Fluorescence investigations of calmodulin hydrophobic sites

Vol. 119, No. 3, 1984 March 30, 1984

AND BIOPHYSICAL RESEARCH COMMUNICATIONS Pages 1154-1160

BIOCHEMICAL

FLUORESCENCE INVESTIGATIONS

OF CALMODULIN HYDROPHOBIC SITES and Dominique

Army FOLLENIUS*

GERARD

Laboratoire de Biophysique, ERA CNRS 551 rJ.E.R. des Sciences Pharmaceutiques, B.P. No 10 67048 STRASBOURG CEDEX, FRANCE Received

February

6, 1984

SUMMARY - Calmodulin activation of target enzymes depends on the interaction between calmodulin hydrophobic regions and some enzyme areas. The Ca2+ induced exposure of calmodulin hydrophobic sites was studied by means of 2-p-toluidinylnaphthalene-6-sulfonate, a f;Torescent probe. Scatchard and Job plots showed that the calmodulin-Ca4 complex bound two molecules of this hydrophobic probe, with KD N- 1.4 x 10 -4 M. These sites are not totally exposed until calmodulin has bound four Ca2+ per molecule, so the conformational change is not over before the four specific Ca2+ binding sites are saturated with Ca2+.

Calmodulin course

of evolution,

of cyclic

(2).

quitous

protein

ting

the

tein

family

4

primary

When the

to and activate

target

enzymes. the

of techniques.

These

10,

Several

tional

change

authors resulted

Ca

2+

circular and nuclear

(12, in

to the

characterized level

rises,

the

change

that

- dependent

modifications (7)

Copyright All rights

0

has been

(6,

Inc. reserved.

7),

(8) , U.V.

absorption

magnetic

resonance

13) showed

the exposure

pro-

by the

of

that

the

of hydrophobic

1154

Ca

2+

presence

calcium-binding free

Ca

2+

enables

ions

with

in

spectroscopy spectroscopy - induced on the

g-AC, acid; TNS, (hydroxymethyl)-

the

a number

differences

areas

bind

CaM to bind

of enzymes,

investigated

dichroism

$1.50

1984 by Academic Press. of reproduction in cm? form

(1) and

calcium-binding

regulation

* To whom correspondence should be addressed. Abbreviations used: ANS, 1,8 anilinonaphthalene-sulfonate; EDTA, ethylenediaminetetraacetic 9-anthroylcholine; 2-p-toluidinylnaphthalene-6-sulfonate; Tris, tris aminomethane. 0006-291X/84

the

activator

by Cheung

by a EF handlike

2+

change

include

to chemical fluorescence

11).

Ca

to clarify

tyrosine

is

a conformational

conformational

reactivity

CaM belongs

constituted

intracellular

Ca2+ - induced

during

as a Ca2+ - dependent independently

structure

each

inducing

to the proteins,

little

reviews, see 3, 4) showed that this ubiCa 2+ - regulatory roles in cells by modula-

multiple

areas,

very

(for

of many enzymes.

and its

In order

discovered

studies

plays

activity

(5).

has changed

phosphodiesterase,

Further

Ca2+ - binding

loop

which

was first

nucleotide

Kakiuchi

a protein

(CaMI,

(6), (9,

conformasurface

Vol. 119, No. 3, 1984 of the

protein.

enzyme

interface,

BIOCHEMICAL These

and affinity

since

enzymes

is

In the present to CaM, required

areas

and compared

hydrophobic

probe,

and

greatly

with

hydrophobic

to

it

method

we established induce

TNS, whose in apolar

protein of analysis

(14).

yielding

The implications target

the

in the CaM-

analysis

interaction

(12,

13)

between

CaM

sites

enzymes

complete

exposure is

of the hydrophobic

To do so,

negligible

a

solvents

interactions

By equilibrium

dialysis

and another

a Job plot,

we determined

available

or after

we used

in polar

noncovalent

of these are

of Ca 2+ ions

stoichiometry

of Mg2+ ions.

solvents

areas

- bound

different

probe

that the

fluorescence

hydrophobic

the

the

effect

of TNS-sensitive protein.

indicate

a key role

hydrophobic.

to the

enhanced

must play

fluorescence

(12)

mainly work,

bound

areas

hydrophobic

chromatography

and target

standard

hydrophobic

AND BIOPHYSICAL RESEARCH COMMUNICATIONS

on the studies

surface for

the

the of the

regulation

number 2+

Ca

of

discussed.

MATERIALS

AND METHODS

Chemicals - The compounds were of the purest grade available commercially. TNS (2-p-toluidinylnaphthalene-6-sulfonate, potassium salt) was purchased from Serva Co and solutions were made up just before use. Preparations of CaM and phosphodiesterase - CaM was purified in our laboratory from bovine brain accordinq to Isobe et al. (151, except for gel filtration which was carried out on Ultrogel ACA 54 instead of Sephadex G 75. The amino-acid composition of the purified protein was analyzed gel electrophoresis and corresponded to published data (7). 20 % acrylamide was carried out and staining with Coomassie brilliant blue indicated a single band in the protein-overloaded gel. Fluorescence spectroscopy was another criterion of purity : the emission spectrum was typical of Tyr with a maximum at 302 nm,and was identical to that fluorescence emission, published in (7). Calcium was removed from CaM by three successive trichloroacetic acid precipitations using the method of Haiech et al. (16) calcium per mole of protein was < 0.04 mol. This an g+the amount of residual since the fluorescence quantum Ca - free protein was Ca2+ - sensitive yield was 2.7 times higher after Ca2+ - binding to CaM. All protein solutions were prepared in Tris-HCl 100 mM buffer, pH 7.6 made up with Ultrapure water (Milli Q Instrument from Millipore-Waters Corp.) and stored in acid-washed plastic ware to avoid any ion contamination. The Ca2+ concentrations of CaCl2 solutions used for titrations were checked with atomic absorption spectrophotometry. Cyclic nucleotide phosphodiesterase was isolated from bovine aorta in the absence of bovine serum albumin and the assays of phosphodiesterase activation with CaM were run without bovine serum albumin present in the experimental medium (17). Equilibrium dialysis experiments - They were performed with a Dianorm equilibrium dialyzer (Diachema A.G., Ruschlikon, Switzerland), using cells Hydrated cellulose membranes (Diachema type of 1.4 ml of total volume. 10.14, molecular weight cut-off of 5000) were treated successively with with NaHC03 5 % - EDTA 5 mM, water, EtOH/H20 50 : 50 (V/V) and rinsed water before equilibrating in Tris-HCl 100 mM buffer, pH 7.6. TNS adsorption by the membranes was negligible and solution volumes on each side of the membrane remained constant throughout dialysis. The experiments were run for 12 h at 20°C, long enough to reach equilibrium. Samples were then withdrawn from the cells and free TNS concentrations were determined by spectrophotometry. 1155

Vol. 119, No. 3, 1984

BIOCHEMICAL

AND BIOPHYSICAL RESEARCH COMMUNICATIONS

Spectroscopic measurements - Absorption spectra were recorded with a Cary 219 spectrophotometer and calmodulin concentrations were determined coefby measuring absorbance at 277 nm, with M, = 16 500 and an extinction ficient of S (2JJ m) = 3300 M-lcm-1 (6). Correction for scattered light was made as indicated by Shih and Fasman (18). Fluorescence measurements were performed on an absolute spectrofluorometer FICA 55, at 2O'C. Fluorometric titrations were performed at TN.5 emission maxima (465 nm) and carried out directly in quartz cuvettes by adding aliquots of CaC12 or MgC12 stock solutions to a solution containing CaM and TNS, and correcting for dilution. RESULTS

Interaction

between

The hydrophobic water

calmodulin

probe

(# = O,OOO8(16)1

TNS exhibits

characterized

maximum at approximately bic

areas

involved

we checked

that

in the bound

TNS fluorescence

interaction

earlier

maximum to 440 nm,probably greater

enhancement

of the emission

maximum to 460 nm (data

was also

been

tested

reported

that

in conformation

addition

formational

change

bic

this

ion

induced

of phosphoCaM,

in the

binding than

emission of

of TNS to Ca2+ - free

and a blue

an important

shift

physiological

of CaM, since

the

Ca2+ - induced

of TNS emission

by the binding

a

shown).

mimiked

of Mg2+ to Ca2+ - free

enzymes,

intensity

behaviour

partially

target

adsorption

extensive

plays

hydrophobic

No enhancement

(7).

progressive

on the

not

which

hydropho-

was a shift

of TNS fluorescence

of Mg2+,anotherion

to the

of Ca2+ - free

in TNS fluorescence

CaM : a 25-fold

with

hydrophobic

CaM involves

in

spectrum,

activation

presence

and there

changes

emission

CaM and the

due to non-specific

CaM, which

role,

emission

CaM-induced the

higher

Ca2+ - bound

The effect

fluorescence TN.7 bound

between the

(12).In

TNS by CaM. However, far

low

by a broad

TNS inhibited

was two times

gives

very

520 nm. To see whether

as reported

diesterase,

and TNS

was noted

CaM, which of Mg2+ to

it

changes

during

shows that exposed

CaM

has the no con-

hydropho-

sites. Stoichiometric

mational

because

higher

than

tions,

added

Ca2+/CaM

2+

-binding

sites

involved

of CaC12 solution

in the

to an increase

to a solution

in TNS fluorescence

We performed by a progressive blue shift. 2+. of the Ca - binding sites of CaM involved

titration

illustrates

of aliquots

CaM and TNS leads

accompanied change,

of Ca

confor-

change

The addition - free

titration

the

the

protein

concentration

dissociation

constant

Ca2+ was assumed the

increase

concentrations.

used (KH N

to be totally

in TNS fluorescence TNS fluorescence 1156

a stoichiometric in the

10 -6 M) . Under to the

intensity reached

2+

intensity

was more than

bound

of Ca

conformational 10 times these

condi-

protein.

Fig.

as a function

a plateau

when

4 Ca

1 of 2+

Vol. 119, No. 3, 1984

BIOCHEMICAL

0

AND BIOPHYSICAL RESEARCH COMMUNICATIONS

5

10 Ca*+l

iBound

15

ICaM Fig. 1 : Increase in TNS fluorescence intensity (40 m) dingCaM (40 )&lLM)in 100 mM Tris-HCl, pH 7.5. buffer. ions

were

gents

bound

per

the

linear

from

protein, portions

Some additional a solution

simulating

of the

trations

increase

was very

the buffer

imsf CC4

Number

dent (Fig. ding

binding 2a).

per

mol.

was related

which

dissociation (KH = 4.4.

with

axis

were

CaM-TNS

lower

by a Job plot

binding

(21)

for

where

this the

concen-

concentrations curve.

of the

calcium-depen-

to Scatchard

slope

(20)

of TNS-bin-

mean for

5

of the

constant.

such ex-

Scatchard

The average

complexes With

measured

TNS/protein

x loo5

this

magni-

such as CaM-ANS

x 10s4 M) which

different

maximum number

ratio

10s4 M, was of similar

= 4.9 that

changing

the number

dissociation

: KD = 4.4. 2/l

of Ca2+/CaM Moreover,

was the

by the

dialysis.

must be due to the evidence

indicated

probe

than

The

according

2 0.3

-+ 0.1)

(KD

by equilibrium

titrated.

of CaM

plotted

: 7~2.1

in

150 mM KCl,

were

shown).

measurement

intercept

to the

fluorescence

sites

sites

direct

KD = 1 = (1.4 KA of other GM-hydrophobic

of the tan-

TNS stoichiometric

was characterized

was slightly

Further

not

hydrophobic

x LOS4 M) or GM-g-AC,

discrepancies

: CaM was dissolved

as a function

the

enabled

binding

determined

TNS affinity (13)

affect

constant

to those

out

100 mM to 20 mM or the

of protein

plot

(12)

not

The horizontal

sites

intercept

(20 mM Tris-HCl,

1 (data

of TNS to CaM. Data

Total

al.

from

dialysis

periments.

tude

to Fig.

of TNS-sensitive

Equilibrium

medium

in TNS fluorescence similar

to 3 did

0.5

carried

the

and the Ca 2+ - binding

concentration

from

curve.

were

intracellular pH 7.6)

from

of the

experiments

1mM MgC12 buffer, plot

as inferred

Ca2+ bin = 365 2 Snm)

during (hexc

La Porte

technique,

by Tanaka M, but

these

and Hidaka slight

methods. stoichiometry of binding

sites

et

the CaM-

was given was de-

Vol. 119, No. 3, 1984

BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

b

0

1

2

0

0.5 0.65 ITNSI ,x [TNSI+ICaMl

v

1

=

Fig. 2a : Scatchard plot for !lNS binding to CaM + Ca2+. Binding parameters were obtained by keeping CaM concentrations constant (20 @I) and measuring free TNS absorption (Et317 nm)= 1.89 x lo4 M-1 cm-l (14)) at various concentrations (S-600 @i). V , number of moles of TNS bound per mole of total protein ; c, molar concentration of free TNS. Fig.

2b

Buffer

: Job

termined

of

as follows.

was varied, tant

plot

TNS : CaM interaction. pH 7.6, in the

The mole

while

the

total

fraction,X,

intensity

of the

The maximum was at

ratio

of

1.86.

This

on the protein

of

molecule

solution

X = 0.65,

indicated

of TNS plus

10 mM CaC12.

TNS-CaM that

+[c~M] = 2 x 10m5M. of 10 mM CaC12.

of TNS in TNS-CaM solutions

concentration

at 20 mM, in the presence

apparent tionx.

[TNS] presence

was 100 mM Tris-HCl

which

the

Fig.

CaM was kept 2b plots

the

consrise

in

as a function

of TNS mole

corresponded

to a TNS/CaM

maximum number

frac-

of TNS-binding

sites

was two. DISCUSSION

The TNS spectralmodificationsare bic

regions

on the CaM surface,

This

Ca2+ - induced

other

spectroscopic

dichroism quantum of the

(7,

and the

residues

affinity

binding

10, ll),

showed of

tion

the

CaM

exposure

of the

(9,

sites. that

10,

(7,

11).

subsequent after

change

with

circular in Tyr binding

additional

Ca2'

affecting

2 Ca2+ had bound

by NMH spectroscopy,

conformational

19),

The increase

The conformational

However,

in CaM.

previously

ellipticityoccurredupon

while

seemed to be finished

change

observed

fluorescence

in molar

parameter.

had been

of hydropho-

several

changesoccurredduring

had the

to the

high

authors

(9,

the

sequential

1 to 4 Ca2+ to CaM. TNS,

regions of

such as Tyr

molecule

appearance

conformational

change

variation

two Ca2+ per

tyrosyl

phobic

methods

on either

With

following

conformational

22) or NMH spectroscopy

yield first

no effect

binding

due to the

it

is

possible

to monitor

of the protein, with

target

enzymes.

of two hydrophobic

4 Ca2+ - binding

these

sites

changes

regions

We found sites

did

place

in the

hydro-

being that not

of CaM. In fact, 1158

taking

related to the interac2+ the Ca effect involving

finish previous

before

the

studies

saturation showed

that

Vol. 119, No. 3, 1984

the most

common enzyme-activating

nucleotide another

with

terase

is

the enzyme also different

formation

regions

: it

medium, the

sites

noticeable CaM-Mg2+

fluorescence

complex

(71

; this

(7,

on the activation

surface

The exact

but

would

that

structural

extent

less

than

confirmed

with

are

although

to acti-

Kilhoffer

still

modification, (7,241

of Ca2+

intracellular

not

in

affected

Mg 2' binds

to

no hydrophobic

and consequently,

of phosphodiesterase

function

complex

changes,

why,

phosphodies-

of Ca2'. but

properties

explain

of the protein

the

CaM

of hydrophobic

in the

was affected

But TNS spectral

a slight

(24).

present

local

complex

that

CaM-Ca2+ n

with

For

of hydrophobic

exposure the

(22).

CaM-Ca2+4

clear,

cation

reflected

Ca2'

cyclic

or other

there

is

enzymes

no

by the

complex. of CaM to activate formation,

of hydrophobic

rearrangement. enzymes

22).

to CaM. This

of Ca2+_complex

cases

leading

to CaM was compared Tyr

The ability number

not

4 : thus

reported

complex

or total in

that

and induces

appear

has been

still

physiological

binding

dichroism

EF hands

be involved

a divalent

by Mg2+ - binding the

is partial

the

interaction

enzyme.

so its

circular

it

4 bound

only

the

However,

CaM-complexes

reported

CaM-Ca2+

kinase,

by the CaM-Ca2+3

may thus

Mg2+ is

requires

allowing

may involve

a specific

et al.

(23).

activated

complex

chain

changes

2+

of CaM was CaM-Ca

activation light

significant

areas

vate

myosin

enzyme,

AND BIOPHYSICAL RESEARCH COMMUNICATIONS

form

phosphodiesterase

undergoes

of these

BIOCHEMICAL

sites

The presence

an important

factor

but

target

another

enzymes

specificity

on the CaM surface the

criterion exposed

of 2 hydrophobic to understand

may depend

sites activation

on the

extent

may be the

by the

structural

may constitute mechanism

in some of target

by CaM. ACKNOWLEDGEMENTS

We thank Dr. J. BAUDIER for kindly providing CaM, Dr. C. LUGNIER for the preparation and the activation assays of cyclic nucleotide phosphodiesterase and are indebted to Dr. M.C. KILHOFFER for stimulating discussions. Louis This work was supported by grants to D.G. from CNRS and University Pasteur. REFERENCES 1 - CHEUNG, W.Y. (1970) Biochem. Biophys. Res. Commun. 38, 533-538. 2 - KAKIUCHI, S., YAMAZAKI, R., & NABAJIMA, H. (1970) Proc. Jpn. Acad. 46, 587-592. 3 - CHEUNG, W.Y. (1980) Science 207, 19-27. 4 - MEANS, A.R. & DEDMAN, J.R. (1980) Nature 285, 73-77. 5 - REID, R.E. & HODGES, R.S. (1980) J. Theor. Biol. 84, 401-444. 6 - KLEE, C.B. (1977) Biochemistry 16, 1017-1024. 20, 7 - KILHOFFER, M.C., DEMAILLE, J.G. & GERARD, D. (1981) Biochemistry 4407-4414. 8 - RICHMAN, P.G. & KLEE, C.B. (1978) Biochemistry 17, 928-935. 9 - SEAMON,K.B. (1980) Biochemistry 19, 207-215. 1159

Vol. 119, No. 3, 1984

BIOCHEMICAL

AND BIOPHYSICAL RESEARCH COMMUNICATIONS

124, 619-627. 10 - KREBS, J. & CARAFOLI, E. (1982) Eur. J. Biochem. 11 - IKURA, M., HIRAOKI, T., HIKICHI, K., MIKUNI, T., YAZAWA, M. & YAGI, K. (1983) Biochemistry 22, 2573-2579. 12 - LAPORTE, D., WIERMAN, B.M. & STORM, D.R. (1980) Biochemistry 19, 38143819. 13 - TANAKA, T. & HIDAKA, H. (1981) Biochemistry Int. 2, 71-75. 14 - MC CLURE, W.O. & EDELMAN, G.M. (1966) Biochemistry 5, 1908-1919. 15 - ISOBE, T., NAKAJIMA, T. & OKUYAMA, T. (1977) Biochim. Biophys. Acta 494, 16 17

18 19 20 21

22 23 24

222-232.

20, 3890- HAIECH, J., KLEE, C.B. & DEMAILLE, J.G. (1981) Biochemistry 3897. - LUGNIER, C., STIERLE, A., BERETZ, A., SCHOEFFTER, P., LEBEC, A., WERMUTH, C.G., CAZENAVE, J.P., & STOCLET, J.C. (1983) Biochem. Biophys. Res. Commun. 113, 954-959. - SHIH, T.Y. & FASMAN, G.D. (1972) Biochemistry 11, 398-404. - HAIECH, J., KILHOFFER, M.C., GERARD, D. & DEMAILLE, J.G. (1983) Mol. and Cel. Biochemistry 51, 33-54. - SCATCHARD, G. (1949) Ann. N.Y. Acad. Sci. 51, 660-672. - JOB, P. (1928) Ann. Chim. (Paris) 9, 113-120. - CROUCH, T-H- & KLEE, C-B. (1980) Biochemistry 19, 3692-3698. - BLUMENTHAL, D.K. & STULL, J.T. (1980) Biochemistry 19, 5608-5614. - COX, J.A., MALNOE, A. & STEIN, E.A. (1981) J. Biol. Chem. 256, 32183222.

1160