Fluorescence characterization of the low pH-induced change in diphtheria toxin conformation: Effect of salt

Vol. 120, No. 1, 1984

BIOCHEMICAL

AND BIOPHYSICAL RESEARCH COMMUNICATIONS

April 16, 1984

Pages

286-290

FLUORESCENCE CHARACTERIZATION OF THE LOW pH-INDUCED CHANGE IN DIPHTHERIA TOXIN CONFORMATION: EFFECT OF SALT Michael

G. Blewitt, Raghupathy

Department Received

March

7,

Jian-Min Zhao, Brian McKeever, Sarma and Erwin London*

of Biochemistry, Stony Brook,

S.U.N.Y. at Stony N.Y. 11794

Brook,

1984

We have examined the effect of pH on diphtheria toxin conformation using intrinsic protein fluorescence and a new fluorescence quenching method. In aqueous solutions, fluorescence indicates toxin conformation undergoes a drastic change at low pH. This conformational change is closely associated with a switch from a hydrophilic conformation to a hydrophobic one, as judged by quenching-detected detergent binding. In the absence of NaCl these changes occur around pH 4-4.5. However, in 150 mM NaCl the conformational change occurs in the pH 5-5.5 range, close to the pH the toxin is expected to encounter in endosomes and lysosomes. Therefore, the conformational change observed at low pH is likely to be physiologically signficant.

It

is

now clear

virus

and protein

virus

and influenza

endocytosis, endosomes low

and

(4)

virus then

switch study

binding

into

enter

cells.

cells

penetrating

that

similarities toxin,

to

membrane

latter

of

step

is

Semliki

vesicles,

to

significant

Forest

undergoing

accomplished

hydrophobicity,

becomes

between

receptors,

acidic

conformation

toxin

toxin,

strong

attaching

a hydrophilic

indicated

by normal

by

This

are

Diphtheria

the

(l-4). from

there

either

partly

by a

a hydrophobic

inferred

from

the

pH 4-4.5

below

one. Triton range

*

Our secondary

recent and

concentration

preliminary tertiary

previous

is results

hydrophobic

Copyright All rights

structure

around

the

noted

pH of only

above.

pH 5, close

should

have In

(5).

significant

whom correspondence

0006-291X/84

studies

influences

hydrophobicity

*To

entry

or lysosomes

A previous

in some cases

toxin

pH-induced

X-100

that

the

below

to physiological

be addressed.

$1.50 286

that

this

report,

low

pH 4-4.5, in

150 values.

pH alters we find

conformational

However,

0 1984 by Academic Press. Inc. of reproduction in any form reserved.

shown

switch. in

that Without

agreement

mM NaCl

toxin

toxin

salt NaCl

with

the

becomes

Vol.

120,

No.

1,

BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

1984

MATERIALS AND METHODS Crude diphtheria toxin was purchased from Connaught Laboratories (Ontario, Canada). Trypsin, N-e -p-tosyl-L-lysine-chloromethylkentone, and Brij 96 (an oleyl polyoxyethylene ether detergent) were purchased from Sigma Chemical. All other chemicals were reagent grade. Purified toxin was isolated as described previously (6). The dimer toxin fraction with bound ApUp (a natural dinucleotide ligand (7)) was used for further experiments. Only a few percent of nicked toxin was present as judged by SDS gel electrophoresis:l -fP otein concentration was calculated using a value of E 28C = 55,300 cm M for ApUp-bound toxin derived from published values for aromatic amino acids and ApU (8). (Lowry assay gave roughly 20% higher values). Toxin nicked between the A and B subunits was prepared by incubation of 10.8 mg intact toxin in 0.75 ml of 1.3 ug/ml trypsin /50 mM Tris-Cl pH 7.2/2mM CaCl . After lh 5 ul of 1 mg/ml N-a-p-tosyl-L-lysinechloromethyl i?etone was added. Then toxin was reisolated by chromatography on a 2.5 cm x 95 cm column of Sephacryl S-200 using a buffer of 50 mM Tris-Cl pH 7.2/l mM EDTA 0.02% (w/v) NaN3. SDS gel electrophoresis showed that nicking was complete. Dibrominated Brij 96 was synthesized by titration of an aqueous solution of Brij 96 with concentrated bromine water. The reaction was over in 5 seconds as monitored by loss of bromine color. Following excess bromine by we found a sharp endpoint to the titration indicating that 70% of Brij kg3 molecules contained double bonds in their hydrocarbon chains (the site of bromination). The Brij 96 used for further experiments was brominated to 80% of this maximal value to avoid residual free bromine. Fluorescence was measured in 1 cm quartz cuvettes using a Spex 212 Fluorolog Spectrofluorimeter. Excitation and emission slits were used with nominal bandpass of 4.5 nm. Excitation was at 280 nm and emission was scanned from 300 to 360 nm. Background without toxin was subtracted from all reported values.

RESULTS Ah'D DISCUSSION Figure toxin. In

1 shows

20-25%

toxin

It

nm below

the

is

takes is

of

this

at

at

low

noted

above

pH. low pH.

pH is

that the

These

in

transition

In

toxin

without

shift

in

general,

the

behavior

that

without

gradual

for

some

of

toxin

At present

variability. 287

shows

a gradual

fluorescence

NaCl and at pH 5-5.5 from

the

nicked

preparations we do not

326 nm to

that of

of

intensity.

to a low

indicate

more

our

decreased

Amax

results

solutions

fluorescence

"transition"

reproducibly

pH.

of aqueous

as pH is

by a rapid

by a red

We find

in the

at pH 4-4.5

accompanied transition

drop

fluorescence

followed

similar.

be

in

occurs

place

fluorescence

constant

first,

of pH on fluorescence

is a marked

drop

The transition

mlf NaCl.

should

the

decrease

state.

toxin

effect

At low pH there

intact

change

the

in

about

150 333

a conformational nicked

NaCl

the

toxin.

and intact change

in

However,

it

fluorescence understand

the

is source

Vol. 120, No. 1, 1984

BIOCHEMICAL

AND BIOPHYSICAL RESEARCH COMMUNICATIONS

.6 -

.O i

.6 -

.8 -

OI

02

PH

I

I

I

I

50

I

I

150

IO0

O

[NCICI]

(mM)

Figure 1: Effect of pH upon Fluorescence of Diphtheria Toxin. (A) Intact toxin (4.5 pg/ml). (B) Nicked toxin 5.1 pg/ml. Samples contained toxin in 10 mM buffer with (-) or without (---) 150 mM NaCl. Na phosphate buffer was used at pH 2.3 and 6 to7; Na formate buffer at pH 3 to 4.5. Na acetate buffer was used at pH 5 and 5.5; tris-Cl buffer was used at pH 9. Samples were incubated lh before fluorescence at 33Onm was measured. Figure 2: Effect of NaCl upon Toxin Fluorescence at pH 5. Concentrated was added to a sample of 4.4 ug/ml intact toxin in 10 mM Na acetate Fluorescence was read 5 min. after each addition of salt. Figure

2 shows

gradually

stabilized

the

the

pH of Using

effect

of

binding

transition

brominated

is Brij

detergents,

which

detected

via

fluorescence As shown

as

of

the

Triton

and unbrominated

with

of

brominated

in figure

mild,

Brij

3, detergent

Brij

96 at

With

binding 288

of NaCl

this

each

method

pH.

that

appears

were

Detergent

with at

it

is

class

of

incubated binding

as the

ratio

unbrominated

pH 4-4.5

the

detergent

because

toxin

expressed to

present.

polyoxyethylene of

that

detergent

96 was chosen

Samples

is

implies

quenching

fluorescence, 96 relative

This

amount

non-ionic

X-100. Brij

tryptophan

exact

fluorescence

(5).

conformation

increased.

was examined.

detected

includes

quenching

a

fluorescent

is

to the

binding

member

weakly

concentration

96

can be conveniently

brominated

pH 5 the

sensitive

pH upon detergent

brominated

96.

at

as NaCl

an easily

with

that

NaCl 5.

pH

without

was of Brij NaCl

BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

Vol. 120, No. 1, 1984

0.4

F/FOE

‘Bi$-id

:

0.4 I 2

I 3

I 4

I 5

1 6

I 7

PH

of pH upon Quenching of Toxin Fluorescence by Br,,sinated Figure 3: Effect Detergent. (A) Intact Toxin (B) Nicked Toxin. Samples prepared as in figure 1 except that 0.1% (w/v) of either Brij 96 or dibrominated Brij 96 was present. Symbols same as Figure 1. Ordinate is the ratio of fluorescence in the presence of dihrominated Brij 96 (F) to that in the presence of Brij 96 (Fo).

and around appears

pH5 in is

15OmM NaCl.

close

transition.

to

It

associated

is

or

Therefore,

the

lower

than

slightly

probable

with

either

the

The pH of

4-4.5,

at

that

the

formation

pH at which the

pH of

transition

in

or exposure

detergent the

toxin

binding

fluorescence

conformation

of a hydrophobic

is

site

on the

protein.

close

to

ionic

conditions,

value

of pH 5 found

of

pH is

that

reported

hydrophobicity atoms

are

these

in

what

is attached

experiments

a much more 150 mM NaCl

the

toxin

tryptophan

residues

At low pH, the 150 mM NaCl

to the

are

stronger

could

that

either

appears

Triton

closer

it

sucrose

encounter

in

is

that

likely

exposed quenching involve

to

the

a difference 289

fraction

toxin

relative in

tryptophan

toxin bromine

conclusion the

of to

range

induced

another

region

The

endosomes.

quenching

of

similar

This

or

acid

the

very

method. pH.

the

is

under

lysosomes

chain

hydrophobic

of nicked

binding gradient

Because

hydrocarbon considerable

NaCl,

to physiological

significant.

a

without

X-100

laborious is

detergent's that

for

should

physiologically

is

binding

(4)

we now can conclude

Therefore,

Brij

previously

using

about

which

of

fluorescent the

intact exposure

micelles. toxin

in

or

the

Vol. 120, No. 1, 1984 amount

of

residues

bound

detergent.

come into

contact

Even

with

conformation of

the

BIOCHEMICAL

these is not

the

including

to

these

the

CRM 45

toxin,

receptor-binding is

to toxin

story

of

could

at

may also

free

affect

pH control

example,

exist

of

acids

of

the

neutral

the

various factors

or

with

pH

hydrophobicity.

forms

example, of

(9).

may

reduced

For

C-terminal

pH

ions,

in many other

ApUp

from

toxin

may affect

in hydrophobicity.

near

tryptophan

environmental

can

differ

amino

X-100

For

toxin

subunits,

that

variables

and other

the

likely

--in vivo.

transition.

missing

Triton

is

additional

potential

forms

it

precise

Several

separated

of

binds

lipids

Furthermore

Any

fragment,

membrane the

membrane

aisulfiae. mutant

case

hydrophobic

transition. monomer,

either

results,

a pH gradient,

affect

with

complete.

hydrophilic

lipids,

In

AND BIOPHYSICAL RESEARCH COMMUNICATIONS

its

B

Furthermore,

Clearly,

more work

needed. Fluorescence

lipids

has

several

protein

recently

aspects

brominated

quenching

of

detergents

been

by

introduced

lipid-protein will

spin-labeled as

(10,ll) a powerful

interaction

also

and

be a powerful

tool

(14). tool

This for

brominated for

(12,13)

elucidation

report

shows

of that

understanding

membrane

to E.L.,

19131

behavior. ACKNOWLEDGEMENTS

R.S.

This work was supported and a S.U.N.Y. University

by N.I.H. grant Award to E.L.

GM 31986

A.I.

to

REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14.

Lenard, J. and Miller, D.K. (1982) Cell, 2, 5-6. Simon, M.I. and Draper, R.K. (1980) J. Cell Biol. 87, 849-854. Sandvig, K. and Olsnes, S. (1980) J. Cell. Biol. 3, 828-832. Sandvig, K. and Olsnes, S. (1981) J. Biol. Chem. 256, 9068-9076. London, E., Zhao, J.-M., Chattopadhyay, A., Blewitt, M., McKeever, B. and Sarma, R., Proceedings of the New York Academy of Sciences 1983 Colloquium in Biological Sciences, in press. McKeever, B. and Sarma, R. (1982) J. Biol. Chem. 257, 6923-6925. Barbieri, J.T., Carrol, S.F., Collier, R.J. and McCloskey, J.A (1981) J. Biol. Chem. e, 12247-12251. CRC Handbook of Biochemistry and Molecular Biology, Vol. 1 (1976)(G.D. Fasman ed.) CRC Press, Cleveland 552 pp. Boquet, P., Silverman, M.S., Pappenheimer, A.M. Jr. and Vernon, W.B. (1976) Proc. Nat. Acad. Sci USA 2, 4449-4453. London, E. and Feigenson, G.W. (1981) Biochemistry 3, 1939-1948. London, E. and Feigenson, G.W. (1981) Biochim. Biophys. Acta 649, 89-97. Leto, T.L., Roseman, M.A. and Holloway, P.W. (1980) Biochemistry E,1911-1916 East, J.M. and Lee, A.G. (1982) Biochemistry 21, 4144-4151. London, E., (1982) Molec. and Cell. Biochem. E, 181-188. 290