Hypertonic saline alters ion transport across the
P.G. Middleton, K.A. Pollard, J.R. Wheatley
Hypertonic saline alters ion transport across the human airway epithelium. P.G.
Middleton, K.A. Pollard, J.R. Wheatley. #ERS Journals Ltd 2001.
ABSTRACT: Aerosolized hypertonic saline is currently being investigated as a new
agent for the treatment of impaired mucociliary clearance which occurs in many
respiratory diseases. Mannitol aerosols, in particular dry powder inhalers, have been
proposed as an alternative treatment to saline, offering the same osmotic load with other
benefits. However, the effects of these hypertonic aerosols on airway epithelial ion
transport processes have not been tested in human subjects in vivo. This report examines
the effect of these solutions on airway ion transport using the nasal potential difference
Seven healthy nonsmoking adult volunteers were studied. On different days, a dose-
response curve was constructed for the saline added to Krebs N-[2-hydroxyethyl]
piperazine-N9-[2-ethanesulphonic acid] (HEPES) diluent. The reversibility of this saline
effect was measured, and the response to additional saline (500 mM) and mannitol(1 M) compared.
Hypertonic saline decreased nasal PD in a dose-related manner, with mean (SEM)
decreases in PD (less negative) of 6.6 (1.5), 7.6 (1.6), 10.0 (2.0), 13.1 (2.9) and 14.8(3.2) mV (n~4) for addition of 150 mM, 250 mM, 500 mM, 1,200 mM and 2,000 mM
NaCl to the Krebs HEPES diluent, respectively. The effect of hypertonic saline was
fully reversible with washout for 3 min (presaline 15.9 (0.5) mV, postwashout 15.8
(1.1) mV, (n~4)). The hypertonic saline response was rapid in onset, sustained for at
Fibrosis Association, the GovernmentEmployees Medical Research Fund and
least 4 min, and decreased PD from 13.7 (1.7) mV to 5.1 (1.3) mV (n~7, pv0.001). In
contrast, addition of mannitol to the perfusate did not significantly alter nasal PD, witha nonsignificant trend towards an increase (more negative) in the PD, (premannitol 13.9(1.6) mV, postmannitol 15.3 (2.0) mV, n~7).
As the osmotic stimulus of the 1 M mannitol is similar to that of the 500 mM sodium
chloride, the divergent nasal potential difference responses suggest that the response tothe saline was specific to the sodium chloride itself and not the simultaneous change inosmolarity. This demonstrates that the human airway epithelium in vivo can respond totopical hypertonic saline independent of the altered osmolarity. Eur Respir J 2001; 17: 195–199.
The human airways are covered by a thin layer of
physical removal of the retained secretions or the use of
fluid, the airway surface liquid (ASL), which protects
pharmacological agents. Chest physiotherapy, an
the underlying epithelial cells from drying, and traps
example of physical removal of secretions, is extremely
inhaled airborne particles and bacteria. The ASL is
effective, but expensive and time consuming in the long
thought to comprise a watery periciliary or sol layer,
term. For this reason a number of pharmacological
over which floats the gel or mucous layer. Though the
therapies have been developed to increase MCC.
exact composition and volume of these 2 layers is
Recently, interest has focussed on the use of
currently debated, the ASL is thought to be largely
nebulised therapy, in particular hypertonic saline and
regulated by the ion transport processes across the
mannitol, as potential new treatments to increase
airway epithelium, namely sodium absorption and
MCC. Topical application of hypertonic saline via
chloride secretion. Optimal functioning of the ASL is
nebuliser has been shown to increase MCC in patients
required to maximise mucociliary clearance (MCC),
with cystic fibrosis (CF) [1] and a 2 week clinical trial
preventing accumulation of mucous and inhaled
has demonstrated that 10 mL of 6% saline improved
particles. Retained mucous and particles are a site for
symptoms and lung function [2]. Similarly inhaled
bacterial infection, which can then lead to airflow
mannitol has been shown to increase MCC in subjects
limitation, chronic cough and airway damage.
with CF [3] and idiopathic bronchiectasis [4]. However,
As impaired MCC is thought to be involved in the
the mechanism of action of both hypertonic saline and
pathogenesis of a number of respiratory disorders,
mannitol remains speculative. Whilst it has been
treatments have been devised to improve MCC through
hypothesised that the aerosols provide an osmotic
load to the ASL, no studies have examined the effects of
subjects had undergone nasal surgery, and all tests were
hypertonic saline and mannitol on airway epithelial ion
performed at least 4 weeks following an upper
respiratory tract infection. The study was approved
The nasal potential difference (PD) technique, which
by the hospital ethics committee and all subjects gave
involves passage of a small exploring electrode along
the floor of the nose, allows in vivo measurement of iontransport across the human airway epithelium, bypass-ing the difficulties of in vitro techniques. The nasal PD
technique has previously been used to study the effectof topical antibiotics on human airway epithelial ion
The two-tailed paired t-test was used for comparison
transport [5], and to investigate potential new aerosol
and the null hypothesis was rejected at pv0.05. For
therapies for CF [6]. In the current study, the effects of
discussion purposes, increases and decreases refer to the
topical application of hypertonic saline and mannitol
absolute magnitude of the PD, which was lumen
on nasal PD have been investigated, demonstrating that
saline, but not mannitol, rapidly and reversiblydecreases nasal PD in normal human subjects.
Initial studies measured the response of the nasal
epithelium to different concentrations of added NaCl.
Nasal PD was measured using previously described
Following stabilisation of the PD in the Krebs HEPES
methods [5, 7, 8]. Briefly, the exploring electrode
diluent, the nose was perfused for 3 minutes with each
consisted of a double lumen silicone rubber tube with
of the following solutions: Krebsz150, Krebsz250,
the openings of both lumens at the same site, 3 mm
Krebsz500, Krebsz1,200 and Krebsz2,000 mM
from the tip. One lumen was filled with an equal
NaCl. This was associated with significant (p
mixture of saline and electrocardiography electrode
cream, connected to a high impedance voltmeter via a
(1.6), 10.0 (2.0), 13.1 (2.9) and 14.8 (3.2) mV (n~4)
silver/silver chloride electrode. The second lumen was
(fig. 1). As the higher concentrations of saline present in
perfused with the different solutions as outlined below,
using a peristaltic pump which provided a continuous
resulted in subjective discomfort in the nose during
flow of 4 mL.min-1 throughout the perfusion period.
the testing, a solution of Krebsz500 mM NaCl was
The reference electrode consisted of a second silver/
selected for further investigation. This gave approxi-
silver chloride electrode placed over an area of abraded
mately 70% of the maximal response without causing
skin on the forearm, again connected to the voltmeter.
any discomfort. All subsequent studies of hypertonic
Prior to recordings, the offset of the electrodes was
saline in this paper were then performed with
recorded values. In all cases, the tests started with aninitial period of stabilization using standard Krebs N-[2-hydroxyethyl]piperazine-N-[2-ethanesulphonic acid](HEPES) solution. Following stabilization of the nasal
PD, the diluent was changed to the various hypertonicsolutions as indicated below, with different hypertonic
responses measured on different days, in random order.
Perfusion commenced with Krebs HEPES solution
of composition (mM): Naz 140, Kz 6, Ca2z 2, Mg2z
1, Cl- 152, glucose 10 and HEPES 10, titrated topH 7.4. Hypertonic solutions were produced volume-
trically, so that the total concentrations of Naz and Cl-
in the Krebsz500 mM saline were 640 mM and
652 mM, respectively. In the dose response testing,
the solutions were perfused sequentially as indicated.
Due to the 3 mL dead space of the perfusion system,
the new perfusate reached the catheter tip approxi-
mately 45 s following solution change. Fresh stock
solutions were prepared daily and diluted as required.
All solutions were perfused at room temperature
Fig. 1. – The effect of additional saline on the nasal potentialdifference (PD) in four normal subjects. Each new solutionreached the nose approximately 45 s after the perfusate was
changed, reflecting the dead space of the system. The solutionstested are indicated in the bar at the top of the graph. Data are
Nonsmoking control subjects (4 female and 3 male,
expressed as mean¡SEM. The PD is expressed in absolute terms,and was lumen negative. Arrows indicate KrebszNaCl solutions
aged 18 – 40 yrs) with no history of respiratory disease
of the following concentrations; A: 150 mM NaCl; B: 250 mM;
were recruited from the hospital staff for testing. No
nasal PD, with a trend to an increase in the PD,
(2.0) mV, n~7, p~0.14)), as shown in figure 3.
This study has demonstrated that topical application
of hypertonic saline, Krebsz500 mM NaCl, corre-
sponding to a stimulus of y3% saline, decreases nasal
potential difference. The response to the hypertonicsaline was rapid, reversible and dose related, suggestinga direct ion transport effect rather than nonspecific
epithelial toxicity. In contrast, application of a similar
osmotic load through hypertonic mannitol did not
decrease the nasal PD, with a nonsignificant increase
Fig. 2. – The effect of additional saline (500 mM) added to the
(more negative). This suggests that the ionic composi-
Krebs HEPES N-[2-hydroxyethyl]piperazine-N9-[2-ethanesulpho-
tion of the saline, and not the osmotic load, was
nic acid], on the nasal potential difference (PD) in four normal
involved in the saline response. As amiloride pretreat-
subjects. The new solution reached the nose approximately 45 s
ment can block the saline response in normal subjects
after the perfusate was changed (indicated by arrows), reflecting
(unpublished data), passive diffusion of Cl- ions down
the dead space of the system. After 1 min of perfusion with thehypertonic saline solution, the perfusate returned to Krebs
the paracellular pathway would not explain the changes
HEPES diluent to assess the effect of washout. Data are
in nasal PD. Therefore it is hypothesized that the saline
expressed as mean¡SEM. The PD is expressed in absolute terms,
directly alters ion transport processes across the normal
human airway epithelium. This is the first demonstra-tion in human subjects in vivo that increasing the NaClconcentration on the surface alters airway epithelial ion
transport, independent of any effect of osmolarity.
Previous studies measuring the effect of hypertonic
solutions on epithelial cells in vitro have found varying
results. Early studies in canine trachea, a predominantly
Cl- secreting epithelium, have shown that additionalluminal NaCl (75 mM) decreased short-circuit current
(Isc) and increased tissue conductance [9]. The decreasein Isc predominantly reflected an increase in mucosal to
serosal Cl- movement, which was proportional to the
increased Cl- concentration on the mucosal surface. Interestingly, addition of (150 mM) mannitol to theluminal surface decreased Naz absorption across the
canine trachea without significant changes in either Cl-
movement or tissue conductance. MAN et al. [10]
demonstrated that addition of 100 mM mannitol to theluminal fluid decreased PD across the canine trachea by
Fig. 3. – The effect of additional saline (500 mM) or mannitol
3.9 mV, with a similar decrease in Isc. Interestingly
(1 M) added to the Krebs HEPES N-[2-hydroxyethyl]piperazine-N9-[2-ethanesulphonic acid], on the nasal potential difference
there was a small increase in tissue resistance, opposite
(PD) in 7 normal subjects. The new solution reached the nose
to the increased conductance found by YANKASKAS et
approximately 45 s after the perfusate was changed, reflecting
the dead space of the system. Data are expressed as mean¡SEM.
In monolayers of cultured human nasal polyp cells,
The PD is expressed in absolute terms, and was lumen negative. &
WILLUMSEN et al. [11] measured the effects of raised
: saline; #: mannitol. ***: pv0.001 versus baseline presaline;
luminal and serosal osmolality. Addition of 150 mMmannitol to the luminal surface was associated withdecreases (less negative) in both apical (Va) and
To determine whether this response may be related to
basolateral (Vb) membrane potentials, with small but
epithelial toxicity, the reversibility of the hypertonic
variable effects on the transepithelial membrane
saline response was measured. As shown in fig. 2,
potential (Vt) in different monolayers. Addition of
following 1 min of perfusion with Krebsz500 mM
75 mM NaCl to the luminal solution also decreased
NaCl, washout for 3 min reversed the response (presa-
both Va and Vb, but did not alter Vt [11]. The
line 15.9 (0.5) mV, postwashout 15.8 (1.1) mV, (n~4)).
nonsignificant increase in nasal PD found in the current
In seven subjects the change to Krebsz500 mM
study corresponds with the variable effects on Vt found
NaCl was associated with a significant decrease in nasal
in the nasal polyp monolayers; Va and Vb cannot be
PD from 13.7 (1.7) mV to 5.1 (1.3) mV (pv0.001). This
assessed using the nasal PD technique. Similarly, the
response was rapid in onset and sustained for at least
changes in epithelial cell volume noted in the nasal
polyp cultures cannot be measured by the current
(1,000 mM) to the perfusate did not significantly alter
technique. Further studies will be necessary to deter-
mine whether luminal mannitol does alter ion transport
ESCHENBACHER et al. showed that hypertonicity,
in vivo, with similar but opposite effects resulting in no
irrespective of ionic composition, induced both bronch-
oconstriction and cough [17]. Isosmotic saline solution
However, the responses to luminal saline appear to
induced neither bronchoconstriction or cough, but
differ in the two studies with a significant decrease in
isosmotic dextrose, with a low Cl- concentration,
nasal PD in the current study contrasting with the lack
induced cough in the absence of bronchoconstriction
of change in Vt found by WILLUMSEN et al. [11]. This
[17]. This suggests that hypertonicity and ionic
may simply reflect the different doses used in the
composition can, in certain situations, be stimuli for
respective studies – the only dose reported in the
different responses. The relationship between this and
WILLUMSEN et al. study was 75 mM, whilst the smallest
the nasal PD responses in the current study will require
dose used in the current study was 150 mM. Another
possible explanation for the differences between the two
Finally, the differing nasal PD response to hyper-
studies is the use of nasal polyp cells, which are known
tonic saline and mannitol may have important
to exhibit ion transport processes which differ from
implications for the development of new treatments
those found in normal airway epithelial cells [12]. The
for lung diseases. Clinical trials will be necessary to
nasal PD also avoids the difficulties of cell culture
investigate each hypertonic solution individually, as the
techniques, in particular the submersion culture process
results from hypertonic saline and mannitol are not
and the hormones and antibiotics required to maintain
the cells in a viable state. Finally, another alternative
In conclusion, this study has demonstrated that
may be the presence of hormonal factors which may
topical application of hypertonic saline induces a rapid
control ion transport in vivo, but are not maintained
and reversible decrease in nasal potential difference in
when the cells are grown into monolayers in vitro.
vivo, most likely reflecting altered ion transport across
Irrespective of the basis for the differences between the
the human airway epithelium. The nonsignificant
in vivo and the in vitro studies, the current report
increase in nasal potential difference seen with
provides important new information about the ion
hypertonic mannitol suggests that the two solutions
transport responses of the human airway epithelium in
exert different effects on the epithelium, dependent
on their composition. Further studies are necessary
The nasal PD response to increasing concentrations
to investigate the mechanisms involved in these
of NaCl in the current study corresponds with the dose-
related effects of saline on mucociliary clearance. ROBINSON et al. [1] found that aerosols of 0.9%, 3%,
7% and 12% saline were associated with isotope
to thank the subjects who took part in this
clearance at 90 min of 13%, 20%, 24% and 26%,
respectively. Whilst the MCC response may be related
to the nasal PD effects demonstrated in the currentstudy, there are a number of other possibilities,including increased ciliary beat frequency in vivo [13]and direct effects on sputum [14]. Furthermore, thedivergent effects of hypertonic mannitol and saline on
nasal PD do not correspond with the similar increasesin MCC in a variety of clinical situations [3, 4, 15].
Robinson M, Hemming AL, Regnis JA, et al. Effect of
Hyperosmolarity is also a reliable stimulus to
increasing doses of hypertonic saline on mucociliary
provoke bronchoconstriction in susceptible subjects.
clearance in patients with cystic fibrosis. Thorax 1997;
Hypertonic saline solutions have long been used in the
laboratory as an indirect challenge agent, with more
Eng PA, Morton J, Douglass JA, Riedler J, Wilson J,
recent studies by ANDERSON et al. [16] studying the
Robertson CF. Short-term efficacy of ultrasonically
effects of inhaled mannitol. Whilst for the majority of
nebulized hypertonic saline in cystic fibrosis. Pediatr
subjects saline and mannitol produced qualitatively
similar results, some subjects did show 10-fold higher
Robinson M, Daviskas E, Eberl S, et al. The effect of
inhaled mannitol on bronchial mucus clearance in
et al. [17] compared the effects of different solutions
cystic fibrosis patients: a pilot study. Eur Respir J 1999;
with the same osmolarity on the airways of asthmatic
subjects. They demonstrated that 4% saline solution
Daviskas E, Anderson SD, Eberl S, Chan H-K,Bautovich G. Inhalation of dry powder mannitol
was significantly more potent at inducing bronchocon-
striction than 18.3% dextrose in 1% saline, though both
bronchiectasis. Am J Respir Crit Care Med 1999;
solutions had the same osmolarity. Similarly MAGYAR
et al. [18] has reported that 1.3 M KCl was more potent
than 1.7 M NaCl, (both solutions prepared at 10%
weight/volume). This suggests that as well as the
epithelial sodium absorption. Am J Respir Crit Care
osmotic stimulus, the ionic composition may also be
involved in bronchoconstrictor responses.
Smith SN, Middleton PG, Chadwick S, et al. The in
It is well recognized that both the osmolarity and
vivo effects of milrinone on the airways of cystic
ionic composition of an aerosol are important in the
fibrosis mice and human subjects. Am J Respir Cell
induction of cough. In an elegant series of experiments,
Middleton PG, Geddes DM, Alton EWFW. Effect of
Wong LB, Miller IF, Yeates DB. Pathways of
amiloride and saline on nasal mucociliary clearance
substance P stimulation of canine tracheal ciliary
and potential difference in cystic fibrosis and normal
beat frequency. J Appl Physiol 1991; 70: 267 – 273.
subjects. Thorax 1993; 48: 812 – 816.
Wills PJ, Hall RL, Chan W, Cole PJ. Sodium chloride
Middleton PG, Caplen NJ, Gao X, et al. Nasal
increases the ciliary transportability of cystic fibrosis
application of the cationic liposome DC-Chol:DOPE
and bronchiectasis sputum on the mucus-depleted
does not alter ion transport, lung function or bacterial
bovine trachea. J Clin Invest 1997; 99: 9 – 13.
growth. Eur Respir J 1994; 7: 442 – 445.
Daviskas E, Anderson SD, Gonda I, et al. Inhalation
Yankaskas JR, Gatzy JT, Boucher RC. Effects
of hypertonic saline aerosol enhances mucociliary
of raised osmolarity on canine tracheal epithelial ion
clearance in asthmatic and healthy subjects. Eur Respir
transport function. J Appl Physiol 1987; 62: 2241 –
Anderson SD, Brannon JD, Spring J, et al. A new
Man SFP, Hulbert W, Park DSK, Thomson ABR,
method for bronchial-provocation testing in asthmatic
Hogg JC. Asymmetry of canine tracheal epithelium:
subjects using a dry powder mannitol. Am J Respir
osmotically induced changes. J Appl Physiol 1984; 57:
Crit Care Med 1997; 156: 758 – 765.
Willumsen NJ, Davis CW, Boucher RC. Selective
Alteration in osmolarity of inhaled aerosols cause
response of human airway epithelia to luminal but not
bronchoconstriction and cough, but absence of a
serosal solution hypertonicity. Possible role for
permeant anion causes cough alone. Am Rev Respir
transducer. J Clin Invest 1994; 94: 779 – 787.
Magyar P, Dervaderics M, To´th A. Bronchial
Bernstein JM, Yankaskas JR. Increased ion transport
challenge with hypertonic KCl solution in the
Otolaryngol Head Neck Surg 1994; 120: 993 – 996.
DR COLIN K. W. WATTS Watts Scientific Consulting Sydney, Australia ABN 16 091 051 856 www.understandingcancer.com.au/WattsScientificConsulting.html QUALIFICATIONS BSc Hons (First Class), University of Otago, New Zealand PhD, University of Sydney, Australia AREAS OF EXPERTISE • Molecular and cellular biology of cancer; molecular pharmacology; gene cloning and protei
Evolution and Human Behavior 29 (2008) 19 – 25Women's body morphology and preferences forBoguslaw Pawlowskia,b,c,⁎, Grazyna Jasienskad,eaDepartment of Anthropology, University of Wroclaw, 50-138 Wroclaw, PolandbDepartamento de Ecologia Humana CINVESTAV-IPN, Unidad Mérida, 97310 Mérida, MexicocInstitute of Anthropology, Polish Academy of Sciences, 50-951 Wroclaw, PolanddDepartment o