Nitric oxide mediates mucin secretion in endotoxin-induced otitis media with effusion

Student Research Award 1996

Nitric oxide mediates mucin secretion in endotoxin-induced otitis media with effusion AUSTIN S. ROSE,MSIV,JIRI PRAZMA, MD, PhD,SCOTt H. RANDELL,PhD, HENRYC. BAGGETr,MSIV,ANDREW P. LANE, MD, and HAROLD C. PILLSBURY,MD, Chapel Hill, North Carolina The mechanisms that regulate mucin release in chronic otitis media with effusion, a leading cause of hearing loss in children, remain largely unknown. We developed an animal model using Sprague-Dawley rats to determine the factors responsible for mucin production in chronic otitis media with effusion. N-nitro-L-arginine methyl ester (L-NAME), a competitive inhibitor of nitric oxide synthase, was used to investigate the role of nitric oxide in mucin secretion by the middle ear epithelium. All rats underwent eustachian tube obstruction. In the first set of rats, the middle ear was then injected transtympanically with 35 gl of either 300 mOsm Krebs-Ringer bicarbonate buffer (control group) or I mg/ml lipopolysaccharide in Krebs:Ringer (experimental group I). In a second set of rats, the middle ear space was injected with lipopolysaccharide and then infused at a continuous rate for 7 days with either Krebs-Ringer (experimental group 2) or I mmol/L L-NAME in Krebs-Ringer (experimental group 3) through an osmotic infusion pump. After 7 days the volume of effusion and the quantity of mucin collected were significantly greater in lipopolysaccharideexposed ears than in controls. In addition, antimucin immunostaining demonstrated mucous cell hyperplasia in response to lipopolysaccharide. The lipopolysaccharideinduced production of mucin and mucous cell hyperplasia was inhibited in ears treated with lipopolysaccharide and t-NAME. These results suggest that nitric oxide is a mediator in the pathway of mucin secretion in chronic otitis media with effusion. ( O t o l a r y n g o l H e a d Neck Surg 1997; 116:308-16,)

O t i t i s media is one of the most common pediatric infections. Despite aggressive treatment with antimicrobial agents, approximately 5% to 10% of aCute otitis media progresses to chronic otitis media with effusion (OME), a leading cause of hearing loss in children. 1 O M E is associated with a 20- to 30-dB hearing loss and

From the Division of Otolaryngology-Head and Neck Surgery, Department of Surgery (Drs. Rose, Prazma, Baggett, Lane, and Pillsbury), and the Cystic Fibrosis/Pulmonary Research and Treatment Center, Department of Medicine (Dr. Randell), University of North Carolina School of Medicine. Supported by the University of North Carolina School of Medicine One-Year Fellowship for Medical Student Research. Presented at the Annual Meeting of the American Academy of Otolaryngology-Head and Neck Surgery, Washington, D.C., Sept. 29-Oct. 2, 1996. Reprint requests: Jiri Prazma, MD, PhD, Division of Otolaryngology-Head and Neck Surgery, Department of Surgery, University of North Carolina School of Medicine, 610 BurnettWomack Building, CB#7070, Chapel Hill, NC 27599-7070. Copyright © 1997 by the American Academy of OtolaryngologyHead and Neck Surgery Foundation, Inc. 0194-5998/97/$5.00 + 0 2311176368

308

is known to cause impairment of speech and language development. 2,3 Several theories have been proposed to explain the cause of OME, including eustachian tube obstruction, allergic reaction, and local immune dysfunction associated with the persistence of bacterial cell wall components, such as endotoxin. 4,5 Lipopolysaccharide (LPS), the principal component of endotoxin, has been isolated from the middle ear effusions of patients with O M E 6,7 and used to induce middle ear effusions in animal models. 4-9 LPS contributes to the development of middle ear effusion by generating a localized inflammatory response through its interaction with numerous cytokines and other immunologic mediators. 1° In airway epithelium LPS appears to promote the storage and release of mucin. 11-14 Typically, the effusions observed in O M E are grouped into two clinical types, which differ significantly in viscosity: serous (thin) and mucoid (thick). It has been shown that differences in viscosity are solely the result of mucin content. 15 Mucoid effusions have a higher concentration of mucin and are associated with a

OtolaryngologyHead and Neck Surgery ROSE et ai. 309

Volume ] ]6 N u m b e r 3

Table 1. S a n d w i c h - t y p e

ELISA p r o t o c o l for measuring mucin concentration

CONTROL GROUP1 (n=6)

EXPERIMENTAL GROUP1 (n=6)

EXPERIMENTAL GROUP2 (n=3)

EXPERIMENTAL GROUP 3 (n=5)

KR

Day 1 1. Primary antibody. Coat wells with 100 Ftl per well of 0.25 gg/pl rabbit antirat mucin IgG in PBS. Wrap in plastic wrap and incubate for 2 hr at 37 ° C. 2. Wash each well with 100 BI TBST x 2. 3. Block each well with 200 gl 15% milk in TBST for 1 hr at 37 ° C. Dump solution, wrap plate in plastic wrap, and store at 4 ° C for up to 1 mo.

I LPS

I

LPS & KR

KR via osmotic infusion pump

LPS & I L-NAM

L-NAME via osmotic infusion pump

0

l 1

I 2

~ 3 Days

L 4

I 5

L 6 t Effusion collection & KR lavage

Fig. 1. D i a g r a m of e x p e r i m e n t a l design illustrating the solutions initially injected into the m i d d l e ear s p a c e a n d a n y subsequent treatments until the time of s a m p l e collection at 7 days, KR, 300 mOsm; LPS, ] m g / m l ,

Day 2 1. Samples and mucin standard. Warm middle ear lavage samptes and mucin standard to 37 ° C. Make serial dilutions in 1% milk-TBST and add 100 gl of each dilution to a separate well. Incubate for 30 min at RT. 2. Wash each well with 100 gl TBST x 2. 3. Labeled secondary antibody. Add 100 gl peroxidaselabeled rabbit antirat mucin (diluted 1:500 in 1% milkTBST) to each well and incubate 30 min on rotomix at RT. 4. Wash each well with 100 ~1TBST x 2. 5. Initiate peroxidase reaction. Add 100 gt TMB substrate solution (t00 gl TMB, 10 ml citrate buffer, 33 gl H202) to each well and incubate in dark for 5 minutes at RT. 6. Stop peroxidase reaction. Add 50 t.tl of 4 mol/L H2SO4 to each well and read spectrophotometrically at 450 to 575 nm. TBST,Tris-buffered saline/Tween 20; RT, room temperature; TMB, tetramethylbenzidine.

greater degree of hearing impairment. 16 Mucin is clearly an important factor in hearing loss caused by chronic OME. However, the mechanisms that regulate mucin production in chronic OME remain largely unknown. It has recently been suggested that nitric oxide (NO) is a possible mediator in the pathway of mucin secretion. ~7 NO, a free radical synthesized by NO synthase (NOS) from the amino acid L-arginine, is a biologic mediator with many known cellular functions. Our laboratory previously demonstrated the importance of NO in the development of middle ear effusion. ~s We hypothesized that NO in part mediates the release of mucin in chronic OME. To test this hypothesis, we developed an animal model of LPS-induced chronic OME. We then examined the effect of N-nitro-L-arginine methyl ester (L-NAME), a competitive inhibitor of NOS, on LPS-induced production of mucin.

to LPS and then implanted with osmotic infusion pumps that delivered either KR (EXP 2) or L-NAME (EXP 3) at a constant rate for 7 days. The purpose of EXP 2 was to control for any effects of fluid infusion by the osmotic pump. The experimental design is illustrated in Fig. 1. This study was performed in accordance with the Public Health Service Policy on Humane Care and Use of Laboratory Animals, the National Institutes of Health Guide for the Care and Use of Laboratory Animals, and the Animal Welfare Act (7 U.S.C. et seq.). The animal use protocol was approved by the Institutional Animal Care and Use Committee of The University of North Carolina.

METHODS A N D MATERIAL

Eustachian Tube Obstruction

Twenty Sprague-Dawley rats, weighing between 310 and 470 gm, were randomly divided into one control and three experimental groups. In the control group (CON 1), six rats were exposed to Krebs-Ringer solution (KR) only. In experimental group 1 (EXP 1), six rats were exposed to LPS. At that point rats in CON 1 and EXP 1 were allowed to recover from anesthesia and received no further treatment until the time of sample collection at 7 days. In experimental groups 2 and 3, three and five rats, respectively, were initially exposed

Each rat was anesthetized with an intramuscular injection of 0.1 ml/100 gm of equal parts of 20 mg/ml xylazine and 100 mg/ml ketamine hydrochloride. The right tympanic membrane of the rat was inspected to rule out the presence of an effusion. Hair was shaved from the ventral neck region, and the rat was placed in the supine position. After decontamination of the skin, a ventral midline incision was made. While anesthetized, each rat received 0.1 ml/hour of 300 mOsm KR (140 mmol/L Na +, 120 mmol/L CI-, 5.2 mmol/L

OtolaryngologyHead and Neck Surgery March 1997

310 ROSEet al.

Tympanic bulla

~

200 gl osmotic infusion pump (1 gl/hour x 7 days)

0.8 m m hole in rostromedial process, occluding eustachian tube

0.6 m m diameter catheter tip

or 1 mmol/L L-NAME in 300 mOsm KR (EXP 3), was then implanted subcutaneously between the scapulae. The osmotic pump was connected to the tympanic bulla by 1.22 mm diameter polyethylene tubing with a 0.6 mm diameter catheter tip, as shown in Fig. 2. The tip was inserted into the bulla through the previously made hole and secured in place with 6-0 coated Vicryl suture. The ventral midline incision was then sutured closed, and the rat was allowed to recover from anesthesia.

Recovery Period Fig. 2. Diagram of osmotic infusion pump apparatus.

K +, 25 mmol/L HCO3-, 1.1 mmol/L Ca2 +, 1.2 mmol/L Mg 2+, 0.4 m m o g L HPO42-, and 5.6 mmol/L glucose in distilled water) subcutaneously to maintain systemic blood pressure and renal perfusion. Body temperature was maintained at 37 ° C with a thermistor-controlled heating pad. Through the midline incision, the ventral surface of the fight tympanic bulla was approached. The bulla's rostromedial process, through which the eustachian tube passes, was isolated. Middle ear clearance through the eustachian tube was eliminated by drilling an 0.8 mm diameter hole into the process and packing it w i t h dental wax. Care was taken to avoid entering the m i d dle ear cavity during this procedure.

Transtympanic Middle Ear Fluid Instillation (CON I and EXP 1) After eustachian tube obstruction, the ventral midline incision was sutured closed. The rat was placed in the left lateral recumbent position. A 0.5-mm perforation was made in the pars tensa of the right tympanic membrane with a microscalpel to equilibrate middle ear pressure during fluid instillation. Next, 35 gl of either 300 mOsm KR (CON 1 ) o r 1 mg/ml LPS (L 9143; Sigma Chemical Co., St. Louis, Mo.) in 300 mOsm KR (EXP 1) was injected transtympanically into the middle ear space. The rat was then allowed to recover from anesthesia.

Middle Ear Fluid Instillation/Osmotic Implantation (EXP 2 and EXP 3)

Pump

After eustachian tube obstruction, a 0.6-mm hole was carefully drilled in the ventral surface of the right tympanic bulla. Through this hole, 35 ~tl of either 1 mg/ml LPS in 300 mOsm KR (EXP 2) or 1 mg/ml LPS and 1 mmol/L L-NAME (N 5751; Sigma) in 300 mOsm KR (EXP 3) was injected into the middle ear space. A 200-gl osmotic infusion pump with a flow rate of 1 ~tl/hour (ALZET 2001; A!za Pharmaceuticals, Palo Alto, Calif.), containing either 300 mOsm KR (EXP 2)

After surgery rats were placed on paper towels in the left lateral recumbent position and allowed to recover from anesthesia under a warm lamp. Each rat received 300,000 units/kg/day of intramuscular penicillin G potassium to prevent middle ear infection.

Middle Ear Sample Collection After 7 days the rat was again anesthetized. The tympanic membrane was inspected, and any effusion was removed atraumatically with a small cannula. The volume of this initial collection was measured with calibrated micropipets. The middle ear space was then lavaged with 35 gl of 300 mOsm KR for 10 minutes. Lavage fluid was centrifuged for 10 minutes at 300g, and the cell-free supernatant was frozen and stored at -20 ° C until being analyzed for mucin content by enzyme-linked immunosorbent assay (ELISA).

ELISA The mucin concentration of each middle ear lavage sample was determined with a sandwich-type ELISA. Our ELISA protocol is detailed in Table 1. Lavage samples were diluted 48:1 by mixing 5 ~tl of each sample with 235 gl of 1% milk in Tris-buffered saline/0.05% Tween 20. Serial dilutions of the initial 48:1 dilution of sample were then plated on 96-well microtiter plates, previously coated with antimucin antibody. In addition, serial dilutions of a mucin standard of known concentration were prepared in duplicate and run on each plate. A peroxidase-conjugated antimucin antibody was used to detect bound mucin. Wells containing no sample served as negative controls. The generation of an antimucin antibody and mucin standard has been described previously. 19 After each plate was read, a curve of optical densities of the serial dilutions of mucin standard was generated. The mucin concentration of each sample was determined by regression analysis from the standard curve of known mucin concentrations.

Immunohistochemistry After lavage of the middle ear, each rat was deeply anesthetized and euthanized with an intracardiac injec-

OtolaryngologyHead and Neck Surgery

ROSE et al. 311

Volume ] 16 Number 3

50

40

30

VOLUME (gl) 20

10 ~i?~ ¸

!ii?);i

,.oo

KR

LPS

LPS/KR

LPS/L-NAME

Fig, 3. Mean volume of middle ear effusion after 7 days. KR, Control group of rats exposed only to 300 mOsm KR (CON 1). LPS,Group of rats exposed only to 1 m g / m l LPS (EXP 1). LP$/KR, Group of rats exposed to LPS and then KR by osmotic pump (EXP 2) to control for effects of fluid infusion into the middle ear space. LPS/L-NAME,Group of rats exposed to LPS and then 1 mmol/L L-NAME by osmotic p u m p (EXP 3). Bars indicate group means _+ SEM. Asterisk designates a significant difference between LPS and KR (p < 0.05).

tion of 3 mol/L KC1. The middle ear was dissected free and placed in 4% paraformaldehyde in phosphatebuffered saline solution (PBS; pH 7.2 to 7.4). In groups EXP 2 and EXP 3, catheter tip placement in the tympanic bulla was confirmed at this time. After fixation, the middle ear was decalcified in 10% EDTA in PBS for 48 hours. The tympanic bulla was then separated from the cochlea, dehydrated in a series of increasingly concentrated alcohol solutions, and embedded in a paraffin block. A microtome was used to cut 6-gm sections through the bulla at the origin of the eustachian tube. Sections were mounted on glass slides pretreated in 2% aminoalkylsilane (A 3648; Sigma) in dry acetone. Slides were deparafflnized in xylene, rehydrated in a series of alcohol solutions of decreasing concentration, and placed in 0.6% hydrogen peroxide in methanol for 30 minutes at room temperature. Slides were then washed with PBS and blocked with 5% normal goat serum, 1% gelatin, and 1% bovine serum antigen in PBS/Tween 20 (BSA-PBST) for 30 minutes at room temperature. Next, slides were treated with 5 gg/ml of the primary antibody (rabbit antirat mucin IgG) or negative control antibody (rabbit IgG) in 0.1% gelatin and 1% BSA-PBST for 1 hour at room temperature. The slides were washed in buffer solution and then placed in a 1:400 dilution of peroxidase goat antirabbit antibody in 0.1% gelatin and 1% BSA-PBST

for 60 minutes at room temperature. After being washed three times in PBST and once in 0.05 mol/L Tris-HC1 (pH 7.6), slides were exposed to 3,3"diaminobenzidine (D 5637; Sigma) and 0.06% hydrogen peroxide in 0.05 mol/L Tris-HC1 for 15 minutes at room temperature and then rinsed with H20. Finally, slides were counterstained with 1% methyl green, rinsed in H20, dehydrated, and covered with Permount and glass coverslips.

Statistical Analysis For each group of rats, the volumes of middle ear effusion and mucin concentrations of lavage fluid were averaged and analyzed by one-way analysis of variance and the Student-Newman-Keul's method of pairwise multiple comparison; p < 0.05 was considered statistically significant. Graphs were made of the mean and standard error of the mean for each group. RESULTS Volume of Middle Ear Effusion

The mean volume of middle ear effusion was significantly greater in rats exposed to LPS (EXP 1) than in controls exposed only to KR (CON !) (Fig. 3). The mean volumes were 28.41 +_ 13.80 gl and 1.0 _+0.63 gl, respectively. In rats exposed to LPS initially and then

312

OtolaryngologyHead a n d N e c k Surgery March 1997

ROSE et al.

25

20

15

MUCIN (gtg/gtl) lO

T

5

o.

KR

LPS

LPS/KR

LPS/L-NAME

Fig. 4. Mean concentration of mucin in middle ear lavage fluid after 7 days. KR, Control group of rats exposed only to 300 mOsm KR (CON 1). LPS, Group of rats exposed only to 1 mg/ml LPS (EXP 1). LPS/KR, Group of rats exposed to LPS and then KR by osmotic pump (EXP 2) to control for the effects of fluid infusion into the middle ear space. LPS/L-NAME, Group of rats exposed to LPS and then 1 mmol/L L-NAME by osmotic pump (EXP 3). Bars indicate group means + SEM Asterisk designates a significant difference between LPS and KR (p < 0.05) and between LPS/L-NAME and LPS/KR (p < 0.05).

Fig. 5. Light photomicrograph of middle ear from a rat exposed only to LPS (EXP 1) showing immunohistochemical staining of mucin. Increase in intraepithelial mucin content compared with CON 1 is suggestive of mucous cell hyperplasia. Spaces below epithelial basement membrane represent possible submucosal gland formation. M, Intraepithelial mucosubstances. (Bar = 30 ~m.)

OtolaryngologyHead and Neck Surgery

ROSE et al, 313

Volume 116 Number 3

Fig. 6. Light photomicrograph of middle ear from a rat exposed only to KR (CON 1) showing irnmunohistochemical staining of mucin. Rare goblet cells are positively stained. M, Intraepithelial mucosubstances. (Bar = 30 ~m.)

KR by osmotic pump (EXP 2), the mean volume of effusion, 18.00 _+7.94 gl, was significantly greater than in controis exposed only to KR. In rats exposed to LPS and then L-NAME by osmotic pump (EXP 3), the mean volume of effusion was 9.43 _ 2.20 btl. This volume was significantly less than in rats exposed to LPS, but it was not significantly different from that in rats which received LPS and then KR by osmotic pump. Mucin Concentration of Middle Ear Lavage Fluid The mean concentration of mucin in middle ear lavage fluid was significantly greater in rats exposed to LPS (EXP 1) than in controls exposed only to KR (CON 1), rising fivefold from 3.18 +_ 1.24 btg/btl to 16.16 -+ 4.39 gg/btl, respectively (Fig. 4). In rats exposed to LPS initially and then KR by osmotic pump (EXP 2), the mean concentration of mucin, 12.57 _+ 5.47 Bg/btl, was significantly greater than in controls exposed only to KR. In rats exposed to LPS and then L-NAME by osmotic pump (EXP 3), the mean concentration of mucin was 4.01 _+ 1.92 btg/btl. This concentration was significantly less than in rats exposed to LPS (a fourfold decrease) and rats exposed to LPS and then KR by osmotic pump (a threefold decrease). Immunohistochemistry Antimucin staining of middle ear epithelium was strongly positive in ears exposed to LPS (EXP 1, Fig. 5) compared with controls (CON 1, Fig. 6). Histologically,

LPS-treated ears were characterized by numerous inflammatory cells within the middle ear mucosa and by leukocyte exudation into the bulla lumen. The inflammatory infiltrate contained a few neutrophils but was primarily mononuclear. A large increase in the number of epithelial goblet cells and the appearance of submucosal glands was also observed in LPS ears compared with controls. Ears exposed to LPS and then KR by osmotic pump (EXP 2, Fig. 7) demonstrated immunolabeling and histologic changes similar to those observed in ears exposed to LPS. Ears exposed to LPS and then L-NAME by osmotic pump (EXP 3, Fig. 8) were weakly immunostained, with fewer goblet cells and submucosal glands than in ears exposed to LPS and ears exposed to LPS and then KR by osmotic pump. No immunostaining was noted when normal rabbit IgG was substituted for the primary antibody. Normal rat tracheal epithelium, which showed marked immunostaining for mucin, served as a positive control. DISCUSSION OME is the most common inflammatory disease of young children after the common cold. 2° It follows the resolution of acute otitis media and is characterized by the persistence of fluid in the middle ear without signs or symptoms of infection. Middle ear effusions are commonly classified into clinical types, such as serous and mucoid, on the basis of their viscoelastic properties. In chronic OME there is often a progression from

314 ROSE et al.

OtolaryngologyHead and Neck Surgery March 1997

Fig. 7. Light photomicrograph of middle ear from a rat exposed to LPSand then KR by osmotic p u m p (EXP 2) showing immunohistochemical staining of mucin. Section demonstrates a similar amount of intraepithelial mucin and mucous cell hyperplasia c o m p a r e d with levels in group EXP 1, M, Intraepithelial mucosubstances. (Bar = 30 #m.)

Fig. 8. Light photomicrograph of middle ear from a rat exposed to LPS and then L-NAME by osmotic p u m p (EXP 3) showing immunohistochemical staining of mucin. Intraepithelial mucin and mucous cell hyperplasia are reduced in comparison with levels in groups EXP 1 and EXP 2. M, Intraepithelial mucosubstances. (Bar = 30 Fire.)

OtolaryngologyHead and Neck Surgery Volume 116 Number 3

serous to mucoid effusions with an associated decrease in auditory response. 16 This transformation is the result of an increase in mucin secretion by the middle ear epithelium.15 The mechanisms that regulate mucin production in chronic OME have not been previously described. The results of our study demonstrate that LPS induces middle ear effusion and stimulates mucin production by the middle ear epithelium. At 7 days LPSexposed ears exhibit an influx of mononuclear leukocytes, goblet cell hyperplasia, and mucous gland formation. L-NAME delivered by osmotic infusion blocks the LPS-induced production of mucin and inhibits gland and goblet cell hyperplasia. In rats exposed to LNAME, the volume of middle ear effusion is less than that in rats exposed to LPS alone, but not significantly different from that in controls exposed to LPS and then KR by osmotic pump. These findings suggest that NO is a mediator of LPS-induced mucous cell hyperplasia and mucin production in OME. This study describes an animal model of chronic OME in which LPS was used to induce middle ear effusion and mucin production. Our results corroborate the findings of previous studies in which LPS increased the storage and release o f mucin by airway epithelial cells. H-14 Steiger et al. 11 infused LPS intratracheally into rats and demonstrated a 2.5-fold increase in airway mucin production over that in controls at 7 days after instillation. Histologically, LPS has been shown to induce mucous cell hyperplasia in airway epithelium.14 Gland formation in response to OME has also been previously reported. 2° The findings of these previous studies support the efficacy of our model for investigation of the mechanisms of mucin secretion in chronic OME. LPS contributes to middle ear effusion through its interaction with numerous cytokines. Cytokines are glycoproteins produced by inflammatory cells and tissues with diverse biologic activities in immune and inflammatory processes. These immunologic mediators promote effusion production and persistence through their effects on inflammatory cell chemotaxis, microvascular permeability, 5~S and mucociliary clearance. 2t A variety of inflammatory cytokines has been found in middle ear effusions, including interleukin (1L)-l~3, IL-2, IL-6, IL-8, and tumor necrosis factor (TNF). 1°,22 Analysis of middle ear effusions has revealed a correlation between gram-negative bacterial component concentration and TNF levels, 1° suggesting that LPS stimulates the release of TNE Our laboratory previously demonstrated the role of NO and TNF as mediators of LPS-induced middle ear effusion.iS,23 The LPS-induced release of mucin is likely the result of similar interactions.

ROSE et al. 3 | 8

~+~,~ Inflammatory CellInflux

f

TNF Cytoldnes f

PC-PECk,

]

diacylglycerol Protein Ki. . . . C

Arginine--~ ~

~-~= N A ~/~ E

J

Nitric Oxide / -I~ Arglnine NOSynthase ~" (NO) Citralline

cGMP

/

MUCIN Fig. 9. Proposed pathway of mediators in LPS-induced mucin hypersecretion in chronic OME on the basis of study findings and the known regulatory mechanisms of mucin production in airway epithelium. 1°,17,25

Recently, TNF has been shown to induce both mucin secretion and mucin gene expression in airway epithelium. 17,24 Fischer et al. 17 demonstrated that TNFinduced mucin production is mediated by the enzymes phosphatidylcholine-specific phospholipase C (PCPLC) and protein kinase C (PKC). PC-PLC produces diacylglycerol, which in turn activates PKC. In addition, we used a competitive inhibitor of NOS to show that TNF-induced mucin secretion was in part mediated by NO. Furthermore, TNF-induced guanosine 3",5"cyclic monophosphate (cGMP) production was blocked by NOS inhibitor, suggesting that cGMP is an important factor in the sequential relationship between NO and mucin secretion. A similar mechanism has been suggested for rat gastric mucosal cells in which NO stimulates both cGMP levels and gastric mucus production. 25 TNF has also been shown to stimulate arginine transport in human umbilical vein endothelial cells through a process that requires activation of intracellular PKC. 26 The finding that TNF acts to increase the intracellular supply of the NOS substrate arginine suggests an additional potential mechanism in the pathway of LPS-induced mucin secretion. Our study is the first to demonstrate that LPS induces the hypersecretion of mucin in OME. Furthermore, inhibition of NO synthesis blocks the LPS-induced production of mucin in middle ear effu-

316 ROSE et al.

sion. This finding correlates with previously described mechanisms of NO-mediated mucin secretion in airway 17 and gastric mucosal epithelium. 25 On the basis of our results and the known regulatory mechanisms of mucin production in other physiologic systems, a potential pathway of mucin secretion in OME can be theorized (Fig. 9). Although we have demonstrated for the first time that NO is a mediator of LPS-induced mucin release in OME, further investigation is necessary to confirm the role of other proposed mediators in the regulatory pathways of mucin secretion in middle ear effusion. In conclusion, we have developed an animal model of LPS-induced middle ear effusion and demonstrated that NO is a mediator of mucin production in chronic OME. Our results suggest that NO inhibitors may have potential clinical applications in the management of mucoid middle ear effusion, a leading cause of hearing loss in children. REFERENCES 1. Bluestone CD. Otitis media in children: to treat or not to treat. N Engl J Med 1982;306:1399. 2. Bluestone CD, Beery QC, paradise J L Audiometry and tympanometry in relation to middle ear effusions in children. Laryngoscopy 1973;83:594-604. 3. Reichman J, Healey CH. Learning disabilities and conductive hearing loss involving otitis media. J Learning Disabilities 1983;16:272-8. 4. Ripley-Petzoldt ML, Giebink S, Juhn SK, Aeppli D, Tomasz A, Tuomanen E. The contribution of pneumococcal cell wall to the pathogenesis of experimental otitis media. J Infect Dis 1988;157:245-55. 5. Demaria TE Briggs BR, Lim DJ, Okazaki N. Experimental otitis media with effusion following middle ear inoculation of nonviable H. influenzae. Ann Otol Rhinol Laryngol 1984;93:52-6. 6. Demaria TF, Prior RB, Briggs BR, Lim D J, Birck HG. Endotoxin in middle ear effusions from patients with chronic otitis media with effusion. J Clin Microbiol 1984;20:15-7. 7. Iino Y, Kaneko Y, Takasaka T. Endotoxin in middle ear effusions tested with Limulus assay. Acta Otolaryngol (Stockh) 1985;100:42-50. 8. Nonomura NM, Nakano Y, Satoh Y, Fujioka O, Niijima H, Fujita M. Otitis media with effusion following inoculation of Haemophilus influenzaetype B endotoxin. Arch Otorbinolaryngol 1986;243:31-5. 9. Ohashi Y, Nakai Y, Esaki Y, Ohno Y, Sngiura Y, Okamoto H. Experimental otitis media with effusion induced by lipopolysaccharide from Klebsiella pneumoniae: mucociliary pathology of the eustachian tube. Acta Otolaryngol (Stockh) Suppl 1991;486:105-15.

OtolaryngoiogyHead and Neck Surgery March 1997

10. Johnson MD, Fitzgerald JE, Leonard G~ Burleson JA, Kreutzer DL. Cytokines in experimental otitis media with effusion. Laryngoscope 1994;104:191-6. 11. Steiger D, Hotchkiss J, Bajaj L, Harkema J, Basbaum C. Concurrent increases in the storage and release of mucin-like molecules by rat airway epithelial cells in response to bacterial endotoxin. Am J Respir Cell Mol Biol 1995;12:307-14. 12. Gordon T, Harkema JR. Effect of inhaled endotoxin on intraepithelial mucosubstances in F344 rat nasal and tracheobronchial airways. Am J Respir Cell Mol Biol 1994; 10:177-83. 13. Harkema JR, Hotchkiss JA. In vivo effects of endotoxin on nasal epithelial mucosubstances: quantitative histochemistry. Exp Lung Res 1991;17:743-61. 14. Harkema JR, Hotchkiss JA. In vivo effects of endotoxin on intraepithelial mucosubstances in rat pulmonary airways. Am J Pathol 1992;141:307-17. 15. Carrie S, Hutton DA, Birchall JR Green GGR, Pearson JR Otitis media with effusion: components which contribute to the viscous properties. Acta Otolaryngol (Stockh) 1992;112:504-11. 16. Majima Y, Hamaguchi Y, Hirata K, Takeuehi K, Morishita A, Sakakura Y. Hearing impairment in relation to viscoelasticity of middle ear effusions in children. Ann Otol Rhin01 Laryngol 1988;97:272-4. 17. Fischer BM, Krunkosky TM, Wright DT, Dolan-O'Keefe M, Adler KB. Tumor necrosis factor-alpha stimulates mucin secretion and gene expression in airway epithelium in vitro. Chest 1995;107:133S-5S. 18. Ball SS, Prazma J, Dais CGD, Pillsbury HC. Nitric oxide, a mediator of endotoxin-induced middle ear effusions. Laryngoscope 1996;106:1021-7. 19. Randell SH, Liu JY, Ferriola PC, et al. Mucin production by SPOC1 cells--an immortalized rat tracheal epithelial cell line. Am J Respir Cell Mol Biol 1996; 14:146-54. 20. Bernstein JM. Middle ear mucosa: histological, histochemical, immunochemical and immunological aspects. In: Jahn AF, Santo-Sacchi J, eds. Physiology of the ear. New York: Raven Press, 1988:59-80. 21. Ohashi Y, Nakai Y, Hiroyuki F, et al. Mucociliary disease of the middle ear during experimental otitis media with effusion induced by bacterial endotoxin. Ann Otol Rhinol Laryngol 1989;98:479-84. 22. Takeuchi K, Maesako K, Atsushi Y, Sakakura Y. Interleukin-8 gene expression in middle ear effusions. Ann Otol Rhinol Laryngol 1994;103:404-7. 23. Ball SS, Prazma J, Dais CGD, Triana RJ, Pillsbury HC. Role of tumor necrosis factor and interleukin-1 in endotoxin-induced middle ear effusions. Ann Otol Rhinol Laryngol 1997 (in press). 24. Levine S J, Larivee R Logun C, Angus CW, Ognibene FP, Shelhamer JH. Tumor necrosis factor-alpha induces mucin hypersecretion and MUC-2 gene expression by human airway epithelial cells. Am J Respir Cell Mol Biol 1995;12:196-204. 25 Brown JF, Keates AC, Hanson PJ, Whittle B JR. Nitric oxide generators and cGMP stimulate mucous secretion by rat gastric mucosal cells. Am J Physiol 1993;265:G418-22. 26. Pan M, Wasa M, Lind S, Gertler J, Abbott W, Souba WW. TNFstimulated arginine transport by human vascular endothelium requires activation of protein kinase C. Ann Surg 1995;221:590601.