Airway plumbing Commentary
See related article,
John W. Hanrahan pages 1419–1427.
Department of Physiology and Meakins-Christie Laboratories, McGill University, 3655 Drummond Street, Montréal,
Québec, Canada H3G 1Y6. Phone: (514) 398-8320; Fax: (514) 398-7452; E-mail: firstname.lastname@example.org.
Airway surface liquid (ASL), the thin The variable results obtained in vivo salt gradient than the one reported for
film of fluid covering the luminal aspect may be due to mechanical stimulation airway epithelia (12). Regardless, demon-
of airway epithelial cells, plays a central during sampling, since this could strating that transepithelial Pf can be
but mysterious role in cystic fibrosis induce secretion by submucosal glands studied conveniently in cultured airway
(CF). CF is characterized by abnormal and change the local ASL volume and epithelia is important and provides hope
transepithelial salt transport, viscous composition just as one tries to collect that conflicting in vitro results will be
airway mucus, chronic bacterial infec- it. Filter paper, which has been widely reconciled. Water and ion permeabilities
tions, and inflammation, but precisely used to collect ASL, exerts a capillary need to be compared among different in
how mutations in the cystic fibrosis pressure of about 9 cm H2O (4), which vitro preparations and also with those of
transmembrane conductance regulator may pull macromolecules and isotonic native tissue to allow a full understand-
(CFTR) gene lead to these symptoms is fluid through the epithelium (5). Most ing of the regulation of ASL salinity and
the subject of a vigorous debate. As dis- studies indicate that the salt concen- volume. It may be significant that cul-
cussed one year ago in the Perspective by tration of ASL is somewhat lower than tures producing hypotonic surface liquid
J.J. Wine (1), there are two very different that of plasma, but there is not yet a (3) had higher electrical resistance com-
hypotheses to explain CF pathogenesis consensus. For technical reasons, most pared with those from which isotonic
in the airways. According to the “vol- detailed studies of ASL composition fluid was collected (2).
ume” model, lack of CFTR in the apical and volume will probably be performed The results of Matsui et al. (6) raise
membrane leads to increased salt and in vitro, at least in the immediate interesting questions regarding the
fluid absorption by airway epithelial future, which makes validation of cell molecular basis of Pf in the bronchial
cells, which reduces ASL volume and culture models crucial. epithelium. In the rat, the bronchial sur-
leads to ineffective mucociliary clearance So far, most attention has been focused face epithelium expresses aquaporin-1
of bacteria. According to this hypothesis, on salt transport in these models, rather (AQP1) in both apical and basolateral
sodium transport is increased in CF than apical membrane and transepithe- membranes (13). The basolateral mem-
through upregulation of sodium chan- lial water permeabilities or evaporative brane also contains the mercurial-insen-
nels that would normally be inhibited by loss. The novel confocal microscopy sitive water channel AQP4 (14). If AQP4 is
CFTR (2). The other hypothesis, which experiments reported by Matsui et al. in the dominant pathway at the basolateral
we shall refer to here as the “salt” model, this issue of the JCI (6) show that cul- membrane, it may constitute a rate-limit-
proposes that a lack of functional CFTR tured human bronchial epithelial cells ing step for transcellular water flow, since
chloride channels leads to decreased salt have moderately high osmotic water per- Matsui et al. find that basolateral mem-
absorption, which elevates salt concen- meability (Pf). The value for transepithe- brane Pf is relatively low (6). This could
tration in the ASL, thereby inhibiting lial Pf is consistent with previous reports explain the weak sensitivity to mercurials
activity of antibacterial substances (3). for freshly excised bovine (7) and porcine of transepithelial Pf in cultured bronchial
Both hypotheses are strongly sup- (8) trachea and guinea pig distal airways cells and freshly isolated hamster airways
ported by in vitro experiments, despite (9). The apical membrane Pf of cells (9). On the other hand, the relatively large,
making different predictions concern- grown at the air-liquid interface are mercurial-insensitive component of Pf at
ing the volume and composition of slightly higher than those obtained pre- the apical membrane is surprising and
ASL. Fluid collected from the surface viously for normal and CF tracheal hints that AQP4, AQP5, or perhaps a
of CF airway cell cultures by one group epithelia cultured under conventional, novel apical pathway may exist there in
had elevated salt and diminished abili- submerged conditions (10). The moder- parallel with AQP1.
ty to kill bacteria (3), whereas liquid ately high Pf measured by Matsui et al. (6) High apical membrane water perme-
collected from cultures by the other favors the volume hypothesis since high ability would enable water to leave the
group was isotonic (2). Different Pf, which is characteristic of epithelia cells when the tonicity of ASL increases
methodologies were used for sampling such as renal proximal tubule and gall through evaporation (15). Thus, hyper-
and analysis, but the contrasting bladder that carry out isosmotic water ventilation or breathing dry air would
results may simply reflect phenotypic transport (11), would collapse transep- elevate ASL osmolality and reduce
differences between cell cultures pre- ithelial solute gradients. However, to dis- epithelial cell volume by drawing water
pared in different laboratories. One prove the salt model one needs to show from the cells faster than it enters
would like to use the composition of that airway Pf is too high to be compati- through the basolateral membrane.
ASL in vivo as the gold standard, but ble with the salt hypothesis, and this dis- Matsui et al. make the intriguing sug-
problems associated with sampling a criminating value is not known. Presum- gestion that cell shrinkage could play a
compartment that is only tens of ably Pf could be somewhat higher than in sensory role in signaling hypertonicity
microns thick are even worse in vivo, the renal collecting duct, which, during of the ASL to the microvasculature
and those data are also contradictory. water diuresis, maintains a much larger below (6). Indeed, it was shown years
The Journal of Clinical Investigation | May 2000 | Volume 105 | Number 10 1343
Proposed mechanism for feedback regulation of ASL volume. (a) Evaporation reduces ASL volume, reduces the efficiency of mucociliary clear-
ance, and increases ASL osmolality. (b) High osmotic permeability of the apical membrane allows water to be drawn into the ASL to restore osmo-
lality and causes the cells to shrink. Shrinkage of surface columnar cells triggers release of vasodilator substances such as nitric oxide (NO), which
increase fluid delivery by subepithelial microvasculature.
ago that blood flow to the airway sub- rehydration of the ASL. ty in this region of the airways. The
mucosa is stimulated by breathing dry Understanding the role of ASL in absence of submucosal glands and
air (16) and by hypertonic luminal fluid normal and CF airways will require perhaps other important features of
(17, 18). Moreover, several studies (19, detailed (and preferably simultaneous) the native epithelium in these culture
20) have shown that airway epithelia studies of solute and water flows under systems remains an important limita-
can release smooth muscle–relaxing well controlled conditions, which tion of such studies, and improved
factors such as nitric oxide (21), and could be varied to probe the effects of methods are still needed for sampling
perhaps other vasoactive agents when hormones and inflammatory media- ASL without affecting the volume or
their luminal surfaces are exposed to tors from the CF lung on Pf and AQP ionic composition of this fluid. Never-
hypertonic solution (Figure 1). expression. Comparing cultures pre- theless, the approach described by
Although still speculative, this intrigu- pared from normal and aquaporin Matsui et al. (6) should provide a
ing feedback loop makes sense physio- knockout mice might provide insight wealth of new and practical informa-
logically, since increasing epithelial per- into the relative contributions of tion on epithelial hydrokinetics rele-
fusion would allow more efficient AQP1, 4, and 5 to osmotic permeabili- vant to cystic fibrosis.
1. Wine, J.J. 1999. The genesis of cystic fibrosis lung lium. Am. J. Respir. Crit. Care Med. 157:A556. (Abstr.) Gatzy, J.T. 1981. Regional differences in airway sur-
disease. J. Clin. Invest. 103:309–312. 9. Folkesson, H.G., Matthay, M.A., Frigeri, A., and face liquid composition. J. Appl. Physiol. 50:613–620.
2. Matsui, H., et al. 1998. Evidence for periciliary liq- Verkman, A.S. 1996. Transepithelial water perme- 16. Baile, E.M., Guillemi, S., and Paré, P.D. 1987. Tra-
uid layer depletion, not abnormal ion composi- ability in microperfused distal airways. J. Clin. cheobronchial and upper airway blood flow in
tion, in the pathogenesis of cystic fibrosis airways Invest. 97:664–671. dogs during thermally induced panting. J. Appl.
disease. Cell. 95:1005–1015. 10. Farinas, J., Kneen, M., Moore, M., and Verkman, Physiol. 63:2240–2246.
3. Smith, J.J., Travis, S.M., Greenberg, E.P., and Welsh, A.S. 1997. Plasma membrane water permeability of 17. Deffenbach, M.E., Salonen, R.O., Webber, S.E., and
M.J. 1996. Cystic fibrosis airway epithelia fail to kill cultured cells and epithelia measured by light Widdicombe, J.G. 1989. Cold and hyperosmolar
bacteria because of abnormal airway surface fluid. microscopy with spatial filtering. J. Gen. Physiol. fluids in canine trachea: vascular and smooth mus-
Cell. 85:229–236. 110:283–296. cle tone and albumin flux. J. Appl. Physiol.
4. Landry, J.S., Cowley, E.A., and Eidelman, D.H. 11. Spring, K.R. 1998. Routes and mechanism of fluid 66:1309–1315.
1999. Techniques for ASL collection: role of capil- transport by epithelia. Annu. Rev. Physiol. 18. Wells, U.M., Hanafi, Z., and Widdicombe, J.G.
larity. Am. J. Respir. Crit. Care Med. 159:A678. 60:105–119. 1994. Osmolality alters tracheal blood flow and
(Abstr.) 12. Zabner, J., Smith, J.J., Karp, P.H., Widdicomb, J.H., tracer uptake in anesthetized sheep. J. Appl. Physiol.
5. Erjefält, I., and Persson, C.G.A. 1990. On the use of and Welsh, M.J. 1998. Loss of CFTR chloride chan- 77:2400–2407.
absorbing discs to sample mucosal surface liquids. nels alters salt absorption by cystic fibrosis airway 19. Hay, D.W.P., Muccitelli, R.M., Horstmeyer, D.L.,
Clin. Exp. Allergy. 20:193–197. epithelia in vitro. Mol. Cell. 2:397–403. Wilson, K.A., and Raeburn, D. 1987. Demonstra-
6. Matsui, H., Davis, C.W., Tarran, R., and Boucher, 13. Nielsen, S., King, L.S., Christensen, B., and Agre, P. tion of the release of an epithelium-derived
R.C. 2000. Osmotic water permeabilities of cul- 1997. Aquaporins in complex tissues. II. Subcellu- inhibitory factor from a novel preparation of
tered, well-differentiated normal and cystic fibro- lar distribution in respiratory and glandular tis- guinea-pig trachea. Eur. J. Pharmacol. 136:247–259.
sis airway epithelia. J. Clin. Invest. 105:1419–1427. sues of rat. Am. J. Physiol. 273:C1549–C1561. 20. Munakata, M., Mitzner, W., and Menkes, H. 1988.
7. Durand, J., Durand-Arczynska, W., and Haab, P. 14. Frigeri, A., Gropper, M.A., Turck, C.W., and Osmotic stimuli induce epithelial-dependent
1981. Volume flow, hydraulic conductivity and Verkman, A.S. 1995. Immunolocalization of relaxation in the guinea pig trachea. J. Appl. Physiol.
electrical properties across bovine tracheal epithe- the mercurial-insensitive water channel and 64:466–471.
lium in vitro: effect of histamine. Pflügers Arch. glycerol intrinsic protein in epithelial cell plas- 21. Smith, T.L., Prazma, J., Coleman, C.C., Drake, A.F.,
392:40–45. ma membranes. Proc. Natl. Acad. Sci. USA. and Boucher, R.C. 1993. Control of the mucosal
8. Ballard, S.T., Crews, A.D., and Taylor, A.E. 1998. 92:4328–4331. microcirculation in the upper respiratory tract.
Hydraulic conductivity of porcine tracheal epithe- 15. Boucher, R.C., Stutts, M.J., Bromberg, P.A., and Otolaryngol. Head Neck Surg. 109:646–652.
1344 The Journal of Clinical Investigation | May 2000 | Volume 105 | Number 10