J Appl Physiol 104: 809–820, 2008.
First published January 10, 2008; doi:10.1152/japplphysiol.00846.2007.
Innovative Methodology
Isolation of rat trachea interstitial fluid and demonstration of local cytokine
production in lipopolysaccharide-induced systemic inflammation
Elvira Semaeva, Olav Tenstad, Athanasia Bletsa, Eli-Anne B. Gjerde, and Helge Wiig
Department of Biomedicine, University of Bergen, Bergen, Norway
Submitted 7 August 2007; accepted in final form 8 January 2008
interstitium; capillary filtration; extracellular matrix
SIGNIFICANT INFORMATION ON fluid exchange and transport between the blood and tissue compartments may be gained from
analysis of interstitial fluid, i.e., the fluid portion of the extracellular compartment. As an example, will proteins dissolved
in the interstitial fluid exert a colloid osmotic pressure (COP)
that opposes the corresponding pressure in plasma in such a
manner that changes in protein content may impair the normally well-functioning and autoregulated fluid balance (4).
In the airways, knowledge of the factors contributing to fluid
homeostasis such as interstitial fluid COP is of vital importance
because an edema in this organ may result in increased airway
resistance and dyspnea (37, 44). Accordingly, access to interstitial fluid from the tracheobronchial tree is of major importance and may have therapeutical implications. In this context,
fluid from trachea may be considered as a surrogate for lower
airways. To our knowledge, only one study has dealt directly
Address for reprint requests and other correspondence: H. Wiig, Dept. of
Biomedicine, Univ. of Bergen, Jonas Lies vei 91, N-5009 Bergen, Norway
(e-mail: [email protected]).
http://www. jap.org
with the question of trachea interstitial fluid isolation before. In
rabbits, Nordin et al. (23) divided and clamped the trachea to
measure transvascular fluid exchange parameters. During this
process, the tissue became edematous, and the authors were
able to sample interstitial fluid and lymph. Several authors
have addressed questions in relation to mechanisms behind
fluid extravasation in the airways (for review, see Refs. 27, 35,
36), and the subepithelial (i.e., interstitial fluid) pressure that
drives the inflammatory transudation into airway lumen has
been estimated (30). Furthermore, Persson et al. (27) pointed to
the importance of developing good in vivo models to study
such phenomena. There are, however, no studies on native
interstitial fluid isolation in control conditions and/or in smaller
animals.
Although various techniques have been developed for tissue
fluid isolation (for review, see Ref. 4), none of these seems
suitable for trachea except sampling of lymph. In preliminary
experiments with control rats, we were not able to discern
lymphatics in trachea that we could cannulate and therefore
had to use an alternative approach. In a recent study, our group
(38) validated a centrifugation method for isolation of interstitial fluid in tumors and skin and later showed that this method
is also applicable for this purpose in other tissues, such as bone
marrow (39) and dental pulp (6), where the interstitium is
difficult to access. Our major aim was to isolate interstitial fluid
from the airways. Encouraged by our previous studies, we
asked the question of whether centrifugation would be a
suitable method to isolate tissue fluid from rat trachea, which is
itself of interest but which may also serve as a surrogate for
lower airways. If so, we asked whether the isolated fluid
represented undisturbed interstitial fluid and whether a systemic inflammatory response was reflected in the interstitium
of trachea in rats.
MATERIALS AND METHODS
Experimental Animals
The experiments were performed in Wistar rats (n ⫽ 87) of either
sex (range 195–260 g, median 230 g). The rats had free access to food
(a standard laboratory diet) and water before any experimental procedure. The rats were anesthetized with pentobarbital sodium (50
mg/kg body wt ip). While anesthetized, we maintained the rats’ body
temperature at 37 ⫾ 1°C using a heating pad and lamp. Polyethylene
(PE-50) catheters were placed in the femoral vein for injection of
tracers and substances (see below) and in the femoral artery for
blood sampling. Blood was sampled in heparinized vials for
isolation of plasma unless otherwise specified. Rats were killed by
cardiac arrest induced under anesthesia with an intravenous injecThe costs of publication of this article were defrayed in part by the payment
of page charges. The article must therefore be hereby marked “advertisement”
in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
8750-7587/08 $8.00 Copyright © 2008 the American Physiological Society
809
Downloaded from http://jap.physiology.org/ by 10.220.33.1 on June 15, 2017
Semaeva E, Tenstad O, Bletsa A, Gjerde EB, Wiig H. Isolation
of rat trachea interstitial fluid and demonstration of local cytokine
production in lipopolysaccharide-induced systemic inflammation.
J Appl Physiol 104: 809–820, 2008. First published January 10, 2008;
doi:10.1152/japplphysiol.00846.2007.—Access to interstitial fluid
from trachea is important for understanding tracheal microcirculation
and pathophysiology. We tested whether a centrifugation method
could be applied to isolate this fluid in rats by exposing excised
trachea to G forces up to 609 g. The ratio between the concentration
of the equilibrated extracellular tracer 51Cr-labeled EDTA in fluid
isolated at 239 g and plasma averaged 0.94 ⫾ 0.03 (n ⫽ 14),
suggesting that contamination from the intracellular fluid phase was
negligible. The protein pattern of the isolated fluid resembled plasma
closely and had a protein concentration 83% of that in plasma. The
colloid osmotic pressure in the centrifugate in controls (n ⫽ 5) was
18.8 ⫾ 0.6 mmHg with a corresponding pressure in plasma of 22 ⫾
1.5 mmHg, whereas after overhydration (n ⫽ 5) these pressures fell to
9.8 ⫾ 0.4 and 11.9 ⫾ 0.4 mmHg, respectively. We measured inflammatory cytokine concentration in serum, interstitial fluid, and bronchoalveolar lavage fluid in LPS-induced inflammation. In control
animals, low levels of IL-1␤, IL-6, and TNF-␣ in serum, trachea
interstitial fluid, and bronchoalveolar lavage fluid were detected. LPS
resulted in a significantly higher concentration in IL-1␤ and IL-6 in
interstitial fluid than in serum, showing a local production. To conclude, we have shown that interstitial fluid can be isolated from
trachea by centrifugation and that trachea interstitial fluid has a high
protein concentration and colloid osmotic pressure relative to plasma.
Trachea interstitial fluid may also reflect lower airways and thus be of
importance for understanding, e.g., inflammatory-induced airway obstruction.
Innovative Methodology
810
ISOLATION OF TRACHEA INTERSTITIAL FLUID IN RATS
tion of saturated potassium chloride. All animal experiments were
conducted in accordance with regulations of the Norwegian State
Commission and with preapproval from the ethical committee of
University of Bergen.
Centrifugation Technique
Validation Experiments
Assessment of potential admixture of intracellular fluid. Our aim
was to isolate native, uncontaminated interstitial fluid from trachea.
To estimate any contribution of cellular fluid, we studied the concentration ratio of an extracellular tracer in trachea interstitial fluid and
plasma collected from rats (n ⫽ 20). Addition of cellular fluid not
containing tracer will appear as a reduced concentration in centrifugate relative to plasma. After anesthesia and catheter placement, the
animals were bilaterally nephrectomized to avoid excretion of the
extracellular tracer and thereby to keep the plasma concentration
stable by ligating both kidneys. Then, ⬃0.1 MBq 51Cr-labeled EDTA
in a volume of 0.2 ml was administered intravenously and allowed
time to distribute in the extracellular fluid phase. The dose was chosen
to produce sufficient counts in isolated trachea interstitial fluid samples. After an equilibration phase of 115 min, ⬃0.01 MBq 125I-labeled
human serum albumin (125I-HSA) was injected intravenously to label
the vascular space and was allowed 5 min of circulation time before
blood sampling and cardiac arrest. Thereafter, the rat was transferred
to the incubator, and the trachea was isolated and handled as described
above.
After centrifugation, the tube was brought back into the incubator,
and the fluid was collected with a graded glass micropipette. Trachea
interstitial fluid isolated in these experiments was pipetted into 1 ml of
saline contained in counting vials. Thereafter, the samples were counted
in LKB universal gamma counter (model 1282; Compugamma, Turku,
Finland), using window settings of 240 – 400 keV for 51Cr and 15– 80
keV for 125I. Standards were counted in every experiment to obtain
J Appl Physiol • VOL
Fluid Volumes
In a separate series of experiments (n ⫽ 6 rats), we determined the
extracellular fluid and plasma distribution volumes in trachea utilizing
51
Cr-EDTA and 125I-HSA, respectively. After anesthesia and placement of the catheter in the femoral vein, the animals were bilaterally
nephrectomized to prevent excretion of the extracellular tracer. Volumes were determined with 51Cr-EDTA and 125I-HSA as described
above, except that the dose of 51Cr-EDTA was 0.05 MBq. Tissue
samples taken from the trachea were placed in tared, covered vials and
weighed and counted as described above. Plasma equivalent distribution volumes (V) for 51Cr-EDTA and 125I-HSA were calculated as
follows: V ⫽ (cpm/g tissue)/(cpm/ml plasma).
104 • MARCH 2008 •
www.jap.org
Downloaded from http://jap.physiology.org/ by 10.220.33.1 on June 15, 2017
In our attempt to isolate interstitial fluid from the trachea, we used
a modification of the method described by Wiig et al. (38) in tumors
and skin. The euthanized rats were transferred immediately after
cardiac arrest to an incubator at room temperature and 100% relative
humidity to avoid evaporation from the trachea. The trachea was
rapidly dissected free of surrounding tissue, excised, and split longitudinally. The split trachea was placed with the serosa side facing
downward on a nylon mesh basket (pore size of ⬃15 ⫻ 20 ␮m),
which was placed in an Eppendorf centrifugation tube so that the
sample was kept up from the bottom of the tube (3). The preweighed
tube with nylon basket was capped immediately and reweighed after
we added the trachea sample before centrifugation for 10 min in an
Eppendorf 5417R centrifuge placed in a cold room at 4°C. Immediately after centrifugation, the tube was brought back into the incubator.
Trachea interstitial fluid that accumulated at the bottom was collected in glass capillaries, which were either closed immediately for
later analysis or diluted in buffer for HPLC (see below). Typical
samples were 0.5–2 ␮l.
To reduce the risk of compression of cellular elements of the
trachea, the centrifugation speed was as low as possible to collect an
interstitial fluid sample. In initial experiments, the samples were spun
at 424 g (2,000 rpm) for 10 min. In subsequent experiments, we
reduced the G force successively in 100-rpm steps to explore the
minimal G force needed to produce a sample. We found that it was
always possible to isolate fluid at 239 g (1,500 rpm). Generally, with
increasing speed of centrifugation, an increased amount of erythrocytes accumulated in the centrifugate. Exposing the tracheal tissue to
G forces of 609 g (2,500 rpm) and above resulted in progressive
hemolysis of the centrifugate. Therefore, 239 g (1500 rpm) was used
for sampling of trachea fluid unless otherwise specified. In this study,
this fluid is called trachea interstitial fluid.
spillover corrections, and counts were corrected for background and
spillover.
To estimate whether fluid not containing tracer had been added to
the centrifugate, we calculated the ratio of trachea interstitial fluidto-plasma concentration (R) of the extracellular tracer using the
following formula: R ⫽ (cpm/␮l interstitial fluid)/(cpm/␮l plasma),
where cpm is counts per minute of 51Cr-EDTA or 125I-HSA.
Assessment of potential admixture of trachea surface liquid
and proteins. The surface of the trachea is covered with liquid
containing mucin macromolecules, electrolytes, and water (7, 33). To
evaluate whether such fluid was contained in the trachea interstitial
fluid, we conducted two other series of experiments. In the first series
(n ⫽ 6), we investigated whether, and if so, how much of 51Cr-EDTA
and 125I-HSA present in plasma was delivered to trachea surface
liquid. We would then be able to determine whether the surface liquid
was equilibrated with tracer after a regular extracellular tracer equilibration period. If not, admixture of surface liquid would make the
isolated interstitial fluid more dilute than plasma with respect to tracer.
After a regular tracer equilibration period of 120 and 5 min, respectively, the experiment was terminated. The trachea was isolated and
incised as described above, and the cut proximal and distal ends of the
trachea were wiped carefully with blotting paper to avoid potential
contamination from severed blood vessels. To avoid the use of an
absorbing material, which may disturb the epithelial barrier function
and sample subepithelial (i.e., interstitial) fluid and solutes (10), we
decided to try an alternative approach. With the assumption that some
trachea surface liquid would adhere to a nonabsorbing catheter surface, a preweighed PE-50 catheter was carefully stroked along the
inner circumference of the trachea. Attempt was made to avoid
contact between the catheter and the exposed surface of the trachea
wall. The catheter was transferred to a preweighed tared vial and
reweighed. After 40 ␮l of HPLC buffer were added, the sample was
counted together with plasma and trachea. After counting was completed, the trachea surface liquid protein distribution pattern was
determined by HPLC and compared with that of trachea interstitial
fluid and plasma.
In another series of experiments (n ⫽ 5), we examined the transport
of substances from trachea surface liquid through the interstitium into
the centrifugate to determine the possible contribution of surface
liquid macromolecules to isolated interstitial fluid. In these experiments, 1 ␮l of a tracer mixture containing stock solution of 51CrEDTA (5 ⫻ 10⫺4 MBq) and 125I-HSA (5 ⫻ 10⫺4 MBq) was carefully
deposited onto the trachea surface. The trachea was transferred to a
centrifuge tube for centrifugation at 239 g, and isolated interstitial
fluid and trachea were counted as described above.
Estimation of volume of trachea surface liquid. To calculate the
mucosa surface area of trachea and thus be able to estimate the
volume of the trachea surface liquid, we determined the inner diameter in proximal and distal end of excised tracheas (n ⫽ 6) in
anteroposterior and frontal directions using a stereomicroscope
equipped with a graticule.
Innovative Methodology
ISOLATION OF TRACHEA INTERSTITIAL FLUID IN RATS
HPLC of Plasma and Fluid Isolated From Trachea
Identification of Plasma Protein by HPLC-Mass Spectrometry
After HPLC separation as described above, macromolecular fractions from plasma and trachea interstitial fluid were exchanged into
100 mM ammonium bicarbonate and concentrated to a final volume of
25 ␮l using a Microcon YM-50 centrifugal filter unit (Millipore). The
samples were then denatured by addition of 25 ␮l trifluoroethanol
(Sigma-Aldrich T-8132) and 2.5 ␮l dithiothreitol (Sigma-Aldrich
D-5545) at 60°C for 45 min and digested at 37°C overnight by trypsin
(Sigma-Aldrich T6763, protease-to-protein ratio of 1:20) in accordance with a protocol provided by Agilent Technologies (http://
www.chem.agilent.com; publication no. USHUPO3). Digested proteins from plasma and trachea interstitial fluid were desalted by C18
peptide cleanup spin tubes (Agilent Technologies) and analyzed with
an Agilent 1100 LC/MSD Trap XCT Plus system consisting of a
nanoflow pump, wellplate sampler, capillary pump, HPLC-Chip/MS
interface, and a Trap XCT Plus mass spectrometer. The HPLC-Chip
contained a 0.075 ⫻ 43 mm ZORBAX 300SB C18 5-␮m column and
an integrated 40-nl enrichment column packed with the same material
(part no. G4240-62001). Processing of the MS/MS data was performed with Spectrum Mill MS proteomics workbench software (Rev
A.03.02.060) using identity mode and default settings.
Colloid Osmotic Pressure
The COP levels in trachea interstitial fluid and in plasma were
measured in a colloid osmometer designed for submicroliter samples
(41), using membranes with a cutoff size of 30 kDa. Usually, samples
of 0.5–1 ␮l were applied to the osmometer membrane.
Analysis of Inflammatory Mediators in Serum and Interstitial Fluid
To be able to monitor whether an inflammatory process was
reflected in the interstitial fluid, the proinflammatory mediators IL-␤,
IL-6, and TNF-␣ and the anti-inflammatory mediator IL-10 were
measured simultaneously in samples using Lincoplex kit (Linco
J Appl Physiol • VOL
Research, St. Charles, MO) according to the manufacturer’s instructions. In these experiments, analyses were performed on serum as
recommended by the manufacturer. The samples were prepared for
analysis by dilution with serum matrix diluent provided with the kit.
Serum and interstitial fluid samples were reacted with a mixture of
fluorescent beads bound with specific anticytokine primary antibodies,
resulting in binding of the cytokines in the sample to the beads with
the corresponding antibody. Then, biotin anticytokine secondary antibodies were added and allowed to bind to the cytokine-bead complex, followed by addition of fluorescent phycoerythrin-conjugated
streptavidin. Total surface fluorescence was then measured with a
BioPlex fluorescent flow-based fluorescence detection system (BioPlex multiplex suspension system; Bio-Rad). A broad range of standards (4.8 –20,000 pg/ml) was provided in the multiplex kit, and the
multiplexed assay was analyzed on a flow cytometer (Luminex100,
Luminex). The minimum detection level ranged from 2.3 (IL-1␤) to
9.8 (IL-6) pg/ml. To prevent potential protein degradation of mediators present in low concentrations, we added protease inhibitor cocktail (P8340, Sigma) to these samples. The stock solution was diluted
1:10 in 0.9% saline, and 1 ␮l of the diluted solution was added to the
tubes before centrifugation for an immediate effect, whereas undiluted
protease inhibitor (15 ␮l/ml) was added to serum. After interstitial
fluid isolation, the samples were frozen immediately and stored at
⫺80°C until further analyses were performed.
Bronchoalveolar Lavage
Bronchoalveolar lavage (BAL) is a versatile and much used technique to study, e.g., signaling substances liberated from the airways.
Fluid sampled by BAL will mainly represent airways distally to
trachea and reflect more closely the surface liquid than the present
centrifugation technique. To compare the composition of the interstitial fluid isolated by centrifugation from trachea to that obtained in
more distal airways by BAL method, we added a series of experiments
in which lavage fluid was sampled. In these experiments, PBS totaling
9 ml was instilled into trachea via a tracheal cannula, in three 3-ml
aliquots. Around 5–7 ml of BAL fluid were recovered for each rat.
Protease inhibitor (15 ␮l/ml) was added immediately after sampling,
and the samples were stored at ⫺80°C until analysis.
Immunohistochemistry
Tracheas from rats exposed to LPS for 3 h (n ⫽ 2) and from control
rats (3 h vehicle) (n ⫽ 2) were removed and fixed in 4% paraformaldehyde with 0.2% picric acid. The tissue was then rinsed in 0.1 M
phosphate buffer, soaked in 30% sucrose overnight, and stored at
⫺80°C until they were sectioned. The specimens were then embedded
in Tissue-Tek OCT compound (Sakura, Zoeterwoude, The Netherlands), and 10-␮m-thick saggital sections were cut in a freezing
(⫺20°C) slide microtome. The immunoreactions were performed on
precoated glass slides (SuperFrost Plus, MenzelGlaser, Braunschweig,
Germany). For the immunohistochemical staining of possible cytokine-producing cells, the following antibodies were used: mouse
anti-rat W3/13 (dilution 1:400; AbD-Serotec, Oxford, UK), which
stains all leukocytes except B lymphocytes; and mouse anti-rat ED1
(dilution 1:400; AbD-Serotec), which stains tissue macrophages. For
the cytokine localization, we used rabbit anti-rat IL-1␤ (dilution
1:150; Santa Cruz Biotechnology, Santa Cruz, CA). Alternate serial
sections were double stained with IL-1␤ and W3/13 or ED1 antibodies. Sections were blocked with 5% normal goat serum (Chemicon)
before incubation with primary antibodies (overnight at 4°C). After
sections were washed, samples were incubated with corresponding
secondary antibodies Cy3 goat anti-rabbit IgG (dilution 1:300; Jackson ImmunoResearch, West Grove, PA) and Cy2 goat anti-mouse IgG
(dilution 1:300; Jackson ImmunoResearch) for 1 h at room temperature. The sections were coverslipped with Vectashield mounting
medium (Vector Laboratories, Burlingame, CA) with 4⬘,6-diamidino2-phenylindole, which stains DNA, and viewed with a fluorescence
104 • MARCH 2008 •
www.jap.org
Downloaded from http://jap.physiology.org/ by 10.220.33.1 on June 15, 2017
We wanted to compare trachea interstitial fluid with that eluted into
a buffer to investigate whether the centrifugation process per se
affected the composition of the interstitial fluid. For this purpose,
trachea was soaked in PBS containing 0.02% azide for 2 h, and
trachea eluate was isolated. Furthermore, to induce a situation with
cellular damage and release of cellular proteins to the surrounding
solution, we isolated extracts from trachea after induction of cell
damage. These tracheas (n ⫽ 5) were freeze dried five times and
crushed before being soaked in PBS for 24 h. After centrifugation, the
supernatant was collected for further analysis.
The distributions of macromolecules in isolated trachea interstitial
fluid, eluate, tissue extracts (homogenates), trachea surface liquid, and
plasma were determined by high-resolution size-exclusion chromatography using three 7.8 mm (ID) ⫻ 30.0 cm TSKgel G3000SWXL
or two 4.6 mm ⫻ 30 cm SuperSW3000 columns coupled in series
(Tosoh Biosciences, Stuttgart, Germany) with an optimal separation
range for globular proteins of 10 –500 kDa.
The total protein concentration and the concentration of albumin
and IgG in the eluted fluids were measured by UV detection at 220 nm
on an Ettan HPLC system (GE Healthcare), as described in previous
publications (38, 40). Isolated fluid, 0.2–2 ␮l, was diluted in mobile
phase (0.1 M Na2SO4 in 0.1 M phosphate buffer, pH 7.0) in a total
volume of 40 ␮l and injected onto the HPLC system. The injected
volumes were 10 ␮l (super SW3000 columns) or 35 ␮l (G3000SWXL
columns) using 20- or 200-␮l loops, respectively. The flow was held
constant at 0.2 or 0.5 ml/min. Human IgM, ␣2-macroglobulin, fibrinogen, haptoglobin, rat IgG, BSA, and ovalbumin were purchased from
Sigma-Aldrich and used as standards to calculate the range of hydrodynamic radii and apparent molecular weights of the protein fractions
from plasma and trachea interstitial fluid.
811
Innovative Methodology
812
ISOLATION OF TRACHEA INTERSTITIAL FLUID IN RATS
microscope (Axio Imager, Carl Zeiss MicroImaging). The images
were captured with AxioCam MRm camera (Carl Zeiss), and the
AxioVision 4.4 (Carl Zeiss) imaging system was used for analyses.
The specificity of the immune reaction was tested by omission of
the primary antibodies or substitution with isotype controls.
Experimental Protocol
Statistical Methods
All values are means ⫾ SE unless otherwise stated. Differences
were tested with two-tailed t-test using paired comparison when
appropriate or ANOVA followed by Dunn or Bonferroni test. P ⬍
0.05 was considered statistically significant.
RESULTS
The weight of an isolated trachea used in our experiments
averaged 0.047 ⫾ 0.01 g (n ⫽ 22). Typical trachea fluid
samples were straw colored. Occasionally, some erythrocytes
were collected at the bottom of the centrifugation tube, but
these were avoided on removal of the sample from the tube and
were not included in further analyses. Smears (n ⫽ 2) showed
that the isolated fluid was practically cell free in the control
situation. The amount isolated at the lowest force reported here
(239 g) was typically 1–1.5 ␮l, increasing to 1.5–2 ␮l at 424 g.
If we assume a density of 1.0 g/ml, these volumes suggest
2– 4% of the tissue wet weight was removed by the tissue
centrifugation. There was no clear relationship between rat
body weight and thus trachea weight and the amount of
interstitial fluid that was isolated by centrifugation.
Analyses of Trachea Fluid
To determine whether the interstitial fluid was concentrated
or diluted during the isolation process, we measured recovered
J Appl Physiol • VOL
Fig. 1. Ratio of trachea fluid-to-plasma concentration of an extracellular and
intravascular tracer as a function of applied G force. A: an extracellular tracer
(51Cr-labeled EDTA) was equilibrated in extracellular fluid isolated by centrifugation (centr). Increasing the G force resulted in a reduced trachea
fluid-to-plasma ratio of the extracellular tracer, indicating dilution of the
centrifugate. All ratios differed significantly from 1.0. B: an intravascular
tracer [125I-labeled human serum albumin (125I-HSA)] had equilibrated for 5
min before fluid isolation. The trachea fluid-to-plasma ratio reflects plasma
contribution to the centrifugate. Values are means ⫾ SE; n ⫽ 14 rats for 239
g, n ⫽ 12 rats for 424 g, and n ⫽ 3 rats for 609 g. ANOVA: *P ⬍ 0.05
compared with the ratio at 239 g.
104 • MARCH 2008 •
www.jap.org
Downloaded from http://jap.physiology.org/ by 10.220.33.1 on June 15, 2017
LPS-induced acute inflammation. To study whether a condition
known to induce respiratory symptoms and respiratory distress syndrome resulted in changes in inflammatory mediators in trachea
interstitial fluid, we induced a systemic inflammation by infusion of
LPS from Escherichia coli (serotype 0127:B8; Sigma). The LPS was
diluted in PBS containing 0.1% BSA, and the stock solution had a
concentration of 0.25 mg/ml. The rats received a dose of 3.0 mg/kg
LPS by infusion of stock solution intravenously and were observed for
3 h (n ⫽ 10). In these experiments, the femoral artery was catheterized
for monitoring arterial pressure. Five of these rats were used for
isolation of interstitial fluid by centrifugation; the others were subjected to BAL. In this series, protease inhibitor was added to trachea
and BAL fluid and serum as described above. Samples were stored at
⫺80°C until analysis.
To investigate whether leukocytes migrated to trachea interstitial
fluid during LPS treatment, we made a smear of isolated interstitial
fluid in LPS-treated (n ⫽ 2) and control (n ⫽ 2) rats after dilution in
10 ␮l PBS. All smears were stained with Colorrapid (Lucerna-Chem,
Luzern, Switzerland). To study whether the LPS treatment influenced
the fluid volumes in trachea, we measured extracellular fluid and
plasma volumes using 51Cr-EDTA and 125I-HSA as described above
(see Fluid Volumes above) (n ⫽ 6).
Overhydration. We asked whether a volume load known to induce
a dilution of plasma proteins and tissue overhydration was reflected in
trachea interstitial fluid. To this end, a volume of Ringer solution
corresponding to 15% of body weight was infused intravenously with
an infusion pump over a period 30 min in five rats. After infusion, an
equilibration period of 1 h was allowed. After rats were killed,
interstitial fluid from trachea was isolated, and COP levels in interstitial fluid and plasma were measured as describe above.
tracer that had equilibrated in extracellular fluid (51Cr-EDTA)
or distributed in plasma (125I-HSA), and the results are shown
in Fig. 1.
The average concentration of 51Cr-EDTA in trachea fluid
relative to that in the plasma was 0.94 ⫾ 0.03 (n ⫽ 14), 0.94 ⫾
0.03 (n ⫽ 12), and 0.49 ⫾ 0.04 (n ⫽ 3) at centrifugation speeds
of 239, 424, and 609 g, respectively (Fig. 1A). Although the
ratios found at the two lowest G forces were close to unity,
both differed significantly from 1.0 (P ⬍ 0.05 for both ratios).
We also determined the amount of fluid in the centrifugation
that derived from plasma and the recovery in the centrifugate
of the intravascular tracer 125I-HSA. The ratios for 125I-HSA
corresponding to those of 51Cr-EDTA above was 0.017 ⫾
0.004 (n ⫽ 11), 0.032 ⫾ 0.008 (n ⫽ 8), and 0.058 ⫾ 0.017
(n ⫽ 3) at the centrifugation speeds of 239, 424, and 609 g,
respectively (Fig. 1B).
By careful circumferential stroking of inner surface of the
trachea with a plastic catheter, we were able to isolate
0.2– 0.4 mg (corresponding to 0.2– 0.4 ␮l assuming a density of 1.0 g/ml) of surface liquid from an isolated trachea.
Innovative Methodology
ISOLATION OF TRACHEA INTERSTITIAL FLUID IN RATS
When analyzing for tracer recovery in this fluid after 120
and 5 min of equilibration, we found that the mean concentration of 51Cr-EDTA and 125I-HSA was 0.69 ⫾ 0.05 (n ⫽
5) and 0.004 ⫾ 0.001 (n ⫽ 5) relative to that of plasma,
respectively.
When a mixture of 51Cr-EDTA and 125I-HSA was applied to
the mucosa (n ⫽ 5), we were on average able to recover
0.019 ⫾ 0.011 and 0.011 ⫾ 0.004 of the dose of tracer,
respectively.
The inner diameter (estimated as mean value of proximal and
distal diameter) of trachea (n ⫽ 6) averaged 1.8 ⫾ 0.1 and 2.3 ⫾
0.1 mm in anteroposterior and frontal direction, respectively, and
the mean length of the trachea used for centrifugation was 30.0 ⫾
0.2 mm. These numbers were used to calculate a mean inner
trachea surface area of samples used in our experiments of 194
mm2 and were found as the product of the circumference (C) of an
ellipsis [calculated as C ⫽ ␲ ⫻ 公2 ⫻ (a2 ⫹ b2)] and trachea
length.
We quantified the space of origin of the extracellular fluid
isolated from trachea. Total extracellular fluid volume, calculated as 120-min distribution volume of 51Cr-EDTA, was
0.38 ⫾ 0.02 ml/g wet wt (n ⫽ 8). The intravascular plasma
volume determined as the 5-min distribution volume of
125
I-HSA was 0.027 ⫾ 0.006 ml/g wet wt. Thus the trachea
interstitial volume, calculated as the difference between
extracellular fluid and plasma volume, was 0.35 ⫾ 0.03
ml/mg wet wt.
LPS exposure did not result in any changes in the fluid
distribution volumes in trachea. After exposure, the total extracellular fluid volume was 0.37 ⫾ 0.02 ml/g wet wt, with a
corresponding intravascular plasma volume of 0.023 ml/g
wet wt.
Protein Patterns in Trachea Fluid Analyzed With HPLC
In an attempt to validate that the fluid isolated by centrifugation was of extracellular origin, we subjected the samples
from six rats to HPLC analysis and compared the patterns of
isolated fluids with those of corresponding plasma. Representative patterns of plasma, of trachea fluid isolated by centrifugation, of eluate from intact trachea eluted in PBS for 2 h, and
of eluate from crushed trachea eluted in PBS for 24 h are
shown in Fig. 2.
The protein peaks found in plasma were also found in
trachea centrifugate and eluate (Fig. 2, A–C) (n ⫽ 6), whereas
the pherograms from crushed trachea samples (n ⫽ 5) differed
substantially (Fig. 2D), indicating significant intracellular fluid
contamination in the latter.
For molecules smaller than albumin, the elution pattern of
plasma, trachea interstitial fluid, and eluate differed slightly.
Whereas plasma had practically no peaks eluting after albumin
(Fig. 2A), trachea interstitial fluid contained minor peaks representing proteins not present in plasma (Fig. 2B). These peaks
were more pronounced in eluate (Fig. 2C), indicating some
addition of intracellular fluid and/or fluid from the surface layer
(see below).
To compare the relative distribution of differently sized
macromolecules in plasma, trachea interstitial, and surface
liquid, we normalized all pherograms with respect to albumin
(n ⫽ 5) and plotted the elution curves for interstitial and
surface liquid together with that of plasma (Fig. 3). Plasma and
trachea interstitial fluid from four rats were also analyzed by
on-line reverse-phase nano liquid chromatography coupled
with ion-trap mass spectrometry to establish the identity of the
most abundant proteins derived from a sample corresponding
to ⬃20 nl of undiluted fluid (Table 1). Although 20 –30 of the
most abundant plasma proteins were also found in trachea
Fig. 2. High-resolution size-exclusion chromatography of plasma and tissue fluid samples using
three 7.8 mm ID ⫻ 30.0 cm TSKgel G3000SWXL
columns coupled in series. Representative patterns
are shown for plasma (A), trachea interstitial fluid
isolated by centrifugation at 239 g for 10 min (B),
eluate from intact trachea eluted in PBS for 2 h
(C), and eluate from crushed trachea eluted in PBS
for 24 h (D). Ve, elution volume.
J Appl Physiol • VOL
104 • MARCH 2008 •
www.jap.org
Downloaded from http://jap.physiology.org/ by 10.220.33.1 on June 15, 2017
Fluid Distribution Volumes in Trachea
813
Innovative Methodology
814
ISOLATION OF TRACHEA INTERSTITIAL FLUID IN RATS
(Fig. 3), resulting in an interstitial fluid-to-plasma protein concentration ratio slightly greater than 1 and reflecting minor contamination from epithelial-derived proteins (Table 1). A somewhat different picture was found when we compared plasma
and trachea surface liquid (Fig. 3B). Whereas the total protein
concentration was low (Fig. 4), surface liquid contained additional high-molecular-mass peaks and a significant amount of
proteins eluting in the low-molecular-mass range (⬍60 kDa)
that were not found in plasma (Fig. 3B). Most of these proteins
probably originate from shed epithelial cells and mucus.
The results from the HPLC determination of albumin and
total protein in trachea interstitial and surface liquid and
plasma (n ⫽ 6 for all) are shown in Fig. 4. In trachea interstitial
fluid, the concentration of total protein and albumin relative to
that in plasma averaged 83 and 86%, respectively (P ⬍ 0.05 for
both comparisons). The corresponding average numbers for
trachea surface liquid were 14 and 10%.
COP levels in fluid isolated from trachea in control situations (n ⫽ 5) and after overhydration of 15% of body weight
(n ⫽ 5) are shown in Fig. 5. In the control situation, COP in
trachea interstitial fluid averaged 18.8 mmHg with a corresponding pressure in plasma of 22 mmHg.
Overhydration induced by infusion of saline corresponding
to 15% of body weight resulted in significant reduction in COP
in plasma and in trachea interstitial fluid. Thus COP levels in
fluid isolated from the trachea and plasma after an equilibration
period of 60 min following the infusion averaged 9.8 and 11.9
mmHg, respectively. In control situations and after overhydration, the COP in trachea interstitial fluid was lower than the
corresponding COP in plasma (P ⬍ 0.05, paired t-test).
Cytokine Concentrations in Serum, Trachea Interstitial
Fluid, and BAL Samples
Fig. 3. Direct comparison of high-resolution size-exclusion pherograms
from plasma and trachea tissue fluid samples using two 4.6 mm ⫻ 30 cm
SuperSW3000 columns coupled in series. The peaks are numbered to aid
description. The trachea fluid curves have been normalized with respect to
albumin (peak 8), and the curve normalization factor relative to plasma is
shown. A: representative patterns for plasma (dashed line) and trachea interstitial fluid isolated by centrifugation (IF; solid line). B: representative patterns
for plasma (dashed line) and trachea surface liquid (SL; solid line). Elution
time was at 0.2 ml/min; UV signal was at 220 nm in arbitrary units (AU).
centrifugate, this fluid also contained intracellular enzymes and
proteins known to reside in or on the epithelial cell layer. The
contribution of intracellular proteins was most pronounced in
fractions 7 and 9 (Fig. 3A). When comparing plasma and
interstitial fluid, we observed that plasma contained a significantly higher fraction of proteins from high-molecular-mass
fractions 2– 6, as shown in Table 1 and exemplified in Fig. 3A,
suggesting a certain size selectivity of the trachea capillaries
for these proteins. Of note, the first peak in pherograms from
trachea interstitial fluid clearly exceeded that in plasma (interstitial fluid-to-plasma protein concentration ratio of ⬃7; Table
1), indicating local production of IgM. We also observed that
peak 7 in the pherogram from trachea interstitial fluid did not
superimpose completely with the corresponding peak in plasma
J Appl Physiol • VOL
We asked whether a systemic inflammation resulted in detectable changes in inflammatory mediators in trachea interstitial
fluid and whether there was a local production of cytokines in
trachea. Also, we wanted to compare the present fluid isolation
method to the more commonly used method for demonstration
of potential inflammatory reactions in airway tissue, namely,
the BAL method. It should be noted, however, that substances
released to the lavage fluid will be diluted by the large lavage
volume and that BAL fluid originates mainly from lower
airways (see above). These facts should be kept in mind when
interstitial fluid and lavage data are compared. In control
animals, we found low levels of the proinflammatory cytokines
IL-1␤, IL-6, and TNF-␣ in samples from serum, trachea
interstitial fluid, and BAL fluid (Table 2), whereas the antiinflammatory cytokine IL-10 was not detectable in any of the
fluids examined in this condition.
Induction of a systemic inflammatory response resulted in
significant changes in all of the cytokines that were measured.
In the LPS-treated animals, the levels of IL-1␤, IL-6, IL-10,
and TNF-␣ in serum, trachea interstitial fluid, and BAL samples were all increased compared with results shown in controls (Fig. 6). Of interest, we found that the average concentration of IL-1␤ was significantly higher in trachea interstitial
fluid (4.4 ng/ml) than in serum and BAL samples (0.27 and
0.07 ng/ml, respectively). Furthermore, the level of IL-6 was
104 • MARCH 2008 •
www.jap.org
Downloaded from http://jap.physiology.org/ by 10.220.33.1 on June 15, 2017
COP in Trachea Interstitial Fluid
Innovative Methodology
815
ISOLATION OF TRACHEA INTERSTITIAL FLUID IN RATS
Table 1. Identification of abundant proteins found in plasma and interstitial fluid
Peak No.
Abundant Protein
Molecular Mass, kDa
Stokes Radius, nm
IF/P
IgM
␣2-Macroglobulin
Fibrinogen
Murinoglobulin 1
Haptoglobin (trimer)
⬎600
⬎600
400–600
300–400
200–300
⬎8.0
⬎8.0
6.5–8.0
5.6–6.5
5.1–5.6
6.86⫾3.67
0.47⫾0.12
0.62⫾0.46
0.25⫾0.06
0.65⫾0.33
6
IgG
Complement C4
Inter ␣-trypsin inhibitor
Aldolase*
140–200
4.5–5.1
0.56⫾0.21
7
Ceruloplasmin
Complement C3, C5
␣1-B glycoprotein
Plasminogen
GAPDH*
Transketolase*
␣-Enolase*
100–140
3.9–4.5
1.15⫾0.21
8
Rat serum albumin
Transferrin
␣1-Antitrypsin
Hemopexin
Complement C9
60–100
3.2–3.9
0.84⫾0.20
9
Actin ␤- and ␥1-cytoplasmic*
Hemoglobin
Carbonic anhydrase*
␣2-Heparan sulfate-glycoprotein
Apolipoproteins A, E, and H
Vitamin D-binding protein
Transthyretin
⬍60
⬍3.2
NA
Values for IF/P are means ⫾ SE; n ⫽ 6 rats. Peak numbers refer to corresponding numbers in Fig 3. Apparent molecular mass and Stokes radius were estimated
by size exclusion chromatography. IF/P, interstitial fluid-to-plasma protein concentration ratios; GAPDH, glycerylaldehyde 3-phosphatedehydrogenase; NA, not
applicable. *Trachea fluid-specific protein.
increased in trachea interstitial fluid, serum, and BAL samples,
averaging 87.1, 46.3, and 1.4 ng/ml, respectively. Again, the
increase in trachea interstitial fluid was significantly higher
than that in serum, suggesting that there was a local production
of these inflammatory cytokines in trachea in this condition. Of
interest, the control-to-LPS ratio between the concentrations of
the two proinflammatory cytokines produced locally (IL-1␤,
and IL-6) was practically identical in trachea interstitial and
BAL fluid (62.8 and 62.2, respectively), suggesting a similar
dilution and thus response pattern for these two cytokines in
upper and lower airways.
For TNF-␣, another of the proinflammatory cytokines that were
assayed, the expression pattern was somewhat different. Significantly
higher mean concentrations of this cytokine were found in trachea
interstitial fluid, serum, and BAL samples in LPS-treated animals
(1.7, 4.5, and 1.3 ng/ml, respectively) than in controls. In contrast to
Fig. 4. Concentration of total protein and albumin in plasma (solid bar),
trachea interstitial fluid (IF; open bar), and trachea surface liquid (SL; hatched
bar) estimated using HPLC (n ⫽ 6). Values are means ⫾ SE. *P ⬍ 0.05
compared with corresponding plasma value.
Fig. 5. Colloid osmotic pressure in plasma (COPp) and trachea interstitial fluid
(COPif) isolated by centrifugation at 239 g for 10 min in control situation (n ⫽
5) and after overhydration (n ⫽ 5) induced by infusion of saline corresponding
to 15% of body weight. Values are means ⫾ 1 SE. *P ⬍ 0.05 compared with
corresponding control value; §P ⬍ 0.05 compared with corresponding plasma
value.
J Appl Physiol • VOL
104 • MARCH 2008 •
www.jap.org
Downloaded from http://jap.physiology.org/ by 10.220.33.1 on June 15, 2017
1
2
3
4
5
Innovative Methodology
816
ISOLATION OF TRACHEA INTERSTITIAL FLUID IN RATS
Table 2. Cytokine levels in serum of control rats and in
trachea IF and bronchoalveolar lavage fluid
Immunostaining
Cytokine Level, ng/ml
IL-1␤
IL-6
TNF-␣
Serum (n ⫽ 10 rats)
Trachea IF (n ⫽ 5 rats)
BAL (n ⫽ 5 rats)
0.011⫾0.003
0.342⫾0.096
0.025⫾0.009
0.152⫾0.023*
0.259⫾0.047
0.146⫾0.012*
0.011⫾0.001†
0.040⫾0.012*†
0.053⫾0.012
Values are means ⫾ SE. IF, interstitial fluid; BAL, bronchoalveolar lavage
fluid. *P ⬍ 0.05 compared with serum; †P ⬍ 0.05 compared with trachea IF.
Fig. 6. Concentration of cytokines (IL-1␤, IL-6, IL-10,
TNF-␣) in trachea interstitial fluid (■), serum (), and
bronchoalveolar lavage (BAL; F) samples in experimental endotoxemia induced by intravenous injection of
LPS (3.0 mg/kg). Experiment ended 180 min after LPS
administration. Controls (n ⫽ 10) received vehicle
alone and were kept for 180 min. Serum (n ⫽ 10),
trachea interstitial fluid (n ⫽ 5), and BAL samples (n ⫽
5) from control animals exhibited low or undetectable
levels of cytokines. Except for IL-10 in BAL, LPS
resulted in significant increases in all cytokines in serum, interstitial fluid, and BAL fluid. Values are
means ⫾ SE. *P ⬍ 0.05 compared with experimental
serum, and §P ⬍ 0.05 compared with experimental
BAL (ANOVA).
J Appl Physiol • VOL
104 • MARCH 2008 •
www.jap.org
Downloaded from http://jap.physiology.org/ by 10.220.33.1 on June 15, 2017
what was found for IL-1␤ and IL-6, the concentration of TNF-␣ in
serum was significantly higher than in trachea interstitial fluid.
Clearly, LPS treatment also resulted in an activation of the
anti-inflammatory cytokine IL-10. The level of IL-10 in BAL
samples was detectable, averaging 0.02 ng/ml. Somewhat
higher mean levels of 0.36 ng/ml were found in trachea
interstitial fluid, whereas a significantly higher mean concentration of this cytokine of 0.82 ng/ml was found in serum.
Smears made from trachea interstitial fluid showed that the
fluid contained scattered granulocytes and mononuclear cells,
but the number was insufficient for differential counting. A low
cell count was to be expected because of the fluid isolation
procedure and because a possible pellet was avoided during
sample removal from the tube.
Because the analysis of cytokines in the interstitial fluid
suggested that some cytokines were locally produced, we
aimed at verifying the results by immunostaining and possibly
revealing the cellular source of cytokines. Therefore, we performed double staining with IL-1␤ and W3/13 or ED1. IL-1␤
was chosen because it was the most differentially expressed
cytokine in trachea interstitial fluid and serum. Control trachea
exhibited normal tissue architecture, whereas LPS exposure
caused flattening of the epithelium and enlargement of the
tracheal glands in the submucosa. There was some staining
for IL-1␤ in the tracheal epithelial cells in the control rats;
however, after LPS treatment, the staining for this cytokine
was markedly increased in the lamina propria and submucosa (Fig. 7, B and F). Most of the cytokine-positive cells in
these areas were also ED1 positive (Fig. 7, D and F) or W3/13
positive (data not shown), indicating that macrophages, T
lymphocytes, and cells of the granulocyte lineage (most likely
neutrophils) were producing IL-1␤. In addition, epithelial cells
and chondrocytes in the hyaline cartilage exhibited staining for
IL-1␤ in the LPS-treated rats (Fig. 7, B and F), and they may
both have contributed together with cells in lamina propria and
submucosa to the observed increase in local production of
IL-1␤ after LPS challenge.
All of the negative controls showed a lack of specific
immunostaining.
Innovative Methodology
ISOLATION OF TRACHEA INTERSTITIAL FLUID IN RATS
817
DISCUSSION
Evaluation of the Centrifugation Method
In the present study, we asked whether it was possible to
isolate interstitial fluid from trachea and have shown that tissue
fluid samples can be obtained by exposing isolated trachea
tissue to increased G forces by centrifugation. We have evaluated the method with respect to potential contamination of
fluid from cellular sources, as well as trachea surface fluid, and
found that the isolated fluid is representative for tracheal
interstitial fluid at G forces ⱕ424 g. At higher G forces, a
dilution from cellular fluid is likely to occur. Below, we will
discuss the foundation of our conclusions and the implications
for fluid balance studies in the trachea.
Although we were able to isolate some fluid from the trachea
by centrifugation, the major question was to provide convincing evidence that this fluid derives from the extracellular fluid
phase. By determining the ratio of the extracellular tracer
51
Cr-EDTA in isolated trachea fluid and plasma, we could also
determine the origin of the centrifuged fluid. Addition of
intracellular fluid not containing 51Cr-EDTA would thus dilute
the centrifugate and make the centrifugate-to-plasma ratio
⬍1.0. In contrast, a centrifugate-to-plasma ratio of ⬎1.0 would
indicate that some fluid not containing tracer had been removed
from the centrifugate, e.g., by cell swelling or evaporation. We
J Appl Physiol • VOL
104 • MARCH 2008 •
www.jap.org
Downloaded from http://jap.physiology.org/ by 10.220.33.1 on June 15, 2017
Fig. 7. Effect of LPS on IL-1␤ production in rat
trachea. Shown are photomicrographs of saggital
tracheal sections double stained with IL-1␤ (red)
and ED1 (green; nuclei are counterstained with
blue; DAPI). There is almost no staining for IL-1␤
in the control rat (A). A few tissue macrophages
reside in the lamina propria (C), and the tissue
architecture appears normal (A, C, E). Exposure to
LPS induced flattening of the epithelial cells and
increased staining for IL-1␤ in the lamina propria
and submucosa (B, F) and increased infiltration
with ED1-positive cells in the same areas (D, F).
Thin arrows indicate ED1-positive cells producing
IL-1␤ (B, D). Furthermore, chondrocytes in the
tracheal cartilage were positively stained for IL-1␤
after LPS exposure (thick arrows; B), indicating
that they may have contributed to the local production of IL-1␤. DAPI, 4⬘,6-diamidino-2-phenylindole; lu, lumen; ep, epithelial cells; sub, submucosa; hc, hyaline cartilage.
Innovative Methodology
818
ISOLATION OF TRACHEA INTERSTITIAL FLUID IN RATS
J Appl Physiol • VOL
results, suggest that intracellular fluid without tracer accounted
for the fluid that was added to the centrifugate discussed above.
During perturbations in tissue fluid balance, the composition
of the interstitial fluid will change, and we could demonstrate
that such changes occurring during overhydration and systemic
inflammation were reflected in the isolated trachea interstitial
fluid. Together, the tracer experiments, HPLC data, and induced interventions all suggest that samples representative for
interstitial fluid can be isolated by centrifugation of trachea.
The isolated fluid may also serve as a substrate for studies of
biomarkers originating in the airways.
Comparison With Previous Studies
To our knowledge, there are no previous studies in which
interstitial fluid was isolated from trachea in control conditions
and in rats. Of major relevance, Nordin et al. (23) studied fluid
balance in rabbit tracheal mucosa. To expose the microcirculation and to immobilize the tissue, they made a longitudinal
cut in trachea and clamped the tissue to the table. This procedure generated an inflammation that was actually intended and
resulted in a significant edema. Under these conditions, they
were able to isolate lymph by direct puncture of terminal
lymphatics and from protein concentration calculated an average COP in lymph of 19 mmHg, i.e., 91% of the plasma COP
of 21 mmHg. Whether these values differed significantly is not
known. They pointed out that a more gentle tissue handling
resulted in a lower protein concentration in the isolated lymph,
suggesting that the high protein concentration in lymph was the
result of tissue manipulation. Although the experimental conditions differed, our finding of an interstitial fluid COP in
control situation 84% of that in plasma corresponds well to
their data and may suggest that there is a high interstitial fluid
protein concentration in control situation and that inflammation
results in an increased filtration of water and protein (see
Physiological Implications of the Study).
Although there are no other studies on trachea from euhydrated animals, Normandin et al. (24) were able to isolate fluid
from the peribronchial-perivascular space in normally hydrated
dogs using wick catheters, but the volume obtained in the
control situation (⬍4 ␮l) was insufficient for analysis in their
system. For the same tissue, Negrini et al. (22) were able to
isolate fluid from the same space in euhydrated rabbits. They
found a perivascular fluid protein concentration 58% of that in
plasma, which was even lower in peribronchial fluid. The
somewhat lower values may suggest that there are regional
differences in interstitial fluid protein concentration and thus
interstitial fluid COP in respiratory tissue. For lung tissue,
lymph has been sampled and used as representative for interstitial fluid (e.g., Refs. 26, 32). There may be fluid exchange in
the lymph nodes; therefore, afferent lymph will be most representative for lung interstitial fluid (13, 20, 25). From these
studies, a protein concentration ratio between interstitial fluid
and plasma of ⬃0.7 is found, i.e., somewhat lower than our
corresponding ratio between COP in the two compartments of
0.85. Still, these observations suggest that the interstitial fluid
protein concentration is high in respiratory tissues compared
with other tissues (for references, see Ref. 4).
BAL has been used extensively to monitor various processes
occurring in respiratory tissue, mainly of inflammatory origin.
BAL has its obvious advantages because of its versatility and
104 • MARCH 2008 •
www.jap.org
Downloaded from http://jap.physiology.org/ by 10.220.33.1 on June 15, 2017
found a centrifugate-to-plasma ratio of 0.94 at 239 and at 424
g, with both ratios significantly different from 1.0. These data
suggest that ⬃6% of the isolated fluid is of intracellular origin,
possibly deriving from erythrocytes, connective tissue cells,
and/or epithelial cells damaged during centrifugation. This
assumption is supported by the finding of intracellular proteins
(e.g., actin) in the interstitial fluid. Notwithstanding the fact
that the passage of tracers applied to the trachea surface fluid
to centrifugate was low, addition of surface fluid will add to the
interstitial fluid dilution effect because such fluid was not fully
equilibrated with respect to 51Cr-EDTA after 2 h of circulation
time. We tried to minimize this problem by reducing the
applied G force and optimizing the tissue handling, e.g., by
rapid processing of the tissue to avoid degradation. Exceeding
424 g resulted in an even lower centrifugate-to-plasma ratio,
indicating that an increased dilution occurred. This observation
is probably a result of further extrusion of cellular fluid and/or
cell damage and shows that a G force ⱕ424 g should be used
for tracheal fluid isolation, in agreement with previous data for
skin and tumors (38).
It is possible that trachea surface fluid may be part of the
fluid sampled by centrifugation, and we thus attempted to
quantify surface fluid admixture. There are several reasons for
us to conclude that such admixture or contamination is negligible. First, the total volume of trachea surface fluid is small
compared with the interstitial fluid volume. The average trachea sample weight was 47 mg, and with an average extracellular fluid fraction of 38% this results in a total extracellular
fluid volume in our trachea samples of 17.9 ␮l assuming a
density of 1.0 mg/␮l. We estimated that the area of the inner
surface of our trachea sample averaged 194 mm2. The thickness of the surface layer is 5–10 ␮m (45), suggesting a surface
fluid volume of 10 –20 ␮l, i.e., ⬃10% of the extracellular fluid
volume. Second, we measured the passage of tracer from
surface fluid to centrifugate and found a very low recovery
(e.g., 1% of the applied dose of HSA). In addition, we estimated the total protein, albumin, and IgG concentrations in
trachea surface fluid and found these to be 10 –15% of the
concentration in plasma. Thus a combined low-volume fraction
compared with total interstitial fluid, low passage of albumin to
the centrifugate, and low protein concentration will result in a
negligible amount of trachea surface fluid and proteins in the
fluid isolated by centrifugation. Our observed albumin concentration in trachea surface fluid of ⬃10% that in plasma is
higher than that found in humans (16), a difference that may be
due to species differences and experimental conditions.
If we accept that the pattern for tissue eluate is representative
for extracellular fluid, the comparison of the pherograms of
centrifugate and eluate suggests that the addition of possible
intracellular proteins is very low. To show the sensitivity of our
approach to detect intracellular protein contamination to the
centrifugate, we made extracts of crushed trachea. These extracts will, in addition to extracellular fluid-phase proteins,
contain a substantial amount of intracellular proteins. That this
phenomenon is reflected in the pherograms is shown by Fig.
2D. Whereas trachea fluid and eluate had a HPLC pattern
similar to that of plasma, except for a lower content of large
macromolecules, trachea lysate had a much higher fraction of
low-molecular-mass macromolecules. This is to be expected
because a large fraction of the latter consists of intracellular
proteins. These experiments, together with the 51Cr-EDTA
Innovative Methodology
ISOLATION OF TRACHEA INTERSTITIAL FLUID IN RATS
because samples are representative for more distal airways
where the consequences of edema are more dramatic for
airflow than in trachea. Here, we used this technique in LPSinduced inflammation to compare it with the centrifugation
technique with respect to monitoring of inflammatory mediators. Because of the nature of the BAL technique, i.e., a large
dilution volume, the concentration of inflammatory mediators
was substantially higher in trachea centrifugate. Of note, the
cytokine response pattern on LPS stimulation was, however,
similar for the two techniques. This finding suggests that the
response pattern is similar for upper and lower airways, again
suggesting that signaling responses recorded in trachea interstitial fluid are representative for lower airways. There may be
qualitative differences with respect to other proteins found in
the two fluids, but this was not studied here. This question will
be of importance for proteomic studies to identify mediators in
such fluid, for example (34).
The findings of COP levels in trachea interstitial fluid that
were 85% of that in plasma in control situation, corroborated
by the protein concentration data, and 82% after an induced
overhydration were somewhat unexpected. This is a significantly higher fraction than that of ⬃0.7 observed in lung
lymph, as discussed above, as well as skin and muscle measured
previously (42, 43) and gives information on fluid handling in this
organ. A high interstitial fluid COP during control conditions may
be a consequence of a very large capillary surface area in trachea
(35, 36), allowing for a substantial diffusive transport of proteins. A possible low net capillary filtration will add to this
effect (31). Whether such effects are sufficient to explain why
an increased filtration induced by overhydration did not reduce
lymph or interstitial fluid to plasma protein concentration ratio
as observed in several other tissues (31) is not known. Although
the filtration is increased, a large surface area will contribute to a
significant diffusion during overhydration also. Furthermore, the
profuse lymphatic network (5) may contribute to a very effective
removal of filtered fluid and proteins in this situation and to the
maintenance of the relatively high interstitial fluid COP.
Interstitial fluid COP is one of the determinants of fluid
filtration across the capillary wall. The importance, however,
of interstitial fluid COP as measured here (“global” interstitial
fluid COP) as a determinant of transcapillary fluid balance has
been questioned because there may be significant protein
gradients at the filtration barrier. This was initially suggested
and shown for fenestrated capillaries (15, 17). Recently, Adamson et al. (1) verified that the same phenomenon also applies
to nonfenestrated mesenteric capillaries in rats. These gradients
are most pronounced during states of high filtration, whereas,
at normal filtration, their model predicts that the effective COP
is ⬃80% of that in the global interstitial fluid phase (i.e.,
interstitial fluid COP as measured in our study). Trachea has
continuous or nonfenestrated capillaries, and the same phenomenon may apply here. Still, there is no doubt that this
parameter is one of the predictors for the edema-preventive
capacity of a tissue (4, 14). Previous studies by us in rat skin
and muscle have shown that a decrease in interstitial fluid COP
could counteract an increase in capillary pressure or a reduction in plasma COP of 8 –10 mmHg (11, 29). Although we do
not know at what interstitial fluid COP an edema will develop
J Appl Physiol • VOL
in trachea, the high interstitial fluid COP levels in control
situations suggest that the interstitium is able to compensate for
a significantly higher increase in net filtration pressure in
trachea than in skin and muscle. Because avoidance of edema
in the airways is of vital importance, the implication of this
observation may be significant.
One significant finding of our study was the demonstration
of a substantial local production of pro- and anti-inflammatory
cytokines in trachea during a systemic inflammation induced
by LPS. This effect was especially pronounced for IL-1␤,
which increased dramatically and was 16 times higher than the
corresponding concentration in serum after LPS treatment,
whereas IL-6 was twice that of serum in the same situation. A
higher concentration in the tissue than in serum must mean that
these cytokines are produced locally, an assumption that is
supported by immunohistochemistry showing a significant increase in IL-1␤- and IL-6-producing cells in the trachea interstitium. It is possible that the epithelium also contributes to
local cytokine production because the airway epithelium has
been shown to produce inflammatory cytokines on LPS stimulation (8, 28) and is also indicated by our immunohistochemistry data. Although the response to an inflammatory mediator
may be complex and dependant on the species, the primary
response to all of the main inflammatory mediators seems to be
vasodilatation, resulting in thickening of the mucosa with
edema and exudate as the final result (35, 36). Of note, the
close association between cytokine-expressing cells and tracheal glands during LPS treatment suggests a stimulatory effect
on mucous secretion that will add to swelling of the trachea.
Activated macrophages represent the major source of IL-1.
These cells have a role in defense mechanisms following
exogenous invasion of pathogenic stimuli and are most likely
the origin for the observed IL-1␤ increase. They will also
recruit monocytes and neutrophils when stimulated with LPS,
enhancing the inflammatory reaction (19). TNF-␣ was higher
in serum than in trachea fluid, in line with observations in skin
(21). The liberated proinflammatory cytokines will induce
local vascular leakage and edema (9, 18). LPS treatment,
however, did not result in any edema in trachea in our experiments. A likely explanation for this observation is that LPS
resulted in reduced capillary pressure and blood flow (6), and
increased permeability alone is insufficient to cause edema
unless capillary pressure is elevated (12).
The simultaneous increase in tissue fluid and serum IL-10
demonstrates the well-known stimulatory effect of LPS on the
production of this anti-inflammatory cytokine (2). Furthermore, a higher level of IL-10 in serum than in trachea fluid
suggests that this cytokine partly or fully originates from
serum, a finding that is reasonable considering the preferential
effect of LPS on IL-10 production in liver (2). Together, the
present method for trachea fluid isolation may give significant
new information regarding the process of edema generation
during systemic inflammation caused by LPS.
In conclusion, we have shown that fluid can be isolated from
trachea by centrifugation of excised tissue. We have validated
that the isolated fluid is representative for trachea interstitial
fluid providing that G forces ⱕ424 g are used. Our experiments
show that the trachea interstitial fluid has a relatively high
protein concentration and therefore high COP relative to
plasma, again showing a significant potential for prevention of
fluid accumulation and edema generation. During LPS-induced
104 • MARCH 2008 •
www.jap.org
Downloaded from http://jap.physiology.org/ by 10.220.33.1 on June 15, 2017
Physiological Implications of the Study
819
Innovative Methodology
820
ISOLATION OF TRACHEA INTERSTITIAL FLUID IN RATS
systemic inflammation, there is a significant local production
of proinflammatory mediators that may contribute to edema
formation in trachea as well as lower airways. The present
method thus represents a useful tool for studies of the microcirculation in the airways and may also be used in the search
for airway specific biomarkers.
20.
ACKNOWLEDGMENTS
22.
We thank Odd Kolmannskog, Gerd Signe Salvesen, and Bartlomiej Janaczyk for expert technical assistance.
23.
GRANTS
24.
Financial support from Western Norway Regional Health Authority, The
Research Council of Norway and Faculty of Medicine, University of Bergen
(Locus on Circulatory Research) is gratefully acknowledged.
25.
REFERENCES
26.
J Appl Physiol • VOL
27.
28.
29.
30.
31.
32.
33.
34.
35.
36.
37.
38.
39.
40.
41.
42.
43.
44.
45.
104 • MARCH 2008 •
www.jap.org
Downloaded from http://jap.physiology.org/ by 10.220.33.1 on June 15, 2017
1. Adamson RH, Lenz JF, Zhang X, Adamson GN, Weinbaum S, Curry
FE. Oncotic pressures opposing filtration across non-fenestrated rat microvessels. J Physiol 557: 889 –907, 2004.
2. Asadullah K, Sterry W, Volk HD. Interleukin-10 therapy—review of a
new approach. Pharmacol Rev 55: 241–269, 2003.
3. Aukland K. Distribution volumes and macromolecular mobility in rat tail
tendon interstitium. Am J Physiol Heart Circ Physiol 260: H409 –H419,
1991.
4. Aukland K, Reed RK. Interstitial-lymphatic mechanisms in the control of
extracellular fluid volume. Physiol Rev 73: 1–78, 1993.
5. Baluk P, Tammela T, Ator E, Lyubynska N, Achen MG, Hicklin DJ,
Jeltsch M, Petrova TV, Pytowski B, Stacker SA, Yla-Herttuala S,
Jackson DG, Alitalo K, McDonald DM. Pathogenesis of persistent
lymphatic vessel hyperplasia in chronic airway inflammation. J Clin Invest
115: 247–257, 2005.
6. Bletsa A, Berggreen E, Fristad I, Tenstad O, Wiig H. Cytokine
signalling in rat pulp interstitial fluid and transcapillary fluid exchange
during lipopolysaccharide-induced acute inflammation. J Physiol 573:
225–236, 2006.
7. Boucher RC. Human airway ion transport. Part one. Am J Respir Crit
Care Med 150: 271–281, 1994.
8. Cheng DS, Han W, Chen SM, Sherrill TP, Chont M, Park GY, Sheller
JR, Polosukhin VV, Christman JW, Yull FE, Blackwell TS. Airway
epithelium controls lung inflammation and injury through the NF-kappa B
pathway. J Immunol 178: 6504 – 6513, 2007.
9. Daffonchio L, Novellini R, Bertuglia S. Protective effect of ketoprofen
lysine salt on interleukin-1␤ and bradykinin induced inflammatory changes in
hamster cheek pouch microcirculation. Inflamm Res 51: 223–228, 2002.
10. Erjefalt I, Persson CG. On the use of absorbing discs to sample mucosal
surface liquids. Clin Exp Allergy 20: 193–197, 1990.
11. Fadnes HO. Effect of increased venous pressure on the hydrostatic and
colloid osmotic pressure in subcutaneous interstitial fluid in rats: edemapreventing mechanisms. Scand J Clin Lab Invest 36: 371–377, 1976.
12. Gabel JC, Hansen TN, Drake RE. Effect of endotoxin on lung fluid
balance in unanesthetized sheep. J Appl Physiol 56: 489 – 494, 1984.
13. Grimbert FA, Martin D, Parker JC, Taylor AE. Lymph flow during
increases in pulmonary blood flow and microvascular pressure in dogs.
Am J Physiol Heart Circ Physiol 255: H1149 –H1155, 1988.
14. Guyton AC, Granger HJ, Taylor AE. Interstitial fluid pressure. Physiol
Rev 51: 527–563, 1971.
15. Hu X, Adamson RH, Liu B, Curry FE, Weinbaum S. Starling forces
that oppose filtration after tissue oncotic pressure is increased. Am J
Physiol Heart Circ Physiol 279: H1724 –H1736, 2000.
16. Knowles MR, Robinson JM, Wood RE, Pue CA, Mentz WM, Wager
GC, Gatzy JT, Boucher RC. Ion composition of airway surface liquid of
patients with cystic fibrosis compared with normal and disease-control
subjects. J Clin Invest 100: 2588 –2595, 1997.
17. Levick JR. An analysis of the interaction between interstitial plasma
protein, interstitial flow, and fenestral filtration and its application to
synovium. Microvasc Res 47: 90 –125, 1994.
18. Martin S, Maruta K, Burkart V, Gillis S, Kolb H. IL-1 and IFN-␥
increase vascular permeability. Immunology 64: 301–305, 1988.
19. Maus UA, Koay MA, Delbeck T, Mack M, Ermert M, Ermert L,
Blackwell TS, Christman JW, Schlondorff D, Seeger W, Lohmeyer J.
21.
Role of resident alveolar macrophages in leukocyte traffic into the alveolar
air space of intact mice. Am J Physiol Lung Cell Mol Physiol 282:
L1245–L1252, 2002.
McClure DE, Weidner WJ. Comparison of afferent and efferent lung
lymph in the sheep. J Appl Physiol 64: 2340 –2347, 1988.
Nedrebo T, Reed RK, Jonsson R, Berg A, Wiig H. Differential cytokine
response in interstitial fluid in skin and serum during experimental inflammation in rats. J Physiol 556: 193–202, 2004.
Negrini D, Passi A, Bertin K, Bosi F, Wiig H. Isolation of pulmonary
interstitial fluid in rabbits by a modified wick technique. Am J Physiol
Lung Cell Mol Physiol 280: L1057–L1065, 2001.
Nordin U, Kallskog O, Lindholm CE, Wolgast M. Transvascular fluid
exchange in the tracheal mucosa. Microvasc Res 15: 287–298, 1978.
Normandin D, Tung H, Hargens AR, Peters RM. Sampling of lung
interstitial fluid in intact dog. J Surg Res 48: 91–98, 1990.
Parker JC, Crain M, Grimbert F, Rutili G, Taylor AE. Total lung
lymph flow and fluid compartmentation in edematous dog lungs. J Appl
Physiol 51: 1268 –1277, 1981.
Parker RE, Wickersham NE, Roselli RJ, Harris TR, Brigham KL.
Effects of hypoproteinemia on lung microvascular protein sieving and
lung lymph flow. J Appl Physiol 60: 1293–1299, 1986.
Persson CG, Erjefalt JS, Greiff L, Andersson M, Erjefalt I, Godfrey
RW, Korsgren M, Linden M, Sundler F, Svensson C. Plasma-derived
proteins in airway defence, disease and repair of epithelial injury. Eur
Respir J 11: 958 –970, 1998.
Poynter ME, Irvin CG, and Janssen-Heininger YM. A prominent role
for airway epithelial NF-␬B activation in lipopolysaccharide-induced
airway inflammation. J Immunol 170: 6257– 6265, 2003.
Reed RK. Interstitial fluid volume, colloid osmotic and hydrostatic pressures in rat skeletal muscle. Effect of venous stasis and muscle activity.
Acta Physiol Scand 112: 7–17, 1981.
Serikov VB, Jang YJ, Widdicombe JH. Estimate of the subepithelial
hydrostatic pressure that drives inflammatory transudate into the airway
lumen. J Appl Physiol 92: 1702–1708, 2002.
Taylor A, Granger D. Exchange of macromolecular substances across the
capillary wall. In: Handbook of Physiology. The Cardiovascular System.
Microcirculation. Bethesda, MD: Am. Physiol. Soc., 1984, vol. IV, pt. I,
chapt. 11, p. 467–520.
Vreim CE, Snashall PD, Staub NC. Protein composition of lung fluids in
anesthetized dogs with acute cardiogenic edema. Am J Physiol 231:
1466 –1469, 1976.
Wanner A, Salathe M, O’Riordan TG. Mucociliary clearance in the
airways. Am J Respir Crit Care Med 154: 1868 –1902, 1996.
Wattiez R, Falmagne P. Proteomics of bronchoalveolar lavage fluid.
J Chromatogr B Analyt Technol Biomed Life Sci 815: 169 –178, 2005.
Widdicombe J. The airway vasculature. Exp Physiol 78: 433– 452, 1993.
Widdicombe J. The tracheobronchial vasculature: a possible role in
asthma. Microcirculation 3: 129 –141, 1996.
Wiggs BR, Bosken C, Pare PD, James A, Hogg JC. A model of airway
narrowing in asthma and in chronic obstructive pulmonary disease. Am
Rev Respir Dis 145: 1251–1258, 1992.
Wiig H, Aukland K, Tenstad O. Isolation of interstitial fluid from rat
mammary tumors by a centrifugation method. Am J Physiol Heart Circ
Physiol 284: H416 –H424, 2003.
Wiig H, Berggreen E, Borge BA, Iversen PO. Demonstration of altered
signaling responses in bone marrow extracellular fluid during increased
hematopoiesis in rats using a centrifugation method. Am J Physiol Heart
Circ Physiol 286: H2028 –H2034, 2004.
Wiig H, Gyenge CC, Tenstad O. The interstitial distribution of macromolecules in rat tumours is influenced by the negatively charged matrix
components. J Physiol 567: 557–567, 2005.
Wiig H, Halleland EG, Fjaertoft M, Aukland K. Measurement of
colloid osmotic pressure in submicrolitre samples. Acta Physiol Scand
132: 445– 452, 1988.
Wiig H, Heir S, Aukland K. Colloid osmotic pressure of interstitial fluid
in rat subcutis and skeletal muscle: comparison of various wick sampling
techniques. Acta Physiol Scand 133: 167–175, 1988.
Wiig H, Sibley L, DeCarlo M, Renkin EM. Sampling interstitial fluid
from rat skeletal muscles by intermuscular wicks. Am J Physiol Heart Circ
Physiol 261: H155–H165, 1991.
Wilson J. The bronchial microcirculation in asthma. Clin Exp Allergy 30,
Suppl 1: 51–53, 2000.
Yoneda K. Mucous blanket of rat bronchus: an ultrastructural study. Am
Rev Respir Dis 114: 837– 842, 1976.