State of Art

State of Art The Propulsion of Mucus by Cilia 1- 3

MICHAEL A. SLEIGH, JOHN R. BLAKE, and NADAV LlRON

Contents Introductory Background Morphology of the Lining of the Respiratory Tract Mechanics of Propulsion of Liquids by Cilia Movement and Coordination of Cilia of the Respiratory Tract Microscopic Observations and Measurements of Mucus Transport Control of Mucociliary Transport The Composition and Rheologic Properties of Mucus Theoretical Models of Mucociliary Transport Airflow Interactions Overview of Mucociliary Transport and Unsolved Problems

SUMMARY The presence of cilia on epithelia of the respiratory tract was reported more than 150 yr ago, and the two-layer model of mucus transport was put forward more than 50 yr ago. However, it is only in the last 10yr or so that the motion of mucus-propelling cilia of the mammalian respiratory system has been adequately described, and fluid dynamic studies have developed far enough to allow descriptions of the mechanisms by which ciliary movement is coupled to mucus transport. In this review, scientific developments on the study of cilia and mucus, and Interactions between them, are drawn together to further understanding of mucociliary clearance mechanisms of the respiratory tract. The study of the cilia incorporates a discussion of the internal mechanics and biochemistry of the ciliary axoneme, the physical principles of the beat pattern, and the (weak) metachronal coordination of cilia in the lung. Mucus rheology plays a central role in mucociliary transport with the rheologic properties of the mucus determining the effective functioning of this clearance mechanism. Theoretical models provide information on the mechanical principles of the beat pattern as well as providing reliable estimates of the transport rates. Although airflow is not thought to contribute to mucus transport in the normal state, high frequency ventilation and coughAM REV RESPIR DIS 1988; 137:726-741 ing may make significant contributions.

Introductory Background

T here has been a growing appreciation

tive research, notably by such workers as over the last century and a half of the Hilding, Lucas, and Proetz, who studsignificant role played by mucociliary ied ciliary currents and ciliary function clearance in the efficient functioning of in the upper respiratory tract. As well as the respiratory system by the removal of providing detailed confirmation of the foreign substances and agents of disease. role of cilia in nasal clearance, this work Research over the last 10yr has been par- led to the important conclusion that muticularly fruitful in elucidating details of cus is propelled by the tips of the cilia, the mechanism of mucociliary clearance which themselves move in a low viscosity that give new insight into various clini- layer beneath the mucus (4). Advances in many areas of scienceconcal problems and that have much intrintributed to the development of undersic scientific interest. The first comprehensive account of standing of the 2 main components of cilia in the English language appears to this system, the cilia and the mucus. Some be that of Sharpey in 1835 (1), who not significant landmarks in the study of cilia only gave detailed descriptions of the ac- depended upon the introduction of elections of cilia in a wide variety of animals tron microscopy, which was used to dembut also reported the discovery of ciliary onstrate the 9 plus 2 arrangement of inmotion in the reproductive and respira- ternal fibrils in sections of cilia (5), to tory systems of mammals in the previ- discover subsidiary components of the ous year by Purkinje and Valentin (2). axoneme (6), to support a sliding-fibril Sharpey himself confirmed their obser- hypothesis of ciliary bending (7), and, vations, commenting that ciliary motion combined with improved biochemical seems to convey the secretions along the techniques, was used by Gibbons (8) to lining membranes of the air passages of show the localization of ATPase activity both the nose and the trachea. Observa- in dynein arms. When research on these tions on ciliary motion made by many aspects of structure and function and in workers over the next 90 yr are summa- areas concerned with movement, hydrorized in Gray's classic monograph (3), dynamics, coordination, and control was published in 1928. Although this in- reviewed in a volume edited by Sleigh in cluded little information on human re- 1974(9), there was still little specific menspiratory cilia, it led into a period of ac- tion of respiratory cilia. Since that time, 726

the understanding gained from studies of other cilia has been applied to mammalian respiratory cilia, which are more difficult to study, so that now we have considerable knowledge of the pattern of movement of these cilia and of their coordination, control, and hydrodynamic relations with mucus, although some major areas of uncertainty remain. Physical and chemical methods have been combined in studies on mucus, but such studies started later and present many difficulties associated with the nature and variability of the material. Important conceptual and methodologic advances in the study of mucus, associated particularly with the names of Litt and Silberberg in the account that follows, have substantially helped the development of understanding in this area. Weaim in this

1 From the Department of Biology, University of Southampton, Southampton, England; the Department of Mathematics, University of WolIongong, Wollongong, Australia; and the Department of Mathematics, Israel Institute of Technology, The Technion, Haifa, Israel. 1 Supported by Grant No. F85/5994 from the Australian Research Grants Scheme. a Requests for reprints should be addressed to Professor J. R. Blake,Department of Mathematics, University of Wollongong, Wollongong, N.S.w. 2500, Australia.

727

STATE OF ART: THE PROPULSION OF MUCUS BY CILIA

review to draw together those developments in the study of cilia and mucus and the interactions between them that help in the understanding of mucociliary clearance mechanisms of the respiratory tract. Morphology of the Lining of the Respiratory Tract All surfaces of the upper airways are covered by ciliated epithelium, except for the nasal entrance and those parts of the nasopharynx, pharynx, and larynx that are covered by squamous epithelium, and the olfactory area, which has a specialized sensory epithelium; the tracheobronchial tree and pulmonary compartment are ciliated down to the nonalveolar walls of the respiratory bronchioles. The lining epithelium of the respiratory tract is thickest in its upper parts where it is underlain by a thick submucosal zone; these become thinner in the finer airways where the submucosa is lost and eventually the epithelium is made up of the flattened cells lining the alveoli, some of which are thought to be concerned with secretion of the alveolar surfactant fluid (10). Submucosal glands occur in the upper airways and in the airways of the tracheobronchial tree down to the smallest bronchi whose walls contain cartilage, and goblet cells are present in the epithelia of all ciliated regions of human airways, but they become progressively fewer in the finer bronchioles. The epithelia of ciliated regions are known to be complex, at least in part, for in addition to the easily distinguishable columnar ciliated cells, there are 7 other types of epithelial cells, as described by Jeffrey and Reid (11),5 of which have surfaces that line the airways. Three of these are glandular cells, the well-known goblet cells, serous cells, and Clara cells; the latter 2, which appear to be confined to smaller airways, contain small secretory granules and probably produce watery secretions, but either type may transform to goblet cells. Another type of superficial cell is the brush cell, with a border of microvilli as long as 2 J.1m and possibly an absorptive function. Undifferentiated intermediate cells may lie between ciliated, brush, and glandular cells. The submucosal glands are branching tubular structures, their distal ends lined by serous cells whose secretion passes over the more proximal parts lined by mucous cells to enter the main ducts whose proximal section is ciliated. Secretory granules of moderate electron density and as much as 1 to 2 J.1m across oc-

cur near the cell apex in both the goblet cells of the epithelium and the mucous cells of the glands; these cells contain acid glycoprotein, often of several types in mucous cells. Serous cells of the glands contain secretory carbohydrate, particularly sialic acid, and in these cells there are smaller, denser secretory granules with more persistent membranes. In humans the volume of submucosal glands is about 40 times that of goblet cells, but the lesser amounts of go blet secretion in peripheral airways may be of more physiologic importance than the larger glands (12). In the large airways, the apices of ciliated cells may form a more or less continuous cover at the surface of the epithelium. The tracheal epithelium may have only 1 goblet cell for every 5 ciliated cells, and even fewer brush cells, but in the smaller airways the proportion of brush cells increases, and there may be numerous intermediate cells. The openings of the submucosal glands (50 J.1m or so across) and of airway branches also

a

m

~ II

--.-;::;.

-:=-

~

~.~-

Fig. 1. The structure of a cilium from the respiratory tract. In the longitudinal section (a), the ciliary shaft is surrounded by a membrane continuous with that around the cell (m) and terminates in a crown of "claws" attached to a dense cap at the tips of the longitudinal rnlcrotubuies of the axoneme; these microtubules continue into the basal body, which lies in the cell cortex and bears rootlet fibers. Details of structures seen in cross section are seen at (e), including the arrangement and conventional numbering of the 9 doublets (as seen looking from the base towards the tip). The doublets are linked by strands of nexin (n), and the A subfibers of each doublet carry outer (0) and inner (i) dynein arms, projecting towards the next doublet, and radial spokes (r) with dilated heads (h), which can attach to projections (p) associated with the central microtubules. Changes in microtubule pattern at different levels are shown in b, c, and d. Modified from Sleigh (15).

interrupt the ciliated surface. Each ciliated cell has a diameter of 5 J.1m or so at its apex and carries some 200 cilia at a density of 6 to 8 J.1m- 2 , interspersed with short microvilli (13, 14). These cilia are about 6 J.1m long in the larger airways reducing to 5 J.1m or a little less in the smaller bronchioles (15). They have the normal 9 plus 2 axonemal structure (figure 1) (16), but are unusual in the possession of a crown of 3 to 7 short 'claws' 25 to 35 nm long, projecting from a dense cap at their tips (11, 17). The ciliary basal body is of a common type with a basal foot, short striated rootlets, and attached cytoplasmic microtubules, which together provide anchorage (18). The basal foot tends to be at the side of the cilium towards which the effective stroke occurs (19), and since all of the basal feet on a single cell are normally aligned in approximately the same direction (20), the effective strokes of all cilia on the cell have a common orientation. However, neither the orientation of the basal feet (20) nor the orientation of the beat (21, 22) are precisely identical on adjacent cells. A thin layer of secreted mucus overlying these epithelia traps particulate matter and absorbs chemicals from the air; ciliary action clears the mucus with these bound materials from the ciliated surfaces. The bulk of the inhaled particulate matter is deposited on the nasal mucosa and soluble gases dissolve there (23-26); less soluble gases and many of the particles in the range of 0.1 to 1.0 J.1m penetrate deep into the lower airways before deposition (25). Particles trapped on the mucous surface are carried by mucociliary transport to the pharynx; particles that reach the surfactant lining of the alveoli may be drawn onto the mucociliary escalator or may be phagocytosed by macrophages, which themselves are probably removed by mucociliary transport (27). Mechanics of Propulsion of Liquids by Cilia Cilia propel fluids because the cyclical movements that they perform are asymmetric. There are two active parts in the ciliary beat cycle: in the effective (or power) stroke of the cycle the cilium remains fully extended and moves through an arc in a plane approximately perpendicular to the cell surface, whereas in the recovery (or preparatory) stroke, a bend is propagated along the length of the cilium from base to tip, and the cilium swings around near the cell surface to

728

SLEIGH, BLAKE, AND L1RON

a , if

C

¥·c~

\:-....

•

.~

~""'" .'

•.•.. !'

.

Fig. 2. The beat cycle of a typical water-propelling cilium consists of 2 parts, an effective stroke (a) and a recovery stroke (b), both seen from the side, with profiles at equal time intervals; the dotted line shows the path taken by the ciliary tip. When the cilium is seen from above (c), the sideways swing of the shaft during the recovery stroke is evident.

reach the starting position for the next effective stroke (figure 2). Some types of cilia show a rest period in every cycle, and stop moving briefly before the effective stroke or before the recovery stroke. Bending movements are produced as the outer 9 microtubule doublets of the ciliary axoneme activelyslide against one another when propelled by molecular bridges of dynein that project from one doublet towards the next. Active sliding of the doublets at one side of the axoneme bends the cilium in one direction and active sliding of doublets at the other side bends it back again. Dynein is an ATPase protein that uses energy from ATP in performing cyclical shape changes that produce the active sliding movements. The machinery of motility is spread along the whole length of the cilium, and different patterns of sliding along the length of the nine doublets are responsible for the differences in shape of the cilium in the effective and recovery strokes. Radial connections between axonemal fibrils are assumed to resist the sliding and contribute to the formation of bends. A detailed description of the mechanism of motility is beyond the scope of this review, but a recent review of present understanding has been given by Gibbons (28). To understand how the bending cycle

of a cilium results in a unidirectional flow of water, it is necessary to appreciate the dominance of viscosity in this situation (29). An indication of the physical characteristics of a particular example of fluid motion is given by a nondimensional index used by fluid dynamicists called the Reynolds number: R = p£u/~, where £ is a linear dimension of the structure moving in fluid at speed u, and p and u are, respectively, the density and viscosity of the fluid. A large structure moving quickly through a fluid of relatively low viscosity, e.g., a ship or large fish in water, will have a high Reynolds number where inertial forces dominate its motion. A microscopically small structure moving slowlythrough a medium of relatively high viscosity, e.g., a cilium in mucus or bacterium in water, will have a low Reynolds number and its motion will be dominated by viscous forces. Typically, the Reynolds number for a cilium in the respiratory tract is of the order of 10-3 or smaller. In discussing the fluid mechanical principles of mucociliary transport, it should be impressed that there are severallength scales of particular relevance to this study. They are (1) molecular length scales relative to the biochemical structure of the mucus; (2) length scales associated with the cilium tip (0.1 to 1 urn); (3) a length associated with the cilium length, cell size, ciliary wavelength, and coherence of a mucous plaque (5 to 50 urn); (4) a length scale associated with the length of an airway (5 to 10mm). Of

particular importance with regard to cilium-mucus interaction is the scale of the cilium tip with respect to the entangled network of molecules constituting the mucus. Does the cilium "see" a continuous viscous liquid or a loosely coiled network of molecules? A qualitative understanding of the ciliary beat cycleis now well known. The difference in motion of the cilium between the effective and recovery strokes can be wellunderstood if we consider the motion of a needlelike body in a viscous liquid. The force acting on 2 identical needlelike bodies, one moving parallel to the axis of symmetry, the other perpendicular to the axis, is that the force acting on the needlelike body moving in the perpendicular direction is almost twice that of the one moving in the axial direction. Thus, it would be highly desirable for the cilium to have a "perpendicular mode" of motion in its effective stroke (thus generating a greater force) as against a "tangential mode" of motion during the recovery stroke. In addition, the force is linearly dependent on the velocity; thus, a larger velocity implies a larger force, which is clearly evident in the ciliary beat pattern with the fast effective stroke. In addition, there is the very important role of the epithelial surface. When a body moves in a liquid close to a rigid boundary, 2 discernible regions of flow exist: one close to the body where the motion of the liquid is strongly influenced by the motion of the body, and the other

(b)

(a)

Fig. 3. The "volume of influence" during the (a) effective stroke, (b) recovery stroke, and (c) the combined effect of all the cilia in a metachronal wave. In (d), the effect of a "solidlike" mucous layer is incorporated.

(e)

(d)

STATE OF ART: THE PROPULSION OF MUCUS BY CILIA

729

region exterior to this where the retarding motion of the wall dominates. Furthermore, the size of this "volume of influence" depends on the height of the body above the wall; the farther away, the greater the volume of influence. This is clearly emphasized in figure 3 where a much larger volume of liquid is under the direct influence of the cilium during the effective stroke (figure 3a) than in the recovery stroke (figure 3b). The difference in extent of these entrained layers accounts for the net transport of water by a ciliary beat cycle. Cilia are coordinated into waves (see below), and when the effects of all the cilia in a wavelength are combined together, a fairly constant flow in the upper sections of the ciliary sublayer is obtained because this volume is only being acted upon by cilia in their effectivestroke, whereas fluid in the lower part of the ciliary sublayer is under the 4. Scanning electron micrograph of rabbit tracheal epithelium fixed in 2% OsO..and postfixed in 4% glutaraldeinfluence of all of the cilia in the wave- Fig. hyde. The 2 active patches contain cilia that would propel mucus towards the right, and cilia between these active length, including cilia in their recovery patches lie at rest with their tips directed towards the right. Micrograph by M.J. Sanderson, from reference 22. strokes that are densely pushed together, so that a predominantly oscillatory flow can be expected in this region (figure 3c). ing almost parallel to one another as a the activity of the cilia are dependent The above is clearly evident in fast swim- result of the viscous forcesacting between upon their state of activation. If ciliary ming ciliates with relatively long cilia (> the shafts of the moving cilia. Accord- excitabilityis depressed, the cilia willbeat 9 11m) that can develop a highly asym- ing to the positional relationship between slowly, the interaction between them may them relative to their plane of beating, be weakened, and the metachronal waves metric beat pattern. However, for the shorter cilia in mu- they may end up beating in phase with may move more slowly,whereas if ciliary cus-propelling environments where there one another, or with a constant phase excitation is increased they may beat is not such a clear distinction between difference. When many cilia interfere more quickly with faster metachronal effective and recovery strokes, further with their neighbors in such a way, their waves. Control of ciliary activity is thus mechanisms must be employed to pro- beating will become organized into coor- exerted directly on the ciliary mechanism, duce a continuous steady motion instead dinated metachronal waves. Changes in often by some conducted "impulse," but of the oscillatory motion that would re- the viscosity of the medium or in the effects of this control upon the coordisult. Clearly, an elastic, almost rigid, length or spacing of the cilia will have nation only occur as a consequence of structure on the macro scale, but a more a profound influence on the characteris- changes in the beating activity. liquidlike response on the microscale of tics of the metachronal waves. the cilium tip, would be the most desirSuch coordination does not depend Movement and Coordination of Cilia able attributes for the surface layer of upon conduction of any "impulse" at the of the Respiratory Tract mucus. Viscoelastic properties such as cell surface, since it is a process depenoccur in coiled macromolecular networks dent upon forces in the fluid between the The cilia of respiratory epithelia, in comprovide the consistency and integrity that moving cilia. However, characteristics of mon with other mucus-propelling cilia, is required for transport in mucociliary systems. Clearly then, the rheologic properties of the liquids in the lung are censide tral to any scientific discussion of muco, , ciliary transport. An illustration of these -' . ... , Fig. 5. The beat cycle of a rabbit tra• ". I ideas can be found in figure 3d. ~ I cheal cilium seen from the side and from The presence of an entrained layer of above. In the recovery stroke (left), the r fluid around a cilium has implications cilium starts from the rest position (r) and for the coordination of the rhythm of unrolls clockwise (in top view)to the left, beat of adjacent cilia (30). If 2 cilia lie and in the effective stroke (right), it remains extended and bends over to reach close enough together for the zones of the rest position at the right. Mucus is top entrained fluid around them to overlap propelled towards the right (mp), and the