An" Otol RJoiIwlLaryngollOl:l992

ERYTHROMYCIN INHIBITION OF LIPOPOLYSACCHARIDESTIMULATED TUMOR NECROSIS FACTOR ALPHA PRODUCTION BY HUMAN MONOCYTES IN VITRO YUKIKO IINO, MD MINORU TORIYAMA, MD

KOICHIRO KUDO, MD

YASUHIRO NATORI, PHD

AKIRA YUo, MD TOKYO, JAPAN

The mechanism of clinical effectiveness of low-dose and long-term erythromycin (EM) treatment for diffuse panbronchiolitis, sinobronchial syndrome, and associated otitis media with effusion was investigated by studying the effects ofEM on tumornecrosis factor alpha (TNF-a) production by cultured human monocytes stimulated with lipopolysaccharide. At concentrations ofO.ll1g1mL or more, EM inhibited TNF-a release from human monocytes stimulated by lipopolysaccharide in a dose-dependent manner. Of the other macrolides tested, roxithromycin, an EM derivative, also showed significant inhibition ofTNF-a production, whereas josamycin failed to inhibit TNF-a release from monocytes. Nonmacrolidic drugs such as minocycline hydrochloride, ofloxacin, or penicillin G had no significant effect on TNF-a production. These results suggest that the clinical improvement of chronic respiratory diseases by EM may depend on the suppression of production of inflammatory cytokines such as TNF-a. KEY WORDS - erythromycin, lipopolysaccharide, monocytes, sinobronchial syndrome, tumor necrosis factor alpha.

competent cells such as macrophages, lymphocytes, and neutrophils.6-1S Thesecells secrete various inflammatory mediators, such as arachidonic acid metabolites, enzymes, and cytokines, that control or aggravate inflammation. Tumor necrosis factor alpha (TNFa), a cytokine produced by activated monocytes and lymphocytes, may play an important role in potentiating and prolonging chronic inflammation. However, there have been no reports on the effect of EM on TNF-a release. In the present investigation, we studied the effects ofEM compared to other antimicrobial compounds on TNF-a production by human peripheral monocytes primed with bacteriallipopolysaccharide (LPS) in vitro.

INTRODUCTION

Low-dose, long-term erythromycin (EM) treatment has recently been reported to be very effective for patients with sinobronchial syndrome.F' a clinical entity in which chronic sinusitis is accompanied by chronic lower respiratory tract diseases such as chronic bronchitis, bronchiectasis, diffuse panbronchiolitis, and bronchial asthma. Otitis media with effusion is also frequently found in cases ofsinobronchial syndrome, and we have shown that this is also improved dramatically with EM treatment.' The mechanisms underlying improvement of chronic inflammation in the upper and lower respiratory tracts are still unclear. However, the clinical efficacy of low-dose, long-term EM treatment has been suggested to have little to do with the antimicrobial effect of EM on the pathogens in the respiratory tract, because the treatment is also effective in patients with pathogens insensitive to EM.s On the other hand, some patients with sinobronchial syndrome show a high titer of cold hemagglutinin response and a positive rheumatoid reaction, which in a study by Kudoh et al2 returned to normal after EM treatment. Kudoh et al suggested that EM treatment normalized the hyperimmune response that might be the cause of chronic intractable inflammation. In accordance with this hypothesis, it has been reported that EM influences the cellular function of immuno-

MATERIALS AND METHODS

Materials. Erythromycin base was obtained from Shionogi Pharmaceutical Co, Osaka, Japan; roxithromycin(RXM)fromEisaiPharmaceuticaICo, Tokyo, Japan;josamycin (JM) from YamanouchiPharmaceutical Co, Tokyo; and ofloxacin (OFLX) from Daiichi Pharmaceutical Co, Tokyo. Each antibiotic sample was suspended in distilled water (l mglmL) and sonicated. Minocycline hydrochloride (MIND, Nippon Ledare Pharmaceutical Co, Tokyo) and penicillin G (PC-G, Meiji Pharmaceutical Co, Tokyo) for injection were dissolved in distilled water atconcentrations 00 mglmL and 1 x 106 U/mL, respectively. Es-

From the Departments of Otolaryngology (Iino, Toriyama) and Respiratory Diseases (Kudo) and the Institute ofC1inical Research (Natori, Yuo), National Medical Center, Tokyo, Japan. Presented at the conference on Immunologic Studies on the Ear, Oita, Japan, October 8-9,1991. REPRINTS - Yukiko Iino, MD, Dept of Otolaryngology, Teikyo University School of Medicine, Kaga 1-21-1, Itabashi-ku, Tokyo 173, Japan.

16

lino et al, Erythromycin Inhibition ofTNF-a Production TNF-a PRODUcnON OF HUMAN MONOCYTES INDUCED BY LPS WITHANDWITHOUT EM TNF-a in Medium (pg/mL)

LPS Dose (ng/mL)

o 25 100

Without EM (n 4) 16.2 z 6.6 379.7 z 12.0 409.6 z 14.4

=

With EM (I ~g/mL) (n 4) 0.8 z 1.0 204.9 z 11.7* 111.8 z 14.6*

=

Data are mean ± SEM. TNF-a - tumor necrosis factor alpha, LPS - lipopolysaccharide, EM - erythromycin. *Signif"lC8Dtly different from control culture without EM (p < .01).

cherichia coli 0111:B4 LPS was obtained from Difco Laboratories, Detroit, Mich, and was suspended in distilled water (1 mglmL) and sonicated. Fresh serial dilutions of drugs were prepared on the day of each experiment in RPMI 1640 culture medium (Nissui, Tokyo). Human Monocyte Cultures. Heparinized blood obtained from healthy adults was diluted twice with phosphate-buffered saline. The blood was stratified on Ficoll-Hypaque (Separate-L, Muto Pure Chemical Co, Tokyo) and then centrifuged to isolate mononuclear cells. The mononuclear cells thus obtained were washed three times with Hanks' balanced salt solution and were suspended in RPMI 1640 medium containing 10% fetal calf serum (1 x 1()6 cells per milliliter). Then 270 JAL of the cell suspension was preincubated with 10 JAL of EM, other antibiotic preparations, or RPMI medium as a control at 3rC for 24 hours in culture wells under 5% carbon dioxide. Afterpreculture, 10 JALofLPS suspension (final concentration 25 or 100 nglmL) and a further 10 JAL of antibiotic preparation, together with interferon-y (final concentration 500 U/mL, Shionogi Pharmaceutical Co) to prime the cells, were added to each well, and then incubated for 24 hours under the same conditions. The cells were removed by centrifugation and the resulting supernatant was stored at -20°C until the TNF-a measurement. Cell Viability. Determination of cell viability as measured by trypan blue exclusion was carried out at the end ofthe culture. Cell viability was not altered by any concentration of the drugs tested. Tumor Necrosis Factor Alpha Determination. The amount ofTNF-a in each supernatant was assayed by an enzyme-linked immunosorbent assay. Briefly, each well was coated with 10 ~glmL (in 0.1 mollL carbonate buffer) of monoclonal antirecombinant TNF-a at 4°C overnight. After being washed several times, the well was blocked with 1% bovine serum albumin (in phosphate-buffered saline) for at least 2 hours at room temperature. Individual samples, negative controls, or serial dilutions of recombinant hu-

-

500

.e

Q

400

.5! U

r:::

300

..

200

E

::l '0 0 Do

U.

z

17

•• ....

100

.... 0

o

0.001 0.01 0.1

1

10

dose of EM (,ug/mi) Fig 1. Effect of erythromycin (EM) on tumor necrosis factor alpha (TNF) production by cultured human monocytes stimulated by Escherichia coli lipopolysaccharide. Monocyte suspension (270 ~L at 1 x l()6 cells per milliliter) was cultured with interferon-y (final concentration 500 U/mL) and 10 ~L of lipopolysaccharide suspension (final concentration 100 ng/mL). Various doses of EM (final concentration 0.001 to 10.0 ~g/mL) were added to each culture medium at onset of preculture and culture. AmountofTNF-a in medium was determined by enzymelinked immunosorbent assay. Results are means z SEM of four replicates. Single asterisk- p < .05 as compared with cultures without EM, double asterisk - p < .025 as compared with cultures without EM.

man TNF-a were added to the wells and incubated at 4°C overnight. Rabbit antirecombinant TNF-a (1 to 5,000) and then horseradish peroxidase monoclonal rat anti rabbit immunoglobulin G (1 to 1,000, Zymed Laboratories, Inc, San Francisco, Calif) were added and incubated at room temperature for at least 2 and 3 hours, respectively. Finally, 0.4 mg/ml, of o-phenylenediamine dihydrochloride and hydrogen peroxide solution (0.0006%) in 0.1 mollL citrate phosphate buffer (pH 5.0) were applied to each well and left for 30 minutes in the dark. The reaction was terminated by the addition of 40 JAL per well of2.5 mollL sulfuric acid. The absorbance of each well was measured at 490 nm by using a microplate reader (Bio-Rad Laboratories, Richmond, Calif). Between each step, the plates were gently washed eight times with 0.05% Tween-20 in phosphate-buffered saline.

Statistical Analysis. All values were expressed as the mean ± SEM. Statistical differences were determined by Student's t test. RESULTS

Escherichia coli LPS at final concentrations of both 25 and 100 nglmL induced human peripheral monocytes to release TNF-a into the culture medium (see Table). The TNF-a production was slightly higher with LPS at the dose of 100 nglmL than at 25 nglmL. On the other hand, little TNF-a was released into the culture medium without the addition ofLPS. Erythromycin itselfdid not stimulate TNF-a produc-

18

lino et al, Erythromycin Inhibition ofTNF-a Production 800

-E

1000

0,

.e:

800

c:

600

.~

§

Q

E

0,

.e: c:

0

0

:;;

U

:::l

-g 0

...

z

I-

400

(J

:::l

400

-g 0

...

Q.

ll.

600

Q.

200 0

O+-----.---.-----.----,------.r---,

o

200

u..

10

20 time

30

(hr)

Fig 2. Tumor necrosis factor alpha (TNF) concentrations in medium of cultured human monocytes at time of addition (0 hour) or at various intervals (I, 3, 6, 12, and 24 hours) after addition of lipopolysaccharide (final concentration 100 nglmL) and interferon-y. Cells were precultured with 1.0 ug/ml, of erythromycin for 24 hours. Results are expressed as mean ± SEM (n = 4).

tion of cultured monocyte without LPS, even if the cells were printed by interferon-yo If the cells were incubated with 1 ~glmL (final concentration) of EM during the period of the culture as well as preculture, TNF-a production of monocytes induced by LPS was significantly inhibited, although the suppression by EM was greater in cultures stimulated by 100 ngl mLofLPS. Figure 1 shows the effect of EM concentration on TNF-a production induced by 100 nglmL of E coli LPS. Without EM treatment, 436.5 ± 40.8 pglmL of TNF-a was detected in the supernatant of monocyte culture. The TNF-a production was inhibited by EM in a dose-dependent manner, the reduction being statistically significant at or above an EM concentration of 0.1 ug/ml., The time course of TNF-a production by monocytes stimulated with LPS and interferon-y was studied by using cells precultured with 1.0 ug/rnl, of EM for 24 hours. One hour after the addition of LPS and interferon-y, little TNF-a release was observed, but the TNF-a concentration of medium increased rapidly between 1 and 6 hours, as shown in Fig 2. When 1.0 ~glmL of EM was further added into the culture medium at 0, 1, 3, 6, 12, or 24 hours from the onset of the culture, the TNF-a production measured at 24 hours was significantly inhibited in the cases of addition times 0 and 1 hour. On the other hand, no significant effect on TNF-a production was observed when EM was added at or after 3 hours (Fig 3). We then studied whether other antibiotics would have the same effect on TNF-a production by cultured monocytes. The concentration ofthe antibiotics tested was 1 ~glmL, except for PC-G, which was

Z I-

o

o

1

3

6

12

24

time of EM addition after culture (hr)

Fig 3. Tumor necrosis factor alpha (TNF) concentrations in culture medium 24 hours after addition oflipopolysaccharide (100 nglmL) and interferon-y. Following preculture with erythromycin (EM; 1.0 l!g1mL)for 24 hours, 1.0 1!g1 mL of EM was added to culture medium at 0, 1,3,6,12, or 24 hours after lipopolysaccharide stimulation. Each bar represents mean ± SEM of six replicates. Asterisk - p < .025 as compared with 24-hour group.

used at 6 U/mL. These values corresponded to the levels found in the blood during the daily administration of therapeutic doses of the various drugs in humans. The results are presented in Fig 4. None of these antibiotics induced TNF-a production by themselves in the absence of LPS. Apart from EM, only RXM showed significant suppression of the release ofTNF from monocytes induced by LPS. Minocycline had a slight depression, but the difference was not statistically significant. There was no pronounced effect of JM, OFLX, or PC-G compared with the control TNF-a production in the absence of antibiotics. DISCUSSION

Erythromycin has been in clinical use for the last 30 years for the treatment ofvarious types ofbacterial infections. In recent years, EM has been accepted as the drug ofchoice in many clinical situations, particularly in infections caused by Mycoplasma species, Chlamydia trachomatis, Bordetella pertussis, and Legionella pneumophilia that survive after phagocytosis. The effectiveness of long-term and low-dose EM treatment for diffuse panbronchiolitis has been reported.s and EM treatment has been extended to the therapy of sinobronchial syndrome, chronic lower respiratory tract inflammatory disease without sinusitis, chronic intractable sinusitis, and otitis media with effusion. It has been suggested that the improvement of chronic respiratory inflammation is not dependent on the antimicrobial effect of EM, but on its anti-inflammatory effects and the regulation of the hyperimmune response.I The proposed sites of action of EM include 1) chemotactic activity and superoxide production by neutrophils,6-9 2) proliferation

lino et al, Erythromycin Inhibition ofTNF-a Production

500

activated to produce inflammatory cytokines such as TNF-a, resulting in a chronic inflammatory response. The present study demonstrated that EM inhibited TNF-a production significantly at the dose ofO.1 Ilgl mL or more. In low-dose EM treatment, the highest concentration of EM in blood and in sputum reaches around 1.0 IlglmL. Therefore, therapeutic levels of EM should be sufficient to cause inhibition of the TNF-a production by monocytes.

:::E

C!

400

e

300

~ 0

+: u

:::l

"... 200 0

Q.

LI.

...z

100 0

19

none

JM

MINO OFLX PC-G

Fig 4. Effect of antibiotics on tumor necrosis factor alpha (TNF) production by cultured human monocytes stimulated by lipopolysaccharide. Open bar represents amount ofTNF in medium of monocytes stimulated by lipopolysaccharide (100 nglmL) without any antibiotic treatment. Closed barrepresents that with erythromycin (EM; 1.0 1!g1 mL) treatment, and stippledbars that with other antibiotics (1.0 l!g1mLexcept for penicillin G [PC-G], which was 6 UI mL). Values are expressed as mean :t SEM (n = 4). Asterisk - p < .025 as compared with group without antibiotics, RXM - roxithromycin, 1M - josamycin, MINO - minocycline, OFLX - ofloxacin.

and interleukin-I (IL-I) production of monocytes and macrophages,10-12 3) activation of T lymphocytes and natural killer cells, 13, 14 and 4) enhancement of the ciliary function of ciliated cells. IS Concerning the effect of EM on cytokine production by monocytes and macrophages, Kita et allO reported that EM treatment enhanced IL-I production in patients with diffuse panbronchiolitis. On the otherhand, Takeshita et al ll found that clarithromycin, an EM derivative, inhibited the IL-I production by murine peritoneal macrophages. In the present study, we investigated the effect of EM on the production of another monokine, TNF-a, and showed that the release ofthis cytokine from human monocytes stimulated with LPS in vitro was inhibited by EM. Lipopolysaccharide is known to stimulate monocytes and macrophages to produce various kinds of inflammatory mediators. Since gram-negative bacteria such as Hemophilus influenzae and Pseudomonas aeruginosa are frequently recovered in chronic respiratory tract infections, the secretion contains a high amount of bacterial LPS. For example, in one study middle ear effusions contained IOO nglmL to I Ilgl mL ofLPS.16 In the present study, 25 and IOO nglmL of LPS stimulated TNF-a production by cultured human monocytes. Therefore, in the respiratory tract, monocytes and macrophages might be expected to be

It is possible that other antibiotics may have an effect on immunocompetent cells. Therefore, we tested RXM and 1M as macrolidic drugs and MINO as a tetracycline, OFLX as a new quinolone, and PCG as a penicillin, all of which are frequently used for the treatment of respiratory infections. The present study demonstrated that only EM and RXM suppressed TNF-a production by monocytes. These are both macrolides consisting of a I4-member lactone ring. On the other hand, neither 1M, which is a macrolide consisting of a Ifi-member lactone ring, nor MINO, OFLX, or PC-G showed any inhibitory effects at the concentrations tested. In accordance with this notion, long-term, low-dose treatment with 1M and OFLX for diffuse panbronchiolitis was clinically unsatisfactory.17,18

In contrast to ~-lactams and aminoglycosides, EM and RXM have been reported to accumulate quickly in phagocytic cells such as macrophages and polymorphonuclearleukocytes.19-21 The uptake of these antibiotics by phagocytes is thought to depend on an active metabolic process.P The highly accumulated drugs in the phagocytes may alter cellular functions. Villa et al22 found that EM, and 1M, inhibited the primary antibody response elicited in vitro by human mononuclear cells, and suggested that the suppression by the macrolides, which can penetrate into phagocytic cells, might depend on an increase in the production of activated oxygen species from antigen-stimulated cells. Furthermore, Fraschini et al 23 demonstrated that EM potentiated phagocytosis by an increase in ingestion of microorganisms and by the production of superoxide anion. Therefore, the suppressive effects on cytokine production might result from increased production of active oxygen species. Our results failed to show a suppressive effect of 1M on TNF-a production, contrary to the results by Villa et al described above. Further study is necessary to elucidate the mechanism of the inhibitory effect of EM and other macrolides on monocyte function.

ACKNOWLEDGMENT - We thankDr R. M. Hopps of the Department of Oral and Maxillofacial Surgery, Institute of Dental Surgery, University of London, England. for valuable comments on the manuscript.

1. Kikuchi S, Suzaki H, Aoki A, Ito

REFERENCES 0, Nomura Y. Clinical sinusitis [in Japanese with English abstract]. Pract Otol (Kyoto)

effect of long-term low-dose erythromycin therapy for chronic

1991;84:41-7.

20

lino et al, Erythromycin Inhibition ofTNF-a Production

2. Kudoh S, Uetake T, Hagiwara K, et al. Clinical effect of low-dose long term erythromycin chemotherapy on diffuse panbronchiolitis [in Japanese with English abstract]. Nippon Kyobu Shikkan Gakkai Zasshi 1987;25:632-42.

12. Roche Y, Gougerot-Pocidalo MA, Fay M, Forest N, Pocidalo JJ. Macrolides and immunity: effects of erythromycin and spiramycin on human mononuclear cell proliferation. J Antimicrob Chemother 1986;17:195-203.

3. Sawaki M, Mikami R, Kikasa K, et al. The long-term chemotherapy with erythromycin in chronic lower respiratory tract infections. I. Comparing with amoxicillin [in Japanese with English abstract]. Kansenshogaku Zasshi 1985;60:37-44.

13. Sugiyama Y, Sugama Y, Takeuchi K, Kudoh S, Kitamura S. Analysis of peripheral lymphocyte subsets and changes due to erythromycin therapy in patients with diffuse panbronchiolitis [in Japanese with English abstract]. Nippon Kyobu Shikkan Gakkai Zasshi 1990;28:1574-80.

4. lino Y, Sugita K, Toriyama M, Kudo K. Low-dose and long-term erythromycin treatment for otitis media with effusion associated with sinobronchial syndrome [in Japanese]. Nippon Jibiinkoka Gakkai Kallio 1991;94:1537. 5. Sawaki M, Mikami R, Kikasa K, et al. The long-term chemotherapy with erythromycin in chronic lower respiratory tract infections. II. Including cases with pseudomonas infections [in Japanese with English abstract]. Kansenshogaku Zasshi 1985; 60:45-50.

14. Mikasa K, Sawaki M, Konichi M, et al. The effect of erythromycin treatment on natural killer cell activity in patients with chronic lower respiratory tract infections [in Japanese with English abstract]. Kansenshogaku Zasshi 1989;63:811-5. 15. TamaokiJ, TakeyamaK, ChiyotaniA, SakaiN, Yamauchi F, Takizawa T. Effect of roxithromycin on ciliary motility of rabbit tracheal epithelium in culture [in Japanese with English abstract]. Kokyu To Junkan 1991;39:481-5.

6. Anderson R, Fernandes AC, Eftychis HE. Studies on the effects of ingestion of a single 500 mg oral dose oferythromycin stearate on leukocyte motility and transformation and on release in vitro ofprostaglandin E2 by stimulated leukocytes. J Antimicrob Chemother 1984;14:41-50.

16. lino Y, Kaneko Y, Takasaka T. Endotoxin in middle ear effusions tested with Limulus assay. Acta Otolaryngol (Stockh) 1985;100:42-50.

7. Abo P, Miinnislli PT. Effects of two erythromycins, doxycycline and phenoxymethylpenicillin on human leukocyte chemotaxis in vitro. J Antimicrob Chemother 1988;21(suppl D):29-32.

18. Yamamoto M. Long term therapeutic effects of erythromycin and new quinolone antibacterial agent on diffuse panbronchiolitis [in Japanese]. Ther Res (Japan) 1990;11:954-7.

8. Miyachi Y, Yoshioka A, Imamura S, Niwa Y. Effect of antibiotics on the generation of reactive oxygen species. J Invest DeonatoI1986;86:449-53. 9. Anderson R. Erythromycin and roxithromycin potentiate human neutrophil locomotion in vitro by inhibition of leukoattractant-activated superoxide generation and autooxidation. J Infect Dis 1989; 159:966-73. 10. Kita E, SawakiM, Nishikawa F, et al. Enhanced interleukin production after long-term administration of erythromycin stearate. Phannacology 1990;41:177-83. 11. TakeshitaK, YamagishiI, HaradaM,Otomo S, Nakagawa T, Mizushima Y. Immunological and anti-inflammatory effects of clarithromycin: inhibition of interleukin 1 production of murine peritoneal macrophages. Drugs Exp Clin Res 1989; 15:52733.

17. Orizu T. Efficacy of macrolides except for erythromycin [in Japanese]. Ther Res (Japan) 1990; 11:973-4.

19. Johnson JD, Hand WL, Francis JB, King-Thompson N, Corwin RW. Antibiotic uptake by alveolar macrophages. J Lab Clin Med 1980;95:429-39. 20. Carlier MB, Zenebergh A, Tulkens PM. Cellular uptake and subcellular distribution of roxithromycin and erythromycin in phagocytic cells. J Antimicrob Chemother 1987;20(suppl B):47-56. 21. Prokesch R, Hand WL. Antibiotic entry into human polymorphonuclear leukocytes. Antimicrob Agents Chemother 1982;21 :373-80. 22. Villa ML, Valenti F, Mantovani M, Scaglione F, Clerici E. Macrolidic antibiotics: effects on primary in vitro antibody responses. Int J ImmunopharmacoI1988;10:919-24. 23. Fraschini F, Scaglione F, Ferrara F, Marelli 0, Graga PC, Teodori F. Evaluation of the immunostimulating activity of erythromycin in man. Chemotherapy 1986;32:286-90.

Erythromycin inhibition of lipopolysaccharide-stimulated tumor necrosis factor alpha production by human monocytes in vitro.

The mechanism of clinical effectiveness of low-dose and long-term erythromycin (EM) treatment for diffuse panbronchiolitis, sinobronchial syndrome, an...
2MB Sizes 0 Downloads 0 Views