+ All Categories
Home > Documents > Comparison of various macrolides on stimulation of human ...€¦ · Comparison of various...

Comparison of various macrolides on stimulation of human ...€¦ · Comparison of various...

Date post: 21-May-2020
Category:
Upload: others
View: 10 times
Download: 0 times
Share this document with a friend
13
Journal of Antimicrobial Chemotherapy (1996) 38, 81-93 Comparison of various macrolides on stimulation of human neutrophil degranulation in vitro H. Abdelghaffar, D. Vazifeh and M. T. Labro* INSERM U294, Service a" Hematologie et cflmmunologie Biologiques, CHU Xavier Bichat, 46 rue Henri Huchard, 75018 Paris, France Macrolide antibiotics are taken up and concentrated by host cells, particularly phagocytes, and are likely candidates to modify cell functions. In this study, we extended our previous work concerning the effect of three 14-membered-ring macrolides (dirithromycin, erythromycin and erythromycylamine) on human neutrophil exocytosis, and found that three other erythromycin A derivatives (roxithromycin, clarithromycin and the azalide, azithromycin) also triggered neutrophil degranulation in a time- and concentration-dependent manner. After 30 min of incubation, the correlation coefficients for concentration-dependence for roxithromycin were 0.885, 0.739 and 0.750 (P < 0.005) and for clarithromycin were 0.795, 0.599, 0.733 (P < 0.02), respectively, for lysozyme, 0-glucuronidase and lactoferrin release. Although the underlying mechanism was not elucidated, these and previous data suggest that intracellular accumulation is a prerequisite. Furthermore, comparison of the characteristics of macrolide-induced exocytosis with those of exocytosis triggered by the synthetic chemotactic stimulus FMLP suggested that different mechanisms are involved. In keeping with this possibility, we showed that combined treatment (macrolides plus FMLP) resulted in totally additive exocytosis of azurophilic but not specific granules. The clinical relevance of our data remains to be ascertained. Introduction Macrolide antibiotics are strongly concentrated within phagocytes (reviewed in Labro 1993). Although the mechanisms underlying this cellular uptake are not fully clear, they may involve trapping by protonation inside acidic cell compartments (lysosomes and azurophilic granules) (Carlier, Zenebergh & Tulkens, 1987). The possibility that intragranular uptake interferes with phagocyte degranulation has rarely been explored (Engquist, Lundberg & Peterson, 1984; Carevic & Djokic, 1988; Anderson, 1989; Abdelghaffar, Mtairag & Labro, 1994). We have previously shown that three 14-membered-ring macrolides (dirithromycin, erythromycylamine and the prototype drug, erythromycin) stimulate human neutrophil degranulation in a time- and concentration-dependent manner (Abdelghaffar et al., 1994). In this study, we extended our investigations to various 14- and 16-membered-ring macrolides, by comparison with erythromycin. We also tested the azalide azithromycin, which is chemically derived •Corresponding author. Tel: +33 (I) 40 25 85 21; Fax: +33 (1) 40 25 88 53. 81 0305-7453/96/070081 + 13 $12.00/0 £, 1996 The British Society for Antimicrobial Chemotherapy
Transcript
Page 1: Comparison of various macrolides on stimulation of human ...€¦ · Comparison of various macrolides on stimulation of human neutrophil degranulation in vitro H. Abdelghaffar, D.

Journal of Antimicrobial Chemotherapy (1996) 38, 81-93

Comparison of various macrolides on stimulation of human neutrophildegranulation in vitro

H. Abdelghaffar, D. Vazifeh and M. T. Labro*

INSERM U294, Service a" Hematologie et cflmmunologie Biologiques,CHU Xavier Bichat, 46 rue Henri Huchard, 75018 Paris, France

Macrolide antibiotics are taken up and concentrated by host cells, particularlyphagocytes, and are likely candidates to modify cell functions. In this study, weextended our previous work concerning the effect of three 14-membered-ringmacrolides (dirithromycin, erythromycin and erythromycylamine) on humanneutrophil exocytosis, and found that three other erythromycin A derivatives(roxithromycin, clarithromycin and the azalide, azithromycin) also triggeredneutrophil degranulation in a time- and concentration-dependent manner. After30 min of incubation, the correlation coefficients for concentration-dependence forroxithromycin were 0.885, 0.739 and 0.750 (P < 0.005) and for clarithromycin were0.795, 0.599, 0.733 (P < 0.02), respectively, for lysozyme, 0-glucuronidase andlactoferrin release. Although the underlying mechanism was not elucidated, these andprevious data suggest that intracellular accumulation is a prerequisite. Furthermore,comparison of the characteristics of macrolide-induced exocytosis with those ofexocytosis triggered by the synthetic chemotactic stimulus FMLP suggested thatdifferent mechanisms are involved. In keeping with this possibility, we showed thatcombined treatment (macrolides plus FMLP) resulted in totally additive exocytosisof azurophilic but not specific granules. The clinical relevance of our data remainsto be ascertained.

Introduction

Macrolide antibiotics are strongly concentrated within phagocytes (reviewed inLabro 1993). Although the mechanisms underlying this cellular uptake are not fullyclear, they may involve trapping by protonation inside acidic cell compartments(lysosomes and azurophilic granules) (Carlier, Zenebergh & Tulkens, 1987). Thepossibility that intragranular uptake interferes with phagocyte degranulation hasrarely been explored (Engquist, Lundberg & Peterson, 1984; Carevic & Djokic, 1988;Anderson, 1989; Abdelghaffar, Mtairag & Labro, 1994). We have previously shownthat three 14-membered-ring macrolides (dirithromycin, erythromycylamine and theprototype drug, erythromycin) stimulate human neutrophil degranulation in a time- andconcentration-dependent manner (Abdelghaffar et al., 1994). In this study, we extendedour investigations to various 14- and 16-membered-ring macrolides, by comparison witherythromycin. We also tested the azalide azithromycin, which is chemically derived

•Corresponding author.Tel: +33 (I) 40 25 85 21; Fax: +33 (1) 40 25 88 53.

81

0305-7453/96/070081 + 13 $12.00/0 £, 1996 The British Society for Antimicrobial Chemotherapy

Page 2: Comparison of various macrolides on stimulation of human ...€¦ · Comparison of various macrolides on stimulation of human neutrophil degranulation in vitro H. Abdelghaffar, D.

82 H. Abdelghaffar, D. Vazifeh and M. T. Labro

from erythromycin and displays an outstanding ability to concentrate within humancells (Gladue et a!., 1989). Part of this work was presented at the thirty-thirdInterscience Conference on Antimicrobial Agents and Chemotherapy, New Orleans,USA (Labro, Abdelghaffar & Bryskier, 1993).

Materials and methods

Macrolides

Erythromycin and roxithromycin (Roussel Uclaf, Romainville, France), clarithromycin(Laboratoire Abbott, Rungis, France), spiramycin (Rhone-Poulec-Rorer, Vitry surSeine, France), josamycin (Pharmuka, Neuilly, France), rokitamycin (LaboratoirePierre Fabre, Paris, France), oleandomycin and azithromycin (Pfizer, Orsay, France),dirithromycin (Eli Lilly, Indianapolis, USA) were kindly provided by the respectivemanufacturers. The drugs were dissolved (1 g/L) in dimethylsulphoxide (DMSO) andfurther diluted in Hanks buffered salt solution (HBSS-Institut Pasteur-) to the desiredconcentrations. DMSO was also diluted in HBSS to serve as a control.

Human polymorphonuclear neutrophils

Human neutrophils were obtained from heparinized venous blood of volunteers byFicoll-Paque centrifugation followed by 2% dextran sedimentation and hypo-osmoticlysis of residual erythrocytes.

Neutrophil viability

Neutrophil viability was measured in all the experimental conditions used here(incubation with macrolides, DMSO or stimuli for up to 180 min, pH 7-9) by the releaseof lactate dehydrogenase (LDH), a cytoplasmic marker enzyme (Bergmeyer, 1963).In all conditions, LDH release represented less than 10% of total enzyme activity,indicating no alteration of cell viability compared to the control (8 ± 1.5%).

Neutrophil degranulation

Neutrophil degranulation was assessed by the release of three granular components:/?-glucuronidase (located within azurophilic granules); lactoferrin (a marker of specificgranules) and lysozyme (located in both granule subsets). /?-glucuronidase release wasassessed by the release of phenolphthalein from phenolphthalein glucuronic acid,monitored spectrophotometrically at 540 nm (Talalay, Fishman & Huggins, 1946).The release of lactoferrin was determined by an enzyme-linked immunosorbent assaywith an antihuman lactoferrin antibody (Hetherington, Spitznagel & Quie, 1983);lysozyme release was measured in terms of the lysis of Micrococcus lysodeikticus,monitored spectrophotometrically at 450 nm (Litwack, 1955).

In standard experiments, neutrophils were incubated for 5-180 min in thepresence of macrolides (1-100 mg/L) or formyl-methionyl-leucyl-phenylalanine (FMLP5 x 1 0 " 6 M ) or control DMSO solutions, and cytochalasin B (5 mg/L). Afterincubation, cells were centrifuged at 400g for 10 min, and enzyme activities weremeasured in the pellet and supernatant. Enzyme release was expressed as the percentage

Page 3: Comparison of various macrolides on stimulation of human ...€¦ · Comparison of various macrolides on stimulation of human neutrophil degranulation in vitro H. Abdelghaffar, D.

Macrolides and neutrophil exocytosis 83

of enzyme activity in the supernatant relative to that in the pellet plus supernatant. Thepositive control was activity in a cell lysate obtained by sonication (three 15-sec burstsat 30% power from a Fisher sonic dismembranator set at 4°C) in the presenceof 0 .1% Triton X-100, which induced optimal enzyme recovery (results not shown).Activity in the pellet plus supernatant did not differ from overall activity measured inthe cell lysate. We also checked that the macrolides (up to 100 mg/L) did not directlymodify the enzyme activities of a cell lysate. Experimental conditions includedvariations in the pH of the medium from 7 to 9 and omission of cytochalasin B.

Effect of macrolides on neutrophil degranulation triggered bv FMLP

Neutrophils were incubated for 30 min in the presence of macrolides (100 mg/L) orDMSO and cytochalasin B, and further stimulated with FMLP (5 x 1 0 ~ 6 M ) . Enzymerelease was measured as indicated above.

Statistical analysis

Results are expressed as the mean ± S.E.M. of n experiments performed with neutrophilsfrom different human volunteers. Experiments with macrolides were paired withcontrols performed in the presence of the corresponding DMSO solution. Analysis ofvariance (ANOVA) test was used for multiple comparisons. Paired data were analyzedby using Student's / test for normal distributions. All statistical tests were performedwith the Statworks program, version 1.2 1985 (Cricket software).

Results

Effect of macrolides on human neutrophil degranulation

At the concentration of 100 mg/L, roxithromycin, clarithromycin, azithromycinand erythromycin significantly triggered neutrophil exocytosis of both lysozyme(Table I) and /J-glucuronidase (Table II) in a time-dependent manner. Althoughthe effects of roxithromycin and clarithromycin appeared earlier, the overall amountof enzyme released did not significantly differ at the end of the 3 h incubationperiod among the four drugs. By contrast, neither the 16-membered-ring macrolides(josamycin and spiramycin) nor oleandomycin promoted neutrophil degranulation,even after a long incubation period. As already observed by us (Abdelghaffarel al., 1994) and other authors (Niessen et al., 1991), FMLP-induced degranulationwas rapid, with maximal enzyme release after 10 min and no further change.Lactoferrin release by macrolide-treated neutrophils was analyzed at 30 and180 min. Only roxithromycin and clarithromycin (100 mg/L) induced significantrelease of lactoferrin by neutrophils at 30 min (36 ± 5.0 and 26 ± 4.7%, P < 0.001respectively for roxithromycin (15 experiments) and clarithromycin (12 experiments)vs 9 ± 1.5% for the control DMSO solution (15 experiments) and 51 ± 6 . 3 % forFMLP (14 experiments). At 180 min, roxithromycin and clarithromycin (twoexperiments) promoted about 50% lactoferrin release by neutrophils, whereasazithromycin was slightly less effective (31 ± 5.1%; 4 experiments; P < 0.05 vs controlDMSO—13 ± 0.8%). None of the other macrolides induced lactoferrin release, evenafter a 180 min incubation period (data not shown).

Page 4: Comparison of various macrolides on stimulation of human ...€¦ · Comparison of various macrolides on stimulation of human neutrophil degranulation in vitro H. Abdelghaffar, D.

Tab

le I

. E

ffec

t of

mac

roli

des

(lO

Om

g/L

) on

lys

ozym

e re

leas

e by

neu

trop

hils

Mac

rolid

es

Rox

ithro

myc

in

Cla

rith

rom

ycin

Azi

thro

myc

in

Ery

thro

myc

in

Ole

ando

myc

in

Spira

myc

in

Josa

myc

in

Con

trol

sF

ML

P(5

x 1

0"M

)

DM

SO (

1%)

•P <

0.0

1. A

nova

fo

llow

ed*N

umbe

r of

exp

erim

ents

is

5

14 ±

2.7

(6)»

16 ±

3.9

(6) 6

(1)

— — — — —

13±2

.1(6

)

10

15 ±

2.2

(9)

18 ±

2.4

(8)

— — — — —

36 ±

5.4

°(4

)12

±

1.4

(9)

by S

tude

nt's

r-t

est

(pai

red

data

)gi

ven

in p

aren

thes

es.

Incu

bati

on

time

15

21 ±

2.5

'(7

)29

± 2

.4'

(8)

12 (1)

15 ±

0.3

(2)

— — —

33 ±

0.1

(2)

13 ±

1.6

(8)

30

38 ±

2.9

°(1

4)35

± 2

.4'

(13)

26 +

0.9

-(3

)24

± 0

.9(2

)13

± 1

.3(6

)10

±0.

8(6

)10

± 1.5

(6)

47 ±

4.7

"(4

)14

± 1.5

(14)

vs c

ontr

ol

DM

SO

sol

utio

ns.

(min

)60

56 ±

5.4

"(1

2)56

± 4

.4°

(11)

40 ±

3.7

°(7

)29

±

1.0

(2)

17 +

4.3

(4)

10 ±

1.0

(2)

11 +

1.0

(2)

42 ±

5.6

°(4

)16

±

1.8(1

2)

120

61 ±

4.4

°(9

)62

± 4

.0°

(8)

42 ±

6.2

°(4

)38

± 6

.6(2

)20

± 1

.8(4

)20

± 2

.5(2

)14

±0

.5(2

)

40 ±

3.5

°(3

)20

± 3

.0(9

)

71 71 54 48 25 25 17 41 21

180

±4.0

"(1

0)±5

.6°

(9)

±2.4

°(6

10.3

°(4

)±0

.6(4

)±2

.5(4

)+

4.0

(4)

±4.0

(2)

±2.6

(10)

> n f I i 2 H r

Page 5: Comparison of various macrolides on stimulation of human ...€¦ · Comparison of various macrolides on stimulation of human neutrophil degranulation in vitro H. Abdelghaffar, D.

Tab

le I

I. E

ffect

of

mac

rolid

es (

lOO

mg/

L)

on /

J-gl

ucur

onid

ase

rele

ase

by n

eutr

ophi

ls

Mac

rolid

es

Rox

ithro

myc

in

Cla

rith

rom

ycin

Azi

thro

myc

in

Ery

thro

myc

in

Ole

ando

myc

in

Spira

myc

in

Josa

myc

in

Con

trol

sFM

LP

DM

SO

1015

Incu

batio

n tim

e (m

in)

3060

120

•/><

0.0l

A

nova

fol

low

ed b

y St

uden

t's /

-tes

t (p

aire

d da

ta)

vs c

ontr

ol D

MSO

sol

utio

ns.

'Num

ber

of e

xper

imen

ts i

s gi

ven

in p

aren

thes

es.

180

1.6(4

/6

± 1.3

(4) 7

(1)

— — — —

5± 1

.5(4

)

1.8(6

)8

± 1.6

(5)

— — — — —

12 ±

3.4°

'(3

)5±

1.2

(6)

2(4

)9

+ 2

(4)

7 +

2(3

) 6(1

)— — —

0(4

)

.4 .3 .3 .3

19 ±

3(9

)17

±2

(4)

12±

1(3

)11

±4

(3)

2(3

)4

±0

(3)

2(3

)

20 ±

1(4

)7±

1(9

)

.1'

.3°

.2 .9 .1 .5 .6 .9°

.9

37 30 38 16 5 17 5

+ 5

.8°

(7)

±4.9

*(5

)±5

.5"

(7)

±6.1

(2)

± 1.0

(2) 6

(11) 5 (1)

± 1.8

°(3

)±0

.9(7

)

48 ±

4(8

)49

±4

(7)

44

±8

(4)

19 ±

4(4

)

1(2

)11 (1

) 9(1

)

16±

2(3

)9

±2

(8)

.8°

.7"

.8"

.3"

.0 .7°

.8

52 ±

7.2

°(9

)52

± 7

.2"

(8)

66 ±

6.5

°(7

)34

± 1

.3'

(3)

8 ±

2.0

(4)

10 ±

3.8

(4)

9 ±

2.3

(3)

20 ±

2.7

°(4

)10

± 3

.1(9

)

2 to n 3 Si § 1 n X 5?

Page 6: Comparison of various macrolides on stimulation of human ...€¦ · Comparison of various macrolides on stimulation of human neutrophil degranulation in vitro H. Abdelghaffar, D.

86 H. Abdelghaffar, D. Vazifeh and M. T. Labro

Concentration-dependence of enzyme release

This was analyzed at 30 min with roxithromycin and clarithromycin and at 60 min withazithromycin (Figure 1A and B). By regression analysis, there was a good correlationbetween roxithromycin, clarithromycin and azithromycin concentrations and therelease of lysozyme (roxithromycin, slope 0.29, r 0.885, P< 0.001; clarithromycin,slope 0.25, r 0.795, P< 0.001; azithromycin, slope 0.30, r 0.945, P < 0.01) and/?-glucuronidase (roxithromycin, slope 0.22, r 0.739, P<0.0l; clarithromycin,slope 0.16, r 0.599, P<0.0\; azithromycin, slope 0.28, r 0.804, P < 0.001). Theconcentration-dependence of lactofernn release by neutrophils incubated for 30 min inthe presence of roxithromycin and clarithromycin was also significant (roxithromycin,slope 0.45, r 0.750, P 0.005; clarithromycin, slope 0.35, r 0.733, P 0.016).

The kinetics of enzyme release was also analyzed with roxithromycin andazithromycin at a therapeutic concentration (achievable in tissues) of lOmg/L(Figure 2). Both macrolides induced significant, time-dependent release of lysozyme and/?-glucuronidase. By regression analysis, we found a close correlation between theincubation time and neutrophil exocytosis compared to control DMSO solutions(P < 0.05 azithromycin and P < 0.01 roxithromycin). After a 3 h incubation period,lysozyme and ^-glucuronidase release was respectively 23 + 2.2% (P 0.039) and18 ± 3.5% (P 0.032), azithromycin, 6 and 7 experiments, and 31 + 2.0% (P 0.004)and 23 + 1.2% (P 0.015), roxithromycin, three experiments, vs control exocytosis in0.1% DMSO (15 ± 1.1 and 9 ± 1.1, six and seven experiments).

Effect of pH on macrolide-induced degranulation

It is widely acknowledged that macrolide antibiotics enter cells better when themedium is alkaline (Labro, 1993). With the dibasic macrolides (dirithromycin anderythromycylamine), uptake measured at 30 min increases over a wide pH range up topH 9 (Mtairag, Abdelghaffar & Labro, 1994). The uptake of other erythromycin Aderivatives such as roxithromycin and clarithromycin has been reported to increase upto pH 8 (Carlier et al., 1987; Ishiguro et al., 1989) and Hand, King-Thompson &Holman (1987) have reported that maximal uptake of roxithromycin occurs atpH values around 8 and then decreases. We thus investigated whether a modificationof the medium pH also affecting the degranulating effect of macrolides, as we hadpreviously observed with dirithromycin and erythromycylamine (Abdelghaffar et al.,1994). Roxithromycin- and clarithromycin-induced exocytosis was little affected bythe external pH (Figure 3) (Anova: roxithromycin: P 0.062, lysozyme; P 0.199,/J-glucuronidase; clarithromycin: P 0.119, lysozyme; P 0.145, /?-glucuronidase).By contrast, azithromycin-promoted degranulation (60 min incubation) was signifi-cantly increased over the pH range (Anova: P 0.039). This effect was similar to thatobserved with the other dibasic compound dirithromycin (60 min incubation), used asa control.

Effect of cytochalasin B on macrolide-induced exocytosis

Cytochalasin B, a fungal alkaloid which disrupts actin filaments, permits thedegranulation of azurophilic granules in neutrophils stimulated with FMLP, andstrongly enhances that of specific granules (Dewald et al., 1989; Niessen et al.,

Page 7: Comparison of various macrolides on stimulation of human ...€¦ · Comparison of various macrolides on stimulation of human neutrophil degranulation in vitro H. Abdelghaffar, D.

Macrolides and neutrophil exocytosis 87

20 40 60 80

100

Macrolide concentration (mg/L)

Figure 1. Correlation between macrolide concentration and enzyme release, (a) Lysozyme release' • ,roxithromycin, 30 min, slope 0.29, r 0.885; D, clanthromycin, 30 mm, slope 0.25, r 0.795, A, aathromycin,60 mm; slope 0 30, r 0.945. (b) /J-glucuronidase release: • , roxithromycin, 30 min, slope 0.22, r 0.739; • .clanthromycin, 30 min, slope 0.16, r 0.599; A . azithromycin, 60 min; slope 0.28, r 0 804.

1991; Roos et al., 1993). We thus studied whether this agent also interfered withthe degranulation process triggered by macrolides. Neutrophils were incubated(30-180 min) in the presence of macrolides (100 mg/L), DMSO or FMLP (5 x 10 "6 M),with or without cytochalasin B. The stimulating effect of FMLP on lysozyme and^-glucuronidase release was completely abolished in the absence of cytochalasin Bthroughout the incubation period (Table III); by contrast, roxithromycin stimulationwas significantly more rapid in the presence of this agent but still occurred and increasedin a time-dependent manner in its absence (Table III). Similar data were obtainedwith clarithromycin (data not shown). This partial cytochalasin B-dependence ofmacrolide-induced exocytosis has also been observed with dirithromycin, erythro-mycylamine and erythromycin (Abdelghaffar et al., 1994).

Effect of macrolides on FMLP-stimulated neutrophil degranulation

As macrolide- and FMLP-promoted neutrophil exocytosis differ with regard tokinetics and cytochalasin B dependence, suggesting that the underlying transductionmechanism(s) or target granule subset differ, we studied whether the combination of

Page 8: Comparison of various macrolides on stimulation of human ...€¦ · Comparison of various macrolides on stimulation of human neutrophil degranulation in vitro H. Abdelghaffar, D.

88 H. Abdelghaffar, D. Vazifeh and M. T. Labro

60 90

Time (min)

120 150 180

Figure 2. Kinetics of lysozyme and /J-glucuronidase release induced by azithromycin and roxithromycinat 10 mg/L. Lysozyme (full lines and closed symbols) and ^-glucuronidase (dashed lines and open symbols).DMSO 0.1% ( • , D), four to six experiments; roxithromycin (A, A), three experiments; azithromycin ( • .O), five or six experiments. *P < 0.05 versus control DMSO solutions.

FMLP plus macrolides had an additive effect on neutrophil exocytosis. Neutrophilswere incubated with roxithromycin (100 mg/L) or DMSO (1%) for 30 min andthen stimulated with FMLP (5 x 10~6M) for 10 min, before measurement of/?-glucuronidase, lactoferrin and lysozyme release. Control (DMSO-treated andFMLP-stimulated) neutrophils released 19 ±2.1 (22 experiments), 54 + 8.0 (7experiments) and 30 ± 2.0% (15 experiments) of the respective granule markers.Roxithromycin-treated and non-stimulated neutrophils released 25 ± 2.4, 42 ± 7.9 and38 ± 2.3% of the granule markers. When roxithromycin-treated neutrophils werefurther stimulated with FMLP, there was a significant increase in ^-glucuronidaserelease: 35 ± 4.0 (P 0.006 vs FMLP alone and P 0.035 vs roxithromycin alone) andlysozyme release: 50 ± 2.5% {P 0.006 vs FMLP alone and P 0.001 vs roxithromycinalone). Lactoferrin-induced degranulation by the drug combination (67 ± 6.9%) didnot differ significantly from that triggered by each stimulus alone (P 0.063).Similar results were obtained with clarithromycin (data not shown).

Discussion

We found that two 14-membered-ring macrolides, roxithromycin and clarithromycinand the azalide azithromycin, but not various 16-membered-ring macrolides oroleandomycin, promoted neutrophil exocytosis in a time- and concentration-dependentmanner (Tables I, II and Figure 1). In a preliminary study (Abdelghaffar el al.,1994), we observed a similar effect with erythromycin, erythromycylamine and dirithro-mycin, three 14-membered-ring macrolides. Although it is too early to propose astructure-activity relationship for this degranulating effect, it is noteworthy that all thecompounds endowed with this property are chemically derived from erythromycin A

Page 9: Comparison of various macrolides on stimulation of human ...€¦ · Comparison of various macrolides on stimulation of human neutrophil degranulation in vitro H. Abdelghaffar, D.

Macrolides and neutrophil exocytosis 89

5 x $

9.0

Figure 3. Effect of pH on macrolide (100 mg/L)-induced enzyme release, (a) Lysozyme-release; (b)/f-glucuronidase-release. Mean of four to nine experiments: 1% DMSO (O) and roxithromycin (O); two tosix experiments: clarithromycin (•) ; three experiments azithromycin ( • ) and dirithromycin (A). Theincubation times were respectively 30 min: roxithromycin and clarithromycin and 60 min: DMSO,azithromycin and dirithromycin.

and thus possess similar carbohydrate components (L-cladinose and D-desosamine).It remains to be determined whether these sugars alone, or their structural combinationwith a lactone ring, confers this property.

The effect of macrolides on neutrophil degranulation has rarely been explored.Engquist, Lundberg & Petterson (1984) observed that neutrophils incubated for 30 minin the presence of erythromycin (125 mg/L) released significantly more proteasesafter FMLP stimulation than control cells, whereas erythromycin alone did notpromote protease release. Anderson (1989) observed no changes in the exocytosis ofneutrophils, regardless of FMLP stimulation, after 30 min of incubation in the presence

Page 10: Comparison of various macrolides on stimulation of human ...€¦ · Comparison of various macrolides on stimulation of human neutrophil degranulation in vitro H. Abdelghaffar, D.

90 H. Abdelghaffar, D. Vazifeh and M. T. Labro

Table III. Effect of cytochalasin B on roxithromycin-induced exocytosis

Incubation time (min)Stimulus Cytochalasin B 30 60 120 180

Lysozyme release (% of total)DMSO + 15 ±4.8 (If 18 ±5 .7 (7) 19 ± 6.6 (4) 18 ±3 .1 (4 )

16 ±5.6 (7) 16 ±5.7 (7) 18 ±6.3 (4) 16 ±2.0 (4)Roxithromycin + 41 ± 4.2 (7) 60 ± 9.1 (7) 55 ± 5.8 (4) 67 ± 9.2 (4)

25 ± 4.1 {If 44 ± 7.0 (7)* 43 ± 7.5 {Af 49 ± 10.0 (4)r

FMLP + 30 ±6.8 (4) 26 ±8 .1 (4) 26 ± 8.6 (4) 28 ± 9.6 (4)16 ±3.7 ( 4 / 15 ±8.2 (4)* 14 ± 6.2 (4)* 16 ± 7.3 (Af

/?-glucuronidase release (% of total)DMSO + 5 ±0.7 (3) 6 ±1 .5 (3) 5 ± 0.6 (3) 8 ± 2.8 (3)

5 ±1.4 (3) 5 ±2 .8 (3) 5 ± 1.8 (3) 9 ± 0.2 (3)Roxithromycin + 22 ± 4.6 (3) 36 ± 8.7 (3) 44 ± 4.9 (3) 60 ±1 .6 (3)

14 ± 4.2 (3)" 20 ±4 .4 (3)* 32 ± 5.4 (If 34 ± 3.6 QfFMLP + 12 ±4.1 (3) 16 ±5 .0 (3) 15 ±7 .1(3) 18 ±6.1 (3)

6 ±1.4 (3)* 7 ±3 .3 (3)* 7 ± 2.4 (3)* 7 ± 2.5 ( 3 /

•Number of experiments is given in parentheses.*/>< 0.001.'P < 0.05 Anova followed by Student's /-test (paired data).

of roxithromycin or erythromycin at concentrations of <20mg/L. By contrast,Carevic & Djokic (1988) have reported that incubation for 15 min with azithromycinand erythromycin (<10mg/L) decreased the release of intragranular enzymes byimmune-complex-stimulated neutrophils. Our data do not conflict with these results, aswe observed that the neutrophil exocytosis induced by erythromycin A-derivedmacrolides is clearly dependent both on the concentration and on the incubation time,and that low concentrations of macrolides (10-25 mg/L) require a minimum of2-3 h to be effective on exocytosis. Furthermore, neither Engquist et al. (1984) norCarevic & Djokic (1988) used cytochalasin B, an agent which appears to favourmacrolide-induced exocytosis (Table III). The apparent decrease in the external releaseof enzymes in the study by Carevic & Djokic (1988) could be because these authors usedimmune complexes as the stimulus, and that part of the degranulation could occurinside rather than outside the cells.

The mechanism underlying the degranulating effect of macrolides remains to beelucidated. Intracellular accumulation appears to be necessary, as suggested by theobservation that experimental conditions which favour macrolide uptake also favourthe degranulating effect. In particular, alkalinization of the medium significantlyenhances the exocytosis stimulated by erythromycylamine and dirithromycin(Abdelghaffar et al., 1994), two dibasic macrolides whose uptake by neutrophils isextremely sensitive to the environmental pH (Mtairag et al., 1994). A similar effectwas shown with azithromycin (Figure 2), which is also a dibasic macrolide. In the caseof roxithromycin and clarithromycin, whose maximal uptake occurs around pH 8(Hand et al., 1987; Ishiguro et al., 1989), we found no significant change in neutrophilexocytosis stimulated by these drugs when the pH exceeded 7.4 (Figure 2). A similareffect has been described with erythromycin A (Abdelghaffar et al., 1994): pH valuesabove 7.4 did not modify ^-glucuronidase release and only slightly affected lysozymerelease.

Page 11: Comparison of various macrolides on stimulation of human ...€¦ · Comparison of various macrolides on stimulation of human neutrophil degranulation in vitro H. Abdelghaffar, D.

Macrolides and neutrophil exocytosis 91

Various authors have reported that drugs which induce intragranular alkalinization,such as monensin and nigericin (two cationic ionophores) (Fittschen & Henson, 1991),chloroquine (an antimalarial weak base which is concentrated within neutrophilgranules) (Fontagne et al., 1989), and bafilomycin (a macrolide antibiotic which inhibitsV-ATPase pumps responsible for intragranular acidification) (Bowman, Siebers &Altendorf, 1988; Swallow et al., 1991) also promote neutrophil degranulation. In thecase of nigericin and monensin, only the exocytosis of azurophilic granules was promotedand the authors proposed that intracellular alkalinization specifically triggered therelease of this granule subset. However, there are no published data on the effect ofmacrolides on intragranular alkalinization. Furthermore, macrolides induce bothazurophilic and specific granule exocytosis, as shown by the release of specific markers(respectively /?-glucuronidase and lactoferrin). The hypothesis that intragranularaccumulation of macrolides is responsible for their effects on neutrophil exocytosis isnot supported either by the results obtained with dirithromycin (Abdelghaffar et al.,1994) and azithromycin (Tables I and II) compared with roxithromycin andclarithromycin. Although the former drugs are significantly more markedly trappedwithin neutrophil granules (Carlier et al., 1987; Gladue et al., 1989; Mtairag et al.,1995), they do not appear to be more effective in inducing neutrophil exocytosis. Thesedata suggest that intragranular uptake is not the only factor underlying the triggeringof neutrophil exocytosis, and that interaction of macrolides with a cytoplasmic target,possibly membranous, is the key event in macrolide-promoted exocytosis.

It is interesting to note that FMLP-induced exocytosis differs from that induced bymacrolides with respect to kinetics, maximal enzyme release and dependence oncytochalasin B, suggesting that macrolides affect a transduction pathway differingfrom that stimulated by FMLP.

The clinical relevance of our findings also remains to be determined. We found thatthe degranulating effect could occur at therapeutic concentrations of the macrolides butrequired long incubation periods; a condition met in tissues. This property could beinvolved in one aspect of macrolide efficacy. These drugs are of recognized value forthe treatment of infections caused by intracellular pathogens, and it is interesting tonote that various microorganisms on which macrolides show intracellular bioactivityin vitro and in vivo, such as Toxoplasma, Mycobacterium avium, Mycobacterium leprae,Legionella spp., Chlamydia spp., reside in phagosomes by inhibiting phagolysosomalfusion (Miller et al., 1984; Bryskier, 1992; Bryskier et al., 1993; Stamm and Suchland,1986). As macrolides are located in both the cytoplasm and the granular-lysosomalcompartment (Labro, 1993), it is tempting to deduce that these drugs, by promotingphagolysosomal fusion, counteract the inhibitory effect of the microorganisms, thusexposing the latter to both the drugs and granular microbicidal components. Furtherinvestigations are required to verify this hypothesis.

Acknowledgement

The expert secretarial assistance of Miss Francoise Breton was greatly appreciated.

ReferencesAbdelghaffar, H., Mtairag, E. M. & Labro, M. T. (1994). Effects of dirithromycin and

erythromycylamine on human neutrophil degranulation. Antimicrobial Agents and Chemo-therapy 38, 1548-4.

Page 12: Comparison of various macrolides on stimulation of human ...€¦ · Comparison of various macrolides on stimulation of human neutrophil degranulation in vitro H. Abdelghaffar, D.

92 H. Abdelghaffar, D. Vazifeh and M. T. Labro

Anderson, R. (1989). Erythromycin and roxithromycin potentiate human neutrophil locomotionin vitro by inhibition of leukoattractant-activated superoxide generation and auto-oxidation.Journal of Infections Diseases 159, 966-73.

Bergmeyer, H. U. (1963). Methods in Enzymatic Analysis, pp. 737-9. Academic Press, NewYork.

Bowman, E. J., Siebers, A. & Altendorf, K. (1988). Bafilomycins: a class of inhibitors ofmembrane ATPases from microorganisms, animal cells and plant cells. Proceedings of theNational Academy of Sciences USA 85, 7972-6.

Bryskier, A. (1992). Newer macrolides and their potential target organisms. Current Opinion inInfectious Diseases 5, 764—72.

Bryskier, A., Butzler, J. P., Neu, H. C. & Tulkens, P. M. (1993). Macrolides: Chemistry,Pharmacology and Clinical Use. Arnette-Blackwell, Paris.

Carevic, O. & Djokic, S. (1988). Comparative studies on the effects of erythromycin A andazithromycin upon extracellular release of lysosomal enzymes in inflammatory processes.Agents and Actions 25, 124-31.

Carlier, M. B., Zenebcrgh, A. & Tulkens, P. M. (1987). Cellular uptake and subcellulardistribution of roxithromycin and erythromycin in phagocytic cells. Journal of AntimicrobialChemotherapy 20, Suppl. B, 47-56.

Dewald, B., Thelen, M., Wymann, M. P. & Baggiolini, M. (1989). Staurosporine inhibits therespiratory burst and induces exocytosis in human neutrophils. Biochemistry Journal 264,879-84.

Engquist, S., Lundberg, C. & Petterson, C. (1984). Effect of phenoxymethyl penicillin,erythromycin, lymecycline and doxycycline upon the release of proteases from humanleukocytes. European Journal of Clinical Microbiology 2, 221-3.

Fittschen, C. & Henson, P. M. (1991). Selective secretion of azurophil granule contents inducedby monovalent cation ionophores in human neutrophils: evidence for direct ionophore effectson the granule membrane. Journal of Leukocyte Biology 50, 517-28.

Fontagne, J., Roch-Arveiller, M., Giroux, J. P. & Lcchat, P. (1989). Effects of some antimalarialdrugs on rat inflammatory polymorphonuclear leukocyte function. Biomedical Pharma-cotherapy 43, 43-51.

Gladue, R. P., Bright, G. M., Isaacson, R. E. & Newborg, M. F. (1989). In vitro and in vivouptake of azithromycin (CP-62, 993) by phagocytic cells: possible mechanism of delivery andrelease at sites of infection. Antimicrobial Agents and Chemotherapy 33, 277-82.

Hand, W. L., King-Thompson, N. & Holman, J. W. (1987). Entry of roxithromycin (RU 965),imipenem, cefotaxime, trimethoprim and metronidazole into human polymorphonuclearleukocytes. Antimicrobial Agents and Chemotherapy 31, 1553-7.

Hetherington, S. V., Spitznagel, J. K. & Quie, P. G. (1983). An enzyme-linked immunoassay(ELISA) for measurement of lactoferrin. Journal of Immunological Methods 65, 183-90.

Ishiguro, M., Koga, H., Kohno, S., Hayashi, T., Yamaguchi, K. & Hirota, M. (1989). Penetrationof macrolides into human polymorphonuclear leucocytes. Journal of AntimicrobialChemotherapy 24, 719-29.

Labro, M. T. (1993). Intraphagocytic penetration of macrolide antibiotics. In Macrolides:Chemistry, Pharmacology and Clinical Use (Bryskier, A., Butzler, J. P., Neu, H. C. &Tulkens, P. M., Eds), pp. 379-88. Arnette-Blackwell, Paris.

Labro, M. T., Abdelghaffar, H. & Bryskier, A. (1993). Effect of macrolides (M) on humanneutrophil (PMN) degranulation. In Program and Abstracts of the Thirty-Third InterscienceConference on Antimicrobial Agents and Chemotherapy, New Orleans, 1993. Abstract 309.American Society for Microbiology, Washington, DC.

Labro, M. T., Abdelghaffar, H. & Mtairag, E. M. (1994). Extracellular Ca+ + is required formacrolide uptake by human neutrophils (PMN) and subsequent PMN exocytosis.In Abstracts of the Ninety-fourth Annual Meeting of the American Society for Microbiology,Washington, DC. 1994. Abstract El 10. American Society for Microbiology, Washington,DC.

Litwack, G. (1955). Photometric determination of lysozyme activity. Proceedings of the Societyfor Experimental Biology and Medicine 89, 401-3.

Miller, M. F., Martin, J. R., Johnson, P., Ulrich, J. T., Rdzok, E. J. & Billing, P. (1984).Erythromycin uptake and accumulation by human polymorphonuclear leukocytes and

Page 13: Comparison of various macrolides on stimulation of human ...€¦ · Comparison of various macrolides on stimulation of human neutrophil degranulation in vitro H. Abdelghaffar, D.

Macrolides and neutrophil exocytosis 93

efficacy of erythromycin in killing ingested Legionella pneumophila. Journal of InfectiousDiseases 149, 714-8.

Mtairag, E. M., Abdelghaffar, H., Douhet, C. & Labro, M. T. (1995). Role of extracellularcalcium in vitro uptake and intraphagocytic location of macrolides. Antimicrobial Agents andChemotherapy 39, 1676-82.

Mtairag, E. M., Abdelghaffar, H. & Labro, M. T. (1994). Investigation of dirithromycin anderythromycylamine uptake by human neutrophils in vitro. Journal of AntimicrobialChemotherapy 33, 523—36.

Niessen, H. W. M., Kuijpers, T. W., Roos, D. & Verhoeven, A. J. (1991). Release of azurophilicgranule contents in fMLP-stimulated neutrophils requires two activation signals, one ofwhich is a rise in cytosolic free Ca2 + . Cellular Signalling 3, 625-33.

Roos, D., Kuijpers, T. W., Mascart-Lemone, F., Koenderman, L., de Boer, M., Van Zwieten,R. et al. (1993). A novel syndrome of severe neutrophil dysfunction: unresponsivenessconfined to chemotaxin-induced functions. Blood 81, 2735—43.

Stamm, W. E. & Suchland, R. (1986). Antimicrobial activity of U-70138F (paldimycin),roxithromycin (RU965) and ofloxacin (ORF 18489) against Chlamydia trachomatis in cellculture. Antimicrobial Agents and Chemotherapy 30, 806-7.

Swallow, C. J., Sudsbury, R. A., Grinstein, S. & Rotstein, O. D. (1991). Human monocytedegranulation is independent of secretory granule acidification. In Program and Abstracts ofthe Thirty-First Interscience Conference on Antimicrobial Agents and Chemotherapy, Chicago,1991. Abstract 1046. American Society for Microbiology, Washington, DC.

Talalay, P., Fishman, W. H. & Huggins, C. (1946). Chromogenic substrates. II. Phenolphtaleinglucuronic acid as substrate for the assay of glucuronidase activity. Journal of BiologicalChemistry 166, 756-72.

(Received 12 April 1995; returned 19 May 1995; revised 20 November 1995;accepted 20 February 1996)


Recommended