Basic Res Cardiol 98: 295 303 (2003)

DOI 10.1007/s00395-003-0419-6

Martyn P. Kingsbury
Wenxin Huang
J. Leo Donnelly
Emma Jackson
Emma Needham
Mark A. Turner
Desmond J. Sheridan

Received: 12 November 2002

Returned for 1. revision: 11 December 2002
1. Revision received: 5 March 2002
Returned for 2. revision: 21 March 2002
2. Revision received: 16 April 2003
Accepted: 28 April 2003
Published online: 16 May 2003

Dr. M. Kingsbury () W. Huang

J. L. Donnelly E. Jackson E. Needham
M. A. Turner D. J. Sheridan
Academic Cardiology Unit
Imperial College School of Medicine
St. Marys Campus
10th oor QEQM Wing
South Wharf Road
London W2 1NY, UK
Tel.: +44-20/7886-6233
Fax: +44-20/7886-6732
E-Mail: m.kingsbury@imperial.ac.uk

ORIGINAL CONTRIBUTION

Structural remodelling of lungs

in chronic heart failure

Abstract In order to determine whether morphological changes could

account for a previously reported reduction in pulmonary capillary ltration
in heart failure, we studied pulmonary morphology in lungs from a guineapig chronic heart failure model. Heart failure was induced by banding the
ascending aorta with sham operated animals serving as controls; all animals
were studied at 158 6 days post-operation. Following banding, a reduction
in aortic ow, increased peripheral vascular resistance and raised left ventricular end diastolic, left atrial and right ventricular pressures together with
increased right ventricle to body weight ratio (all p < 0.05) are indicative of
established pulmonary hypertension and heart failure. This was associated
with an increase in pulmonary septal volume fraction (38.1 3.1% vs 24.6
2.3 %, p < 0.01) and reticulin fibre density. There was also evidence of
siderophage infiltration and examination of pulmonary ultra structure
revealed a signicantly thicker alveolar-capillary barrier in heart failure (1278
76 vs 638 32 nm, p < 0.001), thickening of both the alveolar (89%, p < 0.01)
and capillary (69%, p < 0.05) basal laminae with pericyte and collagen inltration of the alveolar-capillary barrier. We hypothesise that these pulmonary
adaptations provide protection from oedema formation, but whilst initially
protective, are also likely to confer major long-term disadvantages in chronic
heart failure.

Key words Heart failure pulmonary circulation membrane permeability
remodelling extracellular matrix

Introduction

BRC 419

Heart failure is a common condition (25) that has a poor

prognosis, with only 30% surviving for 5 years from initial diagnosis (13). The pulmonary pathophysiology
associated with heart failure is important, as not only is
dyspnoea and impaired gas exchange characteristic of
heart failure, pulmonary oedema is often the initial
major clinical manifestation and can be fatal. Paradoxically, although congestive heart failure is associated with
chronic elevation of pulmonary capillary pressure with

the risk of pulmonary oedema, patients with longstanding heart failure are less prone to pulmonary oedema for
given levels of pulmonary venous pressure. Thus, elevation of pulmonary venous pressure in patients with acute
myocardial infarction frequently results in frank pulmonary oedema (20) while patients with chronic rheumatic mitral valve disease may tolerate even higher levels of pulmonary venous pressure without developing
oedema (36). Indirect measurement of pulmonary
microvascular permeability suggests that this is reduced
in patients with chronic heart failure (8). Furthermore,
direct measurement of pulmonary microvascular per-

296

Basic Research in Cardiology, Vol. 98, No. 5 (2003)

Steinkopff Verlag 2003

meability in a canine rapid pacing heart failure model

showed while this was normal at resting pulmonary
venous pressures, there was a signicant increase in the
threshold for pressure induced increases in permeability
in failure (34). We have recently demonstrated a signicant reduction in pulmonary capillary ltration in lungs
adapted to chronic heart failure (14), and found this
reduction in pulmonary ltration conferred protection
against volume loading induced pulmonary oedema in
anaesthetised animals (19).
Our studies have demonstrated an increase in arterial
and venous compartment resistances that could conceivably attenuate transmission of elevated pressure
waves to the pulmonary microcirculation (14). While this
could protect the pulmonary microcirculation it would
also increase the right ventricular burden with long-term
disadvantages. Preliminary histological analysis of lungs
adapted to chronic heart failure in our model showed
pulmonary vascular remodelling (14) similar to that
known to occur in pulmonary hypertension (10) and
hypoxia (37). This may be associated with altered vasomotor function, which also occurs in left ventricular dysfunction (9) and heart failure (18, 35). There was also
some evidence of thickening of the interalveolar septa
(14). As water transport across the pulmonary microvacular endothelium is predominantly via a paracellular
route (5), this thickening may form an increased barrier
to water transudation. Thickening of the pulmonary capillary basement membrane has been reported in chronic
heart failure (21) and could act to limit both pulmonary
oedema and gas exchange.
In order to investigate the pulmonary morphological
changes in more detail we used a model of chronic heart
failure in guinea pigs. Lung morphology was studied ve
months following banding of the ascending aorta. We
report parenchymal remodelling with alterations in cell
content and reticulin deposition. The alveolar capillary
barrier was markedly increased as was both the capillary
and alveolar basal lamina. These changes have important
implications for the long-term prognosis in patients with
chronic heart failure.

space and a small high-density plastic clip with an internal diameter of 1.99 mm placed on the ascending aorta.
Following extubation, animals were given subcutaneous
injections of 0.006 mg/kg buprenorphine (Temgesic,
Reckitt & Coleman) for analgesia and allowed to recover
in a warm single cage. Sham-operated animals (n = 18)
underwent identical operative procedures but the plastic
clip was not placed on the aorta. Animals were housed
at 20 2 C with a relative humidity of 50 5% and a 13
hour light, 11 hour dark, light/dark cycle. Animals received Biosure RGP diet and fresh water ad libitum. Experiments were carried out 158 6 days post-operation in
order to study effects of chronic heart failure. All animal
work and surgery was performed in accordance with
United Kingdom legislation, the Home Ofce Guidance
on the Operation of the Animals (Scientic Procedures)
Act 1986 (Her Majestys Stationary Ofce, London).

Haemodynamic assessment
Animals (sham, n = 6 and banded, n = 6) were anaesthetised with a bolus intra-peritoneal dose (60 mg/kg) of
pentobarbitone. Heart rate was calculated from ECGs
recorded using needle electrodes inserted subcutaneously. Lead II was used to measure R wave voltage and
QRS and QTc intervals (calculated as QT interval/(R-R)
intervals). Left atrial, right ventricular and left ventricular pressures were measured by direct puncture and the
right carotid artery was cannulated to measure systemic
blood pressure using pressure transducers (SensoNor
840). Aortic ow was measured using a ow probe placed
on the descending thoracic aorta connected to a Transonic T108 ultrasonic blood flow meter. Amplified
signals (frequency response 5 kHz) were displayed on a
Lectromed recorder (frequency response 200 Hz) and
recorded and analysed using Po-Ne-Mah Acquire Plus
data acquisition software (12 bit resolution; sampling
rates were 250 Hz for arterial pressure and ow and 1 kHz
for ECGs and LV pressure).

Morphology

Materials and methods

Induction of chronic heart failure
Heart failure was induced by banding of the ascending
aorta in male Dunkin-Hartley guinea pigs (600 700 g),
as previously described (14) using a modication of the
method described by Ling and deBold (22). In brief, animals (n = 18) were anaesthetised, intubated and ventilated with 0.4 L/min O2 at a rate of 100 breaths/min. A
thoracotomy was performed in the third left intercostal

Lungs were set up for perfusion as previously described

(14). Briey, sham (n = 6) and banded (n = 6) animals
were anaesthetised, the trachea was cannulated and the
animal ventilated with room air at 70 strokes/min and a
tidal volume of 6 ml (small animal respirator, 7025, Ugo
Basile, Comerio, Italy). The pulmonary artery was cannulated and drainage facilitated by a cannula placed in
the left atrium via the left ventricle. The lungs were gently flushed with heparinised Krebs-Henseleit solution
and were immediately removed and mounted. The
isolated, ventilated lungs were perfused at 10 ml/min
with modied Krebs-Henseleit solution (pH 7.4, 37 C)

M. P. Kingsbury et al.
Structural remodelling of lungs in chronic heart failure

containing (mmol/l): 118.0 NaCl, 4.7 KCl, 1.2 MgSO4,

1.1 KH2PO4, 24 NaHCO3, 2.5 CaCl2, 9 glucose, 2 pyruvate
and 0.25% bovine serum albumin (BSA) saturated with
95 % O2/5 % CO2. This perfusion maintained pressure
within the physiological range (20.4 2.9 mmHg). All
chemicals were of analytic grade and were obtained from
Merck Ltd., Lutterworth, UK. After 10 minutes equilibration, they were perfusion xed with 10% formal saline
at 10 ml/min for 5 minutes. Tissue cubes (0.5 cm3) were
taken from a point between the bottom and middle thirds
of the right and left caudal lung lobes, cut perpendicular
to the cranio-caudal axis and immersed in the same xative for 24 hours. They were then dehydrated in graded
ethanol solutions and placed in chloroform overnight
before being wax embedded. Tissue sections (5 m) were
cut and mounted onto poly-1-lysine coated slides and
dried overnight before staining. Tissue sections were
stained with Haematoxylin and Eosin to show general
structure, Perls stain to highlight siderophages and Gordon and Sweets stain to show reticulin following standard histology protocols (2). The density of reticulin
bres was blindly assessed by ten observers who graded
the reticulin bre density on a scale of 0 to 5, having previously been shown examples of dense and sparse reticulin bres, on 5 sections of each sham and banded lung.
Alveolar septal volume fraction and siderophage density
were measured in six randomly assigned, standard sized
(2.0
105 m2) frames for each section. Mean septal volume fraction was expressed as a ratio of tissue volume
and siderophage density as the number of blue-stained
siderophages in each frame. Morphometric analysis of
the sections was carried out using an image analysis system (Seescan Solitaire Plus, Cambridge, UK).
Lungs from sham (n = 6) and banded (n = 6) animals
were isolated and set-up as previously described and

297

perfusion fixed for electron microscopic examination

with half strength Karnovskys xation buffer (10 ml/
min, 5 minutes) and rinsed in phosphate buffer. Small
1 mm3 cubes of xed tissue were washed, post xed in 1%
osmium tetroxide, 1.5% potassium ferricyanide mixture,
dehydrated in graded alcohol solutions and propylene
oxide, inltrated with three changes of Araldite over 12
hours and then embedded in fresh Araldite. Semithin
sections were cut on a Reichert Ultracut E ultramicrotome and stained with toluidine blue aqueous solution.
Thin sections (50 70 nm) were mounted on formvarcoated nder grids; stained with uranyl acetate and lead
citrate. Grid squares were examined and microvessels
photographed consecutively in a Zeiss electron microscope. Morphometric analysis of electron microscope
images was carried out using an image analysis system
(SigmaScan Pro 5.0, SPSS Science software, Birmingham,
UK).

Statistical analysis
Values were expressed as mean standard error of the
mean; percentage increase or decrease, when given in
parentheses, refers to the percentage change in mean
value. Statistical analysis of data using Students
unpaired t-test enabled the comparison of groups (an F
value calculation was also performed to test for unequal
variance between the groups. If signicant variance was
found, t-tests with Welchs correction for unequal variance were used). The reticulin bre density scores were
ranked and analysed using a Mann-Whitney U test. All
statistical analysis was performed using Prism analysis
software (v3.00, GraphPad Software Inc, San Diego, CA,
USA). In all statistical evaluations, p < 0.05 were taken as
an indication of statistical signicance.

Results

Gross morphology

Fig. 1 Organ weight as a percentage of bodyweight, values are plotted as mean

S.E.M. Sham control (n = 12), Heart failure (n = 12). There are significant increases
in heart, left ventricle (LV), right ventricle (RV) atria and lung weight ratios in heart
failure. While left (LK) and right kidney (RK) weight ratios remain unchanged.
** p < 0.01, ***p < 0.001

Following 158 6 days of aortic banding, there was evidence of severe cardiac hypertrophy with a 54% increase
in heart weight to body weight ratio compared with sham
controls (Fig. 1). This reected marked increases in the
left ventricular (47%), right ventricular (50%) and atrial
(108%) mass. In addition, lung weight to body weight
ratio was increased by over 59% (p < 0.001), while kidney weights remained unchanged. These changes represent an actual increase in organ weight as the body weight
of banded and sham control groups were not significantly different (1292 42 vs. 1368 41 g).

298

Basic Research in Cardiology, Vol. 98, No. 5 (2003)

Steinkopff Verlag 2003

Fig. 2 Representative photomicrographs of sections of lung tissue, magnification
80. A Sham control and
heart failure (B) show tissue stained
with haemotoxylin and eosin, AV Alveolus, IS Interalveolar Septum. Sham
control (C) and heart failure (D) show
tissue stained with Gordon and
Sweetss stain to visualise reticulin
fibres (RT). Sham control (E) and heart
failure (F) show tissue stained with
Perls stain for siderophages (SP)

Systemic haemodynamics
Haemodynamic data at 158 6 days are shown in Table
1. Aortic banding resulted in a reduction (49%) in aortic
ow and resulted in a systolic pressure gradient of 17.4
3.4 mmHg between the left ventricle and the carotid

artery. Increased peripheral resistance (60 %) and

increased left ventricular systolic (13%, p < 0.05) and end
diastolic pressures (188%, p < 0.01), accompanied this
pressure gradient. Left ventricular dP/dt was reduced
(40%), suggesting a reduction in LV contractility in the
banded group. There was a substantial increase in left

M. P. Kingsbury et al.
Structural remodelling of lungs in chronic heart failure

299

Table 1 Haemodynamic and electrocardiographic data from anaesthetised aortic

banded (n = 6) and sham operated animals (n = 6)
Sham control
Systolic Carotid Pressure (mmHg)
Diastolic Carotid Pressure (mmHg)
LV Systolic Pressure (mmHg)
LV End Diastolic Pressure (mmHg)
LV +dP/dt
LV dP/dt
LA Systolic Pressure (mmHg)
LA End Diastolic Pressure (mmHg)
RV Systolic Pressure (mmHg)
RV End Diastolic Pressure (mmHg)
Peripheral Vascular Resistance
(dynes s cm5)
Heart Rate (beats/min)
R-wave height (mV)
QRS interval (ms)
QTC interval (ms)

Heart failure

51.9  3.0
39.2  3.7
53.6  2.6
4.0  1.1
3404  496
2381  284
3.51  0.82
1.13  0.46
9.0  1.9
2.3  0.9

43.0  2.9
31.1  3.1
60.4  5.4
11.5  1.2**
2026  446
1460  120*
7.58  0.38**
4.00  0.47**
13.0  0.8
4.5  0.4*

81883  4504
255  5
0.76  0.08
49.83  7.05
307.5  18.5

130957  16997*
245  11
1.28  0.07***
73.72  6.83*
371.3  7.1*

Values are given as mean S.E.M; difference between banded heart failure animals
and corresponding sham control animals as indicated. *p < 0.05, **p < 0.01,
***p < 0.001

atrial systolic (116 %) and diastolic pressures (254 %),

p < 0.01 in each case and increased right ventricular end
diastolic (96%) pressures, p < 0.05. Heart rate derived
from ECGs was unchanged. Analysis of the QRS complex
showed signicantly increased R wave voltage (68%, p <
0.001) and QRS duration (48%, p < 0.05) compared with
controls. QTc intervals were also signicantly increased
(21%) in banded animals, p < 0.05.

Light microscopy
Perfusion xation resulted in well-stained tissue sections
with well supported alveolar structure. The most striking
changes observed in heart failure lungs were marked
increase in tissue cellularity and an increase in septal
thickness (Fig. 2). This was conrmed by morphometric
analysis of haemotoxylin and eosin stained sections (Fig.
2A and B), which showed a signicant 55% increase in
septal volume fraction in lungs from animals with heart
failure (Fig. 3A). Lung tissue sections stained with Gordon and Sweets stain to visualise reticulin bres (Fig. 2C
and D) clearly showed an increase in the density of septal reticulin bres in heart failure. The median reticulin
density score calculated from ten observers blindly scoring ve sham control lungs was 0.6 (0.41 0.75), while
that for ve heart failure lungs was 4.0 (3.79 4.32): a
highly significant (p < 0.001) increase (Fig. 3B). Pulmonary siderophage inltration was assessed in tissue
sections stained with Perls Stain, which clearly showed
blue-stained siderophages in lungs from animals with

Fig. 3 Morphometric data from sham control and heart failure lung sections.
A Shows the mean S.E.M. septal volume fraction, which is significantly (p < 0.05)
increased in heart failure, Sham control (n = 9), Heart failure (n = 6). B Shows septal
reticulin content as the density score for sham control (open circles) and heart
failure (closed circles) lungs; the overall median is indicated with a line. **p < 0.001

heart failure (Fig. 2E and F). Indeed, while there was only
evidence of siderophages in sections of lung from just
one sham control animal, there were siderophages present in all of the lung sections from animals with heart
failure. The siderophage density was therefore highly
signicantly (p < 0.001) greater in heart failure (37.5
14.9 per frame) than in control lungs (0.6 0.6 per
frame).

300

Basic Research in Cardiology, Vol. 98, No. 5 (2003)

Steinkopff Verlag 2003

Fig. 4 Electron micrographs (magnification
76710) of the alveolar capillary

barrier showing basal laminae (BL) and interstitium (IS) between an endothelial
cell (EC) and type 1 pneumocyte (PN). A control lung (A) and a heart failure lung (B)
are shown. AV alveolus, CP capillary, EC endothelial cell, PN type I pneumocyte,
IS interstitium, CF collagen fibres, PC pericyte

Electron microscopy
Alveoli, capillaries and alveolar-capillary barriers were
visualised under the electron microscope. Gross structural changes were seen that were compatible with those
seen under light microscopy. The alveolar-capillary barrier (including alveolar and capillary endothelial cells
and the intervening interstitium), as illustrated in Fig. 4,
was found to be signicantly thicker in lungs from animals with heart failure (1278.3 76.4 nm) compared with
sham control lungs (637.5 32.4 nm) (Fig. 5A). The interstitium was also found to be signicantly thicker (377.7
43.1 nm cf. 109.0 7.5 nm) (Fig. 5B), and the proportion of the total alveolar-capillary barrier made up by the
Fig. 5 Basal membrane thickening at the alveolar capillary junction measured from
electron micrographs. Values are plotted as mean S.E.M. Sham control (n = 6),
Heart failure (n = 6). A Alveolar capillary barrier thickness; B the thickness of
the interstitium between capillary endothelial cell and adjoining pneumocyte.
C Alveolar basal laminae thickness; D The thickness of the capillary basal laminae.
*p < 0.05, **p < 0.01, ***p < 0.001

M. P. Kingsbury et al.
Structural remodelling of lungs in chronic heart failure

interstium was greater in heart failure (29.1 2.3%) than

in sham control lung tissue (17.8 0.6%). Furthermore,
there was evidence of signicant pericyte inltration of
the alveolar-capillary barrier of animals with heart failure. While pericytes were present in pulmonary tissue, in
the interstium adjacent to the endothelial cells, of all of
the heart failure animals none were observed in the alveolar-capillary interstitium of the controls. There was also
evidence of signicant collagen inltration of the alveolar-capillary barrier in each of the heart failure animals
but in only one sham control animal, although the size of
the collagen bundles were similar in all cases. In addition
to the increased interstitial thickness and collagen and
pericyte inltration in heart failure lungs, there was also
evidence of thickening of both the alveolar (89%, p <
0.01) and capillary (69%, p < 0.05) basal laminae (Fig. 5C
and D).

Discussion
Following chronic aortic banding there was evidence
of heart failure with a reduction in cardiac output,
increased peripheral vascular resistance and raised left
ventricular end diastolic and left atrial pressures in
banded guinea pigs. The signicantly increased left atrial
and right ventricular pressures together with the
increased right ventricle to body weight ratio are indicative of established pulmonary hypertension. Characteristic increases in right ventricular, atrial and lung mass
together with evidence of siderophage inltration in the
lungs of aortic-banded animals further conrm the presence of heart failure. This is supported by our previous
findings of raised plasma catecholamines and atrial
natriuretic peptide in this model (18).
The principal observations from this study are that in
lungs adapted to chronic heart failure there is gross pulmonary septal thickening with increased reticulin deposition, resulting in a signicantly increased alveolar-capillary barrier. This increased barrier thickness reects
not only increased pneumocyte and capillary cell thickness but also a thicker interstitium. Furthermore, there
is evidence of greater collagen deposition within, and
pericyte inltration of, this thickened interstitium and of
signicant thickening of both the alveolar and capillary
basal laminae.
Thickening of the alveolar septa may contribute to the
attenuated alveolar capillary water exchange in heart failure lungs by acting as a functional barrier. Studies in
patients with chronic heart failure (8) have demonstrated
reduced pulmonary microvascular permeability and
thickening of pulmonary capillary basement membrane
(21), which could similarly act to limit the permeability
of proteins and water. Our data demonstrate signicant
increases in septal volume fraction in lungs from animals

301

with chronic heart failure. The morphological changes

would be expected to reduce capillary:alveolar exchange
area providing a further mechanism for reduced water
ltration.
The increased alveolar-capillary barrier was also evident when pulmonary ultra structure was examined.
There was a signicant increase in basal laminae thickness in heart failure which is consistent with previous
clinical work suggesting that basal membrane thickness
was positively related to pulmonary artery wedge pressure and duration of heart failure in human disease (21).
However, the effect of a thickened basal membrane on
permeability in clinical heart failure is not clear. Some
work has suggested a decreased permeability as indicated
by a reduction in mean transferrin index (5), while others have reported no signicant change in pulmonary
transcapillary transferrin escape rate in heart failure (16).
Furthermore, in ve patients with heart failure and a capillary wedge pressure of less than 30 mmHg pulmonary
permeability measured using albumin flux was unaltered, while in one heart failure patient with a capillary
wedge pressure of 40 mmHg it was increased (32).
These apparent increases in permeability in clinical
heart failure may suggest that the thickened basement
membrane did not result in decreased permeability,
perhaps reecting changes in active transcytosis (15) or
changes in calcium dependant, protein kinase Ca (PKC
)
mediated disruption of VE-cadherin junctions (28, 33) to
increase endothelial permeability, especially as PKC
has
been shown to be up-regulated during the development
of right ventricular hypertrophy (3). Permeability may
also be inuenced by changes in the vectoral transport of
salt and water by Na,K-ATPase (1) and ux through pulmonary aquaporin channels (17). However, it is important to understand that clinical measures of permeability are inuenced not only by the permeability of the
exchange barrier, but also by the rate of transbarrier uid
and protein ux. Thus, clinical evidence of increased barrier thickness and apparently unchanged or increased
permeability are not necessarily inconsistent with the
functional pulmonary remodelling that we report in our
model.
Our results showed that not only was the total barrier
thicker, but also there was thickening of the interstitial
layer between adjacent pneumocyte and capillary endothelial cells with increased collagen and some evidence of
pericyte inltration. Pericytes, are heterogeneous cells
associated abluminally with all vascular capillaries and
post capillary veins (12). They regulate endothelial proliferation and differentiation; exacerbate endothelial cell
junctional inflammatory leakage and synthesise, and
secrete vasoactive agonists and structural constituents of
the basement membrane and extracellular matrix (30).
The role of pericytes in forming part of the microvascular barrier is complex. The in situ ratio of pericytes to
endothelial cells as well as the endothelial area covered by

302

Basic Research in Cardiology, Vol. 98, No. 5 (2003)

Steinkopff Verlag 2003

pericytes has been hypothesised as being related to the

degree of tightness of the interendothelial junctions and
to the level of microvascular blood pressure. The ratio of
pericytes to endothelial cells tends to be greater where
microvascular pressure is higher and there are tight
interendothelial junctions such as in the retina and lung,
and less in skeletal striated muscle where microvascular
pressure is lower and interendothelial junctions looser
(30). Other research, however, suggests that in the lung
the pericytes may enhance movement across the microvascular barrier by increasing pore size of endothelial
intercellular junctions under the control of leukotrienes
(23). Indeed, the siderophage inltration we observed in
lungs from animals with heart failure could be a potential source of such leukotrienes. Inammatory mediators
can also increase endothelial permeability via PKC

mediated disruption of VE-cadherin junctions (28, 33).
In addition to their role in the alveolar-capillary barrier, pericytes may also have an inuence in modulating
the pulmonary remodelling process. Pericytes have been
shown to form a high frequency of intimate junctions
with the endothelium and are preferentially localised
near paracellular clefts (7, 29). It has been proposed that
vessel wall cross-talk exists between pericytes and the
microvascular endothelium (4). Co-culture of pericytes
with endothelial cells leads to the activation of transforming growth factor  (TGF ) which in turn inuences growth and differentiation (12). The effects of
pericyte inltration and any subsequent TGF  release in
heart failure lungs are not clear. Studies with an in vivo
wound healing model show newly formed capillaries stop
growing when pericytes migrate into the basement membrane (6). However, TGF  added directly at low concentrations can in fact stimulate endothelial cell proliferation (26). Our nding of increased pericyte numbers
in the lungs of animals with chronic heart failure lends
indirect support to the evidence that they help to stimulate remodelling of the interstitium (4, 12, 30).
This pulmonary remodelling increases the tissue barrier to water transudation and may help to explain the
reduced pulmonary capillary ltration that we have previously described in this model (14). However the
increased barrier and tissue stiffness is also likely to limit
gas exchange. The implications of the changes in alveo-

lar capillary membrane and the resultant increased

resistance to gas transfer was recently reviewed by Guazzi
(11). In patients with heart failure, pulmonary gas diffusion is impaired at rest (27) and during exercise (31), and
inefcient ventilation is associated with poor prognosis
(24). As resting blood volume was similar in control and
heart failure subjects while gas exchange was severely
impaired in heart failure, Puri et al. concluded that the
reduced gas diffusion was due to a reduction in alveolarcapillary exchange surface area and/or structural changes that impair exchange (27). These mechanisms were
also implicated in the reduced gas exchange during exercise in heart failure subjects, together with a reduction in
pulmonary capillary distension acting to attenuate pulmonary recruitment during exercise (31). Our results
show not only an increased alveolar volume fraction,
which suggests a reduced alveolar capillary surface area
available for gas exchange, but also structural changes
that are liable to limit gas exchange by increasing diffusion distances. Furthermore, the increased wall thickness of small pulmonary vessels and increased pulmonary vascular resistance we previously reported (14)
are likely to further limit capillary distension, and thus
reduce capillary recruitment during increased respiratory demand. Thus while the structural changes observed
in the lungs of animals with chronic heart failure are
likely to confer some protection against pulmonary
oedema, they are likely to limit respiratory efciency and
result in the dyspnoea and impaired gas exchange
observed clinically in heart failure.
In conclusion, this study has demonstrated pulmonary remodelling in lungs adapted to chronic heart
failure. We hypothesise that these pulmonary adaptations provide protection from oedema formation. These
changes, while initially protective in helping to prevent
oedema formation, are also likely to confer major longterm disadvantages: by a) augmenting total vascular
resistance and b) contributing to impaired gas diffusion,
known to occur in chronic heart failure (11).
Acknowledgments The authors would like to express thanks to Lorraine
Lawrence for her valuable help with the histology. Grant support from
the British Heart Foundation and Servier Laboratories is gratefully
acknowledged.

M. P. Kingsbury et al.
Structural remodelling of lungs in chronic heart failure

303

References
1. Azzam ZS, Dumasius V, Saldias FJ, Adir
Y, Sznajder JI, Factor P (2002) Na, KATPase overexpression improves alveolar uid clearance in a rat model of elevated left atrial pressure. Circulation 105:
497501
2. Bancroft JD, Stevens A (1977) Theory
and practice of histological techniques.
Churchill Livingstone, Edinburgh
3. Braun MU, LaRose P, Schn S, Borst
MM, Strasser RH (2002) Differential
regulation of cardiac protein kinase C
isozyme expression after aortic banding
in rat. Cardiovasc Res 56: 5263
4. Chakravarthy U, Gardiner TA (1999)
Endothelium derived agents in pericyte
function/dysfunction. Prog Retin EyeRes 18 (4): 511527
5. Clough G (1991) Relationship between
microvascular permeability and ultrastructure. Prog Biophys Molec Biol 55:
4769
6. Crocker DJ, Murad TM, Geer JC (1970)
Role of the pericyte in wound healing an
ultrastructural study. Exp Mol Path 13:
5165
7. DAmore PA (1992) Capillary growth: a
two cell system. Seminars in Cancer Biology 3 (2): 4956
8. Davies SW, Bailey J, Keegan J, Balcon R,
Rudd RM, Lipkin DP (1992) Reduced
pulmonary microvascular permeability
in severe chronic left heart failure. Am
Heart J 124: 137142
9. Deuchar GA, Hicks MN, Cobbe SM,
Docherty CC, MacLean MR (1998) Pulmonary responses to 5-hydroxytryptamine and endothelin-1 in a rabbit model
of left ventricular dysfunction. Cardiovasc Res 38: 500507
10. Edwards WD (1990) The pathology of
secondary pulmonary hypertension. In:
Fishman AP (ed) The pulmonary circulation: normal and abnormal. University of
Pennsylvania University Press, Philadelphia, pp 329342
11. Guazzi M (2002) Alveolar-capillary membrane dysfunction in chronic heart failure: pathophysiology and therapeutic
implications. Clin Sci 98: 633641
12. Hirschi KK, DAmore PA (1996) Pericytes
in the microvasculature. Cardiovasc Res
32 (4): 687698
13. Ho KKL, Anderson KM, Kannel WB,
Grossman W, Levy D (1993) Survival after
the onset of congestive cardiac failure in
Framingham heart study Patients. Circulation 88: 107115

14. Huang W, Kingsbury MP, Turner MA,

Donnelly JL, Flores NA, Sheridan DJ
(2001) Capillary ltration is reduced in
lungs adapted to chronic heart failure:
morphological and haemodynamic correlates. Cardiovasc Res 49: 207217
15. John TA, Vogel SM, Minshall RD, Ridge
K, Tiruppathi C, Malik AB (2001) Evidence for the role of alveolar epithelial
gp60 in active transalveolar albumin
transport in the rat lung. J Physiol 533 (2):
547559
16. Kaplan JD, Calandrino FS, Schuster DP
(1991) A positron emission tomographic
comparison of pulmonary vascular permeability during the adult respiratory
disress syndrome and pneumonia. Am
Rev Respir Dis 143 (1): 150154
17. King LS, Nielsen S, Agre P (2000) Aquaporins and the respiratory system: advice
for the lung investigator. J Clin Invest 105:
1516
18. Kingsbury MP, Huang W, Giuliatti S,
Turner M, Hunter R, Parker K, Sheridan
DJ (1999) Investigation of distal aortic
compliance and vasodilator responsiveness in heart failure due to proximal
aortic stenosis in the guinea pig. Clin Sci
(Colch) 96: 241251
19. Kingsbury MP, Huang W, Turner MA,
Flores NA, Sheridan DJ (2001) Resistance
to pulmonary oedema in anaesthetised
guinea-pigs with chronic heart failure.
J Physiol 533: 90P
20. Lassers BW, George M, Anderton JL,
Higgins MR, Philp T (1970) Left ventricular failure in acute myocardial infarction.
Am J Cardiol 25: 511522
21. Lee Y-S (1979) Electron microscopic
studies on the alveolar-capillary barrier
in the patients of chronic pulmonary
edema. Jpn Circ J 43: 945954
22. Ling ET, deBold AJ (1976) An improved
method for the production of experimental congestive heart failure in the guinea
pig. Lab Animals 10: 285289
23. Lonigro AJ, Mc Murdo L, Stephenson
AH, Sprague RS, Weintraub NL (1996)
Hypotheses regarding the role of pericytes in regulating movement of fluid,
nutrients, and hormones across the
microcirculatory endothelial barrier.
Diabetes 45: S38S43
24. Meyer FJ, Zugck C, Haass M, Otterspoor
L, Strasser RH, Kubler W, Borst MM
(2000) Inefcient ventilation and reduced
respiratory muscle capacity in congestive
heart failure. Basic Res Cardiol 95 (4):
333342
25. Mc Donagh TA, Morrison CE, Lawrence
A, Ford I, Tunstall-Pedoe H, McMurray
JJ, Dargie HJ (1997) Symptomatic and
asymptomatic left ventricular dysfunction in an urban population. Lancet 350:
829833

26. Myoken Y, Kan M, Sato GH, McKeehan

WL, Sato JD (1990) Biofunctional effects
of transforming growth factor b on
endothelial cell growth correlates with
phenotypes of TGFb binding sites. Exp
Cell Res 191: 299304
27. Puri S, Baker BL, Dutka DP, Oakly CM,
Hughes JMB, Cleland JGF (1995) Reduced
alveolar-capillary membrane diffusing
capacity in chronic heart failure. Its
pathophysiological relevance and relationship to exercise performance. Circulation 91: 27692774
28. Sandoval R, Malik AB, Minshall RD,
Kouklis P, Ellis CA, Tiruppathi C (2001)
Ca2+ signalling and PKCa activate
increased endothelial permeability by
disassembly of VE-cadherin junctions. J
Physiol 533 (2): 433445
29. Schulze C, Firth JA (1993) Junctions
between pericytes and the endothelium
in rat myocardial capillaries: a morphometric and immunogold study. Cell
Tissue Res 271 (1): 145154
30. Shepro D, Morel NM (1993) Pericyte
physiology. FASEBJ 7 (11): 10311038
31. Smith AA, Cowburn PJ, Parker ME,
Denvir M, Puri S, Patel KR, Cleland JG
(1999) Impaired pulmonary diffusion
during exercise in patients with chronic
heart failure. Circulation 100: 14061410
32. Sugerman HJ, Tatum JL, Burke TS, Strash
AM, Glauser FL (1984) Gamma scintigraphic analysis of albumin flux in
patients with acute respiratory distress
syndrome. Surgery 95 (6): 674682
33. Tiruppathi C, Freichel M, Vogel SM, Paria
BC, Mehta D, Flockerzi V, Malik AB
(2002) Impairment of store-operated
Ca2+ entry in TRCP(-/-) mice interferes
with increase in lung microvascular permeability. Circ Res 91 (1): 7076
34. Townsley MI, Fu Z, Mathieu-Costello O,
West JB (1995) Pulmonary microvascular
permeability. Responses to high vascular
pressure after induction of pacinginduced heart failure in dogs. Circ Res 77:
317325
35. Townsley MI, Pitts VH, Ardell JL, Zhao Z,
Johnson WH Jr (1994) Altered pulmonary
microvascular reactivity to norepinephrine in canine pacing-induced heart
failure. Circ Res 75: 347356
36. Weaver LJ, Carrico CJ (1984) Congestive
heart failure and edema. In: Staub NC,
Taylor AE (eds) Edema. Raven Press, New
York, pp 543562
37. Wilkins MR, Zhao L, al-Tubuly R (1996)
The regulation of pulmonary vascular
tone. Br J Clin Pharmacol 42: 127131

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.