INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Erythromycin (EM) is known as a potent antibiotic for the treatment of various microbial infections. Considerable attention has been given to the multiple biologic actions of EM, such as inhibition of neutrophil chemotaxis (1), inhibition of interleukin-8 (IL-8) generation (2), inhibition of Cl-secretion by tracheal epithelial cells (3), and antilymphocytic activity (4). In view of its unique pharmacologic actions, EM may function as an immunomodulator. In fact, low-dose and long-term EM treatment has been reported to be effective against diffuse panbronchiolitis (DPB), proposed as a clinicopathologic disease entity characterized by chronic inflammation with inflammatory-cell infiltration, which is predominantly localized in the respiratory bronchioles (5, 6). EM is also used as a therapeutic agent for other chronic inflammations of the lower respiratory tract, such as chronic bronchitis and bronchiectasis, irrespective of its antimicrobial activity.

In our previous reports on influenza-virus-induced pneumonia in mice, we proposed that free radicals such as the superoxide anion (O₂^·) and nitric oxide (NO) are the primary pathogenic molecules in viral lung injury, because the elimination of O₂^· by superoxide dismutase, a scavenger of O₂^·, or the inhibition of NO generation by N^G-monomethyl-L-arginine (L-NMMA), an inhibitor of nitric oxide synthase (NOS), significantly reduced the mortality rate of infected mice (7- 12). We have also recently shown that the expression of inducible NOS (iNOS), and the generation of NO in conjunction with O₂^· participated in the pathophysiology of this model (11).

NO is a simple inorganic radical, produced by different isoforms of NOS in a wide range of cells (13, 14). It is now well documented that iNOS can be induced to produce excess amounts of NO in vascular smooth-muscle cells, bronchial epithelial cells, microglia, and murine macrophages after stimulation with proinflammatory cytokines, (e.g., interferon- [IFN-], tumor necrosis factor- [TNF-], and interleukin-1 [IL-1]), lipopolysaccharide (LPS), lipoteichoic acid, and bacteria themselves (13). Excessive production of NO seems directly related to the hypotension and shock observed in endotoxemia and sepsis (17). Further, the involvement of NO in the pathogenesis of inflammatory disorders (e.g., immune-complex alveolitis [18] and arthritis [19] in rats), and in the immunopathogenesis of various viral infections has been suggested (11, 12, 20).

In this study, we evaluated the effect of EM on acute inflammation in mice with influenza-virus pneumonia. To elucidate the mechanism of the therapeutic effect of EM, we measured IFN- production and NO generation.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

EM was a kind gift of the Dainippon Pharmaceutical Co., Ltd. (Osaka, Japan). Other chemicals were of the highest analytical grade commercially available.

Influenza virus A/Kumamoto/Y5/67 (H2N2), adapted to the mouse, was used throughout the experiments. Five-week-old SPF-grade male ddY mice (25 to 30 g body weight), were used in all experiments. Experimental influenza-virus pneumonia was produced in the mice as described in a previous report (8). Briefly, mice were inoculated by the inhalation of a viral suspension at a dose of 1.5 × LD₅₀ in a rotating turntable. This dose routinely causes lethal viral pneumonia.

EM dissolved in 0.9% saline with 0.5% dimethyl sulfoxide (DMSO) was given to mice intraperitoneally at a dose of 1.0 to 3.3 mg/kg/d from Day 1 to Day 6 after the virus inoculation. The survival rate and body weight of animals were recorded until Day 20 after infection. Mice not treated with EM were used as controls.

Serum was prepared from blood obtained from the inferior vena cava, and bronchoalveolar lavage (BAL) was performed after exsanguination by cutting the abdominal aorta under pentobarbital anesthesia, as described previously (8). The lung was lavaged with 1 ml of Krebs-Ringer phosphate buffer (pH 7.4) containing 10 U/ml heparin. The BAL fluid (BALF) was centrifuged at 400 × g for 10 min at 4° C, and the resultant BALF supernatant (s-BALF) was filtrated with a 0.22-µm-pore-size filter. The concentrations of cytokines in s-BALF were measured with enzyme immunoassay (EIA) kits for mouse cytokines (Endogen Inc., Cambridge, MA) according to the manufacturer's instructions. In some experiments, BAL was repeated five times, and the pellet of cells collected by centrifugation of the BALF was analyzed for total cell counts with a hemocytometer, and for differential cell counts by morphologic examination of cells after Giemsa staining.

Further, the effect of EM treatment on virus replication in the lungs was examined by quantitating infectious virus in the homogenate of virus-infected lungs with a plaque-forming assay, as described earlier (8).

The concentrations of nitrite (NO₂) and nitrate (NO₃) in the serum were quantified with an autoanalyzer system (TCI-NOX 1000; Tokyo Chemical Industry, Tokyo, Japan), based on detection of a diazo compound formed in a flow reactor with Griess reagent, as described previously (21).

The NOS-inducing potential of the s-BALF from mice was tested by stimulating a mouse macrophage cell line (RAW 264 cells) with s-BALF obtained from virus-infected mice with or without EM treatment. Specifically, we added 50 µl of s-BALF obtained from the infected mice with or without EM (6 d postinfection) to the culture of RAW 264 cells at saturation density (1 × 10⁶ cells/well) in 24-well plates (diameter: 1.5 cm) in 450 µl of Dulbecco's minimal essential medium supplemented with nonessential amino acids and 0.2% bovine serum albumin (BSA). The culture supernatant was obtained 48 h after stimulation and was subjected to the NO₂/NO₃ assay as just described. In some experiments, s-BALF (50 µl) was incubated with 10 µl of antibody to murine IFN- ( 1.0 × 10⁵ neutralization units/ml; Biosource International, Camarillo, CA) at 37° C for 1 h, and was added to the cells in culture.

Also, expression of iNOS mRNA in RAW 264 cells was examined with Northern blot analysis as recently described (22). Briefly, total RNA was extracted from the cells by using the guanidine thiocyanate lysis method. Each RNA extract (10 µg) was electrophoresed on agarose gels and transferred to a Hybond^TM-N⁺ nylon membrane (Amersham International, PLC, UK), followed by hybridization with a DNA probe for rat iNOS mRNA. The DNA probe was radiolabeled with the random primer technique, using [-³²P] deoxycytosine triphosphate. An iNOS complementary DNA (cDNA) fragment of 538 bp was used for the hybridization, as was reported previously (11), and a cDNA fragment for glyceraldehyde-3-phosphate dehydrogenase (G3PDH) was used as a gene-expression control. The radioactive bands derived from hybridized probes were detected with BAS2000 a bioimage analyzer (Fuji Photo Film, Tokyo, Japan), and the relative intensity of iNOS mRNA was quantified by densitometry through comparison with that of G3PDH.

All values are expressed as mean ± SEM. The statistical significance of differences in the survival rate of mice treated with EM was determined with Fisher's exact test. Other statistical analyses were done with by two-tailed t tests for unpaired data. Values of p < 0.05 were considered significant.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

To investigate the effect of EM on influenza-virus-induced pneumonia in mice, we observed the time course of the survival rate and body weight of mice after virus infection with or without intraperitoneal administration of EM. Fourteen mice were used for each experimental group with or without EM treatment. As shown in Figure 1A, EM markedly reduced the lethal effect of influenza-virus-infected mice in a dose-dependent manner at doses of up to 3.3 mg/kg/d. A significant difference was found in the survival rate of mice treated with control vehicle (saline + 0.5% DMSO) and those treated with EM (3.3 mg/kg/d). EM also inhibited body weight loss later than 8 d after infection (Figure 1B).

View larger version (22K): [in this window] [in a new window]   Figure 1.   Effect of EM on survival rate (A) and body weight (B) of mice infected with influenza virus. At each time point after influenza virus infection [1.5 × LD50 of influenza virus (A/Kumamoto/ Y5/H2N2)], the survival rate of infected mice was evaluated and the body weight of mice was measured. The mice were given EM intraperitoneally (solid squares: 1.0 mg/kg/d, solid circles: 3.3 mg/ kg/d in saline/0.5% DMSO) every 24 h from Day 1 to Day 6 after virus infection. The control group (open circles) was injected intraperitoneally with 0.5 ml saline/0.5% DMSO. Fourteen mice were used in each experimental group: p < 0.05, control versus EM-treated groups.

Our previous study showed that a large amount of IFN- was induced in lungs infected with influenza virus, and that TNF-, though in a quantity 1,000 times less than IFN-, was also produced in the virus-infected lungs (11, 12). IL-1, however, was not detected in the serum or s-BALF of this model. On the basis of these results, we measured the production of IFN- in this model with or without intraperitoneal administration of EM. Figure 2A shows that EM reduced the production of IFN- in s-BALF on Day 6 (IFN-: control = 425.5 ± 12.8 pg/ml; EM 3.3 mg/kg/d = 181.2 ± 14.3 pg/ml). These data indicate that EM attenuated production of IFN- in the influenza-virus-infected lung.

View larger version (29K): [in this window] [in a new window]   Figure 2.   Reduction of IFN- generation and number of inflammatory cells in virus-infected lung by administration of EM. The experimental protocol for production of viral pneumonia and its treatment was the same as in Figure 1. (A) The amount of IFN- in s-BALF with (solid circles) or without EM (open circles) (3.3 mg/kg/d) was assayed with an enzyme immunoassay kit. (B) Total number of inflammatory cells including neutrophils, lymphocytes, and monocytes/macrophages recovered in BALF on Day 6 after virus infection; (C ) Differential cell counts in the BALF with or without EM treatment. All data are expressed as means ± SEM (n = 4). *p < 0.05; **p < 0.01, control versus EM-treated groups.

The antiinflammatory action of EM was further examined by measuring the number of inflammatory cells recovered in BALF. The result, shown in Figure 2B and C, revealed that EM treatment of the virus-infected animals resulted in a significant reduction in numbers of inflammatory cells, particularly lymphocytes and monocytes/macrophages infiltrating the lung after virus infection. This suggests that the decreased level of IFN- in the virus-infected lungs with EM treatment may have been due to suppression of inflammatory-cell infiltration and/or proliferation in the lung.

Because the generation of IFN- was reduced by EM, it was envisaged that EM inhibited the generation of NO. Therefore, we quantified the level of NO₂/NO₃both metabolites of NO in the serum of infected mice with and without EM treatment. NO₂/NO₃ generation in the serum on Day 6 after infection was significantly reduced in a dose-dependent manner with EM administered to the mice (Figure 3).

View larger version (18K): [in this window] [in a new window]   Figure 3.   Inhibitory effect of EM on NO2/NO3 formation in serum of mice infected with influenza virus. The animals infected with the virus (1.5 × LD50 dose) were treated with EM as in Figure 1. The level of NO2/NO3 in the serum was measured with or without EM (1.0 to 33 mg/kg/d of EM given intraperitoneally). All data are expressed as means ± SEM (n = 4). *p < 0.05; **p < 0.01, control versus EM-treated groups.

In order to elucidate whether EM indeed inhibited the NOS induction pathway in the virus-infected lung, we determined the NOS-inducing potential of s-BALF of virus- infected mice with or without EM treatment. The potential of s-BALF from the virus-infected lungs of animals treated with EM was significantly abrogated, as evidenced by a decrease in production of NO₂/NO₃ by RAW 264 cells incubated with s-BALF from these lungs after treatment with EM (Figure 4A). Northern blot analysis indicated that the decrease in NO production of the cells stimulated with s-BALF from EM-treated animals was due to lower levels of iNOS mRNA expression in the cells (Figure 4B and C). Specifically, although induction of iNOS mRNA expression was clearly demonstrated with RAW 264 cells after stimulation with s-BALF obtained from the virus-infected mice, the iNOS mRNA-inducing activity of s-BALF was strongly inhibited by EM treatment throughout the time course of stimulation of RAW 264 cells. It should be noted that the induction of iNOS by s-BALF was almost completely suppressed by treatment with an anti-IFN- antibody, indicating the major role of IFN- in iNOS upregulation. However, EM had no effect on NO₂/NO₃ generation by RAW 264 cells when it was introduced to the culture system together with s-BALF from the lungs of virus-infected mice without EM treatment (data not shown). These results indicate that EM inhibits NO generation indirectly, by regulating IFN- production.

View larger version (36K): [in this window] [in a new window]   Figure 4.   iNOS-inducing potential of s-BALF obtained from mice infected with influenza virus with or without EM treatment. Mice were infected with the virus (1.5 × LD50 dose) and treated with EM as in Figure 1. After stimulation of RAW 264 cells in culture with s-BALF obtained on Day 6 after infection from the virus-infected animals with or without EM treatment, NO2/NO3 generation in the culture medium was measured (A). Similarly, iNOS mRNA expression was analyzed by Northern blotting after stimulation of cells with s-BALF (B, C ). (a) Control cells stimulated with s-BALF from virus-infected mice without EM treatment; (b) and (c) Cells stimulated with s-BALF from infected mice with 1.0 mg/kg/d and 3.3 mg/kg/d EM, respectively; (d ) s-BALF from the virus-infected mice treated with anti-IFN- antibody. *p < 0.05; **p < 0.01; ***p < 0.001; control (a) versus (b), (c), and (d ). All data are expressed as means ± SEM (n = 4 ).

We evaluated the replication of influenza virus in this model with or without the administration of EM. The yields of influenza virus in the mouse lung on Day 5 after virus infection were 5.16 ± 0.38 log₁₀ plaque forming units (pfu)/lung and 5.43 ± 0.13 log₁₀ pfu/lung, and those on Day 7 were 4.14 ± 0.47 log₁₀ pfu/lung and 4.59 ± 0.28 log₁₀ pfu/lung, in the control and EM-treated groups (3.3 mg/kg/d), respectively (data are expressed as means ± SEM, n = 4). In this study, four different mice were used for virus titration at each time point. Although we observed a trend toward a slight increase in virus yield with EM treatment, there was no statistically significant difference between the group with and that without EM treatment.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Our present results show that EM markedly ameliorated the lethal effect of influenza-virus-induced pneumonia in mice, possibly through its inhibitory action against virus-induced inflammatory responses in the lung. This observation may suggest a great benefit of clinical use of EM for various acute and chronic inflammatory diseases.

It is unlikely that there was a secondary bacterial infection in the mouse model we used, because another antimicrobial agent (cefazoline) had no effect on mortality or body weight in this model, and because no bacteria were isolated by colony-formation assay on trypticase soy agar inoculated with lung homogenates from influenza-virus-infected mice (data not shown).

In contrast, it may be possible that not only EM in 0.5% DMSO solution, but also the DMSO vehicle alone injected intraperitoneally in mice, causes a nonspecific inflammation in the peritoneal cavity, which may modulate the immunopathogenesis of influenza-virus infection. However, treatment with the 0.5% DMSO solution did not affect the time courses of such pathologic parameters as the lethality of infection, virus replication in the lung, or consolidation of the lung, as compared with these parameters in virus-infected mice given saline alone. Also, cefazoline, an antibiotic unrelated to EM, did not have any appreciable effect on this influenza model, as just mentioned. Therefore, nonspecific inflammatory responses caused by EM, if any, appear not to contribute the antiinflammatory actions of EM observed in influenza-virus-induced pneumonia in mice.

Recent investigations have revealed a diverse array of biologic affects of EM, suggesting that EM may function as an immunomodulator (1). It has been reported that NO generation by inflammatory cells and O₂^· generation by inflammatory cells and the xanthine oxidase system induce tissue injury in mouse models of influenza-virus-induced pneumonia; the mortality among infected mice is significantly reduced by the administration of superoxide dismutase (an O₂^· scavenger) (7) and L-NMMA (an NOS inhibitor) (11).

Among a series of proinflammatory cytokines, such as IFN-, TNF-, and IL-1, which are known to induce iNOS mRNA in a variety of cells and tissues, production of IFN- in the lung was significantly deceased by the administration of EM to influenza-virus-infected mice. It is of considerable interest that IFN- recovered in the BALF in our study indeed showed strong induction of iNOS mRNA and NO metabolite production, as seen in the data in Figure 4, which demonstrate abrogation of NOS induction by anti-IFN- antibody. More importantly, it was clearly demonstrated that s-BALF obtained from EM-treated mice after virus infection showed a significantly weaker iNOS-induction potential than did s-BALF from the vehicle-treated group with virus infection, as was verified by measuring NO₂/NO₃ production and iNOS mRNA expression in RAW 264 cells stimulated with s-BALF from the two groups of mice. However, EM did not directly affect iNOS activity or cytokine-induced NO generation in RAW 264 cells in vitro (data not shown).

Furthermore, inhibition of inflammatory-cell infiltration by EM treatment was apparent upon analysis of the cells in BALF obtained from the virus-infected mice. Therefore, it is possible that suppression of IFN- and of iNOS induction is caused by the antiinflammatory effect of EM on virus-induced acute inflammatory responses in the host.

Of relative importance is our finding that EM administered at a therapeutic dose (3.3 mg/kg/d) did not significantly influence the virus yield in the lung. This result may substantiate the notion that EM exhibits a potent therapeutic action against influenza-virus-induced lung injury by ameliorating virus-induced inflammation and modulating the pathogenesis of influenza virus infection.

In any event, complicated interactions between virus and host, including immunologic effects of the host, have been noted in many viral diseases (7). It has been documented that soluble factors, such as proinflammatory cytokines, proteases, kinins, and free radicals produced during the host immune response play an important role in the process of viral pathogenesis (7). Thus, the therapeutic effect of EM on influenza-virus-induced pneumonia in mice may be attributed to its regulatory action against a series of inflammatory mediators of the host. The antiinflammatory action of EM might provide an approach to therapeutically modulating the pathologic events in various inflammatory diseases.