Round window membrane permeability to human serum albumin in antigen-induced otitis media

Am ] Oto|aryngol 8:34--40, 1988

Round Window Membrane Permeability to Human Serum Albumin in Antigen-Induced Otitis Media YUFaYOSHI HAMAGUCHI, MD, TETSUO MORIZONO, MD, AND STEVEN K, JUHN, M D The round window membrane (RWM) permeability to human serum albumin (HSA) was investigated in both normal chinchillas and chinchillas sensitized with HSA. The effect of a corticosteroid agent (triamcinolone) on the RWM permeability was also analyzed, it was found that HSA could not be detected in either the perilymph or the cerebrospinal fluid of normal chinchillas within 1 hour after instillation into the middle ear bulla. Perilymph levels of HSA peaked 24 hours after instillation. In antigen-induced otttis media, the HSA level in the perilymph at 12 hours after challenge was significantly higher than that in normal chinchillas (P < .01), but it did not increase with the development of otitis media. A significant difference of HSA level in the perilymph of animals with and without steroid treatment was noted (P < .025 at 24 hours and P < .05 at 48 hours). It is concluded that little HSA passes through the normal RWM in a short time and that the RWM permeability to HSA increases in the early stage (about 12 hours) of antigen-induced otitis media. Corticosteroids can affect the RWM permeability to HSA by reducing the level of antigeninduced otitis media.

Permeability of the round window membrane (RWM) is an important aspect of inner ear damages such as sensorineural hearing loss associated with otitis media and following ototopical drug exposure. Morphologic study of the RWM in animals suggests that the RWM has an intermediate or tight type of cell junction in the tympanic surface, and impermeability of the outer layer would be suspected. 1,2 However, recent study has shown the possibility of passage of macromolecular proteins through the RWM, with the tracer protein taken up by pinocytosis. 3 In experimentally induced acute serous otitis media (SOM) models, the epithelial layer of the RWM is c o n s i d e r a b l y d e g e n e r a t e d during middle ear inflammation, and the RWM seems

to become permeable against the passage of macromolecular tracer proteins. 4,5 Human temporal bone studies have shown that infiltration of leukoc3/tes into the RWM and obliteration of the RWM niche by thick granulation tissue are frequently (8.9%) observed in the chronic, active type of otitis media. 6 Because acute middle ear inflammation and active, chronic otitis media may affect RWM permeability, s t u d y of the RWM permeability in experimentally induced acute otitis media model is important to further understanding of RWM permeability during otitis media. Previous study has s h o w n that tritiated human serum albumin (HSA) instilled into the middle ear bulla of the normal and serous otitis media model in chinchillas could be detected in the perilymph and cerebrospinal fluid, suggesting the possibility of the passage of macromolecular proteins. 4,7 Changes of RWM permeability during the course of middle ear inflamm a t i o n h a v e n o t b e e n w e l l s t u d i e d . The relationship between RWM permeability and immunologic defense mechanism of the middle ear must be clarified to consider the passage of antigenic proteins or peptides. This study was

Received 5 March 1987, from the Department of Otolaryng01ogy, University of Minnesota School of Medicine, Minneapolis, Minnesota. Revision accepted 25 June 1987. Supported by NINCDS grant NS-14538 for Otitis Media Pathogenesis Research Program. Presented at the Meeting of the American Academy of Otolaryngology, September 14-17, 1986, San Antonio, Texas. Address reprint requests to Dr. Hamaguchi: Department of Otolaryngology, Research East Building, 2630 University Avenue S.E,, Minneapolis, MN 55414. 0196-0709/88 $0.00 + .25

34

HAMAGUCHI ET AL. 1) Non-sensitized Animals (Group A) Local 10m ~lk

No. of ears

I

4 Tracer:

60m I

6hr I

4 4 HSA 5gr/dl 1 ml

12hr

24hr

48hr

I

I

I

4

10

7

Figure I. Experimental procedures for nonsensitized animals (group A) and sensitized animals (groups B and C).

undertaken to investigate the changes of RWM permeability to HSA in both normal chinchillas and chinchillas sensitized with HSA and to clarify the relationship between RWM permeability and degree of middle ear inflammation. The effect of steroid agents on RWM permeability was also studied. MATERIALS AND METHODS

Nonsensitized Animals (Group A) Twenty-two normal chinchillas weighing between 400 and 600 g were anesthetized with ketamine-HC1 (20 mg/kg, intramuscularly). One milliliter of human serum albumin (50 mg/ml, Sigma Chemical Co., St. Louis, Missouri) dissolved with sterile saline was instilled into the middle ear bulla of the chinchillas. The middle ear bulla was exposed surgically, and middle ear fluid retained in the middle ear bulla was recovered at 10 minutes and 1, 6, 12, 24, and 48 hours after instillation. After the middle ear bulla was washed several times with sterile saline, the perilymph and cerebrospinal fluid were collected with capillary glass tubes. The middle ear fluid samples were centrifuged at 2,000 rpm for 10 minutes at 4~ Blood samples were collected by cardiac puncture and put into test tubes containing 3.8% sodium citrate (1:10). After centrifugation, plasma samples were recovered.

Sensitized Animals (Groups B and C) Fifteen normal chinchillas were systemically immunized by 0.1 ml of HSA (10 mg/ml) containing the same volume of complete Freund's adjuvant (Sigma). The emulsion was injected subcutaneously under the back skin of the animals, and the animals were boostered twice: 3 days and 10 days after the first injection. When circulating antibody titers reached between x 24 and x 28, which is high enough to induce local inflammation, the left middle ear bulla was challenged by I ml of HSA (50 mg/ml) dissolved with sterile saline (group B), and the right middle ear bulla was challenged by I ml of HSA (50 mg/ml) containing 25 mg/ml of triamcino-

lone diacetate (Aristocort | Lederle Parenterals Co., Carolina) (group C, Fig. 1), The middle ear fluid retained in the middle ear bulla was collected at 12, 24, and 48 hours after HSA challenge. After the middle ear bulla was washed several times with sterile saline, the perilymph was collected. A blood sample was also collected by cardiac puncture. These samples, except for the perilymph, were stored at -70~ until use. Concentration of HSA was measured by electroimmunodiffusion8 using a commercial goat anti-HSA antibody (Cooper Biomedical, Malvern, Pennsylvania). Its detection limit was 10 ~g/ml. To avoid the cross reaction of chinchilla albumin against anti-HSA antibody, commercial antibody was preincubated with the same volume of normal chinchilla serum for I hour at 37~ The,HSA level in the perilymph was measured immediately after sampling. Concentration of HSA in recovered middle ear fluid was measured by electroimmunodiffusion as described earlier. The HSA level was expressed as the concentration ratio against the original HSA level (50 mg/ml) instilled into the middle ear hullo. Concentration of alpha-l-antitrypsin (alAT) in the middle ear fluid was measured by single radial immunodiffusion9 and that of alpha-2-macroglobulin (c~IM) in the recovered middle ear fluid was measured by electroimmunodiffusion using specific rabbit antibodies against purified chinchilla alAT and o~2M.10

The immune complex level in the middle ear fluid was measured by using a commercial assay kit (ELICA, | Cooper Biomedical). This kit is based on an enzyme-linked immunosorbent assay using horseradish-peroxidase-conjugated staphylococcal protein A that can bind to the IgG portion of immune complexes precipitated by polyethylene glycol. ~ Immune complex levels were expressed as the absorbance at 492 nm against the control, which did not contain samples. After collection of the middle ear fluid, an aliquot of each sample was plated directly onto both sheep blood and chocolate agar plates for culture. Antibody tiler against HSA was measured by the Ouchterlony method and expressed in the m a x i m u m dilution ratio (x 2x) of plasma samples that make a precipitin line. Statistical significance was determined by either the Student's t test or paired t test (between groups B and C).

Volume 9 Number 1 January 1988

35

RWM PERMEABILITY TO HSA IN OTITIS MEDIA 100

RESULTS All of the samples c o u l d be divided into three groups: nonsensitized animals (group A), sensitized animals w i t h local challenge of HSA into the left ear (group B}, and sensitized animals w i t h local challenge of HSA-containing triamcinolone diacetate into the right ear (group C). The bacterial c u l t u r e test w a s negative for all the m i d d l e ear fluid samples. Four of the seven sensitized animals in the experiment at 48 hrs after c h a l l e n g e s h o w e d a p p a r e n t v e s t i b u l a r s y m p t o m s (ataxic gait). T h e results are s h o w n in Figure 2. Purified chinchilla a l b u m i n as well as H S A developed a precipitin line against commercial anti-HSA antibody. With pretreatment of the absorption with chinchilla serum, no precipitin line could be detected in chinchilla albumin, suggesting that the interference of H S A m e a s u r e m e n t by chinchilla a l b u m i n could be prevented b y this treatment. The HSA level in p e r i l y m p h is s h o w n in a logarithmic scale against different times after instillation from 10 m i n u t e s to 48 hours in Figure 3. In group A, no H S A c o u l d be detected in the

American Journal

of Otoloryngology

36

Figure 2. Electrophoretic patterns of h u m a n serum alb u m i n (HSA).and chinchilla albumin before and after pretreatment with chinchilla plasma. A, before pretreatment; B, after pretreatment; a, HSA (50 mg/dl); b, HSA (25 mg/dl); c, c h i n c h i l l a albumin (80 mg/dl); d, chinchilla albumin (40 mg/dl); e, a + c; f, b + d, g, p e r i l y m p h in chinchilla of group A.

9

[ ] Control 50~" ~ GrOUp.

~

9

Q ~"I ~M

E

1 --h=--

10m ( 1/ 6)

,=A_

lhr

6hr

12h,r

24hr

_,=_==,~

__

48hr

Time

Figure 3. HSA levels in the perilymph at different times after instillation (geometric mean -+ SD). [3, group A (control}; [], group B (otitis media}, [], group C (otitls media + steroid); , significant difference from the value in group A (P < .01); , significant difference from the value at different times after instillation in group A (P < .01).

perilymph at 10 minutes and 1 hour after instillation. The peak value, with a geometric mean of 20.4 mg/dl (range, 11.5 to 36.3) was observed at 24 hours after instillation. This value was significantly higher than the values o b t a i n e d from other time points (P < .01). In plasma and cerebrospinal fluid, HSA could not be detected by the present method. In b o t h groups B and C, the h i g h e s t mean value of HSA could be observed at 12 hours after challenge and was significantly higher than that in group A (P < .01). The mean value of the HSA level decreased with the time after challenge, d e s p i t e large i n d i v i d u a l v a r i a t i o n s in both groups B and C. No HSA could be detected in plasma of animals in either group B or C b y the present method. Comparing HSA levels in each case b e t w e e n groups B and C, H S A level in group B was significantly higher than that in group C at 24 hours (P < .025) and 48 hours (P < .05) after challenge (Fig. 4). There was no significant correlation b e t w e e n HSA level in the PL and antibody titer in plasma in either group. The results are s u m m a r i z e d in Figure 5. In group A, HSA level in the recovered middle ear f l u i d s i g n i f i c a n t l y d e c r e a s e d at 24 a n d 48 hours after instillation (P < .05), w h e n compared to the value at 6 hours after instillation. A similar trend could be observed in both groups B and C, and HSA levels in both groups were significantly lower than those in group A at 12, 24, and 48 hours after HSA challenge (P < .01). The results of both cqAT and e~zM levels in the middle ear fluid are s h o w n in Figure 6. In group A, ohAT in the middle ear fluid could be de-

HAMAGUCHIET AL. ii

100

50

1.21,1 1

,...1

O,9 o

v

0~

,.J r

5 Z

8

0.4 0.3

0,2 0.1 12hr

24hr

48hr

IS

12

24

48

IHSTILI, ATIOH TIME (HR'J

Time

1006

at-Antltrypr i

i-, <

Figure 4 [left). Human serum albumin (HSA) levels in the perilymph (PL) between groups B and C. e, group B; O, group C; A, animals in group B with vestibular symptoms. There was a significant difference between groups (P < .025 at 24 hours and P < .05 at 48 hours after instillation).

ff 5O6

Figure 5 (top right). HSA level in recovered middle ear fluid at different times after instillation, (mean -4- SD). [], group A; [], group B; I , group C. v, significant difference from the value at 6 hours after instillation (P < .05}; ~ , significant difference from the value in group A (P < .01}. 0

Figure 6 (bottom right). Alpha-l-antitrypsin (cqAT) and alpha-2-macroglobulin (a2MG) levels in the middle ear fluid at different times after instillation (mean _+ SD). D, group A; @, group B; m, group C; *, significant difference from the value in group C (paired t test; P < .025 at 12 hours, P < .05 at 24 hours, and P < .05 at 48 hours; V, significant difference from the value in group C (P < .01).

I a ~ I~er oglo~ulln

0 6hr

12hr

24hr

48hr

Time

tected initially at 6 h o u r s after instillation. The level i n c r e a s e d w i t h t i m e after instillation, and the p e a k level w a s o b s e r v e d at 24 hours after instillation. C o n v e r s e l y , a2M in the m i d d l e ear fluid c o u l d be d e t e c t e d first at 24 hours after instillation a n d d e c r e a s e d by 48 h o u r s after instillation. In group B, a l A T level increased remarkably after HSA challenge. The highest level c o u l d be o b s e r v e d at 48 h o u r s after HSA challenge. In g r o u p C, t h e a l A T l e v e l was signific a n t l y l o w e r t h a n that in group B (12 hours, P < .025; 24 hours, P < .O5; a n d 48 hours, P < .05) a n d s i g n i f i c a n t l y higher t h a n that of group A (48 h o u r s , s < .01). A similar t r e n d w a s observed in the level of ot2M in b o t h g r o u p s B and C. The level of cr in g r o u p C was significantly lower

than that of group B (48 hours, P < .01) and significantly higher than that of group A (12 and 48 hours, P < .01). In group A, there was a significant correlation b e t w e e n H S A levels in the p e r i l y m p h and chAT levels in the middle ear fluid (r - .61, P < .01) (Fig. 7), but there was no significant correlation b e t w e e n H S A levels in the p e r i l y m p h and cqAT levels in the middle ear fluid of groups B and C. I m m u n e c o m p l e x levels in all middle ear fluid samples of group A were less than 0.015 (0.005 _+ 0.004). Conversely, m e a n values of i m m u n e c o m p l e x levels in the m i d d l e ear fluid w e r e 0,127 -+ 0.028 in group B and 0.104 _+ 0.010 in group C. There was no significant difference bet w e e n the groups.

Volume 9 Number 1 January 1988

37

RWM PERMEABILITY TO HSA IN OTITIS MEDIA

Y 50--

DISCUSSION

Low-molecular-weight substances such as electrolytes (sodium12.13), dye solutions (rhodamine~4), and antibiotics [gentamicin and neomycin,15; streptomycin,18; tetracycline, 17 and cefmetazole18) are reported to pass through the r o u n d w i n d o w m e m b r a n e into the inner ear within a short period of time. However, the passage of react 9 proteins such as horseradish peroxidase (45K3), human serum albumin (67K7}, and ferritin [445K 19) through the RWM seems to depend on the time after instillation, the molecular weight, and the state of the RWM. The HSA (MW, 67 kilodaltons), used as a tracer in this experiment, is a well-known spherical protein distributed widely in human tissues and body fluids. It has no cytotoxic effect on the host ceils, and its protein structure is very stable during middle ear inflammation. 2~ Moreover, albumins (HSA, bovine serum albumin, egg white albumin) are usually used as the antigenic proteins in antigen-induced middle ear inflammation in sensitized animals, 2~,2z Five g/dl of HSA instilled into the middle ear bulla is similar to the albumin concentration in plasma, although .d.. osmotic pressure effect and foreign-body reaction of a high concentration of HSA may affect the epithelial cells of the RWM. No HSA in the per 9 was detected at 10 minutes and 1 hour after instillation in group A. This finding indicates that the contamination by instilled HSA in the per 9 would be negligible, or only a small amount of HSA that is below the detection level can pass through the RWM. In group A, the HSA level in the per9 l y m p h and alpha-l-ant 9 and alpha-2react 9 levels in the recovered middle ear fluid reached the peak at 24 hours after instillation, and significant correlation between HSA level in the per 9 and cqAT level in the recovered m i d d l e ear fluid could be observed. Both a l A T and a2 M are acute-phase reactants, 2~ derived from the blood, and are at a very low level in the normal middle ear. ~~Their increased levels may reflect an increase of vascular permeability in the middle ear mucosa during inflammation. The results of %AT and ct2M level~ in the middle ear fluid suggest that instillation of HSA into the middle ear bulla could enhance the vascular permeability of the m i d d l e ' ear illUCOSa to lead the leakage of these proteins. It is possible that the existence of American HSA in the perilymph is due to the direct lymJournal phatic communication between the middle ear of Otolaryngology membrane and the perilymph of the scala tyro38

10-r-

9

5-

: >./.: ~

-1-

1

i

100

200

~1 - Ant 9

1

_1

x

300 400 (mg/dl)

Figure 7. The relationship between human serum albumin levels in the per 9 and alpha-l-ant 9 levels in the middle ear fluid of group A. A significant correlation could be observed (Log Y = 0.0027X + 0,175, r = .61, P < ,01).

pant. 24 However, the results of the present study suggest that HSA could pass through RWM at a longer period of time after instillation because nonspecific inflammation can develop by a high concentration of HSA, which can enhance RWM permeability by damaging the epithelial layer of the RWM. The decrease of HSA level in middle ear fluid at 24 and 48 hours after instillation may be attributed to the excretory function of eustachian tube. The decrease of HSA levels in the p e r 9 at 48 hours after instillation could be explained by distribution of HSA into other compartment of the inner ear, recovery of the tympanic surface of the RWM from nonspecific inflammation, and the decrease of HSA levels in the middle ear fluid by eustachian tube clearance. The results of IgG-related immune complex levels in the middle ear fluid of both groups B and C indicate that HSA challenge into the middle ear bulla forms immune complexes with specific IgG leaking from the vessels. Immune complex seems to be a major factor inducing the middle ear i n f l a m m a t i o n observed in both groups B and C. A significant decrease of antigenically intact HSA levels in the middle ear fluid of both groups B and C compared to group A may be due to both the elimination by eustachian tube clearance and immune complex formation to break the antigen9 of HSA in the

HAMAGUCHIET AL, (MIDDLE

EAR)

....,

)

i ol/o /

I /

/INNER

s,

S

I1 o l o J o ~

9

sper f I r

-tr H

.,. / - ". ; " / _ l ~ i . ~ : ~ 0 1 . t

_

H::Ae "

T

. /

4"0 ....

c3, 9

~ago~yt~i5 J L

EAR)

,f~

specific * ' ~ E ' \

- k

/

,s

RWM

"-"~'~'1- - - PI

L

+

9

HSA

llC /Iormatlon

) i

Arthus t y p e ~nflammat|on cl~mlCal

e I'

..

rxJ.i., ~ ip/T,~o,~,~,j

9 ,,m,..,,o.

\~\

"

--

" _

".

9

/r"~

Ii~

1 5

I

..L

Ihl IYI

"'-- ~,

eeH

inner ~a r dysfunotlon

II

I I

I-{

"*."

'-,

'

.Io_,~ ~

:.

Figure S. The relationshipbetweenantigen-inducedotitis mediaand the passage of human serum albumin into the inner ear through the round window membrane. middle ear fluid, which is shown schematically in Figure 8. The levels of chAT and aeM in the middle ear fluid increased remarkably after local challenge of HSA. It is conceivable that otitis media develops rapidly after HSA challenge by enhancing vascular permeability in the middle ear mucosa. The significantly higher levels of HSA in the perilymph found in group B, than in group A at 12 hours after chal]enge suggest that RWM permeability to HSA increases considerably in the very early stage of antigen-induced otitis media. However, the HSA level in the perilymph had large individual variation (some of the peril y m p h samples had very low HSA levels) at 24 and 48 hours after challenge, and had no significant correlation with the levels of cqAT and (x2M. These results indicate that RWM permeability does not increase proportionately with the d e v e l o p m e n t of a n t i g e n - i n d u c e d otitis media. Unfortunately, we could not measure immune complex levels in the perilymph because of small amounts of perilymph samples (5 to 10 ~l per each sample). Immunologic studies of the inner ear have shown that chinchillas sensitized by egg white albumin had specific IgG2 antibody in the PL, most of which was derived from serum IgG2 antibody leaking from vessels in the perilymphatic space. 25 The possibility of local production of IgG antibody in the perilymphatic space and the passage of IgG through the RWM is also suspected. 28,27 Therefore, when HSA instilled into the middle ear bulla of chinchillas

sensitized with HSA passes through the RWM into the inner ear, HSA m a y form a complex with anti-HSA IgG in the perilymph (Fig. 8). The significant differences of HSA level in the perilymph, and chAT and (x2M levels in the middle ear fluid between groups B and C indicate that triamcinolone, which w e used in the present study as an anti-inflammatory agent, could affect the RWM p e r m e a b i l i t y by suppressing the degree of antigen-induced otitis media. Corticosteroid agents like triamcinolone have various suppressing functions against allergic inflammation, such as the inhibition of the release of chemical mediators. Inhibition of the release of vascular-permeability-enhancing factors such as histamine, bradykinin, and p~ostaglandins, which may increase RWM permeability by loosening intercellular junction, 2s may be related to the suppression of the RWM permeability. B e c a u s e t r i a m c i n o l o n e c a n pass e a s i l y through the RWM,2Qit may be able to play a role in the suppression of antigen-induced inflammation taking place in the inner ear. This speculation can be supported by the fact that four sensitized chinchillas showed vestibular symptoms on the left side without the steroid treatment. We reached the following conclusions: 1. Only a small amount of HSA could pass through the normal RWM within 1 hour, and a certain amount of time is necessary for HSA to pass into the perilymphatic space. 2. A high concentration of HSA in the middle

Volume 9 Number 1 January 1988

39

RWM PERMEABILITYTO HSA IN OTITIS MEDIA

ear bulla can induce an increase of the RWM permeability to HSA by nonspecific inflammation in the middle ear. 3, The RWM permeability to HSA reaches its m a x i m u m level about 24 h o u r s after middle ear bulla instillation in nonsensitized animals. 4. In antigen-induced.otitis media, the RWM permeability to antigenic protein (HSA) increased considerably in the very early stage of inflammation (about 12 hours), and the individual variation also increased in proportion to the development of middle ear inflammation. 5. Triamcinolone could affect the RWM permeability by suppressing the degree of antigen-induced inflammation in the middle ear.

References 1, Franke K: Freeze-fracture aspects of the junctional complexes in the round window membrane. Arch Otorhinolaryngol 1977;271:331-337 2. Schacherin PA, Paparella MM, Duvall AJ: The normal chinchilla round window membrane. Arch Otolaryngol 1982;108'.550-654 3. Rhodes T: Round window permeability to horseradish peroxidase. An electron microscopic study. Thesis, University of Minnesota Graduate School, Minneapolis, Minnesota, 1980 4. Goycoolea MV, Paparella MM, Goldberg B, et el: Permeability of the round window membrane in otitis media. Arch Otolaryngol 1980;106:430-433 8. Goycoolea MV, Paparella MM, Goldharg B, et al: Permeability of the middle ear to staphylococcal pyogenic exotoxin in otitis media. Int J Pediat Otorhinolaryngol 1980;1:301-308 6. Meyerhoff WL, Kim CS, Paparella MM: Pathology of chronic otitis media. Ann Otol Rhinol Laryngol 1978;87:749-760 7. Goldberg B, Goycoolea MV, Schleivert PM, et al: Passage of albumin from the middle ear to the inner ear otitis media in chinchilla. Am J Otolaryngol 1981;2:210214 8. Laurell CB: Quantitative estimation of proteins by electrophoresis in agarose-gel-containing antibodies. Anal Biochem 1986;15:45-52 9. Mancini G, Carbonara AO, Heremans JF: Immunological quantitation of antigens by single radial immunodiffusion. Immunochemistry 1965;2:235-544 10. Hamaguchi Y, Juhn SK: Protease inhibitors in middle

American Journal of Otolaryngology

4O

ear effusions from experimental otitis media with effusion. In preparation 11. Male DK: Detection of immune complexes. In Roitt IM, Brostoff J (ads): Immunology. New York, Gowar, 1985, pp 9-10 12. Wever EG, Bray CW: Application of sodium chloride to the round window membrane. Ann Otol Rhinol Laryngol 1937;46:289-302 13, Brady DR, Pearce JP, Juhn SK: Permeability of round window membrane to Na or RISA. Arch Otorhinolaryngol 1976;214:183-184 I4. Kaupp H, Geibel W: Distribution of marked peritymph to the subarachnoidal space. Arch Otorhinolaryngol 1980;229:245-253 15. Smith BM, Myar MG: The penetration of gentamicin and neomycin into perilymph across the RWM, Otolaryngol Head Neck Surg 1979;87:888-8gl 16. Schuknecht HF: Pathology of the ear. Cambridge, Massachusetts, Harvard University Press, 1974, pp 276278 17. Breuninger H, Giebel W: Untersuchungen uber die Durch gangigkeit dar Membran des runden Fensters fur Tetracycline. Arch Otorhinolaryngol 1975;210:362 18. Okuno T, Nomura Y: Permeability of the round window membrane. Arch Otorhinolaryngol 1984;240:103-106 19. Nakai Y, Kaneko M: Round window membrane. Submicroscopic structure and permeability. Pract Otol Kyoto 1975;68:223-232 20. Carlsson B, Lundberg C, Ohlsson K: Protease inhibitors in middle ear effusions. Ann Otol Rhinol Laryngol 1981;90:38-41 21, Ryan AF, Catanzaro A, Wasserman SI, et al: Secondary immune response in the middle ear: Immunological, morphological and physiological observation. Ann Otal Rhinol Laryngol 1986;95:242-249 22, Dake Y, Nakanishi H, Jinnin T, et al: Experimental studies on immunogenesis of otitis media with effusion. Auris Nasus Larynx 1985;12(suppl 1):186-187 23. Pepys MB, Balz M, Gomar K, et el: Serum amyloid Pcomponent is an acute phase reactant in the mouse. Nature 1979;278:259-261 24. Arnold WJ: Role of perilymph in the early stage of serous otitis. Ann Otol Rhinol Laryngol 1976;85(suppl 25) 73-80 25. Mogi G, Lim DJ, Watanabe N: Immunologic study on the inner ear. Arch Otolaryngol 1982;108:280-285 26. Harris JP: Immunology of the inner ear. Evidence of local antibody production. Ann Otol Rhinol Laryngol 1984;93:157-162 27. Harris JP, Ryan A: Perilymph antigen and antibody levels following an immunologically-inducedmiddle ear effusion. Auris Nasus Larynx 1985;12(suppl 1): 194-196 28. Boucher RC: Effect on histamine and methacoline on guinea pig tracheal permeability. J Appl Physiol 1978;45:939-948 29. Nomura Y: Otological significance of the round window. In Pfultz CR (ed): Advances in Otorhinolaryngology. Basel, Karger, 1984