Acta Oto-Laryngologica
ISSN: 0001-6489 (Print) 1651-2251 (Online) Journal homepage: http://www.tandfonline.com/loi/ioto20
Clearance of Middle Ear Effusions by the Mucociliary System J. Sade, F. A. Meyer, M. King & A. Silberberg To cite this article: J. Sade, F. A. Meyer, M. King & A. Silberberg (1975) Clearance of Middle Ear Effusions by the Mucociliary System, Acta Oto-Laryngologica, 79:3-6, 277-282, DOI: 10.3109/00016487509124685 To link to this article: http://dx.doi.org/10.3109/00016487509124685
Published online: 08 Jul 2009.
Submit your article to this journal
Article views: 8
View related articles
Citing articles: 2 View citing articles
Full Terms & Conditions of access and use can be found at http://www.tandfonline.com/action/journalInformation?journalCode=ioto20 Download by: [McMaster University]
Date: 09 April 2016, At: 20:24
Acta Otolaryngol 79: 277-282, 1975
CLEARANCE OF MIDDLE EAR EFFUSIONS BY THE MUCOCILIARY SYSTEM J. Sadt,l F. A. Meyer, M. King2 and A. Silberberg
Downloaded by [McMaster University] at 20:24 09 April 2016
From the Polymer Department, Weizmann Institute of Science, Rehovot, Israel
Abstract. There are two extreme types of middle ear effusion leading to hearing loss (a) a rubber-like effusion seen in secretory otitis media and (b) a water-like effusion seen in serous otitis media. The possibility is considered that the degree of crosslinking in these two extreme cases is the basis of an altered mucus transport rate that leads to an accumulation of effusions and hence impaired hearing. It has been shown (King et al., 1974) that the requisite rheological property for transport activity is not unique to mucus structural macromolecules but is found with other polymeric systems that are loosely crosslinked e.g. guaran, polyacrylamide, gelatin and agarose. Studies on one of these systems, guaran, indicate that the transport rate is dependent on the degree of crosslinking with a maximum rate found close to the gel point, i.e. in a region where there are very few crosslinks per macromolecule. The finding that mucus from different mucociliary epithelial sources involves a chemically similar structural glycoprotein suggests that differences observed in transport rate between various mucus samples are more likely due to differences in crosslinking than chemical variations of the glycoprotein units. By analogy with the model guaran system, it is suggested that the two types of middle ear effusions represent extremes in crosslinking and that their transport rate lie to either side of the optimum prcsumably represented by the normal secretion. Factors such as charge, concentration and the influence of temperature on mucus crosslinking are discussed.
Clearance insufficiency of effusions from the middle ear can be considered from the’point of inadequacies of the middle ear mucociliary system. Theoretically, a ciliated epithelium can malfunction due t o various causes and, therefore, be responsible for middle ear clearance deficiency. These causes could be an inadequate This investigation was supported in part by Grant No. NS 10048-04from the National Institutes of Health. One of us (M. K.) is indebted to the N.R.C. (Canada) for a postdoctoral fellowship. Established investigator of the Chief Scientists Bureau, Ministry of Health. The Meir Hospital, Kfar Saba and the Weizmann Institute of Science. Present Address: Meakins-Christie Laboratories, McGill University, Montreal, Canada.
number of cilia, ciliary dysfunction, or an excessive mucus mass. A change in mucus characteristics might possibly be considered as well. For decades it was accepted that a ciliated respiratory epithelium lines the Eustachian tube and in the last few years such an epithelium was described as lining important parts of the middle ear as well (SadC, 1966; Lim & Hassel, 1969). This epithelium in the middle ear can function actively in clearing foreign material, as demonstrated experimentally by the rapid clearance of radio-opaque material (Compere, 1958), and as observed directly through tympanic membrane perforations (Sadt, 1967). Normal mucosa of the respiratory tract and the middle ear has a layer of ciliated cells with mucus-filled goblet cells scattered among them (Gray, 1928). Below these, there is a stratum of mucus-forming glands. The ciliary layer is topped by a micron-thin mucus blanket (Fig. 1); it is this blanket which moves in response to the ciliary beat, and carries with it such foreign particles as may be deposited on it. These events have been observed in laboratory animals as well as in the middle ear (Sadt, 1967). The ciliated mucosa of the frog and toad palates and of the trachea of the cat are so similar in structure and function to human, including middle ear mucosa, that for convenience it was decided by us to use them for many of our experimental studies (Sad6 et al., 1970). The universality of the involved principle and its applicability to middle ear problems, while not certain, is highly probable. The similarity of the epithelia used and the interchangeability of their mucus has been demonstrated. In our clearance experiments, as already deActa Otolaryngol 79
Downloaded by [McMaster University] at 20:24 09 April 2016
278 J. Sad6 et nl.
Fig. 1. Histologic section of the mucociliary system: specimen is from a toad’s palate mucosa. The PAS-positive mucus blanket (arrow) is seen to come from goblet cells (G). Below the mucus blanket cilia are seen ( C ) .
monstrated by Hilding (1957), ciliary activity was maintained for many hours even though the preparation was isolated and such preparations can be considered to be representative of in vivo conditions. It follows that the question as to the precise role of the mucus blanket could be approached through experiments with these model preparations. For this purpose correlations were sought between clearance speed and mass of various particles, such as charcoal dust, glass and steel balls, and also mercury drops, applied to the active mucociliary epithelial model system described, i.e., the isolated palates of the frog and toad or cat trachea (Sad6 et al., 1970). Such foreign particles were cleared from the preparations at an average speed of 1 cm per minute, the inclination to the vertical at which the preparation was held playing no role. Clearance of such foreign materials on each preparation was found to occur at a constant speed in the range from lop7 to 1 g of particle weights tested. The upper limit represents the biggest mercury drop which it was possible to handle and the lowest the smallest charcoal dust particle which could still be discerned using an operating microscope (magnification up to x 64). This constant speed apparently reflects the rate of travel of the mucus blanket, the movement of which was observed Acta Otolaryngol79
to be triggered by the particles without being affected by their weight. The foreign particles may thus be regarded as passive travellers on top of a biologic conveyer belt. It was also observed that when performing several trials on the same preparation, transport time increased until no clearance of foreign particles was achieved (Fig. 2). Cilia, however, continued to beat for hours and at times days after transport had ceased. The time from the beginning of the experiment until transport standstill differed greatly according to the type of preparation and the season. On preparations where transport of particles had stopped, but cilia continued to beat, samples of mucus (accumulated at the cut edge of the preparation) could be placed on the tissue with the result that the mucus blob started moving immediately upon contact with the tissue. The velocity attained was, in this case, similar to the initial transport velocity of foreign particles, and its velocity did not depend on the load that it carried. When mucus samples taken from heterologous epithelia was placed on a depleted preparation, transport was also restored but the velocity obtained was not necessarily the same as with mucus samples from the preparation itself. For example, midcycle bovine cervical mu-
Downloaded by [McMaster University] at 20:24 09 April 2016
Clearance of middle ear effusions
cus restores transport velocity but the rate of transport attained is only about 30% of that with the frog's own palatal mucus. The fact that solid particles are no longer transported after a certain stage is reached, unless mucus is added, would indicate that the amount of mucus which can be secreted by the isolated system is limited. Mucus must be applied externally in order to enable the cilia to bring about transport. The condition of the epithelium at this stage is thus a state of mucus depletion. A far reaching conclusion about the role of mucus can be drawn from these experiments. Mucus is not only a vehicle for various substances such as immunoglobulins (Gottschalk, 1966), and a moving blanket to which foreign bodies adhere, but it is also an essential mechanical coupler whose presence is imperative for translating ciliary beat into effective clearance. Without mucus, clearance does not take place in spite of a regular ciliary beat. What then is mucus and what is its structure to make it suit these physiological requirements. Mucus associated with ciliated epithelia is a gel, which in contact with physiological saline, does not disperse but remains as a distinct phase. The structural components of mucus must thus be intermolecularly crosslinked or at least very heavily entangled with each other. Studies performed, using, for example, midcycle cervical mucus which can be obtained in large
I
'
I
> 01
'0
10
20
30
40
50
60
Time a f t e r death, minutes
70
48 hrs
Fig. 2. Velocity of transport of foreign particles (curved lower lines) and of added mucus (horizontal) on frog and toad palates as functions of time after death. --, Average velocity in 30 experiments; ---, range of velocities observed.
bv the mucociliavy system 279
Fig. 3. Analyses for the acidic and neutral amino acids (a.a.); the basic amino acids represent less than 7 % of the total amino acids.
quantities, indicate that the structural macromolecule is a glycoprotein of high molecular weight ( > lo6) (Gibbons & Glover, 1959). There are sugar side chains attached via 0-glycosidic links to threosine or serine on the protein backbone. A model consistent with the chemical composition indicates that 1 out of every 4-6 amino acids is involved in linkage to sugar. The average carbohydrate side chain contains some 9 sugar units, 1 of sialic acid, 1 of fucose, 2 galactosamine, 2 of glucosamine and 3 of galactose (Meyer et al., 1973). A comparison of structural molecules from various mucus sources was performed (unpublished results). The study included mucus samples isolated from patients with secretory otitis media and chronic bronchitis as well as from the cow's czrvix at midcycle and the frog's palate. In each case the mucus was brought into solution by use of dithiothreitol-a reagent that breaks disulphide bonds (Vered et al., 1972). Fractionation of the solubilized material by centrifugation in a caesium chloride density gradient resulted in the isolation of a glycoprotein fraction containing the bulk of the sugars of the secretion and a fraction containing mainly serum proteins. In gel electrophoresis and in sedimentation studies, the glycoprotein fraction behaved as a single entity, though the so-called microheterogeneity in these materials is evident. Similarities were noted in the gross chemical composition. For example, the glycoproteins could be banded in a caesium chloride density Acta Otolaryngol 79
280 J. Sad&et al. 1.5
I I I
~.* :
.
I
I
I
i G Gunran concentrotion “%I
Downloaded by [McMaster University] at 20:24 09 April 2016
Fig. 4 . The ratio of the transport rates of guaran gel to frog palate mucus versus guaran concentration. Each point represents 5 to 10 readings of both guaran and mucus velocities on a single depleted palate.
gradient. Analyses for amino sugars, neutral sugars, sialic acid and protein were performed on material representing the fraction where the peak occurred in the gradient. For the three mammalian glycoproteins the percentages expressed relative to the total material analyzed were in the range amino sugars 36-38 %, neutral sugars 30-34 76, sialic acid 15-18 and protein 14-18 %. The analyses for the frog material were in this range except for sialic acid which was significantly lower (7 %). It has been shown that this sugar is not of great significance in the biological transport function of mucus (Meyer et al., 1975). Similarities between the four isolated glycoproteins were also seen in the amino acid composition (Fig. 3). This study indicates that a glycoprotein of chemically similar structure is common to mucus from different sources and that the gel nature of mucus presumably arises as a result of the organization of these chains into large three-dimensional structures and their stabilization by intermolecular crosslinks. This chemical evidence supports the result that from the point of view of transport mucus samples from different sources function exchangeably on epithelia. It is obvious that there must indeed be a rheological requirement for mucociliary transport. A recent study (King et al., 1974) has shown that the requisite rheological properties can be specified and that these are not only not specific to mucus but can be obtained with materials of quite different chemical structure. A number of Acta Otolaryngol79
macromolecular materials e.g., guaran, polyacrylamide, agarose and gelatin which are chemically quite dissimilar to mucus have been studied. I t was found that these materials are capable of performing the transport function on a ciliated epithelium provided that these systems possess a slight degree of crosslinking. It should also be noted that foreign systems displaying transport activity generally did so at different transport rates. Indeed it would seem that a qualitative mucus factor is important in influencing the observed rate. While differences in transport rate were observed for foreign systems, differences are also seen for mucus samples from different sources. For example, midcycle cow cervical mucus is transported at only about 30% of the rate of frog’s own palatal mucus. This occurs in spite of the fact that both are based on similar glycoprotein constituents. Thus differences in the physical organization of the structural macromolecules alter the rheological character of the system and determine its effectiveness as mechanical coupler. Particularly by using the same chemically defined system; guaran at various degrees of crosslinking we (King et al., 1974) found that below the gel point (0.5% guaran concentration) no transport is observed (Fig. 4). In the gelation region transport rate rises dramatically to the value obtained with the frog’s own mucus and this rate is maintained in the concentration range 0.5% to 0.9%. As the system becomes more crosslinked and solid-like transport rate falls off and at concentrations greater than 1.8 % the gel behaves like a foreign particle on a depleted mucociliary epithelium and is not transported. Thus the degree of crosslinking and the physical organization of the gel are highly important parameters. It appears that the rheological requirements for mucociliary transport are satisfied only by macromolecular systems which are very weak gels. These observations explain why in earlier experiments (Sad6 et al., 1970), where solutions of macromolecules were used to substitute for mucus, no success was obtained. In fact it is likely that when macromolecular solutions are
Downloaded by [McMaster University] at 20:24 09 April 2016
Clearance of middle ear effusions by the mucociliary system 281 applied to the epithelium, the macromolecules penetrate into the serous layer between cilia and interfere with the ciliary beat. In the physiopathology of hearing loss, middle ear effusions are of prime importance (Sad&, 1974) and are essentially of two extreme kinds: (A) very thick rubber-like effusions, as seen mostly in children and termed “glue ears” secretory otitis media; and (B) water-like effusions, as seen mostly in adults and often termed serous otitis media. The difference in consistency between the two secretions is very likely the result of a difference in crosslinking (Vered et al., 1972) the water-like effusion is probably not a gel. On the basis of the evidence presented, these two types of secretions may in fact represent two extremes in the crosslinking of the glycoprotein structural macromolecules. In “glue ears” the mucus is solid-like and presumably heavily crosslinked and in serous otitis, the glycoproteins are not crosslinked at all. In analogy with the results in Fig. 3, the transport rate may be zero, the two systems being to either side of the optimum represented presumably by the normal secretion. In either of these cases the epithelia would be secreting a mucus which could not be transported away fast enough resulting in a net accumulation of secretions in the middle ear which eventually impairs hearing. In the normal state the degree of crosslinking of the glycoprotein in mucus from different sources is probably matched to the required transport rate for the individual tissue. In pathological states, such as those of middle ear effusions, crosslinking would seem to be altered. The reason for differences in the degree of crosslinking in such glycoprotein-based mucus systems is not clear. Minor changes in p H and ionic strength of the media have little influence. Earlier studies attributed the marked differences in elasticity, a measure of the degree of crosslinking to the sialic acid content of mucus. For example, a difference in sialic acid content has been noted for the glycoprotein isolated from the watery midcycle cervical mucus and that from the more elastic secretion found at other stages of the cycle (Gibbons & Glover, 1959;
Hatcher & Jeanloz, 1973). However, recently it has been shown that removal of sialic acid with a specific enzyme does not alter the elasticity of mucus (Meyer et al., 1975) nor its biological transportability (King et al., 1974). Differences in crosslinking though may arise as a result of different concentrations of glycoprotein component in the mucus and to the passage of time. For example, it has been found that mucus on concentrating does not swell back to its initial volume on standing in buffer. This would suggest that additional crosslinks are formed and indeed the material is found to be more rubberlike. On the other hand, mucus goes spontaneously into solution on standing. The temperature dependence of this process suggests that a relatively low energy of activation is involved indicating disruptions of secondary bonds or complex entanglements (unpublished results).
ZUSAMMENFASSUNG Es gibt zwei extreme Falle von Mittelohr-Ausfluss, die zu einem Gehorverlust fiihren: a) ein gummi-ahnlicher Ausfluss, der bei exsudativer Mittelohrentzundung vorkommt und b) ein wasserahnlicher Ausfluss, der bei seroser Mittelohrentzundung vorkommt. Es wird die Moglichkeit besprochen, dass der Vernetzungsgrad in diesen zwei extremen Fallen die Grundlage der verinderten Transportgeschwindigkeit des Schleimes ist, wodurch es zu einer Anreicherung der Sekrete und demzufolge zur Verminderung des Gehores kommt. Es wurde gezeigt (King et al., 1974), dass diese notwendige rheologische Eigenschaft fur eine Transportaktivitat nicht auf strukturierte Makromolekule im Schleim begrenzt ist, sondern auch bei anderen Polymer-Systemen gefunden wird, die locker vernetzt sind, wie z. B.: Guaran, Polyacrylamid, Gelatine und Agarose. Untersuchungen an einem dieser Systeme, Guaran, zeigen, dass die Transportgeschwindigkeit von dem Vercetzungsgrad abhangt mit einem Maximum nahe dem Gelierungspunkt; das heisst in einem Bereich mit sehr wenigen Vernetzungen pro Makromolekul. Die Beobachtung, dass der Schleim von verschiedenen Flimmerschleimhautepithelien eine chemisch ahnliche Glukoproteinstruktur hatte, spricht dafur, dass die beobachteten Unterschiede der Transportgeschwindigkeit zwischen verschiedenen Schleimproben mehr auf Unterschieden im Vernetzungsgrad beruhen, als auf Veranderungen der Glycoprotein-Einheiten. In Analogie ZLI dem Guaran-Model1 wird angenommen, dass die zwei Typen von Mittelohr-Ausfluss Extreme der Vernetzung darstellen und dass deren Transportgeschwindigkeiten zu beiden Seiten eines Optimums liegen, das bei normaler Acta Otolaryngol 79
282 J. Sade‘ et al. Sekretion vorkommt. Faktoren der Schleimvernetzung, wie Ladung, Konzentration und der Einfluss der Temperatur werden diskutiert.
Downloaded by [McMaster University] at 20:24 09 April 2016
REFERENCES Compere, W. E., Jr. 1958. Tympanic cavity clearance studies trans. A A O O 62, 444. Gibbons, R. A. & Glover, F. A. 1959. Physicochemical properties of two mucoids from bovine cervical mucin. Biochem J 73, 217. Gottschalk, A. 1966. Glycoproteins. Elsevier Publ. Co., Amsterdam and London. Gray, J. 1928. Ciliary mouement. Cambridge Univ. Press. Hatcher, V. B. & Jeanloz, R. W. 1973. The use of glycosidases in the structure elucidation of the carbohydrate chains of a,-acid glycoprotein from human plasma and cervical mucin from Macaca Radiata. Colloque International du C.N.R.S. sur les glycoconjugues,Lille. Hilding, A. C. 1957. Ciliary streaming in the bronchial tree element in carcinogenesis. New Engl J Med 256, 634. King, M., Gilboa, A,, Meyer, F. A. & Silberberg, A. 1974. On the transport of mucus and its rheological simulants in ciliated systems. Am Rev Resp Dis (accepted for publication). Lim, D. J. & Hassel, B. 1969. Human middle ear epithelium. Arch Otolaryngol89, 835. Meyer, F. A., Eliezer, N., Silberberg, A., Vered, J., Sharon, N. & Sade, J. 1973. An approach to the biochemical basis for the transport function of epithelial mucus. Bull Physiopathol Resp Nancy 9, 259. Meyer, F. A,, King, M. & Gelman, R. A. 1975. On the role of sialic acid in the rheological properties of mucus. Biochim Biophys Acta (submitted for publication). Sadt, J. 1966. Middle ear mucosa. Arch Otolaryngol 84, 137. - 1967. Ciliary activity and middle ear clearance. Arch Otolaryngol86, 128. - 1974. The biopathology of secretory otitis media. Ann Otol Rhinol Laryngol83, 59. Sade, J., Eliezer, N., Silberberg, A. & Nevo, A. 1970.
Acta Otolaryngol 79
The role of mucus in transport by cilia. Am Rev Resp Dis 102, 48. Vered, J., Eliezer, N. & Sade, J. 1972. Biochemical characterization of middle ear effusions. Ann Otol Rhinol Laryngol81, 394.
J. Sade, M.D. Dept. of Otolaryngology Meir Hospital Kfar Saba Israel
DISCUSSION D. Hilding: Did you have a chance to test the effect of the mucoid material from a “glue ear” on the cilia of a cat trachea or frog palate? B. McCabe: It is clear that this disease is not a pure Eustachian tube disease, but we have all witnessed the phenomenal resolution of symptoms upon simply ventilating the ear. I wonder what role you feel that the Eustachian tube does play in this disease. G. Zechner: Mucus in the middle ear comes from goblet cells, ciliated cells and normal epithelial cells under certain conditions. In addition, some mucus comes from the Eustachian tube. For ciliary action, I want to give credit to Messerklinger’s findings: the ciliary organ needs for active transport a gel formation layer within which to beat. This gel, which we call mucus, is influenced by pH, ionic concentration and other facts. J. Sad6 (Reply) to Mr Hilding: You are of course right and this should be done, we have not done it so far because of technical difficulties in transporting the whole lab nearer to the operating theatre and we did not want to carry the mucus to the Weizmann Institute 30 miles away. To Mr McCabe: The Eustachian tube is obviously not functioning as it should in secretory otitis media, i s . it probably does not pump air in sufficient amounts into the middle ear but this does not mean that it is mechanically obstructed. To Mr Zechner: The p H and ionic concentration were indeed measured and their range is rather wide before cilia cease beating.
ISSN: 0001-6489 (Print) 1651-2251 (Online) Journal homepage: http://www.tandfonline.com/loi/ioto20
Clearance of Middle Ear Effusions by the Mucociliary System J. Sade, F. A. Meyer, M. King & A. Silberberg To cite this article: J. Sade, F. A. Meyer, M. King & A. Silberberg (1975) Clearance of Middle Ear Effusions by the Mucociliary System, Acta Oto-Laryngologica, 79:3-6, 277-282, DOI: 10.3109/00016487509124685 To link to this article: http://dx.doi.org/10.3109/00016487509124685
Published online: 08 Jul 2009.
Submit your article to this journal
Article views: 8
View related articles
Citing articles: 2 View citing articles
Full Terms & Conditions of access and use can be found at http://www.tandfonline.com/action/journalInformation?journalCode=ioto20 Download by: [McMaster University]
Date: 09 April 2016, At: 20:24
Acta Otolaryngol 79: 277-282, 1975
CLEARANCE OF MIDDLE EAR EFFUSIONS BY THE MUCOCILIARY SYSTEM J. Sadt,l F. A. Meyer, M. King2 and A. Silberberg
Downloaded by [McMaster University] at 20:24 09 April 2016
From the Polymer Department, Weizmann Institute of Science, Rehovot, Israel
Abstract. There are two extreme types of middle ear effusion leading to hearing loss (a) a rubber-like effusion seen in secretory otitis media and (b) a water-like effusion seen in serous otitis media. The possibility is considered that the degree of crosslinking in these two extreme cases is the basis of an altered mucus transport rate that leads to an accumulation of effusions and hence impaired hearing. It has been shown (King et al., 1974) that the requisite rheological property for transport activity is not unique to mucus structural macromolecules but is found with other polymeric systems that are loosely crosslinked e.g. guaran, polyacrylamide, gelatin and agarose. Studies on one of these systems, guaran, indicate that the transport rate is dependent on the degree of crosslinking with a maximum rate found close to the gel point, i.e. in a region where there are very few crosslinks per macromolecule. The finding that mucus from different mucociliary epithelial sources involves a chemically similar structural glycoprotein suggests that differences observed in transport rate between various mucus samples are more likely due to differences in crosslinking than chemical variations of the glycoprotein units. By analogy with the model guaran system, it is suggested that the two types of middle ear effusions represent extremes in crosslinking and that their transport rate lie to either side of the optimum prcsumably represented by the normal secretion. Factors such as charge, concentration and the influence of temperature on mucus crosslinking are discussed.
Clearance insufficiency of effusions from the middle ear can be considered from the’point of inadequacies of the middle ear mucociliary system. Theoretically, a ciliated epithelium can malfunction due t o various causes and, therefore, be responsible for middle ear clearance deficiency. These causes could be an inadequate This investigation was supported in part by Grant No. NS 10048-04from the National Institutes of Health. One of us (M. K.) is indebted to the N.R.C. (Canada) for a postdoctoral fellowship. Established investigator of the Chief Scientists Bureau, Ministry of Health. The Meir Hospital, Kfar Saba and the Weizmann Institute of Science. Present Address: Meakins-Christie Laboratories, McGill University, Montreal, Canada.
number of cilia, ciliary dysfunction, or an excessive mucus mass. A change in mucus characteristics might possibly be considered as well. For decades it was accepted that a ciliated respiratory epithelium lines the Eustachian tube and in the last few years such an epithelium was described as lining important parts of the middle ear as well (SadC, 1966; Lim & Hassel, 1969). This epithelium in the middle ear can function actively in clearing foreign material, as demonstrated experimentally by the rapid clearance of radio-opaque material (Compere, 1958), and as observed directly through tympanic membrane perforations (Sadt, 1967). Normal mucosa of the respiratory tract and the middle ear has a layer of ciliated cells with mucus-filled goblet cells scattered among them (Gray, 1928). Below these, there is a stratum of mucus-forming glands. The ciliary layer is topped by a micron-thin mucus blanket (Fig. 1); it is this blanket which moves in response to the ciliary beat, and carries with it such foreign particles as may be deposited on it. These events have been observed in laboratory animals as well as in the middle ear (Sadt, 1967). The ciliated mucosa of the frog and toad palates and of the trachea of the cat are so similar in structure and function to human, including middle ear mucosa, that for convenience it was decided by us to use them for many of our experimental studies (Sad6 et al., 1970). The universality of the involved principle and its applicability to middle ear problems, while not certain, is highly probable. The similarity of the epithelia used and the interchangeability of their mucus has been demonstrated. In our clearance experiments, as already deActa Otolaryngol 79
Downloaded by [McMaster University] at 20:24 09 April 2016
278 J. Sad6 et nl.
Fig. 1. Histologic section of the mucociliary system: specimen is from a toad’s palate mucosa. The PAS-positive mucus blanket (arrow) is seen to come from goblet cells (G). Below the mucus blanket cilia are seen ( C ) .
monstrated by Hilding (1957), ciliary activity was maintained for many hours even though the preparation was isolated and such preparations can be considered to be representative of in vivo conditions. It follows that the question as to the precise role of the mucus blanket could be approached through experiments with these model preparations. For this purpose correlations were sought between clearance speed and mass of various particles, such as charcoal dust, glass and steel balls, and also mercury drops, applied to the active mucociliary epithelial model system described, i.e., the isolated palates of the frog and toad or cat trachea (Sad6 et al., 1970). Such foreign particles were cleared from the preparations at an average speed of 1 cm per minute, the inclination to the vertical at which the preparation was held playing no role. Clearance of such foreign materials on each preparation was found to occur at a constant speed in the range from lop7 to 1 g of particle weights tested. The upper limit represents the biggest mercury drop which it was possible to handle and the lowest the smallest charcoal dust particle which could still be discerned using an operating microscope (magnification up to x 64). This constant speed apparently reflects the rate of travel of the mucus blanket, the movement of which was observed Acta Otolaryngol79
to be triggered by the particles without being affected by their weight. The foreign particles may thus be regarded as passive travellers on top of a biologic conveyer belt. It was also observed that when performing several trials on the same preparation, transport time increased until no clearance of foreign particles was achieved (Fig. 2). Cilia, however, continued to beat for hours and at times days after transport had ceased. The time from the beginning of the experiment until transport standstill differed greatly according to the type of preparation and the season. On preparations where transport of particles had stopped, but cilia continued to beat, samples of mucus (accumulated at the cut edge of the preparation) could be placed on the tissue with the result that the mucus blob started moving immediately upon contact with the tissue. The velocity attained was, in this case, similar to the initial transport velocity of foreign particles, and its velocity did not depend on the load that it carried. When mucus samples taken from heterologous epithelia was placed on a depleted preparation, transport was also restored but the velocity obtained was not necessarily the same as with mucus samples from the preparation itself. For example, midcycle bovine cervical mu-
Downloaded by [McMaster University] at 20:24 09 April 2016
Clearance of middle ear effusions
cus restores transport velocity but the rate of transport attained is only about 30% of that with the frog's own palatal mucus. The fact that solid particles are no longer transported after a certain stage is reached, unless mucus is added, would indicate that the amount of mucus which can be secreted by the isolated system is limited. Mucus must be applied externally in order to enable the cilia to bring about transport. The condition of the epithelium at this stage is thus a state of mucus depletion. A far reaching conclusion about the role of mucus can be drawn from these experiments. Mucus is not only a vehicle for various substances such as immunoglobulins (Gottschalk, 1966), and a moving blanket to which foreign bodies adhere, but it is also an essential mechanical coupler whose presence is imperative for translating ciliary beat into effective clearance. Without mucus, clearance does not take place in spite of a regular ciliary beat. What then is mucus and what is its structure to make it suit these physiological requirements. Mucus associated with ciliated epithelia is a gel, which in contact with physiological saline, does not disperse but remains as a distinct phase. The structural components of mucus must thus be intermolecularly crosslinked or at least very heavily entangled with each other. Studies performed, using, for example, midcycle cervical mucus which can be obtained in large
I
'
I
> 01
'0
10
20
30
40
50
60
Time a f t e r death, minutes
70
48 hrs
Fig. 2. Velocity of transport of foreign particles (curved lower lines) and of added mucus (horizontal) on frog and toad palates as functions of time after death. --, Average velocity in 30 experiments; ---, range of velocities observed.
bv the mucociliavy system 279
Fig. 3. Analyses for the acidic and neutral amino acids (a.a.); the basic amino acids represent less than 7 % of the total amino acids.
quantities, indicate that the structural macromolecule is a glycoprotein of high molecular weight ( > lo6) (Gibbons & Glover, 1959). There are sugar side chains attached via 0-glycosidic links to threosine or serine on the protein backbone. A model consistent with the chemical composition indicates that 1 out of every 4-6 amino acids is involved in linkage to sugar. The average carbohydrate side chain contains some 9 sugar units, 1 of sialic acid, 1 of fucose, 2 galactosamine, 2 of glucosamine and 3 of galactose (Meyer et al., 1973). A comparison of structural molecules from various mucus sources was performed (unpublished results). The study included mucus samples isolated from patients with secretory otitis media and chronic bronchitis as well as from the cow's czrvix at midcycle and the frog's palate. In each case the mucus was brought into solution by use of dithiothreitol-a reagent that breaks disulphide bonds (Vered et al., 1972). Fractionation of the solubilized material by centrifugation in a caesium chloride density gradient resulted in the isolation of a glycoprotein fraction containing the bulk of the sugars of the secretion and a fraction containing mainly serum proteins. In gel electrophoresis and in sedimentation studies, the glycoprotein fraction behaved as a single entity, though the so-called microheterogeneity in these materials is evident. Similarities were noted in the gross chemical composition. For example, the glycoproteins could be banded in a caesium chloride density Acta Otolaryngol 79
280 J. Sad&et al. 1.5
I I I
~.* :
.
I
I
I
i G Gunran concentrotion “%I
Downloaded by [McMaster University] at 20:24 09 April 2016
Fig. 4 . The ratio of the transport rates of guaran gel to frog palate mucus versus guaran concentration. Each point represents 5 to 10 readings of both guaran and mucus velocities on a single depleted palate.
gradient. Analyses for amino sugars, neutral sugars, sialic acid and protein were performed on material representing the fraction where the peak occurred in the gradient. For the three mammalian glycoproteins the percentages expressed relative to the total material analyzed were in the range amino sugars 36-38 %, neutral sugars 30-34 76, sialic acid 15-18 and protein 14-18 %. The analyses for the frog material were in this range except for sialic acid which was significantly lower (7 %). It has been shown that this sugar is not of great significance in the biological transport function of mucus (Meyer et al., 1975). Similarities between the four isolated glycoproteins were also seen in the amino acid composition (Fig. 3). This study indicates that a glycoprotein of chemically similar structure is common to mucus from different sources and that the gel nature of mucus presumably arises as a result of the organization of these chains into large three-dimensional structures and their stabilization by intermolecular crosslinks. This chemical evidence supports the result that from the point of view of transport mucus samples from different sources function exchangeably on epithelia. It is obvious that there must indeed be a rheological requirement for mucociliary transport. A recent study (King et al., 1974) has shown that the requisite rheological properties can be specified and that these are not only not specific to mucus but can be obtained with materials of quite different chemical structure. A number of Acta Otolaryngol79
macromolecular materials e.g., guaran, polyacrylamide, agarose and gelatin which are chemically quite dissimilar to mucus have been studied. I t was found that these materials are capable of performing the transport function on a ciliated epithelium provided that these systems possess a slight degree of crosslinking. It should also be noted that foreign systems displaying transport activity generally did so at different transport rates. Indeed it would seem that a qualitative mucus factor is important in influencing the observed rate. While differences in transport rate were observed for foreign systems, differences are also seen for mucus samples from different sources. For example, midcycle cow cervical mucus is transported at only about 30% of the rate of frog’s own palatal mucus. This occurs in spite of the fact that both are based on similar glycoprotein constituents. Thus differences in the physical organization of the structural macromolecules alter the rheological character of the system and determine its effectiveness as mechanical coupler. Particularly by using the same chemically defined system; guaran at various degrees of crosslinking we (King et al., 1974) found that below the gel point (0.5% guaran concentration) no transport is observed (Fig. 4). In the gelation region transport rate rises dramatically to the value obtained with the frog’s own mucus and this rate is maintained in the concentration range 0.5% to 0.9%. As the system becomes more crosslinked and solid-like transport rate falls off and at concentrations greater than 1.8 % the gel behaves like a foreign particle on a depleted mucociliary epithelium and is not transported. Thus the degree of crosslinking and the physical organization of the gel are highly important parameters. It appears that the rheological requirements for mucociliary transport are satisfied only by macromolecular systems which are very weak gels. These observations explain why in earlier experiments (Sad6 et al., 1970), where solutions of macromolecules were used to substitute for mucus, no success was obtained. In fact it is likely that when macromolecular solutions are
Downloaded by [McMaster University] at 20:24 09 April 2016
Clearance of middle ear effusions by the mucociliary system 281 applied to the epithelium, the macromolecules penetrate into the serous layer between cilia and interfere with the ciliary beat. In the physiopathology of hearing loss, middle ear effusions are of prime importance (Sad&, 1974) and are essentially of two extreme kinds: (A) very thick rubber-like effusions, as seen mostly in children and termed “glue ears” secretory otitis media; and (B) water-like effusions, as seen mostly in adults and often termed serous otitis media. The difference in consistency between the two secretions is very likely the result of a difference in crosslinking (Vered et al., 1972) the water-like effusion is probably not a gel. On the basis of the evidence presented, these two types of secretions may in fact represent two extremes in the crosslinking of the glycoprotein structural macromolecules. In “glue ears” the mucus is solid-like and presumably heavily crosslinked and in serous otitis, the glycoproteins are not crosslinked at all. In analogy with the results in Fig. 3, the transport rate may be zero, the two systems being to either side of the optimum represented presumably by the normal secretion. In either of these cases the epithelia would be secreting a mucus which could not be transported away fast enough resulting in a net accumulation of secretions in the middle ear which eventually impairs hearing. In the normal state the degree of crosslinking of the glycoprotein in mucus from different sources is probably matched to the required transport rate for the individual tissue. In pathological states, such as those of middle ear effusions, crosslinking would seem to be altered. The reason for differences in the degree of crosslinking in such glycoprotein-based mucus systems is not clear. Minor changes in p H and ionic strength of the media have little influence. Earlier studies attributed the marked differences in elasticity, a measure of the degree of crosslinking to the sialic acid content of mucus. For example, a difference in sialic acid content has been noted for the glycoprotein isolated from the watery midcycle cervical mucus and that from the more elastic secretion found at other stages of the cycle (Gibbons & Glover, 1959;
Hatcher & Jeanloz, 1973). However, recently it has been shown that removal of sialic acid with a specific enzyme does not alter the elasticity of mucus (Meyer et al., 1975) nor its biological transportability (King et al., 1974). Differences in crosslinking though may arise as a result of different concentrations of glycoprotein component in the mucus and to the passage of time. For example, it has been found that mucus on concentrating does not swell back to its initial volume on standing in buffer. This would suggest that additional crosslinks are formed and indeed the material is found to be more rubberlike. On the other hand, mucus goes spontaneously into solution on standing. The temperature dependence of this process suggests that a relatively low energy of activation is involved indicating disruptions of secondary bonds or complex entanglements (unpublished results).
ZUSAMMENFASSUNG Es gibt zwei extreme Falle von Mittelohr-Ausfluss, die zu einem Gehorverlust fiihren: a) ein gummi-ahnlicher Ausfluss, der bei exsudativer Mittelohrentzundung vorkommt und b) ein wasserahnlicher Ausfluss, der bei seroser Mittelohrentzundung vorkommt. Es wird die Moglichkeit besprochen, dass der Vernetzungsgrad in diesen zwei extremen Fallen die Grundlage der verinderten Transportgeschwindigkeit des Schleimes ist, wodurch es zu einer Anreicherung der Sekrete und demzufolge zur Verminderung des Gehores kommt. Es wurde gezeigt (King et al., 1974), dass diese notwendige rheologische Eigenschaft fur eine Transportaktivitat nicht auf strukturierte Makromolekule im Schleim begrenzt ist, sondern auch bei anderen Polymer-Systemen gefunden wird, die locker vernetzt sind, wie z. B.: Guaran, Polyacrylamid, Gelatine und Agarose. Untersuchungen an einem dieser Systeme, Guaran, zeigen, dass die Transportgeschwindigkeit von dem Vercetzungsgrad abhangt mit einem Maximum nahe dem Gelierungspunkt; das heisst in einem Bereich mit sehr wenigen Vernetzungen pro Makromolekul. Die Beobachtung, dass der Schleim von verschiedenen Flimmerschleimhautepithelien eine chemisch ahnliche Glukoproteinstruktur hatte, spricht dafur, dass die beobachteten Unterschiede der Transportgeschwindigkeit zwischen verschiedenen Schleimproben mehr auf Unterschieden im Vernetzungsgrad beruhen, als auf Veranderungen der Glycoprotein-Einheiten. In Analogie ZLI dem Guaran-Model1 wird angenommen, dass die zwei Typen von Mittelohr-Ausfluss Extreme der Vernetzung darstellen und dass deren Transportgeschwindigkeiten zu beiden Seiten eines Optimums liegen, das bei normaler Acta Otolaryngol 79
282 J. Sade‘ et al. Sekretion vorkommt. Faktoren der Schleimvernetzung, wie Ladung, Konzentration und der Einfluss der Temperatur werden diskutiert.
Downloaded by [McMaster University] at 20:24 09 April 2016
REFERENCES Compere, W. E., Jr. 1958. Tympanic cavity clearance studies trans. A A O O 62, 444. Gibbons, R. A. & Glover, F. A. 1959. Physicochemical properties of two mucoids from bovine cervical mucin. Biochem J 73, 217. Gottschalk, A. 1966. Glycoproteins. Elsevier Publ. Co., Amsterdam and London. Gray, J. 1928. Ciliary mouement. Cambridge Univ. Press. Hatcher, V. B. & Jeanloz, R. W. 1973. The use of glycosidases in the structure elucidation of the carbohydrate chains of a,-acid glycoprotein from human plasma and cervical mucin from Macaca Radiata. Colloque International du C.N.R.S. sur les glycoconjugues,Lille. Hilding, A. C. 1957. Ciliary streaming in the bronchial tree element in carcinogenesis. New Engl J Med 256, 634. King, M., Gilboa, A,, Meyer, F. A. & Silberberg, A. 1974. On the transport of mucus and its rheological simulants in ciliated systems. Am Rev Resp Dis (accepted for publication). Lim, D. J. & Hassel, B. 1969. Human middle ear epithelium. Arch Otolaryngol89, 835. Meyer, F. A., Eliezer, N., Silberberg, A., Vered, J., Sharon, N. & Sade, J. 1973. An approach to the biochemical basis for the transport function of epithelial mucus. Bull Physiopathol Resp Nancy 9, 259. Meyer, F. A,, King, M. & Gelman, R. A. 1975. On the role of sialic acid in the rheological properties of mucus. Biochim Biophys Acta (submitted for publication). Sadt, J. 1966. Middle ear mucosa. Arch Otolaryngol 84, 137. - 1967. Ciliary activity and middle ear clearance. Arch Otolaryngol86, 128. - 1974. The biopathology of secretory otitis media. Ann Otol Rhinol Laryngol83, 59. Sade, J., Eliezer, N., Silberberg, A. & Nevo, A. 1970.
Acta Otolaryngol 79
The role of mucus in transport by cilia. Am Rev Resp Dis 102, 48. Vered, J., Eliezer, N. & Sade, J. 1972. Biochemical characterization of middle ear effusions. Ann Otol Rhinol Laryngol81, 394.
J. Sade, M.D. Dept. of Otolaryngology Meir Hospital Kfar Saba Israel
DISCUSSION D. Hilding: Did you have a chance to test the effect of the mucoid material from a “glue ear” on the cilia of a cat trachea or frog palate? B. McCabe: It is clear that this disease is not a pure Eustachian tube disease, but we have all witnessed the phenomenal resolution of symptoms upon simply ventilating the ear. I wonder what role you feel that the Eustachian tube does play in this disease. G. Zechner: Mucus in the middle ear comes from goblet cells, ciliated cells and normal epithelial cells under certain conditions. In addition, some mucus comes from the Eustachian tube. For ciliary action, I want to give credit to Messerklinger’s findings: the ciliary organ needs for active transport a gel formation layer within which to beat. This gel, which we call mucus, is influenced by pH, ionic concentration and other facts. J. Sad6 (Reply) to Mr Hilding: You are of course right and this should be done, we have not done it so far because of technical difficulties in transporting the whole lab nearer to the operating theatre and we did not want to carry the mucus to the Weizmann Institute 30 miles away. To Mr McCabe: The Eustachian tube is obviously not functioning as it should in secretory otitis media, i s . it probably does not pump air in sufficient amounts into the middle ear but this does not mean that it is mechanically obstructed. To Mr Zechner: The p H and ionic concentration were indeed measured and their range is rather wide before cilia cease beating.
