is to certify that the

dissertation entitled
DYNAMIC MOTIONS OF CALMODULIN IN RESPONSE TO
ALUMINUM BINDING: AN IMPLICATION FOR ALUMINUM

TOXIC ITY
presented by

SH IXING YUAN

has been accepted towards fulfillment
of the requirements for

Jh n degree in W

 

‘/

U4 HIM,»

' VNlajor/rofessor
4/3

Date 02/14/1989

MS U is an Affirmative Action/Equal Opportunity Institulion 0-12771

—_.__—_ _ _ _ _ ,__ ______ _,__

 

 

 

MSU RETURNING MATERIALS:
Piace in book drop to
LIBRARIES remove this checkout from
—_—-. your record. FINES win

 

 

be charged if book is
returned after the date
stamped beiow.

 

 

 

DYNAMIC MOTIONS OF CALMODULIN
IN RESPONSE TO ALUMINUM BINDING:

AN IMPLICATION FOR ALUMINUM TOXICITY

BY

Shixing Yuan

A DISSERTATION
Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of
DOCTOR OF PHILOSOPHY
Department of Microbiology and Public Health

1989

C‘
.2

I
I?

a} t
I

i

1.,

5979

ABSTRACT

DYNAMIC MOTIONS OF CALMODULIN
IN RESPONSE TO ALUMINUM BINDING:

AN IMPLICATION FOR ALUMINUM TOXICITY

BY

Shixing Yuan

The dynamic motions of spinach calmodulin in the nanosecond range
can be studied through a strategically anchored probe, e.g., a
fluorescence probe, bimane, or, a EPR probe, proxyl, attached at
cysteinyl-26 residue. With Ca2+ ions present, fluorescence studies
indicated that the rotational motions of calmodulin is faster relative
to those of apocalmodulin. This may result from compactness of
calmodulin conformation in the presence of calcium, which implies that
When Ca2+ ions bind to calmodulin, local conformational changes can
trigger a largescale molecular conformational rearrangement. This
explanation also matches the studies in binding of mastoparan to
calmodulin molecule. Fluorescence studies showed when mastoparan binds
to calomdulin, the whole mastoparan-calmodulin complex rotates slower,
i.e., the rotational correlation time becomes longer because the total
volumn of the complex becomes larger than calmodulin molecule itself.

On the other hand, mastoparan binding also triggers largescale

conformational rearrangement on calmodulin molecule, which can be
monitored through the microenvironmental changes around the EPR probe.
When aluminum ions were added to calmodulin solution, in the presence
of saturation of Ca2+, motional changes of calmodulin different from
those in the absence of aluminum ions were observed. Fluorescence
energy transfer studies showed that the energy transfer efficiency from
the single tryptophanyl residue on mastoparan-X to the probe attached
at the cysteinyl site is different in the presence of aluminum from
that in the absence of aluminum. This suggests that aluminum ions may
interfere the interactions between calmodulin and mastoparan. Also, the
Y-function studies showed the temperature dependent interchange of two
substates of calmodulin molecule is totally different in the presence
of aluminum from that in the absence of aluminum. With aluminum prsent,
the interchange needs (or releases) more energy. This means that the
conformation of calmodulin becomes less flexible in reponse to
temperature changes in the environment in the presence of aluminum. We
concluded that these conformational changes triggered by aluminum ions

may in part constitute aluminum toxicity mechanisms in the eukaryotic

world.

Dedicated to my dear brother, Shizhong Yuan.

iv

ACKNOWLEDGMENTS

I would like to thank the members of my graduate study guidance
committee Dr. Loren Snyder, Dr. Robert Bandurski, Dr. Robert Hausinger
Dr. James Tiedje, Dr. Ronald Patterson, and Dr. Alfred Haug for their
advice, help and guidance. I owe special thanks to Dr. Alfred Haug who
continuously offered me special guidance, freedom and financial support
without which this dissertation would never have been possible.

I also wish to acknowledge Dr. Christopher Weis and Biao Shi for

their friendship and assistance.

TABLE OF CONTENTS

LIST OF FIGURES .

LIST OF TABLES

CHAPTER

I.

i.
ii.
iii.
iv.
v.
vi.

LITERATURE REVIEW.

Introduction .

General Properties of Aluminum Ions.
Aluminum Interaction with Proteins .
Characteristics of Calmodulin.
Aluminum Effects on Calmodulin .
List of References .

II. LIGAND-TRIGGERED CONFORMATIONAL PERTURBATIONS

i.
ii.
iii.
iv.
v.
vi.

ELICIT CHANGES AT THE SINGLE CYSTEINYL RESIDUE
SPINACH CALMODULIN . . . . . . . . . .

Abstract .
Introduction .
Materials and Methods.
Results.

Discussion . . .
List of References .

III. DYNAMIC PROPERTIES OF CALMODULIN IN RESPONSE
OF ALUMINUM. . . . . . . . .

i.
ii.
iii.
iv.
v.
vi.

Abstract .
Introduction .
Materials and Methods.
Results.

Discussion . . .
List of References .

IV. FRICTIONAL RESISTANCE T0 MOTIONS OF BIMANE-
LABELLED SPINACH CALMODULIN IN RESPONSE TO
LIGAND BINDING . . . . . . . . . .

i.
ii.
iii.
iv.
v.
vi.

Abstract .
Introduction .
Materials and Methods.
Results. . .
Discussion .

List of References . . .
vi

Page

viii

H

\DO\O\O\9N

h‘h‘

25

26
27
29
Al
52
S7

60

61
62
65
69
79
81

8h

85
86
87
90
95
98

 

 

 

 

V. SUMMARY.

APPENDIX.

THE OCCURRENCE AND CORRELATION BETWEEN
CALMODULIN IN RAINBOW TROUT MUCUS AND
LEVELS OF pH, Ca AND A1 . .

. Abstract .
ii.
iii.
iv.
. Discussion .
vi.
vii.

Introduction . . .
Materials and Methods.
Results.

Conclusions. .
List of References .

vii

100

107

108
109
110
114
117
120
121

Figures
1.1.

1.2.

——1-IIIIIIIIIII-IIIIIIIEEEEZ;;;___EL —

LIST OF FIGURES

Amino acid sequence of calmodulin.

2+

EF-hand structure of Ca -related

superfamily molecules.

Stereo drawing of the alpha-C of calmodulin.

Calmodulin-related regulatory pathways .
A standard first derivative EPR spectrum .

Structures and lengths, D, of maleimide spin
labels used in EPR experiments .

An illustration of T- and L-format method for
measuring fluorescence anisotropy.

First derivative EPR spectra of spin- labelled
apocalmodulin. . . . . . . . . . . .

Dependence of the rotational correlation time,
, of spin-labelled spinach calmodulin on
spin label length, D . . . . . . . . .

Perrin plot of spinach calmodulin whose single
cysteinyl residue 26 was labelled with a
fluorescence probe, monobromotrimethyl-
ammoniobimane.

The spectral overlap of the fluorescence emission
of the single tryptophanyl residue in mastoparan-
X, associated with spinach calmodulin, and the
absorption of AEDANS- -conjugated cysteine .

Fluorescence anisotropy values of AEDANS-labelled

spinach calmodulin versus temperature, in the
presence and in the absence of aluminum ions .

Y(T) vs. temperature for bimane-labelled
spinach calmodulin . . .

An illustration of microenvironmental changes
taking place in the vicinity around the EPR
probe attached at Cys-26 . . . . . . .

An illustration of the whole molecular
conformational changes of spinach calmodulin
viii

Page

10
12
14

33

35

38

42

44

50

72

77

92

102

in response to ligand binding. . . . . . . . . . . . 104

ix

Table

2.1.

2.2.

3.1,

3.2.

3.3.

4.1.

A.1.

A.2.

LIST OF TABLES

Fluorescence lifetimes, 1, 2, and associated
preexponential factors, a and a2, of bimane-
1abe11ed spinach calmodulin.

Molecular parameters for bimane-labelled spinach
calmodulin .

Fluorescence properties of AEDANS-labelled spinach
calmodulin at various cation concentrations.

Energy transfer characteristics of AEDANS-labelled
spinach calmodulin in the absence and in the
presence of aluminum ions.

Molecular parameters of AEDANS-labelled spinach
calmodulin . . . . . .

Physical parameters for spinach calmodulin in
the presence of ligands.

Results of CaM [1251]-radioimmunoassay performed
on mucus collected from rainbow trout held a various
pH, Ca, and Al levels for 147 days .

Results of ANOVA procedure to test for relationship
between Ca, pH and the amount of CaM in fish mucus .

Page

48

49

71

75

76

94

115

116

Chapter I

LITERATURE REVIEW

Shixing Yuan

Introduction:

Aluminum is the most abundant metal in the crust of the earth.
Until recently, aluminum was believed to have no biological function.
Recent studies indicate, however, that A13+ can activate G proteins
which are involved in signal transduction across the plasma membrane
(1). Furthermore, an increasing numberof toxic effects has been shown
to be associated with aluminum ions. Aluminum becomes toxic when it is
mobilized under acidic conditions and enters into the biosphere (2, 3,
a, 5). This mobilization was suggested happening at the root-soil
interface (6). Since around 40% of the arable land of the world and
perhaps up to 70% of potentially usable land are acidic and the plants
in those areas are subject to aluminum toxicity, this toxicity
represents a formidable barrier to successful food and biomass
production around the world (7). Plants grown on acidic subsoil are
highly susceptible to drought and have less access to subsoil nutrients
since aluminum.can limit the plant root development and penetration
into subsoils and thus increasing the drought susceptibility of the
plants and limiting their access to mineral nutrients (8). The mobile
aluminum ions exert their toxic effects on the fine roots of afflicted
plants and result in significant losses in root structure and the
eventual dieback of the aerial portions of these plants (9). Also when
higher plants were exposed to A1C13 at 800 or 1200 pmol/l for a long
time the contents of both Mg and Ca in roots and needles of the
'seedlings could be reduced (10). Some aluminum-tolerant plants, e.g.,

aluminumetolerant I;1§1ggn_§gggixyn, can overcome this toxic effect by

preventing aluminum ions from entering symplasm or by some internal
detoxification mechanisms (6). When red spruce are exposed to aluminum
ions and simulated acid precipitation, a symptom of calcium deficiency
will take place (11), resulting in the cessation of plant growth. It is
believed that aluminum can inhibit calcium uptake and translocation
(12). In addition, the deposition of acid rain in the northeast U.S.
and Canada has led to the transport of labile monomeric aluminum
(compared with the complexes existing at basic pH) ions in the soil and
enhanced the concentration of this metal ion in lakes and surface
waters (l3, 14). This implied that under acidified conditions, aluminum
toxicity could be the primary cause of death in animals such as
teleosts (15, 16, 17).

Until recently, aluminum toxic effects upon human body has not been
described. Now, it is commonly accepted that some dialysis diseases,
e.g., dialysis encephalopathy and dialysis osteomalacia, are caused by
aluminum ions (18, 19, 20). Also, accumulation of aluminum has been
related to cellular aging and death (21). Some other diseases like
senile dementia of the Alzheimer's type (22, 23), amyotrophic lateral
sclerosis (24), and idiopathic Parkinsonism (21) are suggested to be
related to aluminum toxicity.

Since the acidification of the environment and its implied aluminum
toxicity problem are drawing more and more attention, it will be
important for us to get a better understanding of the mechanisms of how

2+-related

aluminum affects living systems, especially the Ca
regulatory processes in vivo. Since calmodulin is a key factor in these

processes, and a possible target of aluminum toxicity (3, 21, 25.). It

 

is my aim in this dissertation to study the biophysical aspects of
aluminum toxicity at the molecular level, i.e., aluminum ions affecting

conformation of calmodulin molecule and the protein's interaction with

the targets.
General Properties of Aluminum Ions

Estimates indicate that the average concentration of aluminum in
the crust of the earth is about 81,000 ppm, and the total amount of
aluminum accounts for about 7.5% the weight of the crust (26). In
nature, aluminum exists only in the oxidation state (III). Since the
ion is in a strong charged state (usually as Al3+), the ionic radius
for aluminum is small, only 0.51 A0. With this high ratio of charges
over radius (58.8 for Al3+, 30.3 for Mg2+ and 20.2 for Ca2+), the
nonhydrated aluminum ion has a strong polarizing effect on adjacent
atoms. At highly acidic pH, the aluminum ion is trivalent and
hexahydrated, A1(H20)63+, forming an octahedral configuration
with water molecules, while at neutral pH, the formation of
hydroxo-complex polymers, e.g., A104A112(H20)12(0H)247+,
will be promoted as hydrolysis happens. Acidification leads to a
decrease of polymers while increasing the concentrations of free
single, double and triple charged monomers, A1(H20)63+,
A1(H20)5(0H)2+ and A1(H20)4(OH)2+ (27). In complexes,

A13+ is generally bound to oxygen. Consequently, acidic oxygens of
membrane lipids, carboxyl groups of proteins and chromatin material,

e.g., phosphate groups of DNA, make ideal target ligands for cationic

aluminum. Aluminum can also bind nitrogen, but with lower affinity. The
ligand exchange (e.g., change from Al-O to Al-N) is slow partly due to
a fairly marked covalent contribution to the Al-O and Al-N bonds.
Slowness of the ligand exchange is also due to other things, e.g., the
exchange generally occurs by a dissociative mechanism (28), and the
complexes are often chelates by six ligands. For an aluminum ion, the
coordination number is strictly 6, while the formation of dissociative
intermediate needs the number higher than 6. This makes the formation
unlikely. The slowness explains why the toxicity and other effects of
Al3+ are easily underestimated in short-term experiments.

Aluminum ion is the major component of a large number of minerals.
At neutral pH, the solubility of aluminum minerals is extremely low,
but increases as pH decreases. Acid rain and usage of acidifying
fertilizers acidify the environment and increase the concentration of
soluble A13+ in soil and in lakes and rivers. It is not the low pH
but rather the subsequent increase of dissolved Al3+ in the water
that causes gill damage and death among young fish. Also, the elevated
concentration of A13+ in lakes have caused serious and sometimes
lethal toxic effects on birds living in the immediate surroundings
(29).

Al3+ dissolved in soil is one of the foremost growth limiting
factors in many parts of the world (30). This probably results also
from the ready ability of Al3+ to bind root and to prevent the uptake
of inorganic phosphates. Al3+ seems also to specifically inhibit root
growth of plants (31, 32, 33).

Aluminum content in human body was estimated between 30 - 50 mg

(34, 35). About half of this is in the skeleton and a quarter in the
lungs. Basically, all the aluminum comes from food intake except that
in the lungs. The absorption of A13+ varies when suitable complexing
substances are also present, for example, the presence of phosphate can

prevent A13+ uptake just as Al3+ hydroxide prevents phosphate

uptake (36).
Aluminum Interaction with Proteins

Aluminum has been shown to inhibit hexokinase (37 - 42) which
results from the formation of a complex of ATP-A13+. This complex
acts as a strong competitive inhibitor with respect to ATP-Mg2+. In
this way, Al3+ can inhibit ATP—dependent enzymes, such as, Na+-K+
ATPase (43). Al can also inhibit some other enzymes such as: adenylate
cyclase (44, 45), 3', 5'-cylic nucleotide phosphodiesterase (46) serum
cholinesterase (47), alkaline phophatase (48, 49), acetylcholinesterase
(50, 51), catechol-O-methyl-transferase (52), and ferroxidase (53).

Among the important regulatory proteins interacting with Al3+,
calmodulin is one (46, 54) whose activities can be greatly affected by

aluminum binding. The study of the interaction between calmodulin

molecule and aluminum ions will be the main theme of this thesis.‘

Characteristics of Calmodulin

Calmodulin is a small (Mr - 16,000 - 17,000), acidic protein (pI

- 4.1 - 4.2) existing in all species of the eukaryotic world.

Calmodulin molecule is quite conserved during evolution (Fig. 1.1)
(25). According to reports, calmodulin isolated from the protozoan
Tetrahymena has eleven substitutions and one deletion when compared to
bovine brain calmodulin (55). This homology demonstrates that all the
calmodulins have similar conformations and functions. So far, all
calmodulin molecules are involved in regulating calcium related
processes in vivo. Like all members of troponin C superfamily, a
calmodulin molecule has the characterized helix-loop-helix, the
so-called EF-hand structures (Fig 1.2) for Ca2+ binding (56). Among
its four Ca2+ binding sits, the two near the C-terminal can be

defined as high affinity binding sites (Kd - 10'7M), while the

other two are low affinity ones (Kd - 10’6M) (57). The whole

molecular structure of calmodulin is like a dumbbell, two lobes
connected with an 8-turn alpha-helix rod (58, 59) (Fig. 1.3). Each lobe
contains two Ca2+ binding sites (58, 59). In solution, a calmodulin
molecule is not in the rigid dumbbell conformation, instead, the rod
connecting these two lobes is fairly flexible (60).

In the absence of calcium binding to calmodulin, the molecule is in
the inactive state. But when calcium binds, calmodulin becomes
activated and can interact with numerous regulatory enzymes, which, in
turn, can regulate different processes in the cell (61) (Fig. 1.4). The

activated form of calmodulin can either be Ca2+

2-calmodulin or
Ca2+3_4-ca1modulin (62, 63, 64, 65). Calmodulin exhibits
hydrophobic interactions with its target enzymes or peptides.
Amphiphilicity plays an important role in this interaction. When

calmodulin binds Ca2+, more hydrophobic surface is exposed. This

Fig. 1.1. Amino acid sequence of calmodulin. The basic sequence shown
here is that from bovine brain. Wherever there is a substitution in any

different species, it is shown by an arrow. Otherwise, the sequence is

the same.

e.e~e.e.c . i! . -3 I
2......
cae~o.u~& wow oo.mv —¢¢.L e

. a . . co o...ae..eae.@ 9.2.2.. . a
O I. O sctc (6 8.0-0 0-
Ses-.32.. a .1 9° Gaeoeeece an. 2.2.0 egmwmvo 2.2..
sneeze... % ouou QGOQOQGOOQGQEGOOQAU. I five“ coo-e2...—
eee~e.e. @ a. U 6
L ® @0990 :_ >- 09. Co
””HHo.e.a@TIOQG a 052.04 4 Ma. Bananas...
2e... are: 2.22:. i 2.3T. x e @225
no ee..e;— > ec..u..>g.2a...p a % @o.e.ae:ese.
9 We 2....) > 2:23;: . e @He-ce.-ao.e.a
Gee act—oo—gb h 06.92:: .- AQUA .co.
0 2...... o 2...»... a ..oo 0 ..
e oe.e_o.< .- ee.c-.-.acess n 0 0:30:02:
0 2.2.236 a 22.23:. m
.. 2...... . 22.2.2. .-
GO 2......2..< 2 2.33:. o
GGGQWO 2.2.2.2.. 2 2.22 < 0000060... . O
.3232.— @.\O.. .. . . . .emwu “
eae~e.e.a@ w «00 one“. 0 m
eeoeoeeeeeeoeeeooeoeeeeoeo el© .o
G G 0' a 0 0:32.02: C
@296 @000 2.2.
00 o
:0

. is.

10

2+-re1ated superfamily molecules. 0

Fig. 1.2. EF-hand structure of Ca
stands for a liganded Ca2+. Each Ca2+ here is liganded by six
oxygen atoms. An EF-hand is characterized by a stretch of helix-loop-

helix structure.

12

Fig. 1.3. Stereo drawing of the alpha-C of calmodulin. C)stands for the

liganded Ca2+.

14

Fig. 1.4. Calmodulin-related regulatory pathways. * stands for the
activated form of calmodulin or an specific enzyme. Here, the enzyme is
any of the calmodulin-dependent enzymes, e.g., adenylate cyclase,
guanylate cyclase, phosphodiesterase, phospholipase A2, NAD kinase,
phosphorylase kinase, myosin light chain kinase and glycogen synthase
kinase. The calmodulin-dependent processes include cellular motility,
cytoskeletal function, glycogen metabolism, cyclic nucleotide

metabolism, calcium metabolism and photosythesis.

15

calmodulin nCa2+

Ca2+-calmodulin

n 1
2+ 0 *
Can -calmodulln . ,enzyme

2+ .* *
Can -ca1modulln -enzyme

(ENHANCED ACTIVITY)

1

BIOCHEMICAL REACTIONS

CELLULAR RESPONSES

16

exposed surface could then be essential for the target binding. Several
neural toxic peptides, such as mellitin and mastoparan, are able to

inhibit calmodulin activities by specifically interacting with this

exposed surface (66).
Aluminum Effects on Calmodulin

Aluminum has already been shown to have effects on calcium
regulation. The accumulation of aluminum in plants can inhibit calcium
transportation across the membrane and induce calcium deficiency

syndrome. Increase of calcium concentration can alleviate or eliminate

these effects. Since calmodulin is a key factor in the Ca2+-related

regulatory processes in vivo, the complex formed between calmodulin and
aluminum ions may be a key lesion in the syndrome of aluminum toxicity
(46).

The application of aluminum ions to calmodulin causes pronounced
structural changes in the latter molecule (46). The metal binds
stoichiometrically and cooperatively to calmodulin. Binding of aluminum
to calmodulin at a molar ratio of 2:1 leads to a reduction in the alpha
helix of the protein as monitored by ultra-violet circular dichroism
(46), and an increase in the hydrophobic surface area as indicated by
the fluorescent hydrophobic surface probe 1-anilino-8-naphthalene-
sulfonic acid (ANS) (46). These changes could result in a spatial
geometry rearrangement between calmodulin and its peptide ligand,
mastoparan. Also, binding of aluminum to calmodulin results in a

concomitant decrease in the ability of calmodulin to activate

17

3', 5'-cyclic nucleotide phosphodiesterase (46). The presence or
absence of calcium does not interfere with aluminum binding (46, 67).
Taken together, aluminum binding to calmodulin introduces a
perturbation into the conformation of calmodulin and this perturbation,
in turn, may signal a toxicological effect of this metal. Partial
reversal of the toxic effects of aluminum ions on calmodulin by citrate
has been documented (68). Citrate can prevent complete destruction of
the regulatory role of calmodulin in vitro and this organic acid has
been proposed as a natural protectant against the deleterious effects
of aluminum in maize (69). Estimates from spectroscopic data indicated
that the binding affinity for the first mole of aluminum ions bound to
the protein is about one order of magnitude stronger than that of
calcium.

Aluminum appears to interact directly with calmodulin (46). At a
molar ratio of 3:1 of [Al]:[ca1modulin], the calmodulin-stimulated
membrane-bound ATPase activity in barley root plasma membrane vesicles
was severely inhibited. The metal-induced changes in calmodulin
structure were reflected in a reduced formation of a transmembrane
potential when assayed with a voltage-sensitive fluorescent probe (70).

Application of aluminum to calmodulin may induce breakage of
calmodulin's hydrogen bonds which are associated with enthalpic effects
such as changes in steric and electrostatic repulsions and can results
in dramatic conformational changes (71).

The observed changes in aluminum-induced hydrophobic surface may be
accompanied by changes in calmodulin salvation structure since

solvation effects contribute significantly to protein stability (72,

18

73). Al-calmodulin appears to be different in protein flexibility and
in its ability to interact with target proteins, compared with
corresponding characteristics of the calcium-calmodulin structure.

These changes in calmodulin may explain why the aluminum-calmodulin

complex last part of its regulatory character.

10.

11.

12.

13.

14.

19

LIST OF REFERENCES

Gilman, A.G., (1987) G proteins: transducers of receptor-generated
signals. Ann. Rev. Biochem. 56: 615-649.

Haug, A., (1984) Molecular aspects of aluminum toxicity. CRC Crit.
Rev. Plant Sci., 1(4): 345-373.

Baes, C.F., and Mesmer, R.E., (1976) lbs flynxoiysis of Qntions,
John Wiley and Sons, New York.

Burgess, J., (1978) Estsl_lsn§_in.§212tisa. John Wiley and Sons,
New York.

Fratiello, A., (1972) Nuclear magnetic resonance cation salvation
studies, in: Inogganic Reaction Mgcnnnisms Part II, (Edwards, J.O.
ed.), Interscience, New York.

Taylor, G.J., (1988) The physiology of aluminum tolerance in

higher plants. Commun. in Soil Sci. Plant Anal., 19(7-12):
1179-1194.

Osmond, C.B., Bjorkman, 0., and Anderson, D.J., (1980) in:

Physiological Prncesneg in Einnt Ecgingy, Springer-Verlag, Berlin.
pp. 41-66.

Foy, C.D., Chaney, R.L., and White, M.C., (1978) The physiology of
metal toxicity in plants. Ann. Rev. Plant Physiol. 29: 511-566.

Ulrich, B., Mayer, R., and Khanna, P.D., (1980) Chemical changes

due to acid precipitation in loess-derived soil in central EurOpe.
Soil Sci. 130: 193-199.

Godblod, D.L., Fritz, E., and Huttermann, A., (1988) Aluminum

toxicity and forest decline. Proc. Natl. Acad. Sci. (U.S.) 85:
3888-3892.

Shortle, W.C., and Smith, K.T., (1988) Aluminum-induced calcium

deficiency syndrome in declining red spruce. Science 240:
1017-1018.

Ownby, J.D., and Dees, L., (1985) Growth and mineral nutrient

status in peanut and sorghum in response to acid precipitation and
aluminum. New Phytol. 101: 325-332.

Driscoll, C.T., and Newton, R.M., (1985) Chemical characteristics
of Adirondack lakes - Individual lakes respond differently to acid
deposition. Environ. Sci. Technol. 19(11): 1018-1024.

David, M.D., and Driscoll, C.T. (1984) Aluminum speciation and
equilibria in soil solutions of a haplorthod in the Adirondack

15;

16.

17.

18.

19.

20.

21.

22.

23.

24.

25.

26.

20

Mountains (New York, U.S.A.). Geoderma 33: 297-318.

Cronan, 0.8., and Schofield, C.L., (1979) Aluminum leaching
response to acid precipitation: effects on high elevation
watersheds in northeast. Science 204: 304-306.

Brown, D.J.A., (1983) Effect of calcium and aluminum

concentrations on the survival of Brown Trout at low pH. Bull.
Environ. Contam. Tocicol. 30: 582-587.

Alfrey, A.G., LeGendre, G.R., and Daehney, W.D., (1976) The

dialysis encephalopathy syndrome. Possible aluminum intoxication.
N. Eng. J. Med. 294: 184-188.

Ellis, H.A., McCarthy, J.H., and Herrington, J., (1979) Bone
aluminum in haemodialysed patients and in rats infected with

aluminum chloride: relationship to impaired bone mineralization.
J. Clin. Pathol. 32: 832-844.

Driscoll, C.T., Jr., Bader, J.P., Biosogni, J.J., Jr., and

Schofield, C.L., (1980) Effect of aluminum speciation on fish in
dilute acidified waters. Nature 284: 161-164.

Burnatowska-Hledin, M., Kaiser, L., and Mayer, G.H., (1983)

Aluminum, parathyroid hormone, and osteomalacia. in: Special
Topics in Endocrinology and Metabolism, 2. 201-226.

Ganrot, P.O., (1986) Metabolism and possible health effects of
aluminum. Environ. Health Perspect. 65: 363-441.

Crapper, D.R., Karlik, S., and DeBoni, U., (1978) Aluminum and
other metals in senile (Alzheimer) dementia. in: Alzheimer's
Disease; Senile Dementia and Related Disorders, (Aging, Vol. 7)

(Katzman, R., Terry, R.D., and K.L. Bick, eds.), Raven Press, New
York. pp. 471-495.

Trapp, G.A., Miner, C.D., Zimmerman, R.L., Mastri, A.R., and

Heston, L.L., (1978) Aluminum levels in brain in Alzheimer's
disease. Biol. Psychiat. 13: 709-718.

Garruto, R.M., Fukatsu, R., Yanagihara, R.C., Gajdusek, D.C.,

Hook, G., and Fiori, C.B., (1984) Imaging of calcium and aluminum
in neurofibrillary tangles-bearing neurons in Parkinsonism-
dementia of Guam. Proc. Natl. Acad. Sci. (U.S.) 81: 1875-1879.

Klee, C.B., and Vanaman, T.G., (1982) Calmodulin. Adv. Protein
Chem. 35: 213-321.

Lepp, N.W., (1981) Chapter 2, Other trace metals, aluminum. in:
We.

ec o o 0 Vol. 1., Applied
Science, London. pp. 214-218.

27.

28.

29.

30.

31.

32.

33.

34.

35.

36.

37.

38.

39.

40.

21

Smith, R.W., (1971) Relations among equilibrium and nonequilibrium
aqueous species of aluminum hydroxocomplexes. in: Noneguilibrium

tems a w t e (Adv. Chem. Ser. 106) American
Chemical Society, Washington D.C., pp. 250-279.

Burgess, J., (1978) Chapter 11, Kinetics and mechanisms: solvent

exchange. in: Mn£n1_lnn§_in_§nin§innn Ellis Horwood Ltd.,‘
Chichester., pp. 310-348.

Nyholm, N.E.I., (1981) Evidence of involvement of aluminum in
causation of defective formation of eggshells and of impaired
breeding in wild passerine birds. Environ. Res., 26: 363-371.

Foy, C.D., (1974) Effects of aluminum on plant growth. in: The
giant Rgot and itg Envinonment, (Carson, E.W., ed.), Univ. Press

of Virginia, Charlottesville., pp. 601-642.

Morimura, S., Takahashi, E., and Matsumoto, H., (1978) Association
of aluminum with nuclei and inhibition of cell division in onion
(nilium cnpa) roots. Z. Pflanzenphysiol. 88: 295-401.

Clarkson, D.T., (1965) The effect of aluminium and some other

trivalent.metal cations on cell division in the root apices of
nilium cena. Ann. Bot. N. S., 29: 309-315.

Sampson, M., Clarkson, D., and Davies, D.D., (1965) DNA-synthesis
in aluminum treated roots of barley. Science 148: 1476-1477.

Tipton, I.H., and Cook, M.J., (1963) Trace elements in human

tissues. Part II. Adult subjects from the United States., Health
Phys. 9: 103-145.

Alfrey, A.G., (1984) Aluminum intoxication. N. Eng. J. Med. 310:
1113-1114.

Spencer,.H., and Lender, M., (1979) Adverse effects of aluminum-

containing antacids on mineral metabolism. Gastroenterology 76:
603-606.

Trapp, G.A., (1980) Studies of aluminum interaction with enzymes

and proteins -- the inhibition of hexokinase. Neurotoxicology 1:
89-100.

Harrison, W.H., Codd, E., and Gray, R.M., (1972) Aluminium
inhibition of hexokinase. The Lancet 2: 277.

Womack, F.C, and Colowick, S.P., (1979) Proton-dependent
inhibition of yeast and brain hexokinase by aluminum in ATP
preparations. Proc. Natl. Acad. Sci. (U.S.) 76: 5080-5084.

Viola, R.E., Morrison, J.P., and Cleland, W.W., (1980) Interaction

41.

42.

43.

44.

45.

46.

46a.

47.

48.

49.

50.

51.

52.

22

of metal (III)-adenosine-5'triphosphate complexes with yeast
hexokinase. Biochemistry 19: 3131-3137.

Neet, R.E., Furman, T.G., and Hueston, W.J., (1982) Activation of
yeast hexokinase by chelators and the enzymic slow transition due

to metal-nucleotide interactions. Arch. Biochem. Biophy. 213:
14-25.

Lai, J.C.K., and Blass, J.P., (1984) Inhibition of brain
glycolysis by aluminum. J. Neurochemistry 42: 438-446.

Staurnes, M., Sigholt, T., and Reite, O.B., (1984) Reduced
carbonic anhydrase and Na-K-ATPase activity in gills of salmonids

exposed to aluminium-containing acid water. Experientia 40:
226-227.

Sternweis, P.C., and Gilman, A.G., (1982) Aluminum, a requirement

for activation of the regulatory component of adenylate cyclase by
fluoride. Proc. Natl. Acad. Sci. (U.S.) 79: 4888-4891.

Mansour, J.M., Ehrlich, A., and Mansour, T.E., (1983) The dual
effect of aluminum as activator and inhibitor of adenylate cyclase

in the liver fluke Eascioin nepatica. Biochem. Biophys. Res.
Commun. 112: 911-918.

Siegel, N., and Haug, A., (1983) Aluminum interaction with
calmodulin. Evidence for altered structure and function from
optical and enzymatic studies. Biochim. Biophys. Acta 744: 36-45.

Siegel, N., and Haug, A., (1983) Inorgan, Chim. Acta 79: 230-231.

Marquis, J.K., (1983) Aluminum inhibition of human serum
cholinesterase. Bull. Environ. Contam. Toxicol. 31: 164-169.

Rej, R., and Bretaudiere, J.-P., (1980) Effects of metal ions on

the measurement of alkaline phosphatase activity. Clin. Chem. 26:
423-428.

Bamberger, C.E., Botbol, J., and Cabrini, R.L., (1968) Inhibition

of alkaline phosphatase by beryllium and aluminum. Arch. Biochem.
Biophys. 123: 195-200.

Patocka, J., (1971) The influence of Al3+ on cholinesterase and
acetylcholinesterase activity. Acta Biol. Med. Ger. 26: 845-846.

Marquis, J.M., and Lerrick, A.J., (1982) Noncompetitive inhibition

by aluminum, scandium and yttrium of acetylcholinesterase from
Electiopnorng eieggzigng. Biochem. Pharmacol. 31: 1437-1440.

Mason, L., and Weinkove, G., (1983) Radioenzymatic assay of
catecholamines: reversal of aluminium inhibition of enzymatic

 

 

53.

54.

55.

56.

57.

58.

59.

60.

61.

62.

63.

23

O-methylation by desferrioxamine. Ann. Clin. Biochem. 20: 105-111.

Huber, C.T., and Frieden, E., (1970) The inhibition of ferroxidase
by trivalent and other metal ions. J. Biol. Chem. 245: 3979-3984.

Siegel, N., Coughlin, R., and Haug, A., (1983) A thermodynamic and
electron paramagnetic resonance study of structural changes in
calmodulin induced by aluminum binding. Biochem. Biophys. Res.
Commun. 115: 512-517.

Yazawa, M., Yagi, R., Toda, H., Kondo, R., Narita, R., Yamazake,
R., Sobue, R., Kakiuchi, S., Nagao, S., and Nozawa, Y., (1981) The
amino acid sequence of the Ietranynena calmodulin which
specifically interacts with guanylate cyclase. Biochem. BiOphys.
Res. Commun. 99: 1051-1057.

Moews, P.C., and Kretsinger, R.H., (1975) Refinement of the
structure of carp muscle calcium-binding parvalbumin by model
building and difference Fourier analysis. J. Mol. Biol. 91:

201-228.

Haiech, J., Klee, C.B., and Demaille, J.G., (1981) Effects of
cations on affinity of calmodulin for calcium: ordered binding of
calcium ions allows the specific activation of calmodulin-
stimulated enzymes. Biochemistry 20: 3890-3897.

Babu, Y.S., Sack, J.S., Greenhough, T.J., Bugg, C.B., Means, A.R.,
and Cook, W.J., (1985) Three-dimensional structure of calmodulin.

Nature (London) 315: 37-40.

Seaton, B.A., Head, J.F., Engelman, D.M., and Rocjards. F.M.,
(1985) Calcium-induced increase in the radius of gyration and
maximum dimension of calmodulin measured by small-angle X-ray
scattering. Biochemistry 24: 6740-6743.

Heidorn, D.B., Torney, D., and Trewhella, J., (1987) Modelling
studies of X-ray solution scattering data for calmodulin: a
comparison of the X-ray crystal and solution structures. BiOphys.

J. 51: 453a.

Cormier, M.F., (1983) Calmodulin: the regulation of cellular

function. in: Cnlninn in Biology. (Spiro T.G. ed.), John Wiley and
Sons, New York, pp. 53-106.

Crouch, T.H., Holroyde, M.J., Collins, J.H., Solaro, R.J., and
Potter, J.D., (1981) Interaction of calmodulin with skeletal
muscle myosin light chain kinase. Biochemistry 20: 6318-6325.

Johnson, J.D., Holroyde, M.J., Crouch, T.H., Solaro, R.J., and
Potter, J.D., (1981) Fluorescence studies of the interaction
myosin light chain kinase. J. Biol. Chem. 256: 12194-12198.

64.

65.

66.

67.

68.

69.

70.

71.

72.

73.

24

Cox, J.A., Malnoe, A., and Stein, E.A., (1981) Regulation of brain
cyclic nucleotide phosphodiesterase by calmodulin. J. Biol. Chem.
256: 3218-3222

Huang, C.Y., Chau, V., Chock, P.B., Wang, J.H., and Sharma, R.E.,
(1981) Mechanism of activation of cyclic nucleotige phospho-
diesterase: requirement of the binding of four Ca + to
calmodulin for activation. Proc. Natl. Acad. Sci. (U.S.) 78:

871-874.
Anderson, S.R., and Malencik, D.A., (1986) Peptides recognizing

calmodulin. in: gnininn_nn§_§gil_£nng§inn. (Cheung W.Y. ed.),

Academic Press, Inc. Vol 4, pp. 1-42.

Siegel, N., and Haug, A., (1983) Aluminum-induced inhibition of
calmodulin-regulated phosphodiesterase activity: Enzymatic and
optical studies. Inorgan. Chim. Acta 79: 230-231.

Suhayda, C.G., and Haug, A., (1985) Citrate chelation as a
potential mechanism against aluminum toxicity in cells: the role
of calmodulin. Can. J. Biochem. Cell Biol. 63: 1167-1168.

Suhayda, C.G., (1986) Organic acids reduce aluminum toxicity in
maize root membranes. Doctoral thesis. Chapter VI pp. 52-82.

Siegel, N., and Haug, A., (1983) Calmodulin-dependent formation of
membrane potential in barley root plasma membrane vesicles: a
biochemical model of aluminum toxicity in plants. Physiol. Plant.

59: 285-291.

Hartley, F.R., Burgess, C., and Alcock, R.M., (1980) Chapter 14,
The interpretation of stability constant data. in: Soiution

Eguilibria. Halsted Press, Chichester, p. 249.

Conway, B.E., (1981) d d io h sics.

Elsevier, Amsterdam.

Franks, F., and England, D., (1975) The role of solvent
interactions in protein conformation, CRC Crit. Rev. Biochem., 3:

165-219.

 

CHAPTER II

LIGAND-TRIGGERED CONFORMATIONAL PERTURBATIONS ELICIT CHANGES AT THE

SINGLE CYSTEINYL RESIDUE OF SPINACH CALMODULIN

Shixing Yuan and Alfred Haug

Eur. J. Biochem. 175: 119-124.

1988

25

26

Following application of stoichiometric amounts of Ca2+ or

mastoparan, a specific ligand peptide, to spinach calmodulin, dynamic
changes in the nanosecond range could be monitored at a strategically
anchored fluorescence or spin probe. For these studies the single
cysteinyl residue 26 of spinach calmodulin was labelled with a
thiol-specific proxyl (i.e., 2,2,5,5-tetramethyl-l-pyrrolidinyl-oxyl)

2+

spin probe or with a bimane fluorescence probe. With Ca and

mastOparan present, fluorescence studies (anisotropy, lifetime)
indicated that the rotational motion of the mastoparan-calmodulin
complex is slower relative to the motion of calmodulin in the absence
of mastoparan. The probe's attachment site (residue 26) appears to
reside in a fairly polar microenvironment as reported by a series of
proxyl spin probes varying in label length. The rotational correlation
time of the shortest spin probe markedly changed upon binding of
mastoparan to a calmodulin region distant from that of the monitoring
spin probe. We interpret these observations as indicating that
ligand-triggered conformational perturbations elicit specific responses

at the cysteinyl residue 26 of spinach calmodulin.

27

Introduction

2+

Calmodulin is a representative of Ca binding proteins which

2+-dependent biochemical processes (1).

mediate a multitude of Ca
Calmodulin (Mr-15,700 for spinach calmodulin) is an acidic,
heat-stable protein consisting of a single polypeptide chain whose
primary amino acid residue sequence is highly Conserved during
evolution of eukaryotes. The protein has a pronounced tendency to bind
4 Ca2+ in its helix-loop-helix structures known as EF hands (2) (also

Ca2+ binding to calmodulin is sequential since the

see Chapter I).
first two cations apparently bind to the so-called high-affinity
binding sites, III and IV (Rd are about 10'7M.) (1), located in the
C-terminal half of the dumbbell-shaped protein (3,4). The remaining two
Ca2+ bind to the low-affinity binding sites, I and II (Kd are about
10'6M), which reside in the N-terminal lobe of the dumbbell. Upon
Ca2+ binding calmodulin undergoes conformational changes which expose
a hydrophobic region serving as an interface for the interaction
between calmodulin and its target proteins. Certain pharmacological
agents (5) and aluminum ions (6) possess anticalmodulin activity
probably by interfering with this interfacial region.

Calmodulin's multifunctional and regulatory properties are linked to

2+

the protein's ability to specifically coordinate Ca . Questions of

interest are directed towards elucidating mechanisms by which

2+, in the presence or

conformational changes, triggered locally by Ca
absence of a target protein, are globally transmitted within the

regulatory protein. This transmission of information in the protein

28

probably depends in part on structural fluctuations and motions (7,8)
which occur on a time scale as fast as picoseconds. In such a dynamic
system the potential energy surface seems to comprise a multitude of
thermally accessible minima in the neighborhood of the native protein
structure (9). The functional significance of these molecular motions
is illustrated by findings that myoglobin cannot bind oxygen unless the
protein structure fluctuates to permit oxygen passage (10).
Examination of the putative Ca2+-related transfer of information
requires studies of the relative motion of calmodulin's constituent
regions. Therefore we present in this article results concerning
structural changes monitored at a defined locus on calmodulin in
response to the interaction of the same protein with its specific
ligands (Ca2+, mastoparan) which triggered the response in a region
distant from and close to that of the monitoring site. For this purpose
we selected spinach calmodulin which harbors a single cysteinyl residue
at site 26 of the polypeptide chain (11). A thiol-reactive spin probe
or a fluorophore was anchored at this site where ligand-related
structural modifications of the probe‘s microenvironment could be
monitored in terms of national characteristics. We were able to
demonstrate that ligand-triggered conformational perturbations seem to
elicit specific responses at the site of the strategically anchored

probe.

29

Materials and Methods

Chemicals

The spin labels, 3-ma1eimido-proxyl (designated as spin label I),
3-(male-imidomethyl)-proxyl (II), 3-(2-maleimido-ethyl-carbamoy1)-
proxyl (III), 3-(3-maleimidopropylcarbamoyl)-proxyl (IV), and
3-[2-(maleimido-ethoxy)-ethyl carbamoyll-proxyl (V) were purchased from
Aldrich Chemical Company (Milwaukee, WI). Monobromotrimethyl-
ammoniobimane was purchased from Calbiochem (San Diego, CA) under the
brand name thiolyte MQ. Mastoparan was obtained from Peninsula
Laboratories (Belmont, CA). Affinity chromatography material,
phenothiazine Affigel, was obtained from Bio-Rad Laboratories
(Richmond, CA). DEAE-Sephadex A50 and buffers, Maps and Tris, were

purchased from Sigma Chemical Company (St. Louis, MO).

Purification of spinach calmodulin

Calmodulin was isolated from fresh spinach by using slightly
modified procedures which have been described (12,13). Simply speaking,
2 kilograms of fresh spinach leaves was chopped, buffer H ( 50 mM Tris,
pH 8.0, containing 1 mM EGTA, 1 mM mercaptoethanol, and 1 mM
phenylmethyl-sulfonyl fluoride) 2000 ml was added into it, followed by
homogenization. The homogenized solution was filtrated through cheese
cloth and (NH4)2304 was added to the filtrated solution (351g/1

solution). Centrifugation of the solution was followed and the pH value

 

 

30

of the supernatant solution was readjusted to 4.0 with 50% H2504
saturated with (NH4)2804. The solution was centrifuged again and

the pellet was collected and redissolved in buffer B (10 mM Tris, pH
7.5, 1 mM EDTA, 1 mM mercaptoethanol, and 0.2 M NaCl). The pH value was
readjusted to 8.0, and extensive dialysis against buffer B was

followed. Another centrifugation was performed, the supernatant was
collected and applied to a DEAE-Sephades column (2.5 x 40 cm), which
was preequilibrated with buffer B. The column was washed with salt
gradient (starting buffer was buffer B, and the limiting buffer was
buffer B plus 0.3 M more of NaCl.). The eluant from the DEAE-Sephadex
column was monitered with UV absorption spectra and the calmodulin
containing peak was collected. Extensive dialysis against buffer A
(Tris, 10 mM, pH 7.5, 0.3 M NaCl. 1 mM mercaptoethanol, 5 mM CaC12),
then applied to the phenothiazine affinity column (1.5 x 18 cm)
equilibrated with buffer A. The column was washed with 100 m1 of the
same buffer A, then with 100 m1 of buffer A containing 0.5 M instead of
0.3 M NaCl. Finally the protein was eluted with another Tris buffer (10
mM, pH 7.5, 0.5 M NaCl, 1 mM mercaptoethanol, 10 mM EGTA). 50 ml of the
eluant were collected, into which 10 mM CaC12 were immediately added,
followed by dialysis against 4000 ml of (NH4)HCO3 solution (10 mM),
then by extensive dialysis against double-distilled water. The sample
was finally electrodialyzed against water and lyophilized. The purity
of the sample was assessed as follows: (a) Spinach calmodulin was
tryptophan-free as judged by the ultraviolet spectrum and the absence
of fluorescence emission typical of tryptophan (14). (b) The

ultraviolet absorption spectrum has a partially resolved vibrational

31

structure typical of that known for bovine calmodulin [1], and a molar

1 1, estimated to be half

absorption coefficient, €275 - 1500 M' cm’
that of highly purified bovine calmodulin, because spinach calmodulin
harbors only a single tyrosyl residue as opposed to two tyrosyl

residues in bovine calmodulin. (c) Analysis of the calmodulin
preparation on 15 percent polyacrylamide gels followed by Coomassie
blue staining revealed a single band migrating at an apparent molecular
weight of about 17,000, this result being consistent when up to 40 ug
protein were loaded per gel lane (12). (d) The concentration of spinach
calmodulin was measured according to the Bradford procedure (15) with a
protein assay kit commercially available from Bio-Rad (Richmond, CA)
and using bovine plasma gamma globulin solution as standard. The
concentration of the purified spinach calmodulin determined in this way
was compared with that evaluated from its known molecular absorption
coefficient; both values virtually matched. As a control, the
performance of the Bradford test was also evaluated on highly purified
bovine brain calmodulin. (e) The latter data seem to be a good

indicator that the sample is not contaminated by complex aromatic

compounds from spinach tissue.

Electron paramagnetic resonance studies

EPR spectra were recorded on a Varian X-band EPR Spectrometer, model
E-112, at room temperature. Rotational correlation times (nanosecond),
specifying the rate of rotaional motions, of the spin label bound to

spinach calmodulin were calculated from spectral data (16)

 

 

32

or; - 6.5 x 10‘10 x wo x [(ho/h-1)1/2 -1] (1)

where wo is the midfield line width (gauss), and ho and h_1 are

the respective peak heights of the mid- and high-field lines of the
first derivative EPR spectrum (Fig. 2.1). This relationship is
applicable for the calculation of rotational correlation times falling
within the range of isotropic tumbling of nitroxide spin probes, i.e.,
from 10‘10 to about 5 x 10'8 s (17). The lengths of the proxyl spin
labels employed (Fig. 2.2) have been reported (18).

A proxyl spin label, characterized by a given length, was attached
to the thiol group of cysteine-26 (11) on the spinach calmodulin. For
labelling, calmodulin was dissolved in Tris buffer (10 mM, pH 7.0) up
to a concentration of 0.2 mg calmodulin/ml buffer, then dithiothreitol,
a thiol-reducing agent (19), was added to the solution, up to a final
concentration of 0.2 mM. The reaction was continued under room
temperature for 18 hours. The reacted sample was then labelled with the
appropriate spin label by the method described (18, 20). Simply, the
spin probe was added to the reduced calmodulin solution up to a final
concentration of SO‘pM. Extensive dialysis against Mops buffer (10 mM,
pH 6.5, 0.1 M K01) followed. Under these experimental conditions

maleimide spin probes have been shown to be highly specific for free

thiol groups (20).

33

Fig. 2.1. A standard first derivative EPR spectrum. The direction the

arrow points stands for intensity increase of the magnetic field. Here

w is the midfield line width, h

0 +1, ho and h_1 stand for the

respective peak heights of the lower-, mid and high-field lines. a0

is the isotropic coupling constant.

34

 

 

 

 

 

 

 

 

._..+

35

Fig. 2.2. Structures and lengths, D, of maleimide spin labels used in

EPR experiments.

III

IV

36

 

H
0
Egg-*0 4.4
C2-CHZQ - 5.7

(fig-=0 . 9.3
Q-cH20H2NH-c

O
"(CH2)3M+H.8Q010.5

2:340 1 2 .9
cu,cu,ocn,cn,NH—c

oQoo
=o

37

Fluorescence studies

Fluorescence experiments were performed on an SLM spectrofluori-
meter, model 4800, (Urbana, IL). This instrument was interfaced with a
Hewlett-Packard desk top computer, HP-85, and a plotter to aid in data
acquisition and analysis. Quartz cuvettes with an optical path length
of 1 cm were used. Polarization experiments were performed with
Clan-Thompson calcite prism polarizers characterized by a high
polarization extinction ratio. During the polarization and lifetime
experiments, Corning CS 3-72 filters were used.

The steady-state fluorescence anisotropy, r, is defined as [14],

where I// and Il refer to the intensity of fluorescence emission
measured through the emission polarizers parallel and perpendicular to
the polarized plane of the exciting beam. T-format method (14) (Fig.
2.3) was used to perform anisotropy studies and Corning filters were
placed in the emission beam pathways.

The anisotropy parameter, r, is related to the fluorescence

lifetime, 1?, and the rotational correlation time, 0, according to the

equation (14)
r - ro(l +’t'/¢)'1 (3)

where r6 is the limiting anisotropy in the absence of rotational

38

Fig. 2.3. An illustration of T- and L-format method for measuring
fluorescence anisotropy. Pe is the excitation polarizer, whose
polarized plane can take two positions, one horizontal and one
vertical. PA and PB are emission polarizers in channels A and B,
respectively. PMT stands for photomultiplier. a. Each emission

polarizer can take only one position. b. PA can also take two

positions.

39

 

p Lamp

 

PMT (B)

a. T-format method

PMT (A)

 

Lamp

b. L—format method

40

diffusion. It was obtained by using Perrin plot, extrapolating the line
to 0 of T/y, and the anisotropy value obtained in this way is the ro
(14). Fluorescence lifetimes were calculated from phase shift and
demodulation data (14) by using a desktop computer, HP-85. It was
assumed that the fluorescence decay process could be analyzed in terms
of a biphasic decay, viz., a fast and short-lived decay, both
exponential in time and independent from each other.

At first, spinach calmodulin was reduced by dithiothreitol as
described above, followed by addition of the bimane fluorescence probe
(0.1 mM); this probe is highly reactive with thiol groups (21, 22).
subsequently extensive dialysis against the Mops buffer was employed
after dithiothreitol reduction. Control experiments showed that both
the spin and fluorescence probe-labelled spinach calmodulin had
virtually the same biochemical activity as that of the unlabelled
calmodulin, as judged by an enzyme assay of calmodulin-dependent
phosphodiesterase activity (23).

The molar ratio of bimane probe bound per protein was found to be
unity as determined from the known protein concentration and the dye‘s
absorbance,.£378-5,700 M'lcm'1 (21).

All fluorescence experiments were performed at low protein
concentration to virtually exclude depolarization, energy transfer and
scattering. As lifetime reference 1,4-bis-2-(4-methyl-5-phenyl-
oxazolyl) benzene was used (14). To assess the performance of the
fluorescence instrumentation, control experiments with several quinine

solutions were performed as lifetime standards [24]. As fluorescence

anisotropy standard N-acetyl-L-tryptophanamide in propylene glycol was

41

used (25).

Results

EPR studies

EPR spectra of spinach calmodulin, associated with proxyl spin
probes of different length, are depicted in Fig. 2.4. Qualitatively
speaking, the EPR spectrum of spin probe 1 appears to reflect motional
characteristics of a more immobilized label compared with those of spin
probe V. The rotational correlation time, as determined from eq. (1),
is dependent on the length of the spin label (Fig. 2.5). Apparently
there exists a stepwise dependence of the value of the correlation time
as the length of the spin probe increases (Fig. 2.5). Upon addition of
2 Ca2+ per calmodulin, the correlation times of the shortest spin
probes, i.e., I and II, become faster, while those of the other three
probes remain practically unchanged. At a molar ratio of 5:1 for
[Ca2+]/ [calmodulin], the rotational correlation times of all 5 spin
labels virtually coincide with the correlation times determined in the
absence of added Ca2+ (Fig. 2.5).

Upon binding of the small peptide, mastoparan, to spin-labelled
calmodulin, the correlation time of the shortest spin probe, 1, is
significantly reduced compared with the correlation times of the longer
spin probes. Apart from spin label I, the stepwise dependence of the

correlation time on label length is practically unchanged (Fig. 2.5).

42

Fig. 2.4. First derivative EPR spectra of spin-labelled apocalmodulin
(0.18 mg/ml) in Mops buffer (10 mM, pH 6.5, 0.1 n K01), at 22° 0.

Following treatment with dithiothreitol, at 22°C, spinach

apocalmodulin was labelled with the respective spin probe. 5 different

spin probes (see Fig.1) were employed for labelling.

 

 

 

11 ~
Sj »

 

 

44

Fig. 2.5. Dependence of the rotational correlation time, 7c, of
spin-labelled spinach calmodulin on spin label length, D. ( O )
apocalmodulin; ( I ) [Ca2+]/[calmodulin] - 2:1; ( A )
[Ca2+]/[calmodulin] - 5:1; (9) [Ca2+]/[mastoparan]/[calmodulin] -
5:1:1. The experiments were performed in Mops buffer (10 mM, pH 6.5,

0.1 M KCl), at 22°C. The error of the correlation time is 1; 0.05 ns.

45

 

 

 

 

 

46

Fluorescence studies

The apparent fluorescence lifetimes of the bimane-labelled
calmodulin and those of the free fluorescence probe are listed in Table
2.1. The fluorescence intensity of the free dye seems to decay in a
monoexponential fashion (because the’Th and Th of the free probe,
measured at 18 MHz and 30 MHz, were virtually identical, viz.,
3.85:0.10 ns), but upon dye binding to the protein the fluorescence
decay is no more monoexpotential. From a heterogeneity analysis of the
raw data, the fluorescence decay of bimane-labelled calmodulin can be
described as a superposition of two exponentially decaying components,
a short- and a long-lived one, respectively (Table 2.1). This
assumption is in accord with results from time-resolved anisotropy
studies on bovine calmodulin labelled with 2,6-toluidinylnaphthalene-
sulfonate; the position of this probe on the protein surface is not
known (26). Upon addition of Ca2+ to the protein, at a molar ratio of
2:1 and 5:1, in the absence of mastoparan, the lifetime of the
short-lived component remains practically unchanged, whereas the
lifetime of the long-lived component appears to become shorter
concomitant with an increase in [Ca2+] concentration (Table 2.1).

Comparing the fluorescence parameters listed (Table 2.2), it seems
that the anisotropy value decreases upon addition of 2 Ca2+ per
protein to bimane-labelled calmodulin in solution, but at a molar ratio
of 5:1 for [Ca2+]/ [calmodulin], the anisotropy increases to a value
slightly above that typical of calmodulin in the absence of Ca2+

Upon addition of an equimolar amount of mastoparan to bimane-labelled

47

Table 2.1. Fluorescence lifetimes,'$1,-t§, and associated
preexponential factors, a1 and a2, of bimane-labelled spinach
calmodulin. These values were determined on an SLM spectrofluorimeter,
model 4800, by the phase shift and demodulation methods and analyzed on
a Hewlett-Packard computer, model 85. The sample was excited at 385 nm,
the emission was recorded as light passing a Corning filter, CS 3-72.

The calmodulin concentration was 0.18 mg/ml in Mops buffer (10 mM, pH

6.5, 0.1 M KCl), at 22°C.

 

 

sample 47,1 preexpontial
(n5; factor (81 2)

calmodulin 2.7 0'33
9.5 0.67

[calmodulin]/[C82+]

_ 1.2 2.9 0.34
8.6 °~66

[calmodulin]/[Ca2+]

_ 1.5 2.5 0.29
7.6 0.71

[calmodulin]/[Caz+]/

[mastoparan] - 1:5:1 4.2 0'61

10.9 0.39

 

Table 2.2. Molecular parameters for bimane-labelled spinach calmodulin

The values, ’rhv , were calculated from the data listed in Table 1

according to [14]. The corresponding rotational correlation times,

oav.' were calculated from the eq.(3) in the MATERIALS AND METHODS,

using the anisotropy values, r, listed and a limiting anisotrOpy value

of ro - 0.2142, derived from the Perrin plot (Fig.4).

 

 

sample iggv. 03v r percentage*
(n8) (n8)

calmodulin 8.7 6.9 0.0949 100

[calmodulin]/[Ca2+]

- 1:2 7.8 5.6 0.0894 81

[calmodulin]/[Ca2+]

- 1:5 7.0 5.5 0.0943 80

[calmodulin]/[Ca2+]/

[mastoparan] - 1:5:1 8.4 8.4 0.1072 122

 

*: the rotational correlation time of calmodulin is assigned a value of
100 percent. The other percentage values are the ratios of the
respective rotational correlation times to that of calmodulin.

49

calmodulin, the anisotropy value exceeds that ascertained for
bimane-labelled calmodulin in the presence of [Ca2+]/ [calmodulin]

-5:l (Table 2.2). Corresponding molecular parameters are listed for the
interaction of calmodulin with the peptide (Table 2.2). By '
extrapolation of the anisotropy values to the ordinate in a Perrin plot
(14), r'1 vs TA? , (Fig. 2.6), the limiting anisotropy value, r0,
can be derived.

The viscosity of the sample solution was gradually increased by
adding concentrated sucrose solution. The resulting viscosity value of
the aqueous sucrose mixtures was obtained from the literature [27]. The
ro value, obtained by extrapolation (Fig. 2.6), is lower than that
determined for bimane-labelled 8100 protein, at -20°C, in the
presence of 80/20 (v/v) polypropylene glycol/water (21).

The rotational correlation times of the bimane-labelled calmodulin

2+ to calmodulin in

were calculated from eq. (3). Adding two Ca
solution, the average rotational correlation time decreased to about 81
2+ U li ti of
percent of that measured in the absence of Ca . pon app ca on
additional 3 Ca2+, the rotational correlation time remained virtually
constant. On the other hand, upon mastoparan binding to calmodulin
(Table 2.2), the rotational correlation time increased dramatically to

about 122 percent of the value determined for apocalmodulin in

solution.

50

Fig. 2.6. Perrin plot of spinach calmodulin whose single cysteinyl
residue 26 was labelled with a fluorescence probe, monobromotrimethyl
ammoniobimane. The fluorescence anisotropy value was measured by the
T-format method in an SLM spectrofluorimeter, model 4800. Two CS 3-72
filters were placed in the emission path. The sample was excited at 385
nm. The viscosity, 6?, was increased by adding concentrated sucrose
solution (1 g/ml) into the sample solution. The experiments were
carried out in Mops buffer (10 mM, pH 6.5, 0.1 n K01), at 22°C. The

error in anisotropy is i 0.0004.

51

 

 

160

 

 

 

‘
rrlIt'trrrtttjmifiltrTTrrtmfiujjrtjrwti o

as o a: e-

0 to in a:

3.6

,-(Kdo.nosguo)

T/n (OK/centipoise)

52

Discussion

Calmodulin, like other globular proteins, is a dynamic system
undergoing internal motions on a time scale as fast as picoseconds (7).
These internal motions and their proper time-correlation are probably
crucial for protein functioning (28). In such a dynamic system there
exist a multitude of conformational substates which are accessible by
rapid structural fluctuations of the protein. Ligand-triggered
structural perturbations, e.g., binding of Ca2+ to high-affinity
sites of calmodulin, may therefore be transmitted in a flexible
macromolecule (29) over a certain distance. Indeed, we have
demonstrated in this article that a response to ligand-induced
perturbations is elicited at a specific site of calmodulin.

The dynamic interplay of calmodulin's constituent parts can be
better understood in terms of calmodulin's crystal structure. X-ray
diffraction experiments on bovine calmodulin indicate that the protein
has the shape of a dumbbell (3) with an overall length of about 6.2 nm
in the presence of calcium (4). The calcium-binding regions I and II
are located in the lobe containing the amino terminus, whereas the
high-affinity regions III and IV (1) reside in the lobe with the
carboxyl terminal end. The two lobes are interconnected by an
eight-turn-helix.

Spinach calmodulin harbors a single cysteinyl residue at site 26
which is located in region I of the protein, replacing a threonyl
residue present in bovine brain calmodulin (11). The rotational

correlation times of spin-labelled spinach calmodulin are those of

53

rapidly tumbling nitroxide labels. This suggests that the binding site,
i.e., cysteine-26, is not buried inside the protein but seems to be
exposed at the protein‘s surface. The label tested seems to reside in a
polar microenvironment as evidenced by an isotropic coupling constant,
a0, of 16.0 $0.2 gauss; this value was unaffected by the presence or
absence of Ca2+. The isotropic coupling constant measured is
comparable to that found for nitroxide in ethanol as solvent [30]. A
constant label microenvironment is also indicated by our fluorescence
studies, where the fluorescence maximum (wavelength and intensity) did

2+ and a peptide

not change in response to the application of Ca
target.

The rotational correlation times measured for spin-labelled spinach
apocalmodulin show that the shorter spin probes, i.e., I and II, have
less motional freedom than the longer spin labels. However, binding of
2 Ca2+ per protein somewhat alleviates the motional restrictions
imposed by the microenvironment around the shorter spin probes.
Monitored at cysteine-26 in region I, located in the amino terminal
lobe of spinach calmodulin, these motional changes probably do not
result from direct interactions of the cations with this site because
the first two Ca2+ added to calmodulin seem to bind to the
high-affinity binding regions (1) located in the carboxyl terminal lobe
of the molecular dumbbell. Rather, as a result of calcium coordination
at the high-affinity sites, structural perturbations are produced which
are being transmitted over a considerable distance to the other lobe.

2+

Upon further addition of Ca to the protein, up to a molar ratio of

5:1, calcium binding takes place in regions I and II, i.e., in the

54

immediate vicinity of the spin probe. Consistent with our results on
the motional characteristics of the shorter spin labels one wouold
expect a decrease in the label's mobility when calcium is specifically
anchored in its coordination cage of region I when compared with those
without calcium binding at any sites at all. Our findings about the
calcium dependence of spin label motion at site 26 are also in accord
with spectroscopic experiments indicating that Ca2+ binding by bovine
calmodulin takes place in at least two distinct stages (31).

As monitored at site 26, marked motional changes are reported by the
shortest spin label upon binding of mastoparan to calmodulin. The
amphiphilic tetradecapeptide, mastoparan, has an association constant
of 3.3 nM'1 with calmodulin bound with calcium (32), and undergoes a
conversion from a conformation which is low in -he1ix to a highly
helical structure (33) upon formation of the calmodulin-mastoparan
complex, probably in region III of calmodulin's C-terminal half [31],
when Ca2+ is present. This binding region seems to partially match
that of cyclic nucleotide phosphodiesterase because the wasp venom,
mastoparan, is a potent inhibitor of this enzyme (34). Compared with
the rotational correlation time of the shortest spin label in
apocalmodulin or with that in the presence of Ca2+, this time is
reduced by about 30 percent when spinach calmodulin interacts with a
partner peptide, mastoparan. Apparently a specific response is elicited
at the cysteine-26 site following transmission of structural
perturbations triggered in the specific target region where calmodulin
directly communicates with its partner protein. Our observations are in

accord with NMR data on the binding of a 27-residue peptide (denoted

55

M13) to bovine calmodulin. Derived from skeletal muscle myosin light
chain kinase, M13 peptide binding affected many NMR resonances arising
from residues in both moieties of calmodulin (35).

The fluorescence data (anisotropy, apparent lifetimes) obtained from
experiments on the bimane probe, attached to cysteine-26 of spinach
calmodulin, are complementary to results derived from EPR measurements.
The former data can be interpreted as follows: the average rotational
carrelation time seems to be associated with the protein's rotational
motion, probably that arising from a major portion of the calmodulin

2+

molecule. When two Ca are added to calmodulin in solution, the

protein's rotational motion is enhanced compared with that in the

2+, consistent with the compaction reported (36); the

2+

absence of Ca
rotational motion remains unchanged upon further addition of Ca
which bind to regions I and II of the protein. Following peptide
binding to calmodulin, the rotational motion of the protein-ligand
complex slows down because the total volume of the complex is larger
than that of the apoprotein. Calmodulin (M - 16 800 for bovine
calmodulin) has a rotational correlation time of about 6.8 ns, assuming
a partial specific volume of 0.73 cm3/g and a hydration of 0.2,
respectively. Given this simple picture of a spherical molecule, it
appears that the relatively short bimane fluorescence probe at
cysteine-26 serves as a reporter group for the rotational motion of the
protein, characterized by an average rotational correlation time,

either in the absence or presence of Ca2+

. When a specific complex is
formed between calmodulin and mastoparan, the correlation time for the

overall rotation of the complex is becoming slower, as monitored by the

56

bimane reporter group at cysteine-26.

In conclusion, we have demonstrated that a ligand-triggered, C82+-
dependent response can be monitored at a site distant from the trigger
region. Moreover, the methodology introduced of exploiting the presence
of a single cysteinyl residue in spinach calmodulin may be beneficially

applied for the analysis of more detailed dynamic aspects of calmodulin

in relation to specific partner proteins.

This research was supported by a grant from the National Institutes
of health, No. 1 R01 ESO4468-01. The authors thank Dr. K. 'Poff, Plant
Research Laboratory, for use of the SUM spectrometer. The assistance of

Chris Weis in these fluorescence studies is greatly appreciated.

10.

ll.

12.

13.

57

LIST OF REFERENCES
Klee, C.B. and Vanaman, T.G.

, (1982) Calmodulin. Adv. Protein Chem.
35: 213-321.
Kretsinger, R., (1975) Hypothesis: Calcium modulated proteins
contain EF hands. in: C c Tra o n Contraction and
Secgetion (Carafoli, E., Clementi, F., Drabikowski, W. & Margreth,

A., eds.), North-Holland Pub1., Amsterdam. pp.469-478.

Babu, Y.S., Sack, J.S., Greenhough, T.J., Bugg, C.E., Means, A.R.,

and Cook, W.J., (1985) Three-dimensional structure of calmodulin.
Nature (London) 315: 37-40.

Seaton, B.A., Head, J.P., Engelman, D.M., and Richards, F.M.,.
(1985) Calcium-induced increase in the radius of gyration and

maximum dimension of calmodulin measured by small-angle X-ray
scattering. Biochemistry 24: 6740-6743.

Sellinger-Barnette, M. and Weiss, 8., (1984) Interaction of various

peptides with calmodulin. Adv. Cycl. Nucleot. Protein Phosphoryl.
Res. 16: 261-276.

Weis, C. and Haug, A., (1987) Aluminum-induced conformational
changes in calmodulin alter the dynamics of interaction with
melittin. Arch. Biochem. Biophys. 254: 304-312.

Karplus, M. and McCammon, J. A.,

(1981) The internal dynamics of
globular proteins. CRC Crit. Rev.

Biochem. 9: 293-349.

Chou, K.C. and Kiang, Y.S., (1985) The biologiCal functions of

low-frequency vibrations (phonons) 5. A phenomenological theory.
Biophys. Chem. 22: 219-235.

Elber, R. and Karplus, M., (1987) Multiple conformational states of

proteins: a molecular dynamics analysis of myoglobin. Science 235:
318-321.

Beece, D., Eisenstein, L., Frauenfelder, H., Good, D., Marden,
M.C., Reinisch, L., Reynolds, A.H., Sorensen, L.B., and Yue, K.T.,
(1980) Solvent viscosity and protein dynamics. Biochemistry 19.
5147-5157.

Vanaman, T. C., (1983) Chemical approaches to the calmodulin
system. Methods Enzymol. 102: 296-310.

Suhayda, C.G. and Haug, A., (1985) Citrate chelation
mechanism against aluminum toxicity in cells:
calmodulin. Can. J. Biochem.

as a potential
the role of
Cell Biol. 63: 1167-1175.

Watterson, D.M., Iverson, D.B. and Van Eldik, L.J., (1980) Spinach

14.

15.

l6.

l7.

l8.

19.

20.

21.

22.

23.

24.

25.

26.

-Lai, C.-S., and Tooney, N.M.

58

calmodulin: isolation, characte

rization and comparison with
vertebrate calmodulins.

Biochemistry 19: 5762-5768.
Lakowicz, J.R., (1984) Principles pf Eluorescenpe Spectroscopy.
Plenum Press, New York.

Bradford, M.M., (1976) A rapid and sensitive method for the
quatitation of microgram quantities of protein utilizing the
principle of protein-dye binding. Anal. Biochem. 72: 248-254.

Keith, A., Bulfield, G. and Snipes, W., (1970) Spin-labeled
Ngurospppn mitochondria. Biophys. J. 10: 618-629.

Mehlhorn, R.J. and Keith, 11.0., (1972) in: Membrnne Molecular
Biology (Fox, C.F. & Keith, A.D., eds.), Sinauer Assoc., Stamford,
CT. pp. 192-227.

Lai, C.-S., Tooney, N.M. and Ankel, E.G.,

(1984) Spin label studies

of sulfhydryl environment in plasma fibronectin. FEBS lett. 173:

283-286.

Cleland, W.W., (1964) Dethiothreitol, a n

ew protective reagent for
SH groups. Biochemistry 3: 480-482.

, (1984) Electron spin resonance spin
label studies of plasma fibronectin: effect of temperature. Arch.
Biochem. Biophys. 228: 465-473.

Kosower, N.S., Kosower, E.M., Newton, C.L., and Ranney, R.M.,

(1979) Bimane fluorescent labels: labeling of normal human red
cells under physiological conditio

ns. Proc. Natl. Acad. Sci. (U.S.)
-76: 3382-3386.

Baudier, J., Glasser, N. and Duportail, G.

acrylodan- labeled $100 proteins: role of cysteines-85 and -84 in
the conformation and calcium binding properties of 8100 and
$100b( ) proteins. Biochemistry 25: 6934-6941.

, (1986) Bimane- and

Tea, T.S., Wang, T.H. and Wang, J.H., (l973)Purification and
properties of the protein activator of bovine heart cyclic

adenosine 3', 5'-monophosphate phosphodiesterase. J. Biol. Chem.
248: 588-595.

Chen, R.F., (1974) Fluorescence lifetime reference standards for
the range 0.189 to 115 nanoseconds. Anal. Biochem. 57: 593-604.

Lakowicz, J.R. and Weber, G., (1980) Nanosecond segmental mobility
of tryptophan residues in proteins observed by lifetime-resolved
fluorescence anisotropies. Biophys. J. 32: 591-601.

Steiner, R.F. and Norris, L., (1987) Rotational modes of

27.

28.

29.

30.

31.

32.

33.

34.

35.

36.

59

Ca2+-liganded calmodulin, as determined by time-domain

fluorescence. Biophys. Chem. 27: 27-38.

Weast, R.C. (1983/1984) 939 finnnpook pf Cheniggyy and Physics, p.
D-266, CRC Press, Boca Raton, FL.

Careri, G. (1984) dipr nnn Digprde; in flatten, Benjamin/Cummings
Publ., Menlo Park, CA.

Lambooy, P.K., Steiner, R.F., and Sternberg, H., (1982) Molecular

dynamics of calmodulin as monitored by fluorescence anisotropy.
Arch. Biochem. Biophys. 217: 517-528.

Morrisett, J.D., (1976) The use of spin labels for studying the

structure and function of enzymes. in: Spin Labeling; Theory and
Applications, (Berliner, L.J., ed.), Academic Press, New York, NY.

pp. 273-338.

Muchmore, D.C., Malencik, D.A. and Anderson, S.R., (1986) +H NMR

studies of mastoparan binding by calmodulin. Biochem. Biophys. Res.
Commun. 137: 1069-1078.

Malencik, D.A. and Anderson, S.R., (1983) Binding of hormones and
neuropeptides to calmodulin. Biochemistry 22: 1995-2001.

Cachia, P.J., Van Eyk, J., Ingraham, R.H., Mccubbin, W.D., Kay,
C.M., and Hodges, R.S., (1986) Calmodulin and troponin C: a
comparative study of the interaction of mastoparan and troponin I
inhibitory peptide [104-115]. Biochemistry 25: 3553-3562.

Barnette, M.S., Daly, R., and Weiss, B., (1983) Inhibition of

calmodulin activity by insect venom peptides. Biochem. Pharm. 32:
2929-2933.

Klevit, R.E., Blumenthal, D.K., Wemmer, D.E. and Krebs, E.G.,
(1985) Interaction of calmodulin and a calmodulin-binding peptide
from myosin light chain kinase: major spectral changes in both

occur as the result of complex formation. Biochemistry 24:
8152-8157.

LaPorte, D.C., Keller, C.B., Olwin, B.B. and Storm, D.R., (1981)
Preparation of a fluorescent-labeled derivative of calmodulin which

retains its affinity for calmodulin binding proteins. Biochemistry
20: 3965-3972.

CHAPTER III

DYNAMIC PROPERTIES OF CALMODULIN IN RESPONSE TO ALUMINUM

Shixing Yuan and Alfred Haug

The Environmental Chemistry and Toxicology of Aluminum,
(Lewis, T.E., ed.)

Lewis Publ., Chelsea, MI. in press

60

61

As monitored by a fluorescence dye attached to the single cysteinyl
residue 26 of spinach calmodulin, application of aluminum ions to

2+, results in motional

calmodulin, in the presence of saturating Ca
changes of the fluorescence probe different from those observed in the
absence of aluminum. In the presence of aluminum, the extent of
hydrophobicity of the probe's microenvironment resembles that observed
in apocalmodulin. The average rotational correlation times of
calmodulin are dependent on temperature and the type of cation, i.e.,
calcium and aluminum. Energy transfer studies from the single
tryptophanyl residue of a calmodulin target peptide, mastoparan-X, to
the fluorescence probe as acceptor residing at cysteine-26 of spinach
calmodulin, show a higher transfer efficiency in the aluminum-altered
protein. We therefore conclude that structural and dynamic changes take
place in aluminum-altered spinach calmodulin, even in the presence of

2+

saturating Ca concentrations.

62

Introduction

Aluminum, the most abundant metal in the earth's crust, had been
implicated as early as 1918 as a cause of root-growth retardation in
crop plants grown on acidic soil (1). Today, aluminum toxicity is
recognized as a serious global problem for crop productivity because
vast areas of the world suffer from soil acidity which mobilizes soil
aluminum available for plants (2, 3). Besides natural soil acidity,
anthropogenic perturbation of hydrogeo-chemical processes has further
contributed to the acidification of terrestrial and aquatic regions
previously not suffering from aluminum toxicity (4). For example,
so-called acid rain has substantially increased over wide parts of the
eastern United States, Canada, and western Europe (5). As to human
health, clinical and experimental findings suggest a link between
aluminum uptake and a variety of disorders in man, e.g., osteomalacia-
type diseases and dialysis encephalopathy (6).

Little is known about molecular mechanisms regarding entry of
aluminum species into cells via the plasma membrane, about primary
targets for these species, or about aluminum-induced cellular lesions.
Deeper insight into cellular and molecular processes involving aluminum
ions is crucial for the biotechnological development of plant
resistance to aluminum toxicity. Despite this dearth of information,
cellular membranes and calcium-regulated biochemical processes appear
to emerge as potential targets for toxic aluminum ions.

Concerning aluminum interactions with membranes, a body of evidence

suggests pronounced aluminum-induced phase changes in the plasma

63

membrane of thermophilic Archaebacteria (7) and in plasma membrane-
enriched microsomal fractions isolated from maize roots (8).
Furthermore, micromolar aluminum concentrations induced membrane
rigidification and membrane fusion in unilamellar lipid vesicles
containing phosphatidylserine (9).

Besides interfering with the plasma membrane, physiological studies
indicated that aluminum ions effect changes in calcium regulation (2,
6). Therefore, a potential target for aluminum interaction might be
calmodulin, a pivotal, calcium-dependent regulatory protein
(Mr-16,000-l7,000). This acidic protein has been highly conserved
during evolution of eukaryotic cells where it participates in a
multitude of cellular processes (10). Given the significance of
calmodulin in integrating calcium-dependent processes, aluminum-induced
changes of the protein's structure would therefore have severe
repercussions on cellular functions. Indeed, recent investigations
indicate that binding of aluminum ions to bovine brain calmodulin
results in pronounced structural and dynamic changes which, in turn,
lead to the inhibition of crucial calcium- and calmodulin-dependent
enzymes (3, ll, 12).

As to aluminum-induced structural changes, the metal binds
stoichiometrically and cooperatively to the protein, as determined by
fluorescence and circular dichroism studies. Binding of aluminum occurs
irrespective of the presence or absence of calcium ions (11, 13). When
bound to the protein, aluminum ions trigger a helix-coil transition
whereby the helix content decreases by about 30 percent. It is

noteworthy, that addition of calcium ions promotes helix formation

64

(10), with aluminum absent. Atomic absorption experiments indicated
that the application of excess citrate to an [aluminum]/[ca1modulin]-
(3:1) complex resulted in a time-dependent, quantitative removal of the
metal from the protein (14). As judged by spectroscopic techniques,
this citrate-mediated protein restoration is somewhat incomplete in
terms of helical content. The incomplete rearrangement of calmodulin
structure is probably related to aluminum-induced changes in salvation.
The rearranged water molecules may compete for hydrogen bonding sites
on calmodulin when the protein refolds as a result of citrate-mediated
aluminum removal.

Concomitant with aluminum-induced changes in helix content, the
hydrophobic surface exposure of calmodulin is enhanced (3), relative to
the area observed in the presence of stoichiometric quantities of
calcium ions (10). This hydrophobic region or its immediate vicinity
probably play a key role in the interaction of calmodulin with specific
partner proteins, e.g., an ATPase (10). Aluminum-related changes at the
interface between calmodulin and a partner protein were illustrated in
experiments with melittin (15), a small protein which has a high
affinity for calcium-calmodulin (16). As monitored at the single
tryptophanyl residue of melittin, the fluorophore's average
microenvironment is modified suth that its apparent lifetime is
shortened when aluminum is present. Moreover, the interface between the
two proteins seems to become more polar when aluminum is bound to
calmodulin. In the presence of aluminum, calmodulin's salvation
structure is possibly altered, which may be unfavorable for a proper

fit between calmodulin and the target protein.

65

We present in this article further results on aluminum-related
structural and dynamic changes monitored by a fluorescent dye attached
to the single cysteinyl residue of spinach calmodulin. By measuring
molecular parameters like fluorescence anisotropy, motional parameters

were derived which indicate that aluminum altered calmodulin differs

from native calcium-calmodulin.

Materials and Methods

Mastoparan-X and mastoparan were purchased from Peninsula
Laboratories (Belmont, CA). N-iodoacetyl-N'-(5-su1fo-1-naphthy1)
ethylene-diamine (1,5-I-AEDANS) was obtained from Sigma Chemical
Company (St. Louis, MO). Affinity chromatography material,
phenothiazine Affi-gel, and a protein assay kit were purchased from
Bio-Rad Laboratories (Richmond, CA).

Calmodulin was isolated from fresh spinach using methods previously
described (l4, l7, l8). Spinach calmodulin obtained in this manner was
found to be free of tryptophan as judged by UV absorbance and lack of
fluorescence emission at 340 nm following excitation with 295 nm light.
In addition, the protein's UV absorption spectrum showed a partially
resolved vibrational structure typical of that known for bovine brain
calmodulin (10). Furthermore, analysis of the spinach calmodulin
preparation on 15 percent polyacrylamide gels followed by Coomassie
blue staining indicated a single band migrating at an apparent

molecular weight of 17,000, this result being consistent when up to 45

\

J

 

__...___..

66

ug protein were loaded per gel lane.

To attach the fluorescence probe, 1,5-I-AEDANS, to spinach
calmodulin, the protein was first dissolved in Tris buffer (10 mM, pH
7.0) at a concentration of about 0.2 mg/ml buffer, then a thiol-
reducing agent, dithiothreitol, was added to the solution to a final
concentration of 0.2 mM. The reaction was allowed to proceed at room
temperature for 18 hours. The reduction sample was then extensively
dialysed against the same Tris buffer to remove excess thiol-reducing
agent. The fluorescence label, 1,5-I-AEDANS, was added and reacted for
about 2 hours with the sample as described (19). Finally, the reaction
mixture was extensively dialyzed against Mops buffer (10 mM at pH 6.5
and 0.1 M KCl), which was used in all experiments. The molar ratio of
AEDANS bound per protein was found to be 0.95 to l, as determined by
dye absorbance (8-6,100 cm2 mM'1 at 337 nm) (20) and Bradford
assay (21), respectively. Unless indicated otherwise, the protein
concentrations employed were 10 uM. To occupy all four Ca2+ binding
domains of calmodulin, Ca2+ ions were added to the protein at a molar
ratio of 5:1, to achieve saturating conditions.

As described previously (15,18), fluorescence experiments
(anisotropy, lifetime) were performed on an SLM spectrofluorimeter,
model 4800, manufactured by SLM Instruments, Inc. (Urbana, IL). For
each assay a Beckman cuvette with 1 cm optical pathlength and a sample
volume of 2.5 ml was used. Furthermore, at the low protein concentra-
tions used, the anisotropy and the lifetime were independent of protein
concentration (22). Each data point represents the average value for

100 counts per sample which, in turn, was four times replicated.

67

Fluorescence energy transfer was evaluated between a donor, D,
viz., the single tryptophanyl residue on mastoparan-X (M - 1,557) and
an acceptor, A, viz., the fluorescence label, AEDANS, residing at the
single cysteinyl residue, Cys-26 (23), of spinach calmodulin. As
opposed to mastoparan-X, mastoparan (M - 1,479) does not harbor any
tryptophanyl residue. The energy transfer efficiency was measured by

the donor quenching method (24), i.e.,

E " 1 ' QDA/QD (3)

where E stands for the energy transfer efficiency. QDA and QD refer
to the fluorescence quantum efficiency of the donor, in the presence
and that in the absence of the acceptor. The ratio QDA/QD can be

evaluated from corresponding fluorescence intensities, fDA and fD,

according to

QDA/QD ' [fDA/fplav X [DID/[DIDA- (4)

Here, both fluorescence intensities, fDA and fD (22), are measured
upon excitation of the donor at 299 nm to minimize the excitation of
tyrosin. The quantities, fD and fDA' are the fluorescence

intensities of the donor, measured in the absence of the acceptor, and
in the presence of the acceptor within a wavelength range free from
acceptor emission, respectively. Subsequently the intensity ratios were
averaged, thus yielding [fDA/fDJaV' [DID is the concentration

of donor molecules in the absence of any acceptor, whereas [DJDA is

68

that in the presence of both acceptor and donor molecules.

The distance, R, from the donor, D, to the acceptor, A, was

calculated from the following equation,

5
R - R0 (l/E -1)1/6 ( )

where R0 is the F6rster critical distance, evaluated from

R06 - (8.79 x 10'5) x k2 x n“‘ x 0D xJDA (6)
Here, k2 is the orientation factor and a value of 2/3 was used in
these experiments. Usage of this value is justified if donor and
acceptor molecules randomize prior to energy transfer (25). Since the
aromatic ring systems of the donor (tryptophan) (22) and the acceptor
(AEDANS) have two linear transition moments each, the assumption of
random orientation appears to be reasonable. Moreover, n is the

refractive index and JDA is the spectral overlap integral defined as
Joe. 4;.“ £D<2>tA<m4s3 m 3,: new. 1 (7)

whereggA is the extinction coefficient of the acceptor in

M'lcm'l; 2, is the wavelength in mm.

Four kinds of samples have to be prepared for these energy transfer
studies: (a) Ca-calmodulin plus mastoparan (neither donor nor acceptor
present), (b) Ca-calmodulin plus mastoparan-X (donor present, D); (c)

AEDANS-labelled Ca-calmodulin plus mastoparan (acceptor present, A);

\J

69

and (d) AEDANS-labelled Ca-calmodulin plus mastoparan-X (both donor and

acceptor present, DA).

Results

Our experiments show that the emission of the AEDANS-spinach
calmodulin conjugate reaches a maximum at 490 nm and is considerably
blue-shifted from the emission maximum of 520 nm reported for the
aqueous free probe (20). An emission maximum observed at 490 nm
corresponds to that reported for AEDANS dissolved in 60% ethanol/water
(20). When covalently bound to spinach calmodulin, the probe apparently
resides in an environment less accessible to water. Increased rigidity
or further water removal from the probe‘s microenvironment may be
responsible for the blue shift. The fluorescence spectra of
dye-labelled spinach calmodulin, measured in the presence or absence of
micromolar aluminum or calcium concentrations, virtually coincide.

Table 3.1 lists the values of various spectroscopic parameters of
the dye-labelled calmodulin in the presence and absence of calcium and
aluminum. The average fluorescence lifetime of the probe remains
relatively constant through serial calcium titration and upon addition
of aluminum up to a molar ratio of 3 aluminum ions per protein.

However, the rotational correlation time decreases with increasing
calcium concentration, consistent with reports that the protein adopts
a more compact conformation upon calcium binding (26). At a molar ratio

of [Al]/[calcium saturated calmodulin] - 3:1, the rotational

70

correlation time significantly increases relative to the time
determined in the absence of aluminum. The rotational correlation times
were calculated from eq.(2) using an ro value of 0.0910 i 0.0035.

This value was obtained from a Perrin plot by extrapolation with a
correlation coefficient of 0.988.

Our fluorescence lifetime data can be interpreted in terms of a
biphasic fluorescence decay. The lifetime of the long-lived component
roughly coincides with that measured for free AEDANS in water (20).
Since our instrument does not permit measurements of time-resolved
anisotropy decays, an average rotational correlation time had to be
calculated (Table 3.1).

To assess the impact of aluminum on the interaction between
calmodulin and its target proteins, radiationless energy transfer
experiments were performed between the single tryptophanyl residue on
the partner peptide, mastoparan-X, and the AEDANS label on Cys-26 on
spinach calmodulin. This kind of peptide is known to bind calmodulin
with high affinity in a calcium-dependent manner (16). The wavelength
region of the fluorescence emission band of tryptophan shows excellent
overlap with that of the acceptor‘s absorption spectrum, AEDANS-
conjugated cysteine (Fig. 3.1), and therefore, energy transfer from
tryptophan to the acceptor can be expected. By integrating in steps of
2 nm, the value of the overlap integral, JDA’ was calculated
according to eq.(7). From the reported values of the donor quantum
yield, QD, of 0.155 (27), a refractive index of n - 1.334,
corresponding to that of the 0.1 M KCl solution at room temperature

(28), and an orientational factor of k2 - 2/3, the F6rster critical

71

Table 3.1. Fluorescence properties of AEDANS-labelled spinach

calmodulin at various cation concentrations

 

 

*
8) z;
sample r 82 73 Itav q)av
(ns) (ns) (ns)
AEDANS-calmodulin 0.0579 0.14 0.88 9.7 15.0
0.86 9.86
AEDAN -ca1modulin 0.0524 0.12 0.57 10.4 12.6
+2 Ca + 0.88 10.44
AEDAN -calmodulin 0.0528 0.14 0.55 9.6 11.7
+5 Ca + 0.86 9.71
AEDAN -calmodulin 0.0575 0.13 0.08 9.9 14.9
+5 Ca + + 3 Al 0.87 9.94

 

* the preexponential factors, a1 and a2, and the corresponding
lifetimes,’r1 and ’r2, were analyzed on the basis of a biphasic
decay. The error in lifetimes is i 0.07 ns.
r is the fluorescence anisotropy (error i 0.0004) measured by T-format
method as indicated in legend to Figure 2, except at room temperature.
firav are average lifetimes calculated from the relationship (22)

Tav " (Ell/(‘7)z + 512?}1 )/(317«7 + 3273,)
¢av values are the rotational correlation times calculated from eq.

2.

Figure 3 . l

72

The spectral overlap of the fluorescence emission ( O )
of the single tryptophanyl residue in mastoparan-X
(donor), associated with spinach calmodulin, and the
absorption ( I ) of AEDANS- conjugated cysteine
(acceptor). The samples were dissolved in 4-morpholino-
propane sulfonic acid (Mops) buffer (10 mM, pH 6.5, 0.1 M
KCl and 0.15 mM CaClz). The concentration of
mastoparan-X was 20 mM, while that of the spinach
calmodulin was 30,pM. The concentration of AEDANS-
conjugated cysteine was 9.3,uM. Tryptophan was excited at
299 nm, and both experiments were conducted at room

temperature.

73

325.3053 3

0 a. 8. 7. E 5
L 0 0 O 0 0

 

 

1 d d d a q d H d H d -

 

400

rT'11—F-r-'O.2
440

3201 1 t 380'
Wavelength (nm)

V t r

 

 

05-
.4
3
2

O Q 0.

any 0055.834

280

74

distance, R , can be determined (Table 3.2). The Fdrster critical

0
distance, R0, is in accord with that reported for the radiationless
energy transfer from Trp-l9 of melittin to AEDANS-labelled troponin C
at Cys-98, viz., Ro - 23.6 A0 (29). Since the spectral character-
istics of donor emission and acceptor absorption did not change when
aluminum ions were added to the calcium-containing sample, the overlap
integral and therefore Ro remained the same (Table 3.2). The
efficiency of energy transfer between donor and acceptor was higher
when aluminum was present in the sample, resulting in a distance of
29.4 A0. In the absence of aluminum, this distance was 30.9 A0

(Table 3.2). Mastoparan-X probably binds to calmodulin‘s region III,
when calcium is present (30). Therefore a distance of about 30 Ao
between donor and acceptor appears to be reasonable,considering the
location of AEDANS at Cys-26 in region I.

To further evaluate aluminum-related dynamic changes of spinach
calmodulin, the influence of temperature on the fluorescence anisotropy
of AEDANS-conjugated spinach calmodulin was determined (Fig. 3.2). The
anisotropy decreased as the temperature increased; however, in the
presence of aluminum ions, the anisotropy value was higher at all
temperatures tested. Moreover, temperature-induced structural changes
of calmodulin are also reflected in molecular parameters representative
for those of the protein (Table 3.3). The molecular parameters were
evaluated at two temperatures. At 18.2°C, the rotational correlation
time, 29.1 ns, measured for AEDANS-labelled spinach calmodulin in the
presence of calcium, is in general accord with values measured at

different wavelengths for bovine calmodulin (31). At 18.2°C, the

75

 

 

Table 3.2. Energy transfer characteristics of AEDANS-labelled
spinach calmodulin in the absence and in the presence
of aluminum ions

sample energy transfer overlap in- R0 R

efficiency (E) tegral (JAB) (A°) (1°)

calmodplin 0.194 i 0.014 6.50 x 1013 23.9 30.3

+ 5 Ca +

calm dulin 0.219 t 0.009 6.50 x 1013 23.9 29.4

+5Ca ++ 3A1

 

Ro is the Farster critical distance, as calculated from eq. 6. R is

the distance between donor and acceptor molecule, as calculated from

eq. 5. The experiments were performed at room temperature.

76

 

 

Table 3.3. Molecular parameters of AEDANS-labelled spinach
calmodulin

sample temperature anisotropy 2fav ¢av

(6°) (n8) (n8)
calmodulin +50a2+ 18.2 0.0571 1 0.0003 17.3 29.1
28.5 0.0486 i 0.0003 8.5 9.7
calmodulin 18.2 0.0616 i 0.0003 12.0 25.1
+ 5 032+ + 3 A1 28.5 0.0525 1 0.0003 9.2 12.5

 

Symbols equivalent to those in Table 3.1

 

 

 

Figure 3.2.

77

Fluorescence anisotropy values of AEDANS-labelled
spinach calmodulin versus temperature, in the
presence and in the absence of aluminum ions. The
experiments were performed according to the T-format
method on a spectrofluorimeter, model SLM 4800. Two
identical Corning cut-off filters, 08-72, were
placed in the emission beams. The SLM instrument was
interfaced with a microcomputer, HP-85, for data
analysis. Both samples were excited at 344 nm.
Calmodulin concentration was 10 M. (I),
[calmodulin] : [Ca2+] -1:5; ( O) , [calmodulin] :

[Ca2+] : [aluminum] - 1:5:3.

Anisotropy

78

 

0.07 '-

0.06

0.05

0.04

 

 

 

 

0.03
1 0

2'0 ' :50 40
Temperature (°C)

50

79

rotational correlation time is faster, when aluminum is present,
compared with that in the absence of aluminum. At 28.500, however,

this time pattern is reversed (Table 3.3).

Discussion

X-ray diffraction methods revealed that the equilibrium average
structure of calmodulin consists of a dumbbell-shaped protein in which
the two lobes are interconnected by an eight-turn-helix (32). The
C-terminal lobe of calmodulin harbors the two high- affinity calcium
binding sites (10). Like other globular proteins, calmodulin is also a
dynamic system which comprises a multitude of conformational substates,
easily accessible by rapid structural fluctuations (33, 34). These
conformational substates are in part dependent on the viscosity of the
solvent (35). Our understanding of the molecular dynamics of proteins
is in its infancy, however, evidence is accumulating that at least some
of these conformational substates are crucial for protein function (34,
36).

Indeed, we have demonstrated in this article that aluminum ions
perturb the conformation of calmodulin existing in the presence of
calcium ions. This is illustrated by our findings that application of
aluminum ions to [calcium]/[calmodulin] - (5:1) alters the fluorescence
probe's motional properties different from those with aluminum absent.
Moreover, the probe seems to reside in a hydrophobic area which is not

influenced by the binding of calcium or aluminum. Our data on the

80

influence of temperature on the fluorescence anisotropy suggest that
application of aluminum to spinach calmodulin produces a calmodulin
structure different from that in the absence of aluminum, especially at
the lower temperatures measured. Since aluminum application breaks
hydrogen bonds in calmodulin (11, 15), transitions between certain
energy minima of the protein are facilitated which -- in the absence of
aluminum -- would be less accessible. In the plot of the fluorescence
anisotropy versus temperature, changes may reflect rearrangements in
the solid- and liquid-like microdomain structure of the protein (36).
This kind of intramolecular reorganization in aluminum-altered
calmodulin is probably responsible for the protein‘s inability of

stimulating enzymes such as cyclic nucleotide phosphodiesterase (3).

Acknowledgement

This research was supported by a grant from the National Institutes of

Health, No. 1 R01 ESO4468-01. The authors wish to thank Dr. K. Poff,

Plant Research Laboratory, for use of the SLM instrument.

 

 

10.

ll.

12.

13.

14.

81

LIST OF REFERENCES

Hartwell, B.L., and Pember, F.R., (1918) The presence of aluminum
as a reason for the difference in the effect of so-called acid soil
on barley and rye. Soil Sci. 6: 259-279.

Foy, C.D., Chaney, R.L., and White, M.C. (1978) The physiology and
metal toxicity in plants. Ann. Rev. Plant Physiol. 29: 511-566.

Haug, A, (1984) Molecular aspects of aluminum toxicity. CRC Crit.
Rev. Plant Sci. 1: 345-373.

Stumm, M., Morgan, J.J., and Schnoor, J.L., (1983) Saurer Regen,
eine Folge der Storung hydrogeochemischer Kreislaufe. Naturwiss.
70: 216-223.

 

Evans, L.S., (1982) Biological effects of acidity in precipitation
on vegetation: A review. Environ. Exper. Bot. 22: 155-169.

Ganrot, P.D., (1986) Metabolism and possible health effects of
aluminum. Environ. Health Perspect. 65: 363-441.

 

Vierstra, R., and Haug, A., (1978) The effects of A13+ on the
physical properties of membrane lipids in Thermoplasma
ncidophilum. Biochem. Biophys. Res. Comm. 84: 138-144.

Suhayda, C.G., and Haug, A., (1986) Organic acids reduce aluminum
toxicity in maize root membranes. Physiol. Plant. 68: 189-195.

Deleers, M., Servais, J.P., and Wulfert, E., (1986) Neurotoxic
cations induce membrane rigidification and membrane fusion at
micromolar concentrations. Biochim. Biophys. Acta 855: 271-276.

Klee, C.B., and Vanaman, T.G., (1982) Calmodulin. Adv. Protein
Chem. 35: 213-321.

Siegel, N., and Haug, A., (1983) Aluminum interaction with
calmodulin: evidence for altered structure and function from
optical and enzymatic studies. Biochim. Biophys. Acta 744: 35-45.

Siegel, N., and Haug, A., (1983) Calmodulin-dependent formation of
membrane potential in barley root plasma membrane vesicles: A
biochemical model of aluminum toxicity in plants. Physiol. Plant.
59: 285-291.

Siegel, N., and Haug, A., (1983) Aluminum-induced inhibition of
calmodulin-regulated phosphodiesterase activity: enzymatic and
optical studies. Inorgan. Chim. Acta 79: 230-231.

Suhayda, C.G., and Haug, A., (1985) Citrate chelation as a
potential mechanism against aluminum toxicity in cells: the role of

15.

16.

17.

18.

19.

20.

21.

22.
23.

24.

25.

26.

27.

28.

82

calmodulin. Canad. J. Biochem. Cell Biol. 63: 1167-1175.

Weis, C., and Haug, A., (1987) Aluminum-induced conformational
changes in calmodulin alter the dynamics of interaction with
melittin. Arch. Biochem. Biophys. 254: 304-312.

Cox, J.A., Comte, M., Fitton, J.E., and DeGrado, W.F., (1985) The
interaction of calmodulin with amphiphilic peptides. J. Biol. Chem.
260: 2527-2534.

 

Caldwell, G.R., and Haug, A., (1981) Affinity chromatographic
isolation of calmodulin from bovine brain acetone powder. Anal.
Biochemistry 116: 325-330.

 

Yuan, S., and Haug, A., (1988) Ligand-triggered conformational
perturbations elicit changes at the single cysteinyl residue of
spinach calmodulin. Eur. J. Biochem. 175: 119-124.

Wu, C.W., and Stryer, L., (1972) Proximity relationships in
rhodopsin. Proc. Natl. Acad. Sci. (U.S.) 69: 1104-1108.

Hudson, E.N., and Weber, G., (1973) Synthesis and characterization
of two fluorescent sulfhydryl reagents. Biochemistry 12:
4154-4161.

Bradford, M.M., (1976) A rapid and sensitive method for the
quantitation of microgram quantities of protein utilizing the
principle of protein-dye binding. Anal. Biochem. 72: 248-254.

Lakowicz, J.R., (1984) Prinpiples pf Fluorescence Spectroscopy.

Plenum Press, New York.

Vanaman, T.G., (1983) Chemical approaches to the calmodulin system.
Methods Enzymol. 102: 296-310.

Fairclough, R.F., and Cantor, G.R., (1977) The use of

singlet-singlet energy transfer to study macromolecular assemblies.
Methods Enzymol. 48: 347-379.

Eisinger, J., (1969) Intramolecular energy transfer in adrenocor-
ticotropin. Biochemistry 8: 3902-3915.

Seaman, K.B., (1980) Calcium- and magnesium-dependent
conformational states of calmodulin as determined by NMR.
Biochemistry 19: 207-215).

Steiner, R.F., Marshall, L., and Needleman, D., (1986) The
interaction of melittin with calmodulin and its tryptic fragments.
Arch. Biochem. Biophys. 246: 286-300.

Weast, R.C., (1983) Handbook of Chemistry and Physics, CRC Press,

 

29.

30.

31.

32.

33.

34.

35.

36.

83

Boca Raton, FL, p. D-24.

Steiner, R.F., and Norris, L., (1987) The interaction of melittin
with troponin C. Arch. Biochem. Biophys. 254: 342-352.

Muchmore, D.C., Malencik, D.A., and Anderson, S.R., (1986) +H NMR

studies of mastoparan binding by calmodulin. Biochem. Biophys. Res.
Commun. 137: 1069-1078.

LaPorte, D.C., Keller, C.H., Olwin, B.B., and Storm, D.R., (1981)
Preparation of a fluorescent-labeled derivative of calmodulin which

retains its affinity for calmodulin binding proteins. Biochemistry
20: 3965-3972.

Babu, Y.S., Sack, J.S., Greenhough, T.J., Bugg, C.E., Means, A.R.,
and Cook, W.J., (1985) Three-dimensional structure of calmodulin.
Nature (London) 315: 37-40.

Careri, G., (1984) Order nng Qigornpr in Mnrter. The Benjamin/
Cummings Publ., Menlo Park, CA.

Karplus, M., and McCammon, J.A., (1981) The internal dynamics of
globular proteins. CRC Crit. Rev. Biochem. 9: 293-349.

Beece, D., Eisenstein, L., Frauenfelder, H., Good, D., Marden,
M.C., Reinisch, L., Reynolds, A.H., Sorensen, L.B., and Yue, K.T.,

(1980) Solvent viscosity and protein dynamics. Biochemistry 19:
5147-5157.

Elber, R., and Karplus, M., (1987) Multiple conformational states

of proteins: A molecular dynamics analysis of myglobin. Science
235: 318-321.

 

 

 

CHAPTER IV

FRICTIONAL RESISTANCE TO MOTIONS OF BIMANE-LABELLED SPINACH CALMODULIN

IN RESPONSE TO LIGAND BINDING

 

Shixing Yuan and Alfred Haug

 

FEBS Lett. 234: 218-223.

1988

84

85

The single cysteinyl residue 26 of spinach calmodulin was labelled
with the thiol-specific bimane fluorescence probe. Following

application of stoichiometric quantities of Ca2+ or aluminum ions to

the protein, temperature-dependent fluorescence changes (anisotropy,
lifetime) could be monitored via the label. From these data the
Y-function could be constructed which, as a function of temperature,
seems to consist of two linear regions which intersect at the critical
temperature, Tc. From the Y-function the thermal coefficient, b(T),

of the frictional resistance to fluorophore rotation could be
determined. b(T) was dependent on the type and stoichiometry of
1igand(s) bound to calmodulin. Changes of the thermal coefficient
apparently resulted in part from ligand-triggered structural

perturbations transmitted over a considerable distance to calmodulin

region I, the site of the fluorophore.

 

 

 

86

Introduction

Calmodulin is a small (Mr-16,000-17,000), acidic protein, which
mediates Ca2+-dependent regulatory processes in cells. An intriguing
feature of calmodulin is its capability to interact with and stimulate
numerous specific enzymes (1). The biochemically active form appears to
be (Ca2+)4-calmodulin which has a pronounced tendency to bind to a
basic amphiphilic-helix in a variety of target peptides (2). The
protein's intrinsic flexibility probably contributes to its versatility
(3). Thanks to its flexibility and associated rapid local motions,
calmodulin can apparently control and thus correlate molecular events
(4, 5), e.g., those triggered by specific ligands bound to specific
locations. These motions and associated conformational changes are
thought to be crucial for its biochemical activity (5).

To analyze the conformational changes, a small fluorophore attached
to the protein may be used. For example, by measuring the temperature
dependence of fluorescence polarization and lifetime of the probe, the
Y-function can be established (6). This function describes the
frictional resistance associated with conformational changes to the
probe's rotational motions.

Conformational changes at a defined locus on calmodulin may be the
result of transfer of structural perturbations initially generated by
ligands bound to protein regions distant from or close to the defined
locus (7). Examination of these kinds of ligand-triggered conform-

ational changes via a fluorophore requires studies of the relative

motions of calmodulin's constituent regions.

 

 

87

To monitor ligand-triggered conformational changes through a
fluorophore, we therefore selected spinach calmodulin which harbors a
single cysteinyl residue at site 26, i.e., region I of the polypeptide
chain (7, 8). Our results indicate that through a strategically
anchored fluorophore, conformational changes could be determined in

response to ligand-triggered structural perturbations.

Materials and Methods

The fluorescence probe monobromotrimethylammoniobimane was
purchased from Calbiochem Company (San Diego, CA) under the commercial

name of thiolyte MQ. Phenyl-Sephrose CL-4B was obtained from Pharmacia

Company (Uppsala, Sweden).

Purification of spinach calmodulin

Spinach calmodulin was purified from fresh spinach leaves according
to a method described (7); a slight modification was made by replacing
phenothiazine Affigel with Phenyl-Sepharose CL-4B. The purity of the
protein was judged by the characteristic ultraviolet absorption
spectrum and the absence of tryptophan fluorescence. Moreover,

following electrophoresis on sodium dodecyl sulfate gels, only one band

was observed (7).

 

 

88

Labeling of spinach calmodulin

Freshly isolated calmodulin was labelled with a fluorescence probe,
thiolyte MQ. This probe is highly reactive with reduced thiol groups.
The probe concentration was evaluated by measuring its absorbance
(£5378 - 5700 M cm'l) (9). The protein concentration was determined

by the Bradford assay (10). The molar ratio of [label]/[calmodulin] was
found to be unity (7).'

Fluorescence studies of labelled calmodulin

Fluorescence experiments were performed on an SLM spectrofluori-
meter, model 4800, (Urbana, IL), which was interfaced with a Hewlett-
Packard desk top computer, HP-85, and a plotter to aid in data
acquisition and analysis. The steady-state fluorescence anisotropy, r,
was measured by using the T-format method. Two Corning filters, CS-72,
were placed in the emission pathways. The excitation wavelenth was at

385 nm. Fluorescence lifetimes were determined according to the phase

shift and demodulation methods (11).

2.4. Analysis of the data

The Y-function is dependent on the thermal coefficient of
frictional resistance, b(T), to rotations of the fluorophore [6]. If a
fraction, f1, of the protein molecules is in substate 1, while the
other fraction, f2, occupies substate 2, the corresponding thermal

coefficients are b1 and b2, respectively. The ratio, fz/fl - K,

is the equilibrium constant, and

 

 

89

 

(1)
The Y-function is defined as
Y - lnwo/m - but) x (2 ~ To) ;
- murmur] -1) - luau 7mm (2)

where 0 and 07o are viscosities at temperature T and a reference
temperature, TO (293.10K), respectively. b is the thermal

resistance coefficient. r is the anisotropy measured at temperature, T,
r(0) is the apparent limiting anisotropy (11). For the bimane-labelled
protein, we obtained a value of r(0) - 0.2142 (see Chapter II) from a
Perrin plot (7). V is the molecular volume of the fluorophore probe,
calculated from V-Mr(g/M)/density(g/cm3)-l46.6. ’E’is the

fluorescence lifetime measured, and R is the gas constant. Since the
Y-function is the logarithm of a ratio of viscosities, a low value of
the Y-function is associated with restricted fluorophore rotation

because of a high viscosity.

The standard enthalpy at the critical temperature Tc can be

derived from

AH - [(dY/dT)T -0.5(b1+b2)][4RTc2/tc(b2-b1)] (3)
C

Tc is the critical temperature, where K - 1, and tc - Tc - To.
The critical temperature was obtained from a linear regression analysis

 

90

of the data points. The slope at this temperature was determined after
fitting the experimental data to theoretical curves using a non-linear
regression scheme (Marquardt's algorithm) on an IBM-KT computer. The

correlation coefficient for these kinds of statistical analyses was

better than 0.98.

Results

Fluorescence anisotropy and lifetime studies were performed on
bimane-labelled spinach calmodulin in Mops buffer over the temperature
range 10 - 40°C. From these data corresponding Y-function values
(Fig. 4.1) were calculated using eq. 2.

Upon binding of the first two Ca2+ ions to spinach calmodulin,
the Y-function values increased compared with corresponding values of
apocalmodulin, at all temperatures tested (Fig. 4.1). Following binding
of the first two Ca2+ ions to calmodulin's high affinity binding
sites 111 and IV (1), a perturbational signal was locally elicited,
producing conformational changes which, in turn, can be monitored at
the bimane-labelled region 1. Relative to apocalmodulin, the probe's

rotational motions became apparently less restricted in (Ca2+)2-

calmodulin. In the presence of saturating Ca2+ concentrations, where

Ca2+ also occupies region I, the Y-function values were higher at

temperatures below 2980K, while those above that temperature were

lower, compared with corresponding values derived from apocalmodulin

(Fig. 4.1). Below 298°K and in the presence of saturating Ca2+

 

 

91

concentrations, the probe therefore experienced less rotational
constraint compared with that above this critical temperature. This
temperature dependence is probably a manifestation of conformational
changes occurring at the critical temperature of 2980K. At a molar
ratio of [Al]/[calmodulin] - 3:1, addition of aluminum produced a

change which, in turn, restricted probe rotation more than that

 

observed in apocalmodulin above 2930K. Below this temperature the
rotational characteristics were reversed (Fig. 4.1).

The thermal resistance coefficients, b1 and b2, are listed in
Table 4.1. In the low temperature range, the values of the slopes,

b1, were larger than those in the higher temperature range, b2. All

b1 and b2 values decreased as the Ca2+ concentration of the

solution increased.

Concerning changes of the thermal resistance coefficient (Table
4 1), it appeared that the effective resistance to rotation experienced
by the fluorophore comprises two temperature-dependent contributions,
viz., the protein conformations and the solvent viscosity. It is known

that Ca2+- and Al-binding to calmodulin lead to pronounced conform-
ational changes in the protein (1, 7, 12, 13). Ligand-triggered changes
are also reflected in the enthalpy values (Table 4.1) observed at the
critical temperature, where the protein was being transformed from

substate 1 into substate 2. For apocalmodulin this enthalpy change was

8.0 kcal mol'lK'l, while for (Ca2+)2-calmodulin the enthalpy

value decreased to 2.9 kcal mol'lK‘l. The probe residing in region

I seemingly became more sensitive to temperature changes following

binding of two Ca2+ ions at regions III and IV. Structural

 

92

Fig. 4.1. Y(T) vs. temperature for bimane-labelled spinach calmodulin .
Fluorescence studies were performed on an SLM spectrofluorimeter, model
4800. The labelled calmodulin was dissolved in Mops buffer (10 mM, pH 6.5,

0.1 M KCl) to reach a concentration of 7 pM. ( O ), apocalmodulin; ( O ),

calmodulin + 20a2+; ( I ), calmodulin + 5Ca2+; and ( A ), calmodulin

+ 5Ca2+ + 3A1.

 

 

 

 

 

 

gammaammémp
8a 9a can can a: a...

 

2.” 8o 8.

93

 

Ahv>

 

 

 

94

Table 4.1. Physical parameters* for spinach calmodulin in the

presence of ligands

 

 

calmodulin b1 b2 bl/bZ Tc H A S

apocalmodulin 4.7 3.4 1.4 299 8.0 26.8
CaM + 20a2+ 4.2 3.0 1.1. 297 2.9 9.8
CaM + 5Ca2+ - 3.7 2.5 1.5 298 5.3 17.8
CaM + 5Ca2+ +3A1 2.7 2.2 1.2 299 8.2 27.4

 

* b1 and b2 are listed as percent decrease per degree, Tc is

the critical temperature in °K with an error of i l°K,.¢ H in kcal

per mol,.d S in kcal per mol per degree.

u-- -

95

perturbations at these latter sites produced conformational changes
which could be monitored at the defined site of the fluorophore.
Following further addition of Ca2+ until saturation, one of the
Ca2+ ions binds to Ca2+ binding region 1. Consequently new bonds
are established which render the probe's rotational motions less prone
to temperature changes, as indicated by 23H - 5.3 kcal mol'lK'l.
When aluminum (3:1) is present, the probe's motions appear to be less
disposed to temperature changes as indicated by A.H - 8.2 kcal mol"1

K'l. These Al-related changes of motions may be a consequence of

conformational changes monitored at the probe's residing site.

Discussion

Regarding calmodulin's multifunctional roles (1), the inherent
flexibility of the polypeptide chain, in the presence and absence of
Ca2+, is apparently of great import (3). This flexibility is in part
determined by rotational barriers generated by bonds such as those of
peptide linkages and hydrogen bonds in helix structures. Like that of
some other proteins, calmodulin's flexible structure rapidly fluctuates
around its equilibrium configuration in solution. Associated local
motions are generally thought to be crucial for biochemical activities
(4, 5). Molecular dynamics analysis of proteins indicated that there
exist a multitude of conformational substates which are accessible by

rapid structural fluctuations of the protein (14). When triggered

locally by the first two Ca2+ ions, bound to the carboxyl terminus

 

 

96

lobe of the calmodulin dumbbell (15), conformational changes can be
monitored through the strategically anchored fluorophore. These may
originate from changes of segmental motions and/or global protein
motions which, in turn, are produced by ligand-triggered perturbations
of local protein structure. Therefore, these types of perturbations
become manifest in changes of physical properties of fluorophore
rotation.

We have also shown that spinach calmodulin, both in its apoform and
in presence of metals, undergoes a substate change at a temperature of
25°C. This temperature value is significantly below the major thermal
unfolding of the protein, viz., at 55°C for apocalmodulin and at >
90°C for calmodulin in the presence of Ca2+ (16). In biochemically
active calmodulin, i.e., (Ca2+)4-calmodulin generated in the
presence of saturating Ca2+ (Table 4.1), the enthalpies lie below
those of calmodulin which is biochemically inactive (apocalmodulin) or
whose regulatory capacity has been greatly diminished by aluminum ions
(13). Furthermore, in (Ca2+)4-ca1modulin the lower enthalpy seems
to reflect the sensitive dynamic aspects of calmodulin, essential for
the transfer of information within the macromolecule in response to
regulatory stimuli. Again in biochemically active calmodulin, the
Ca2+-protein interaction characteristics probably remain intact (17)
in both structures associated with the substates below and above the
critical temperature. The temperature-dependent changes observed in
apocalmodulin versus (Ca2+)4-calmodulin (Fig. 4.1) may be related
in part to their respective degree of hydration. This is illustrated,

for example, in studies indicating that water molecules are removed at

 

 

9
as
I

97

each coordination site of Ca2+ (18). Furthermore, in biochemically

2+-induced exposure of hydrophobic surface was

active calmodulin a Ca
observed (19). Similarly, concomitant with the breakage of hydrogen
bonds (13), changes in solvation structure are to be expected upon
application of aluminum ions to (Ca2+)4-calmodulin.

2+-dependent

In summary, thanks to its finely tuned, Ca
flexibility, calmodulin is capabable of fulfilling regulatory

functions.

Acknowledgements

The work was supported in part by grant l-ROl-ESO4468-01 from the
National Institutes of Health. The authors thank Dr. Ken Poff, Plant

Research Laboratory, for use of the SLM spectrophotometer.

 

 

'3";
‘

10.

ll.

12.

13.

14.

98

LIST OF REFERENCES

Klee, C.B., and Vanaman, T. C., (1982) Calmodulin. Adv. Protein
Chem. 35: 213-321.

O'Neil, K.T., Wolfe, H.R., Erickson-Vittanen, S., and DeGrado,
W.F., (1987) Fluorescence properties of calmodulin-binding peptides
reflect alfa-helical periodicity. Science 236: 1454-1456.

Lambooy, P.K., Steiner, R.F., and Sternberg, H., (1982) Molecular
dynamics of calmodulin as monitored by fluorescence anisotropy.
Arch. Biochem. Biophys. 217: 517-528.

Karplus, M., and McCammon, J. A., (1981) The internal dynamics of
globular proteins. CRC Crit. Rev. Biochem. 9: 293-349.

Careri, G., (1984) Order and Disorder in Matter. The Benjamin/
Cummings Publ., Menlo Park, CA.

Scarlata, S., Rholam, M., and Weber, G., (1984) Frictional
resistance to local rotations of aromatic fluorophores in some
small peptides. Biochemistry 23: 6789-6792.

Yuan, S., and Haug, A.,(1988) Ligand-triggered conformational
perturbations elicit changes at the single cysteinyl residue of
spinach calmodulin. Eur. J. Biochem. 175: 119-124.

Vanaman, T.G.,(l983) Chemical approaches to the calmodulin system.
Methods Enzymol. 102: 296-310.

Kosower,N.S., Kosower, E.M., Newton, C.L., and Raney, H.M., (1979)
Bimane fluorescent labels: labeling of normal hyman red cells under
physiological conditions. Proc. Natl. Acad. Sci. (U.S.) 76:
3382-3386.

Bradford, M.M., (1976) A rapid and sensitive method for the
quatitation of microgram quantities of protein ulitizing the
principle of protein-dye binding. Anal. Biochem. 72: 248-254.

Lakowicz, J.R., (1984) P inci 1e 0 uo scence S ectrosco ,

Plenum Press, New York.

Weis, C., and Haug, A., (1987) Aluminum-induced conformational
changes in calmodulin alter the dynamics of interaction with
melittin. Arch. Biochem. Biophys. 254: 304-312.

Siegel, N., and Haug, A., (1983) Aluminum interaction with
calmodulin: evidence for altered structure and function from
optical and enzymatic studies. Biochim. Biophys. Acta 744: 36-45.

Elber, R., and Karplus, M., (1987) Multiple conformational states

it.

 

15.

16.

17.

18.

19.

99

of proteins: a molecular dynamics analysis of myogolobin. Science
235: 318-321.

Babu, Y.S., Sack, J.S., Greenhough, T.J., Bugg, C.E., Means, A.R.,

and Cook, W.J., (1985) Three-dimensional structure of calmodulin.
Nature 315: 37-40.

Magpin, S.R.,and Bayley, P.M.,(l986) The effects of Ca2+ and
Cd

on the secondary and tertiary strcture of bovine testis
calmodulin. Biochem. J. 238: 485-490.

Permyakov, E.A., Shnyrov, V.L., Kalinichenko, L.P., and Orlov,

N.Y., (1985) Effects of cation binding on the termal transition
calmodulin. Biochim. Biophys. Acta 830: 288-295.

s in
Mulqueen, P., Tingey, J.M., and Horrocks, W.DeW., (1985)
Characterization of lanthanide (111) ion binding to calmodulin
using luminescence spectroscopy. Biochemistry 24: 6639-6645.

LaPorte, D.C., Wierman, B.M., and Storm, D.R., (1980)

Calcium-induced exposure of a hydrophobic surface on calmodulin.
Biochemistry 19: 3814-3819.

 

 

 

CHAPTERV

SUMMARY

 

Shixing Yuan

 

100

101

The data presented in this dissertation indicate that aluminum can

drastically change the conformation of the calmodulin molecule, and

this, in turn, can interfere with the interactions of the molecule and

its target enzymes. These results may imply that calmodulin is a

potential target for deleterious aluminum ions and the aluminum

intoxication may result from, at least partly, from this disturbance.

We demonstrated

(a)

(b)

When calmodulin interacts with specific ligands, such as cations,
peptides, enzymes, molecular modifications of the interfacial area
between the protein and its ligand can trigger changes in whole
molecular motion, which can be monitored through a specific probe
attached to the single cysteinyl residue of spinach calmodulin. In
addition, this probe can also sense microenvironmental changes in
response to ligands bound distant from the probe location binding.
These latter changes are somehow elicited at the probe's site
following transmission of information generated by ligand binding,
e.g., mastoparan, to the protein (Fig. 5.1). Studies on
ligand-related processes of structural changes of calmodulin
appear to provide means to demonstrate deleterious effects of
aluminum ions on the dynamic structure of the regulatory protein.
When aluminum ions bind to calmodulin at a molar ratio of [A1]:
[calmodulin] - 3:1, the whole molecular rotational motion becomes
slower. This slowdown may result from the less compact structure
of Al-altered calmodulin compared with that of Ca2+4-

calmodulin (Fig. 5.2). In a looser molecular structure more

surface area, as evidenced by an enhanced hydrophobicity, is

 

 

 

102

Fig. 5.1. An illustration of microenvironmental changes taking place in

the vicinity around the EPR probe attached at Cys-26.

 

 

 

 

 

 

 

103

+ 2Ca2+

[A + mastOparan

V

 

 

 

 

Saturated with
Ca2+

 

 

 

 

 

 

 

 

 

 

104

Fig. 5.2. An illustration of the whole molecular conformational changes

of spinach calmodulin in response to ligand binding.

 

105

(5'33.
..

 

Saturated with
2+

® +. 2Ca2+
Ca
r519 .
a 9e .

 

 

 

 

106

exposed, and more friction is encountered by molecular rotations.
An Al-triggered enhancement of the protein's hydrophobic area is
expected to disturb the normal interaction of the molecule with
its target enzymes or peptides.

(c) The temperature-controlled interchange of molecular substates may
be the result of breakage or reunion of some hydrogen bonds. These k

events can be critical for the molecule to sense the outside

 

environment. Following binding of aluminum ions to the protein,
temperature control becomes less sensitive, more energy is needed
for the interchange. In such an Al-altered protein conformation --
similar to denaturation -- it seems to be more difficult for the

molecule to adjust itself to environmental fluctuations.

 

Taken together, aluminum may exert its toxicity by interacting
with the calmodulin molecule. By changing the molecule's conformation,
aluminum ions can disturb whole calmodulin-related regulatory
processes. Since Ca is so involved in the regulatory processes,
(including phosphatidyl innositol pathway and calmodulin-related
processes) this disturbance can probably interfere with the calcium

hemostasis and cause some serious toxic problems in living organisms.

 

APPENDIX

THE OCCURRENCE AND CORRELATION BETWEEN CALMODULIN IN RAINBOW TROUT

mucus AND LEVELS OF pH, Ca AND Al

T.E. Lewis, S. Yuan

A. Haug and D.R. Mount

107

 

108

An experiment was performed to determine whether calmodulin, a
calcium-regulating protein, may be used as an indicator for
biologically active aluminum in lakes impacted by acidic deposition.
Calmodulin has been shown to selectively bind inorganic monomeric A1.
The presence of calmodulin in fish gills and mucus has been documented
The binding of Al to calmodulin causes a change in the conformation of
the protein, rendering it ineffective in mediating many of the normal
enzyme activities which require calcium. Rainbow trout were exposed to
chronic, low-levels of A1 for 147 days. A significant correlation was

found between the amount of calmodulin in the mucus and aluminum

concentrations in the test solutions. Calcium levels did not affect

calmodulin levels. The use of calmodulin as an indicator of

biologically relevant Al in surface waters is proposed.

 

 

109

Introduction

As a result of increasing acidification in various watersheds
elevated levels of aluminum have been observed in soil and surface
water (1, 2, 3, 4). In both field and controlled laboratory experiments
aluminum has been shown to be highly toxic to fish (5, 6, 7, 8, 9). The
toxicity of A1 to fish has been shown to be positively correlated with
the concentration of inorganic monomeric A1 (5, 6, 10). The exact
mechanism(s) of A1 toxicity is not fully understood. It appears that
the gill is the target organ for the toxic action of Al. 0f the organs
analyzed in smallmouth bass (Micrpprerng nolonieui) exposed to high Al
levels, the gill filaments had the highest concentration of A1 (11).
Visible symptoms of A1 toxicity include coughing response, hyper-
venilation, and excessive mucus clogging of the gills (10).
Accompanying these visible symptoms are more specific responses such as
rapid loss of sodium and chloride from plasma and lowered blood oxygen
tension (10). At the cellular level the toxic mode of action remains
elusive. Cleveland et al. (12) observed a reduction in RNA synthesis
and RNA/DNA ratios in fish exposed to pH 5.5 in the presence of Al.
This observation may be indicative of disruption or deactivation of
metabolic pathways. Reduced carbonic anhydrase and Na+-K+-ATPase
activity in gills of salmonids exposed to Al-containing waters has been
reported (13).

Recently, the presence of calmodulin (CaM), a calcium-regulating

protein, has been reported in fish gills and mucus (14, 15). Calmodulin

selectively binds inorganic monomeric A1 causing conformational changes

 

 

110

in the protein (16,17). Aluminum-induced conformational changes cause a
reduction in the ability of calmodulin to mediate Ca-dependent
phosphodiesterase and ATPase activity (16,18). Calmodulin also plays a
key role in coordinating the effects of secondary messenger systems in
response to cellular stimulation (19). Given the involvement of
calmodulin in numerous biochemical pathways, its interaction with

aluminum may be a key lesion in the broadly defined syndrome of
aluminum toxicity (16).

 

The present study was undertaken to establish a relationship
between A1 concentration in aqueous solution and the presence and
activity of CaM in the mucus of adult rainbow trout. Fish were exposed

to various levels of pH, Ca, and A1. Mucus was collected and the amount

 

of CaM was determined. The ability of the Al-exposed CaM to activate

the phosphodiesterase enzyme system was also evaluated.

Materials and Methods

Exposure Conditions

Test water was artificial soft water, reconstituted from deionized
water (U.S. EPA, 1979). Well water was treated sequentially by iron
oxidation/filtration, acidification (to break carbonate complexes),
reverse-osmosis, forced-draft deaeration and deionization. This water

was delivered to a 700 L fiberglass headtank, where well water and NaCl

solution were added to a Ca concentration of 0.5 mg/L and a Na

111

concentration of 1.3 mg/L. An automatic titration unit (Leeds and
Northrup Model) maintained this water at pH 6.5 by addition of dilute
sulfuric acid or potassium hydroxide. Two 1 horsepower water chillers
(Frigid Units, Inc.) thoroughly mixed and aerated the water. After
reconstitution, water was delivered to each of three 180 L fiberglass
headtanks, one for each experimental pH. Here, the water was adjusted

to nominal pH by automatic titrator using dilute acid, and was

 

thoroughly mixed with a recirculating pump (5 gpm, March, Inc.). Eight
combinations of pH, Ca, and Al were achieved addition of Ca and Al
stock solutions (both as chlorides) from Marriotte bottles into
continuous-flow serial dilution systems, Test chambers were 700 L round
fiberglass tanks, each receiving 5.5 L/min. of test water (11.3 volume
additions/day). Each chamber received moderate aeration. The test

temperature was maintained at 9°C.

Test Organism

Test fish were 2-1/2 yr. old rainbow trout (Salmp gairdneri)
(Belaire strain) obtained from the Colorado State Fish Hatchery in
Crystal River, CO on 15 April, 1986. Shortly after arrival at the
University of Wyoming laboratory, some fish began to show signs of
"caudal peduncle" or “coldwater” disease (Cyrpphaga psychrophilia).
Treatment with antibiotic food (Romet 30) was initiated. Visibly
affected fish were destroyed. The remaining fish responded favorably to
treatment and were held without recurrence of the disease. Fish were

arbitrarily assigned to the eight exposure chambers (18 fish/chamber).

 

112

From 3 to 9 July, the culture water was gradually changed from well
water to artificial soft water (pH 6.5, 2 mg/L Ca, 1.3 mg/L Na) and
acclimation to this water continued until August. Over 11 to 18 August,
pH and Ca concentration in all chambers was gradually reduced to
nominal. Aluminum addition began on 18 August; this was designated as
day 0 of exposure. Mean weight at the start of exposure was 1089 g.

Mortality, pH, temperature, and stock solution delivery were
monitored daily. dissolved oxygen and ammonia were checked periodically
and maintained within acceptable limits. Floating trout chow (Purina)
was fed once daily at 0.8% body weight/day, and all tanks were cleaned
daily. One to two times per week, samples were taken from all tanks for
chemical analyses. Cations (Al, Ca, Na, Mg, K) were measured by atomic
absorption spectrometry (Perkin-Elmer) using flame or graphite furnace
techniques Standard methods (1982) protocols were employed. Aluminum
was fractionated into total, reactive, and inorganic forms after the
method of LaZerte (20). Anions (Cl, F, N032', SO32") were

determined by ion chromatography.
Sample Collection

Mucus samples were collected on day 147 of exposure (12 January
1987), after the method of Flik et a1. (15). Fish were gently netted
from the tank and anesthetized with MS-222 (Sigma Chemical Co.) in
ambient tank water which was titrated back to nominal pH with KOH. Once
anesthetized, each fish was grasped by the lower jaw and held over a

3000 mL beaker. Then the body surface was rinsed with 30 to 50 mL of

 

 

 

 

113

buffer solution. The buffer solution consisted of 0.0137 M Tris, 0.12 M
NaCl, and 0.003 M KCl (pH 7.4). Samples were transferred to poly-
propylene centrifuge bottles, chilled to 4°C, and shipped on ice to

Michigan State University for CaM determinations.

Calmodulin Assays

A calmodulin [1251]-radioimmunoassay kit (New England Nuclear
00., Boston, MA) was used for CaM quantitation in fish mucus samples.
Phosphotase and activator-free 3':5'-cyclic nucleotide-phosphodiester-

ase were obtained from Sigma Chemical Co. (St. Louis, MO). Protein

assay reagents were purchased from Bio-Rad Laboratories (Richmond, CA).

All other chemicals were of the highest quality available.

Mucus samples in solution were centrifuged at 1000 x g for 10 min
at 4°C. The protein content of the mucus in the centrifugate was
measured according to Bradford (21). Bovine gamma globulin was used as
a standard material. Stimulation of cyclic AMP-dependent phospho-
diesterase activity by CaM was determined by a one-step procedure (22)
involving two enzymes, viz., phosphodiesterase and alkaline phosphotase
incubated together to liberate inorganic phosphate in a single step.
The phosphate released by the enzyme was quantitated by the Malachite
green method (23). The centrifugate was then incubated at 100°C for 5
min. Calmodulin content in the incubated sample was determined as
outlined above. Preheated bovine brain CaM served as a standard.
Calcium-binding proteins such as troponin C and parvalbumin have been

shown not to react in the calmodulin assay procedure.

 

qw

‘12.“.x2. - _

 

114

Statistical Analysis

A balanced analysis of variance was performed on the data to
determine whether there was a significant relationship between the
amount of CaM and levels of pH, Ca, and Al. An ANOVA procedure found in
the statistical software package Statgraf was employed for this

purpose. The data were separated into three subsets for ANOVA analysis.

 

The first subset consisted of two levels of Ca with Al and pH fixed.
The second subset consisted of three pH levels with Ca and A1 fixed.
The final subset had two levels of A1 with Ca and pH fixed. Due to
small experimental design not enough degrees of freedom were available

to assess component interactions. 1

Results

Table A.l summarizes the results of the experiment. There are some
general trends apparent in the data. The CaM content per mg of total
protein in mucus collected from fish exposed to aluminum stress is
generally greater than that from nonstressed fish. In mucus derived
from Al-stressed fish, the total protein concentration was found to be
similar to that of mucus from nonstressed fish. In addition, there
appears to be a pH dependence of the average CaM content, as
illustrated by the RIA values at pH 6.5 compared to negligible values
obtained at pH 4.8 and 5.2. These trends were tested statistically. The

results of ANOVA procedures are presented in Table A.2. The effects of

115

Table A.l. Results of CaM [lzsll-radioimmunoassay performed on
mucus collected from rainbow trout held a various pH, Ca,
and Al levels for 147 days.

 

effects of Ca at fixed pH and Al

 

pH Ca Al CaM phosphodiesterase
sample (units) (mg/L) (pg/L) (pg/g) Pi release (pM/mg.hr)
1 5.2 2 30 n.d. n.d.
2 5.2 2 30 1.65 12.2
3 5.2 2 30 n.d. n.d.
4 5.2 2 30 0.43 7.1
5 5.2 4 30 0.67 3.9
6 5.2 4 30 0.68 5.8 I
7 5.2 4 30 1.14 4.3
8 5.2 4 30 0.61 n.d.

 

effects of pH at fixed Ca and Al

10
ll
12
13
14
15
l6
17
18
19
20
21
22
23

e.g.p.a.3:a.a.a.c.

axaxoxasmmmmu‘J-‘SML‘P-P
USNJF‘C>'

Uta!usuauahanahahsa>a>a>a>a>
hon)R>hahah>h>hohohoh>hohoha
OOOOOOOOOOOOOO
55.053539555505555
OOOONPPFPPPPPP
scamwFPPPPPPPF

0-P*£‘h‘u)

 

effects of A1 at fixed pH and Ca

30 n.d.
30 12.2
30

30

¢~C>h3ha
C:- u:-

7.1
16
17
18
19

y1U1uaU1Ua01usu:
h>n>h>h>h>hahaks
h>h>h>hah>hansns
C1c>c>c>
P ? F S C): ha:
p-p.a.o.e-aumii
5 3 5 5
PIG-Ong-

 

Note: n.d. - not detectable

 

Table A.2.

Results of ANOVA procedure to test for relationship
between Ca, pH and Al and the amount of CaM in fish

mucus .

 

response variable: CaM content in fish mucus

 

 

source of sum of degrees of mean probability
variation squares freedom square F-ratio (>F)
total 2.13635 7

Ca 0.13005 1 0.13005 0.389 0.56221
error 2.00630 6 0.33438

total 0.30054 13

pH 0.08578 2 0.04289 2.197 0.15749
error 0.21476 11 0.01952

 

 

117

pH and Ca on CaM content in the mucus were not significant. pH had a
slightly larger F statistic and probability than Ca. The Al/CaM
interaction could not be determined statistically due to the lack of
variance in the 0 pg Al/L treatment, i.e., no CaM was detected in any
of the four replicates. This finding significant, irrespective of
statistical evidence. At constant pH and Ca levels the presence of 30
‘pg Al/L in solution resulted in the secretion of CaM into the fish
mucus. In the absence of A1 at the same pH and Ca levels no CaM was
detected.

The presence of CaM in fish mucus is further supported by data
derived from experiments on the ability of mucus to stimulate CaM-free
cyclic nucleotide phosphodiesterase activity (Table A.l). The
CaM-related stimulatory effect of the mucus was
concentration-dependent, and was therefore more pronounced using mucus
collected from Al-stressed fish. A decrease in phosphodiesterase
activity is expected when Al binds to CaM (16). However the Al:CaM
complex must be intact to observe a decrease in activity. The mucus
collected from a single surviving fish exposed to an Al level of 70
‘pg/L was analyzed for A1 by graphite furnace AA. While this fish had

the greatest Al exposure no Al was detected in the calmodulin fraction.

Discussion

The presence of CaM in the mucus of rainbow trout is supported by

two lines of evidence. Both the radioimmunoassay determination and the

 

118

phosphodiesterase assay provided positive results for the presence of
CaM in mucus. These findings support the observations made by Flik et
a1. (15). With aluminum absent, at a Ca concentration of 50 pH, pH 6.5,
an average CaM content of 0.2,ug/mg total protein in fish mucus in the
present study seems to be of the same order of magnitude as that found
in tilipia mucus preparations obtained from fish held in tapwater where
the Ca concentration varied from 200 to 800‘pM (15). Flik and coworkers
observed less CaM in mucus at lower Ca levels. At the two levels of Ca
tested here no statistically significant difference was observed in CaM
content.

The increased CaM content in Al-stressed fish may be due to greater
total protein content in the mucus. In mucus derived from Al-stressed
fish the total protein concentration was found to be similar to that of
mucus from nonstressed fish. Therefore, the Al-induced enhancement of
mucus CaM content does not seem to result from Al-related breakage of
cells at the fish surface. The enhanced presence of calmodulin in mucus
is probably a response to A1 stress for the purpose of maintaining
membrane integrity and regulatory processes in epithelial cells. Flik
et a1. (15) proposed that the presence of CaM in the mucus may related
to enzymatic control of integumental permeability. Calmodulin in fish
mucus may serve as an extracellular chelator of Ca. Calcium plays a
major role in membrane permeability by its interaction with phospho-
lipids. Aluminum stress in fish is manifested by ionic imbalances (10).
The effects of Al on the ability of CaM to mediate enzymatic control of
membrane permeability cannot be discounted as a possible lesion in

overall Al toxicity. Aluminum has been reported to inhibit the

 

 

 

 

119

Na+-K+-ATPase system in fish (13). Inasmuch as CaM mediates this
enzyme system, Al binding to CaM may be the primary cause for this
inhibition.

The lack of discernable trends in phosphodiesterase activity was
expected. The Tris component of the buffer used to rinse mucus from the
fish body contains potentially active binding sites for Al. Ganrot (24)
reported the tendency of Al to "migrate" from one binding site to
another in biological systems. If the stability of the A1:Tris complex
exceeds that of the Al:CaM complexes, then Al may redistribute to Tris
binding sites. This occurrence would return CaM to its normal
conformation thus restoring its ability mediate phosphodiesterase
activity.

Extremely low concentrations of Al were used in the present study
(0-70‘ug/L). Even at these low concentrations a significant correlation
between the concentration of Al in solution and the amount of CaM in
the fish mucus was apparent. Much higher concentrations of Al can be
expected in lakes and streams impacted by acidic deposition (>200 ug/L)
(4). At elevated A1 levels the increase in CaM may be more pronounced.

The presence of calmodulin in fish mucus may serve as an indicator
of stress induced by biologically active Al in surface waters.
Confounding the response of CaM to Al may be variations in calcium
levels. Calcium has been reported to affect the concentration of CaM in
fish mucus (15). However, the magnitude of increase in CaM over a given
range of calcium may not be as significant as that observed for Al. The
results of ANOVA procedures indicated that calcium had no significant

correlation with CaM in the present study. Given the large increase in

 

120

CaM at low concentrations of A1, Ca may not act as a significant

”interference" in the detection of Al-induced CaM

Conclusions

Rainbow trout exposed to chronic, low-levels of A1 exhibited
elevated levels of CaM in the mucus. Calcium and pH levels also appear
to effect the amount of CaM in the mucus, although this effect was not
significant. The results clearly indicate the relationship between A1
concentration in solution and the amount of calmodulin in rainbow trout
mucus. despite the extremely low Al levels used in the experiment the
secretion of calmodulin into the mucus was significant. Given the
involvement of CaM in membrane permeability and various enzyme systems,
Al binding to this protein may be the first mode of action in A1
toxicity syndrome in fish. It appears that calmodulin may serve as an
indicator of stress induced by the presence of biologically relevant Al

in surface waters impacted by acidic deposition.

Acknowledgments

We would like to thank Dr. Martin Stapanian for statistical

support.

10.

ll.

12.

121

LIST OF REFERENCES

Cronan, C.S., and Schofield, C.L., (1979) Aluminum leaching
response to acid precipitation: effects on high-elevation
watersheds in the northeast. Science 204(20): 304-306.

Johnson, N.M., Driscoll, C.T., Eaton, J.S., Likens, C.E., and
McDowell, W.H., (1981) ”Acid rain" dissolved aluminum and chemical
weathering at the Hubbard Brook Experimental Forest, New Hampshire.
Geochim. Cosmochim. Acta 45: 1421-1437.

David, M.B., and Driscoll, C.T., (1984) Aluminum speciation and
equilibria in soil solutions of a haplorthod in the Adirondack
Mountains (New York, U.S.A.). Geoderma. 33: 297-318.

Driscoll, C.T., Baker, J.P., Bisogni, J.J.Jr., and Schofield, C.L.,
(1984) Aluminum speciation and eqilibria in dilute acidic surface
waters of the Adirondack region of New York state. in: Geological

Aspects of Acid Deponitipn, Acin Precipiration Series. Vol. 7,
(O.P. Bricker ed.), Butterworth Publ., Stoneham, MA. pp55-75.

Driscoll, C.T., Baker, J.P., Bisogni, J.J.Jr., and Schofield, C.L.,
(1980) Aluminum speciation and its effects on fish in dilute acidic
water. Nature 284: 161-163.

 

Baker, J.P., and Schofield, C.L., (1982) Aluminum toxicity to fish
in acidic waters. Wat. Air Soil Pollut. 18: 289-309.

Baker, J.P., (1982) Effects on fish of metals associated with

acidification. in: Erpceedingg pf An lntgrnntional Synposium on
Acidic Rain and Eishgry lnpnprp on Norrhpngrprn North America,
lthaca, NY 2-5 Augusr, l281. (R.E. Johnson ed.), ppl65-176.

Henriksen, A., Skogheim, O.K., and Rosseland, B.O., (1984) Episodic
changes in pH and aluminum-speciation kill fish in a Norwegian
salmon river. Vatten. 40: 255-260.

Grahn, 0., (1981) Fish kills due to high aluminum concentrations in

lake waters. in: Proceedingg pf thg International Conference on the
Ecological lnpact of ncin Precipirntion, March 1980, snsf Project
Report, Oslo, Noryay.

Muniz, I.P., and Leivestad, H., (1981) toxic effects of aluminum on
brown trout (Salmp rrntrn, L.), in: Erppgenings of rhg

Internat ona onfere ce 0 t c m a t of Acid
re itat on r 19 0 F 0 lo Norwa .

Brumbaugh, W.G., and Kane, D.A., (1985) Variability of aluminum
concentrations in organs and whole bodies of smallmouth bass

(Micropterng nolomieui). Env. Sci. Tech. 19(9): 828-831.

Cleveland, L., Little, E.E., Hamilton, S.J., Buckler, D R., and {
”1 ti

l3.

14.

15.

l6.

17.

18.

19.

20.

21.

22.

23.

24.

122

Hunn, J.B., (1986) Interactive toxicity of aluminum and acidity to
early life stages of brook trout. Trans. Am. Fish. Soc. 115:
610-620.

Staurnes, M., Sigholt, T., and Reite, 0.B., (1984) Reduced carbonic
anhydrase and Na -K+-ATPase activity in gills of salmonids

exposed to aluminum-containing acid water. Experientia 40: 226-227.

Fl k, G., Wendelaar Bonga, S.E., and Fenwick, J.G., (1983)

Ca +-dependent phosphatase and ATPase activities in eel gill
plasma membranes -- 1. Identification of Ca +-activated ATPase
activities with non-specific phosphatase activities. Comp. Biochem.
Physiol. 76(8): 745-754.

Flik G., Van Rijs, J.H., and Wendelaar Bonga, S.E., (1984) Evidence
for the presence of calmodulin in fish mucus. Eur. J. Biochem. 138:
651-654.

Siegel, N., Coughlin, R., and Haug, A., (1983) A thermodynamic and
electron paramagnetic resonance study of structural changes in
calmodulin induced by aluminum binding. Biochem. Biophys. Res.
Comm. 115: 512-517.

suhayda, C.G., and Haug, A., (1984) Organic acids prevent aluminum-
induced conformationals changes in calmodulin. Biochem. Biophys.
Res. Comm. 119: 376-381.

Siegel, N., and Haug, A., (1983) Aluminum interaction with
calmodulin: Evidence for altered structure and function from
optical and enzymatic studies. Biochim. Biophys. Acta 744: 36-45.

Rasmussen, H., (1981) Calcinn and CAMP ns Synarchic Messengers.
John Wiley and Sons, New York.

LaZerte, B.D., (1984) Forms of aqueous aluminum in acidified
catchments of central Ontario: A methodological analysis. Can. J.
Fish. Aquat. Sci. 41: 766-776.

Bradform, M.M., (1976) A rapid and sensitive method for the
quantitation of microgram quantities of protein utilizing the
principle of protein-dye binding. Anal. Biochem. 72: 248-254.

Teo, T.S., Wang, T.H., and Wang, J.H., (1973) Purification and
properties of the protein activator of bovine heart cyclic
adenosine 3',5'-monophosphate phosphodiesterase. J. Biol. Chem.

248(2): 588-595.

Penney, C.L., (1976) A simple micro-assay for inorganic phosphate.
Anal. Biochem. 75: 201-210.

Ganrot, P.O., (1986) Metabolism and possible health effects of
aluminum. Environ. Health Perspec. 65: 363-441.

 

[I]ll[NINIll][IHIWINIIHHllllllll(HHIIHIHINHI