Reversible inhibition of intercellular junctional communication by glycyrrhetinic acid

vol.

134,

No.

January

14,

1, 1986

BIOCHEMICAL

AND

BIOPHYSICAL

RESEARCH

COMMUNICATIONS

1986

Pages

REVERSIBLE INHIBITION

OF INTERCELLULAR JUNCTIONAL COMMUNICATIONBY GLYCYRRHETINIC ACID

James S. Davidson, Department

of

Chemical

29-36

Ingrid

M. Baumgarten and Eric

H. Harley

Pathology, University of Cape Tom Medical Observatory 7925, South Africa

School,

Received November 14, 1985

Intercellular gap-junctional communication was measured using metabolic co-operation in co-cultures of argininosuccinate synthetase-deficient and argininosuccinate lyase-deficient human fibroblasts. 18- Q-glycyrrhetinic acid (AGA) was found to inhibit communication by more than 95% at concentrations as up to 100 uM ware not cytotoxic over a period of 2 low as 2 uM. Concentrations hours. Communication inhibition was of rapid onset and was readily reversible. Communication remained continuously yet reversibly blocked in cells cultured in the presence of AGA for 20 days. The related compounds 18-p -glycyrrhetinic acid and carbenoxolone also caused communication inhibition. The effect is probably not mediated via mineralocorticoid or glucocorticoid receptors since aldosterone and glucocorticoids had no effect on communication. AGA thus has properties of a useful inhibitor in the study of intercellular junctional communication. 0 1986 Academic Press, Inc.

Gap junctions

INTRODUCTION: molecules

directly

the cytoplasmic measuring

cells

between

junctional

of the urea cycle

(3).

argininosuccinate

lyase-deficient

citrulline

into into

14C-citrulline via

by

junctions of

ratio

lines

derived

fibroblasts

from patients

simultaneous

able

with

is

no detectable

incorporation

for any added compound.

nor

to convert

of argininosuccinate

types

incorporate

between cells transfer

of 3H-phenylalanine

via

the

Serves

Thus the 14C-citrulline/3H-

of the mutant cells

that compounds could

inborn errors

to incorporate

of the two cell

1) and there

for

human

synthetase-deficient are

linking

system

in cultured

enzymes are required

both

the transfer (Fig.

aqueous channels A convenient

(IJC)

argininosuccinate

in co-cultures

Because of the possiblity

(1,2).

However co-cultures

efficiently

The rate

cells

the exchange of small

forming

by

conrnunication

protein , since

as an index of toxicity phenylalanine

Neither

arginine.

intercellular

medium (3).

in contact ,

uses two mutant cell

14C-citrulline

to mediate

of adjacent

compartments

intercellular

fibroblasts

are thought

inhibit

is a measure of IJC. one of the enzymatic 0006-291X/86 $1.50

29

Copyright 0 1986 AN rights of reproduction

b-v Academic Press, Inc. in any form reserved.

Vol.

134,

No.

BIOCHEMICAL

1, 1986

AND

BIOPHYSICAL

RESEARCH

COMMUNICATIONS

CIT 1 ASA

ASA

1 ARG * protein ASS-

Fig. 1.

steps

ASL-

Principle of assay of intercellular junctional conmunication. ASA = argininosuccinate: ARG = arginine. CIT = citrulline;

in citrulline

metabolism , it

is also

compounds on normal human fibroblasts

(4).

Gap-junctional pharmacological

investigate (4,10), that

intercellular inhibition

by

inhibition retinoids

several

by

classes study

acid

to monitor

has previously pesticides

phorbol

esters

(AGA) and the related

acid and carbenoxolone

the effects

is known to be susceptible of compounds (5-9).

organochlorine

(11) and tumor-promoting

18-a-glycyrrhetinic

18-fl-glycyrrhetinic

of IJC

necessary

communication

model used in the present

fibroblast

CELLS

NORMAL

are novel

of

to

The human

been employed to and related (11).

compounds

Here we report

compounds inhibitors

of IX.

MATERIALS AND METHODS wre obtained from Sigma Chemical Co. : Chemicals : The following acid, prednisolone, lupeol 18-Cr-glycyrrhetinic acid (AGA) , 18-fl-glycyrrhetinic and betulin. Glycyrrhizic acid was obtained from Aldrich Chemical Co. Cholesterol and hydrocortisone were from BDH Chemicals Ltd. Carbenoxolone sodium ws a kind gift from Berk Pharmaceuticals (Pty) Ltd. AGA, in dimethyl 18-/3-glycyrrhetinic acid, lupeol and betulin were dissolved sulfoxide (CNSO). Glycyrrhizic acid and carbenoxolone wre dissolved in water. The remaining chemicals were dissolved in ethanol. Final solvent concentration was 0.5% or less and appropriate solvent controls were included in each experiment. A fresh stock solution of AGA was prepared immediately before each experiment, as it us found that stored solutions had increased cytotoxicity. [Carbamoyl-14C]citrulline and [2,3-3H]phenylalanine were from the Radiochemical Centre, Amersham. of the human mutant fibroblast cell Metabolic co-operation assay : The origins lines were as described (3). They were maintained on Dulbecco’s Minimal Essential Medium (DMEM, Gibco) with 10% fetal calf serum (Gibco) in 4% 03 at of co-cultures and the assay of IJC viere as described t 4,10). 37oc. Preparation Briefly, trypsinised suspensions of the two cell types were mixed in equal proportions, plated at high density (confluent or subconfluent) and incubated for 18-22 hours. The co-cultures ware then incubated in labelling medium containing radioisotopes and test chemical(s). Where the labelling medium was time was 2 hours, hepes-buffered saline (medium B in ref. 4) incubation 14C-citrulline was used at 0.25 uCi/ml and 3H-phenylalanine at 0.6 uCi/ml. Where the cells were labelled in growth medium (DMEM with l0% fetal calf serum) 4C-citrulline at 1 uCi/m.l and 3H-phenylalanine incubation time was 5 hours with 30

Vol.

134,

No.

BIOCHEMICAL

1, 1986

AND

BIOPHYSICAL

RESEARCH

COMMUNICATIONS

After labelling, radioactivity in acid-insoluble cellular material 2 uCi/ml. was determined as described (10). All experiments were performed in duplicate Each point shown is the mean and repeated at least once with similar results. of duplicate wells, Bars on the figures are of a length equal to the difference betwaen duplicates and are omitted where this difference was less than 5%.

at

RESULTS: Exposure incorporation indicating

to AGA caused a dose-dependent

in co-cultures

of mutant cells,

that AGA inhibits

LJC in this

inhibition

but not in normal cells

human fibroblast

medium, IJC as measured by citrulline/phenylalanine inhibited overall

by 97% at 2 uM AGA. protein

synthesis,

as 100 uM over a 2-hour period. contrast

microscopy In the presence

communication

during

No

unaffected

morphological

a 2-hour exposure

of 10% serum, higher

inhibition

(Fig.

of 0.5% bovine serum albumin

2b).

A

similar

(not shown),

similar

dose-response

rounding-up

(20 h) of cultures curves

of the cells

concentrations

2.

or with

effect

suggesting

to AGA before

phenylalanine

fetal 31

phase

to elicit

was seen in the presence that AGA binds

in inhibiting

10% dialysed

by

to up to 100 uM AGA.

to albumin,

IJC. labelling

produced

2 (data not shown). incorporation,

calf

serum (b).

very However

indicating

Inhibition of communicationby AGA. co-cultures (O,O ) and nom1 cells (D ,O ) wre labelled for 2 hours in mediumB containing AGAat the indicated concentrations without serum (a),

ws

as high

a

Fig.

2a),

reflecting

changes were apparent

to those shown in Fig.

and decreased

by

ratio,

doses of AGA were required

and that such bound AGA is not effective Longer exposure

incorporation,

(Fig.

In serum-free

system.

incorporation

3H-phenylalanine

was essentially

of 14C-citrulline

Vol.

134,

No.

1, 1986

BIOCHEMICAL

AND

BIOPHYSICAL

RESEARCH

COMMUNICATIONS

Table 1 Freincubation 0.3-i

Wash

31i-phenylalanine cpm

Labelling 2h

21045 20996

4050 4309

19.2 20.5

AGA 30 uM

19922 20292

59 59

0.3 0.3

wxihed with serun

INSO 0.5%

20661 20497

3532 3845

17.1 18.8

wshed

D4SO 0.5%

19465 19344

148 112

0.8 0.6

not b&5hd

AGA30uM

not

kashad

AGA

uM

30

%/4-l ratio x 100

EMSO0.5%

mso 0.5%

AGA30uM

l*C-citrulline cpn

without serum

Reversibility of conxmmication inhibition by AGA. 20 hours after platin9,co-culture.s were preincubated in medium B(hepes-buffered saline) with or without AGAas indicated. After 30 min the cells were washed 3 times with median B with or without 10% dialysed serum, or not washed, as indicated. Labelling with radioisotopes ma for 2h in serum-free median B with or without XA. Results for duplicate veils are shown.

cellular

toxicity,

occurred

30 uM in serum-free Camnunication with

wshed

signs

inhibition

cultures

of cellular

proliferate

partitions

strongly

into

by

AGA WM insufficient

as control

coxnnunication the lipid

by washing

the monolayers

showed nearly

cultures.

suggesting

phase of cell

after

shown in Fig.

showed no

used,

for

serum-free that

AGA

up to 20 days during 17 days (Fig.

communication

3).

inhibition

32

cells

continuous

For this

type

For example,

10%

serum,

whereas

100 uM, the rate of increase

by 24% compared to control

to

membranes.

3, where the medium contained

to induce complete

since

a degree of

and continued

Washing with

inhibition,

reversible

as great

Such washed cultures

microscopy,

was maintained

WH rapid,

the dose of AGA KM found to be critical.

At the concentration

number was slowed

labelling

of AGA above

10% serum.

The rate of reversal

phase-contrast

inhibition

experiment,

in the experiment

was toxic.

damage

to AGA and was still

of long-term

1).

not exposed to AGA.

to reverse

Communication exposure

before

at the same rate

medium failed

to concentrations

by AGA was reversible

serum (Table

immediately

IJC as control

20 h exposure

medium and above 150 uM in medium with

medium containing

cultures

after

(data not shown).

75

uM

150 uM

of cell

Vol.

134,

No.

BIOCHEMICAL

1, 1986

A number of compounds with to inhibit

water-soluble with

BIOPHYSICAL

RESEARCH

COMMUNICATIONS

Effect of continuous long-term exposure to AM. Co-cUltureS were platad cn day 1 and cultured in growth medium (INEN plus 10% fetal calf serum) with l00 uM AGA (0) or vehicle only (0.3% CtSW, 0 ) from the time of plating. Tha medium was renewed every 3 days. On day 17 sane wells (0) which had been grown on AGA+nadiumwere washed twice with growth medium and replaced with medium without AGA. Caballing 'k~s for 5h in growth medium.

Fig. 3.

ability

AND

IJC.

acid

carbenoxolone

curves

were more toxic

related

18-fl-glycyrrhetinic

derivative,

dose-response

structures

very similar

than AGA at the higher

(Fig.

to AGA. However, doses.

: +\./“-”

Glycyrrhizic

acid,

a glycoside

\

60-

+\o

.

t

40-

1,.

\ I

23 .r Carbenoxolone

O 6O-.-.-I-. . IO ,

2030 ,,

6oKm

0d’j Glycyrrhizic Spironolsctona

AIM

16 acid

3b PM NM

Fig. 4. Effect co co-cultures of carbanoxolone (a), glyccyrrhizic acid (b; 0 ,O) and spironolactone (b: n ,o ). Experimental conditions as for Fig. 2a. 33

IX

both of these compounds

-o\ A

60-

their

and the

b

100. .--+

o-l-ilm o

(not shown)

for

4a) ware both able to inhibit

a

20

to AGA were tested

I&

of

Vol.

134,

No.

1, 1986

BIOCHEMICAL

AND

BIOPHYSICAL

RESEARCH

COMMUNICATIONS

Table 2 Pre-treatment (20h)

Labelling medium (2h)

14C-citru11ine/3H-phenylalanine ratio x 100

Normal medium II II I TPA 1000 rig/ml II II II

vehicle only TPA AGA TPA + AGA vehicle only TPA AGA TPA + AGA

28.2 + 0.7 11.0 + 0.4 0.6 7 0.3 0.37 0.1 30.2 T 0.0 30.1? 1.4 1.2 T 0.4 1.0 T - 0.1 Inhibition or ]unctlonal connnunication by AGA in co-cultures refractory to phorbol ester-induced inhibition. Cultures were made refractory to phorbol esters by exposure to 1000 rig/ml TPA for 20h prior to labelling. They were uashed twice with medium with 10% serum and once with serum-free medium, and labelled for 2h in serum-free medium B in the presence of TPA (66 @nl) and/or AGA (10 uM). Results are mean + difference of duplicate walls.

18-B-glycyrrhetinic (Fig.

acid had no effect

on IJC at concentrations

(50 uM), spironolactone

(100 uN), prednisolone

4b). Aldosterone

hydrocortisone

(100 uM), 7-deoxycholate

(30 UM) and betulin concentrations

addition,

(30 uM) all

showed no tendency

the dose-response

the inhibitory

simultaneous

presence

spironolactone

curve

effect

for

(100 uM), (100 uM), lupeol

to inhibit

IJC at

, which were the highest

spironolactone

is shown in Fig.

of AGA (10 uM) was not diminished

of a tenfold

excess

tested.

of the mineralocorticoid

4b.

by the antagonist

(100 uM) (data not shown).

Tumor-promoting acetate)

(60 uM), cholesterol

up to those shown in brackets

For example,

this

up to 100 uM

phorbol

have been previously

co-culture

system

of TPA results

(11).

esters

shown to induce a partial However,

in a refractory

to phorbol

ester

(11).

refractory

to TPA retained

such as TPA (12~0-tetradecanoylphorbol-13-

state

prolonged in which

exposure

the cells

As shown in Table 2, co-cultures their

sensitivity

inhibition

of IJC in

to high concentrations are no longer which

responsive

had been made

to AGA-induced cotununication

inhibition.

DISCUSSION: related

The present

results

compounds 18-P-glycyrrhetinic

demonstrate

inhibition

of IJC by AGA and the

acid and carbenoxolone, 34

in a human

In

Vol.

f

134,

BIOCHEMICAL

No. 1, 1986

ibroblast

Of these,

system.

were investigated

than that

representative

However AGA wx

less

cytotoxic

synthesis

indicating

that AGA acts

new

for

was readily

medium.

junctions.

20 days with

provided

root,

only a minor slowing

receptors

anti-inflammatory

for which glycyrrhetinic affinity junctional

comnunication.

diminished

by

ability

of AGA to inhibit

mineralocorticoid Another

activate therefore

exposure

protein represent

junction-regulatory

kinase

esters

acids

activity

(12,13). receptors,

also have low but demonstrable

tested

had any inhibitory

communication

inhibition

not to be mediated

of inhibiting

capable

such as TPA (5-7,11),

C (16).

effect by

AGA uas not

antagonist. either

on

Thus the

by

IJC are the which are thought

Ccanmunication inhibition

pharmacological mechanism.

for

receptors.

group of substances phorbol

of peptic

and affinity

to a mineralocorticoid

IJC appears

or glucocorticoid

tumor-promoting

from the

may be mediated via glucocorticoid

In addition,

simultaneous

chosen.

(18-fl-glycyrrhetinic

mineralocorticoid

None of the steroids

(13,14,15).

continuously

acid are derived

structure

of

proliferation,

used in the treatment

and glycyrrhizic

to be due

medium was carefully

glabra , and carbenoxolone

properties

onset

the re-opening

could be maintained

acid and glycyrrhizic

and intrinsic

effect

with

is too rapid

of the rate of cell

is a derivative

by its

the cells

reflect

must

of AGA in the culture

Glycyrrhizia

mineralocorticoid

of IJC.

not only on new junction

by washing

inhibition

as retinoic

AGA uas of rapid

rate of reversal

These compounds have steroid-like

Their

properties

of inhibitors

by

gap junctions, reversible

Communication

sodium hemisuccinate)

ulcer.

inhibition

The observed

AGA, 18-@-glycyrrhetinic

acid

and its

(10) which are the most

classes

alone , and therefore

the concentration

liquorice

COMMUNICATIONS

than these compounds , as assessed

on existing

formation

junction

existing

cytotoxic

of tetraphenylboron

Communication

rate.

Inhibition

albumin-containing to

RESEARCH

of IJC was of the same order

compounds of two other

on protein

formation.

AGA was the least

of AGA as an inhibitor

(11) and greater

potent

BIOPHYSICAL

in detail.

The potency acid

AND

activation

However,

esters

of a physiological

co-cultures 35

by phorbol

to

which had been made

may

Vol.

134,

No.

1, 1986

BIOCHEMICAL

refractory

to TPA remained sensitive

probably

does not act via the protein

We conclude

that

AGA is a novel

AND

to the effect kinase

via steroid

receptors

and lack

of toxicity

of AGA should make it

of junctional

or protein

RESEARCH

COMMUNICATIONS

of AGA, indicating

of IJC, an action

kinase

C.

a useful

which

The reversibility

2. 3. 4. 5.

investigative

tool

cotmnunication.

Hertzberg, E.L., Lawrence/ T.S.1 and Gilula, N.B. (1981) Ann. Rev. Physiol. 43, 479491. Loewanstein, W.R. (1981) Physiol. Rev. 61, 829-913. Davidson, J.S., Baumgarten, I.M., and Harley, E.H. (1984) Exp. Cell. Res. 150, 367-378. Davidson, J.S., Baumgarten, I.M., and Harley, E.H. (1985) Cancer Res. 45, 515-519. Yotti, L.P., Chang, C.C., and Trosko, J.E. (1979) Science 206, 1089-1091.

6. 7. 8. 9. 10. 11. 12. 13. 14.

15. 16.

Fitzgerald, D.J., and Murray, A.W. (1980) Cancer Res. 40, 2935-2937. Enomoto, T., Sasaki, Y., Shiba, Y., Kanno/ Y.# and Yamasaki, H. (1981) Proc. Natl. Acad. Sci. U.S.A. 78, 5628-5632. Williams, G.M., Telang, S., and Tong, C. (1981) Cancer L&t. 11, 339-344. Tsushimoto, G., Trosko, J.E., Chang, C.C., and Austr S.D. (1982) Carcinogenesis 3, 181-185. Davidson, J.S., Baumgarten, I.M., and Harley, E.H. Biochim. Biophys. Acta, in press. Davidson, J.S., Bauqarten, I.M., and Harley, E.H. (1985) Carcinogenesis 6, 645-650. * . Armanini, D., Karbowiak, I., Krozowski, Z., Funder, J.W., and Adam, W.R. (1982) Endocrinology 111, 1683-1686. Armanini, D., Ihrbowiak, I., and Funder, J.W. (1983) Clin. Endocrinol. 19, 609-612. Molhuysan, J.A.# Gerbrandy, J., DeVeries, L.A.# Dejong, J.C., Lenstra, J.B., Turner, K.P., and Borst, J.G.G. (1950) Lancet 259, 381-386. E., Ogihara, Y., Ogawar S., Okada, K., Amagaya, S.# Sugishita, and Aizawa, T. (1984) J. Pharm. Dyn. 7, 923-928. Nishizuka, Y. (1984) Nature 308, 693-698.

36

AGA

is not of action

REFERENCES 1.

that

C pathway.

inhibitor

mediated

study

BIOPHYSICAL

in the