Am J Physiol Regul Integr Comp Physiol 304: R171–R176, 2013.
First published November 28, 2012; doi:10.1152/ajpregu.00444.2012.
Review
Role of the fragility of the pulmonary blood-gas barrier in the evolution
of the pulmonary circulation
John B. West
Department of Medicine, University of California San Diego, La Jolla, California
Submitted 26 September 2012; accepted in final form 16 November 2012
pulmonary capillary; pulmonary vascular pressures; type IV collagen; extreme
exercise; stress failure
MARCELLO MALPIGHI (1628 –1694) has the distinction of discovering
the pulmonary capillaries. He studied frog lungs using the recently
invented light microscope and wrote to his friend, Giovanni
Borelli (1608 –1679), that he saw a series of small blood vessels
connecting the arteries and veins. The existence of these had been
surmised by William Harvey (1578 –1657) to explain the circulation of the blood, but the capillaries could not be seen without a
microscope. Malpighi stated “So great is the branching of these
vessels, after they extend out hither and thither from the vein and
artery, that no larger system of vessels will be served, but a
network appears, formed by the offshoots of the two vessels.”
Malpighi made drawings of the capillaries but of course was not
able to visualize their walls clearly.
Over the next 200 years many investigators examined the
walls of the pulmonary capillaries using improved microscopes, but the structure of the wall, which they realized was
the blood-gas barrier, continued to be disputed. On the one
hand, some investigators claimed that the capillaries were
naked with nothing outside them except possibly a little reticular and connective tissue. For example, the French physician
Albert Policard stated that the alveolar wall is like the “flesh of
an open wound” (La surface respiratoire est assimilable à une
plaie à vif) by which he meant that the capillaries were directly
exposed to the alveolar gas (17). Similar views were held by
Loosli et al. (11) and Potter and Loosli (19). By contrast other
investigators such as Bensley and Bensley (1), Macklin (14),
Address for reprint requests and other correspondence: J. B. West, UCSD
Dept. of Medicine, 0623A, 9500 Gilman Dr., La Jolla, CA 92093-0623
(e-mail:[email protected]).
http://www.ajpregu.org
and Miller (16) argued that there was a continuous membrane that
lined the alveolar wall separating the endothelium of the capillaries from the air spaces. However, agreement on the nature of the
blood-gas barrier was impossible at this time simply because the
resolution of the light microscope is not sufficient to determine
the ultrastructure of the capillary wall.
Electron Microscopy of the Blood-Gas Barrier
The breakthrough came in 1952 and 1953 when Frank Low
published the first images obtained by electron microscopy.
These studies had been made possible by the work of George
Palade who showed how it was possible to fix biological tissue
for electron microscopy by using buffered osmic acid. Low’s
first paper (12) described appearances in the rat lung, but the
images were of poor quality. He was particularly interested in
showing the nuclei of the alveolar epithelial cells in addition to
their cytoplasmic layer covering the capillaries. One reason for
this is that some previous investigators had argued that the
capillaries were covered by “nonnucleated plaques.” This was
an understandable error because the nuclei of the alveolar
epithelial cells are few and far between.
However, in his second report, Low (13) showed excellent
images. One is reproduced in Fig. 1 where we can see a red
blood cell inside a capillary, the capillary endothelium labeled
2, the alveolar epithelium labeled 1, the nucleus of the epithelial cell with its nucleolus labeled 5, and the extracellular
matrix labeled 3, though this is not visible at the location
labeled 4. Note that the capillary endothelial and alveolar
epithelial layers both have a thickness of about 0.15 ␮m, and
that the total thickness of the blood-gas barrier is about 0.3 ␮m.
0363-6119/13 Copyright © 2013 the American Physiological Society
R171
Downloaded from http://ajpregu.physiology.org/ by 10.220.33.1 on June 17, 2017
West JB. Role of the fragility of the pulmonary blood-gas barrier in the evolution of the
pulmonary circulation. Am J Physiol Regul Integr Comp Physiol 304: R171–R176, 2013.
First published November 28, 2012; doi:10.1152/ajpregu.00444.2012.—In 1953 Frank
Low published the first high-resolution electron micrographs of the human pulmonary blood-gas barrier. These showed that a structure only 0.3-␮m thick separated
the capillary blood from the alveolar gas, immediately suggesting that the barrier
might be vulnerable to mechanical failure if the capillary pressure increased.
However, it was 38 years before stress failure was recognized. Initially it was
implicated in the pathogenesis of High Altitude Pulmonary Edema, but it was soon
clear that stress failure of pulmonary capillaries is common. The vulnerability of the
blood-gas barrier is a key factor in the evolution of the pulmonary circulation. As
evolution progressed from the ancestors of fishes to amphibians, reptiles, and
finally birds and mammals, two factors challenged the integrity of the barrier. One
was the requirement for the barrier to become increasingly thin because of the
greater oxygen consumption. The other was the high pulmonary capillary pressures
that were inevitable before there was complete separation of the pulmonary and
systemic circulations.
Review
R172
EVOLUTION OF THE PULMONARY CIRCULATION
This micrograph from a human lung represents one of the most
important advances in lung biology. Modern electron micrographs have greatly improved the resolution, and Fig. 2 shows
that the layer of extracellular matrix, which was interrupted in
Fig. 1, extends around the whole of the capillary. The micrograph also shows the nucleus of the type I alveolar epithelial
cell (as indeed did Fig. 1) and also the nucleus of the capillary
endothelial cell. It was remarkable to find a section showing
both of these nuclei. Nevertheless, the basic ultrastructure of
the blood-gas barrier was clearly shown by Low in 1953.
Slow Recognition of the Vulnerability of the Blood-Gas Barrier
It is remarkable that it took so long to recognize the vulnerability of this extremely thin blood-gas barrier. As we have
Fig. 2. Modern high-resolution electron micrograph of a pulmonary capillary.
Note the general similarity to Fig. 1. From Ref. 25.
seen, in 1953 Low clearly demonstrated that there was nothing
separating the blood in the pulmonary capillary from the gas in
the alveolar space except a structure that was only 0.3 ␮m
thick. Furthermore, this structure consisted of only of two very
thin cells with a small amount of extracellular matrix sandwiched between them. This should have suggested that the
blood-gas barrier was therefore vulnerable to increased stresses
caused by any large increase in capillary pressure. However,
the vulnerability of the capillaries was not recognized until 38
years later (28).
In fact, there was plenty of evidence in the clinical and pulmonary pathology literature to suggest that the capillaries were
leaking red cells under some conditions which must mean that
the walls were damaged. For example, it was well known that
the patients with mitral stenosis who had an increased capillary
pressure frequently developed hemoptysis and that their lungs
contained large amounts of hemosiderin, which could only
come from red blood cells. How did this get into the extravascular tissue? One explanation was that the bleeding occurred at
the anastomoses between the pulmonary and bronchial arteries
that were thought to be situated in the mucosa of the terminal
bronchioles (10). This seems improbable because these blood
vessels have substantially thicker walls than the capillaries. Of
course it was known that the capillaries could leak fluid when
the pressure inside them was increased, but this did not mean
structural failure of the wall. It was explained by the imbalance
of the Starling forces as was originally described in 1896 (22).
What was the reason for the slow recognition of the vulnerability of the pulmonary capillaries? Probably there were three
factors. First, many investigators thought that pulmonary capillaries could withstand a high pressure inside them without
failure because this was apparently the case with capillaries in
AJP-Regul Integr Comp Physiol • doi:10.1152/ajpregu.00444.2012 • www.ajpregu.org
Downloaded from http://ajpregu.physiology.org/ by 10.220.33.1 on June 17, 2017
Fig. 1. First published electron micrograph of the pulmonary blood-gas barrier in the human lung. Note that the total thickness is about 0.3 ␮m. CAP, capillary; RBC,
red blood cell; ALV, alveolar epithelium; TSP tissue space. See text for details. Reprinted with permission from Wiley (copyright©John Wiley & Sons) (From Low, Ref. 13).
Review
EVOLUTION OF THE PULMONARY CIRCULATION
Recognition of Stress Failure of Pulmonary Capillaries
The recognition that stress failure can occur came from an
unexpected source. In the 1970s, the pathogenesis of High
Altitude Pulmonary Edema (HAPE) was a puzzle. It was known
that there was a strong link between pulmonary hypertension
and HAPE. For example, individuals who developed HAPE
had high pulmonary arterial pressures, and people who were
susceptible to the disease had an increased hypoxic pulmonary
vasoconstriction response. However, the pulmonary arterial
wedge pressure was found to be normal in the disease, and this
ruled out left ventricular failure. A breakthrough came when it
was shown by bronchoalveolar lavage that the edema fluid of
HAPE had a large concentration of cells and high-molecularweight proteins and therefore the edema was of the high
permeability type (6, 21). This indicated that the edema must
be associated with structural damage to the walls of the
pulmonary capillaries. However, hypoxic pulmonary vasoconstriction occurs upstream of the capillaries and so it was not
clear how the high pulmonary artery pressure could damage
them. A possible mechanism had been suggested by Hultgren
(8) when he argued that the vasoconstriction might be uneven.
This would allow the high pressure to be transmitted to some
of the capillaries. However, nobody previously had raised the
pressure in pulmonary capillaries in experimental animals and
examined the ultrastructure of their walls. When this was done
in a rabbit preparation, there was clear evidence of breaks in
the capillary endothelium, alveolar epithelium, and, in some
cases, the extracellular matrix as well (28, 23). Figure 3 shows
examples of electron micrographs showing stress failure.
Prevalence of Stress Failure in Pulmonary Capillaries
It soon became evident that stress failure of pulmonary
capillaries occurred not only in HAPE but in many other
conditions. For example, the experiments in the rabbit preparation showed that high states of lung inflation greatly increased the number of breaks in the capillary wall for a given
capillary transmural pressure. This, therefore, explained, at
least in part, the damage to the pulmonary capillaries that is
seen in the intensive care unit when high levels of positive
end-expiratory pressure are employed. One of the most dramatic examples of stress failure is seen in racehorses. For
centuries, bleeding from the lung had been known to occur in
these animals after galloping, and when the horses were examined by bronchoscopy, it was shown that bleeding was
common. Eventually it was found that all racehorses in training
have hemosiderin-laden macrophages in their tracheal washings, which meant that they are all bleeding into their alveolar
spaces. Direct evidence of stress failure in the capillaries was
shown by galloping thoroughbreds on a treadmill and examining their lungs by electron microscopy (29). The reason for
the bleeding is that these animals develop extremely high
pulmonary vascular pressures during exercise with mean pulmonary artery pressures up to 120 mmHg and left atrial
pressures as high as 70 mmHg (4, 9). These high pressures
result from the extremely high cardiac outputs during galloping
and, of course, these animals have been selectively bred for
speed over hundreds of years. Racing greyhounds also bleed
into their lungs (18).
The fact that racehorses break their capillaries during exercise raised the question of whether this ever occurs in elite
human athletes. This was studied in racing cyclists and, indeed,
it was shown that after sprinting at levels near the maximum
oxygen consumption there was an increase in the number of
red blood cells in bronchoalveolar lavage fluid. The controls
were cyclists who remained sedentary (7). Since this demonstration of the vulnerability of the blood-gas barrier in elite
athletes at very high levels of exercise, there have been numerous reports of bleeding from the lungs in athletes under a
variety of conditions (7).
AJP-Regul Integr Comp Physiol • doi:10.1152/ajpregu.00444.2012 • www.ajpregu.org
Downloaded from http://ajpregu.physiology.org/ by 10.220.33.1 on June 17, 2017
the systemic circulation. For example, the capillaries in the
lower leg have a high pressure inside them because of the hydrostatic column of blood in the upright body. However, the
fallacy here is that systemic capillaries are supported by the
interstitial tissue around them. The pressure in the interstitium
of the leg rises along with the pressure in the capillary lumen
and therefore the transmural pressure of the capillaries is not
necessarily increased. By contrast, pulmonary capillaries have
no interstitial tissue to support them over much of their surface.
Another common misconception was that the pressure in the
pulmonary capillaries was always relatively low. There is some
excuse for this error because the pressures within the pulmonary circulation, particularly on the venous side, are difficult to
measure. However, when physiologists started measuring the
pulmonary artery pressure, and particularly the pulmonary
arterial wedge pressure, in subjects during vigorous exercise, it
was immediately appreciated that the pressure in the pulmonary capillaries could rise substantially. For example, in one
study of normal human subjects in the upright position, it was
shown that at an oxygen consumption of 3.7 l/min, which is a
substantial level of exercise but not maximal, the mean pulmonary artery pressure was 37 mmHg and the mean pulmonary
arterial wedge pressure was 21 mmHg (24). The wedge pressure reflects pulmonary venous pressure. Micropuncture studies of pulmonary blood vessels in animals have shown that the
capillary pressure is about halfway between arterial and venous
pressures (2). Based on this, the mean capillary pressure at
mid-lung is about 29 mmHg, and at the bottom of the lung,
because of hydrostatic pressure effects, the value would be
about 36 mmHg. Other measurements of pulmonary vascular
pressures in normal subjects during exercise have given similar
results (5, 20).
A third reason why stress failure of the capillaries was not
suspected was that physiologists invoked the Laplace equation.
This states that the wall stress (S) is proportional to the
transmural pressure (P) of a vessel and its radius (r) but
inversely proportional to the thickness (t) of its wall. In
symbols, S ⫽ P·r/t. The fact that the capillaries have a very
small radius of the order of 3 ␮m clearly reduces wall stress.
However, the wall is extremely thin, and furthermore only the
type IV collagen layer in the wall is believed to be responsible
for its strength. Calculations of the stress in this layer, which is
only about 50 nm thick, indicate that with a capillary transmural pressure of 30 mmHg, the stress approaches the ultimate
tensile stress of type IV collagen (26). Admittedly, these
calculations are imprecise but they make the point that a small
radius of curvature is not necessarily sufficient to avoid stress
failure if the thickness of the wall is very small.
R173
Review
R174
EVOLUTION OF THE PULMONARY CIRCULATION
Blood-Gas Barrier and the Evolution of the Pulmonary Circulation
The vulnerability of the blood-gas barrier is a key factor in
the evolution of the pulmonary circulation. As evolution progressed from the ancestors of the fishes to amphibians, reptiles,
and finally birds and mammals, two factors challenged the
integrity of the barrier. One was the high pulmonary capillary
pressures that existed before a complete separation between the
pulmonary and systemic circulations developed. The other was
the requirement for the barrier to become increasingly thin to
allow the greater oxygen consumption of the amphibia and
reptiles, and especially the endothermic birds and mammals.
The progression of the mean thickness of the blood-gas
barrier from its highest value in amphibia to its lowest in birds
is shown in Fig. 4. Note the very small thickness in birds that
is made possible by the support of the blood capillaries by the
surrounding air capillaries. Birds have a flow-through rather
than reciprocating mode of ventilation as exists in mammals.
The result is that birds have air capillaries with diameters of
about 10 –20 ␮m, much smaller than mammalian alveoli
which, in the human lung, have a diameter of about 300 ␮m.
This arrangement allows the walls of the blood capillaries to be
supported by epithelial bridges that form the walls of the air
capillaries. The situation is very different from that in the mam-
malian lung where the capillaries in the alveolar walls are essentially unsupported at right angles to the wall. The net result is that
because of this additional support, the walls of avian lung capillaries can afford to have much thinner walls (30).
The evolutionary progression of the separation of the pulmonary from the systemic circulation is shown in Fig. 5. The first
panel shows the arrangement of the circulation in fishes. The heart
Fig. 4. Harmonic mean thickness ⫾ SD of the blood-gas barrier in birds,
mammals, reptiles, and amphibians. The data are from Table 1 of Ref. 15. This
includes 34 species of birds, 37 of mammals, 16 of reptiles, and 10 of
amphibians. Note the progression in mean thickness. From Ref. 27.
AJP-Regul Integr Comp Physiol • doi:10.1152/ajpregu.00444.2012 • www.ajpregu.org
Downloaded from http://ajpregu.physiology.org/ by 10.220.33.1 on June 17, 2017
Fig. 3. Examples of stress failure in pulmonary capillaries. A: disruption of the capillary endothelial cell (arrows) but the alveolar epithelium and two basement
membranes are intact. B: disruption of an alveolar epithelial cell at the top (arrows), and disruption of a capillary endothelial cell near the bottom (arrows). A
blood platelet is adhering to the exposed basement membrane below. C: disruption of all layers of the capillary wall with a red blood cell apparently passing
through the opening. D: scanning electron micrograph showing breaks in the alveolar epithelium.
Review
R175
EVOLUTION OF THE PULMONARY CIRCULATION
Fig. 5. Diagram showing changes in the evolution of the pulmonary circulation. As evolution progresses from fishes to amphibians, reptiles, mammals, and birds, there is increasing
separation of the pulmonary and systemic circulations.
aorta. Note that because the ventricle is not divided, the lung
capillaries are exposed to a relatively high pressure. However,
again, these animals are exothermic and therefore have relatively
low maximal oxygen consumptions, and so the blood-gas barrier
can afford to be relatively thick (Fig. 4).
When we progress to the noncrocodilian reptiles, we see a
further difference in the arrangement of the circulations. Now in
addition to having two atria, the ventricle is partly divided. This
means that the streaming that we saw in amphibians is improved
with the result that more of the oxygenated blood returning from
the lungs finds its way into the tissues. However, the fact that the
ventricles are not completely divided means that the pulmonary
capillaries are potentially exposed to the high systemic arterial
pressure. There is an interesting exception to this in some reptiles
such as the monitor lizard Varanus exanthematicus. In this case
there is a ridge in the ventricle that allows the outflow tracts of the
lungs and the dorsal aorta to be separated during systole. As a
result, the pressure in the pulmonary artery is lower than that in
the aorta (3).
Crocodilian reptiles differ from the noncrocodilian in that there
is no communication between the two ventricles. As a consequence, the pressure in the pulmonary capillaries is potentially
relatively low. However, these animals have a shunt between the
pulmonary artery and aorta that allows them to divert variable
amounts of blood flow to the lung or the systemic circulation.
Under normal conditions the entire output of the right ventricle
goes to the lung, but during diving apnea, pulmonary vascular
resistance rises and the result is a shunt around the unventilated
lung.
Mammals and birds are unique in this evolutionary development in that they are fully endothermic and capable of very
high maximal oxygen consumptions. As a consequence the
blood-gas barrier has to be extremely thin to allow the rapid
diffusion of oxygen across it (Fig. 4). This in turn means that
the capillaries are highly vulnerable to stress failure unless the
pressure is kept relatively low by complete separation of the
pulmonary and systemic circulations. Nevertheless, as we have
seen, some mammals under some conditions of very high
oxygen consumption raise the pressure in the pulmonary capillaries so high that stress failure occurs.
AJP-Regul Integr Comp Physiol • doi:10.1152/ajpregu.00444.2012 • www.ajpregu.org
Downloaded from http://ajpregu.physiology.org/ by 10.220.33.1 on June 17, 2017
has a single atrium and single ventricle. Blood is pumped to the
gills where the gas exchange with the water occurs via the ventral
aorta, and the effluent blood from the gills is transported by the
dorsal aorta to the tissues of the body. As a consequence, the
hydrostatic pressure in the capillaries of the gills must be greater
than that in the dorsal aorta. The pressure in the dorsal aorta needs
to be relatively high so that the blood can be distributed to all the
tissues of the body. On the face of it this suggests that the gill
capillaries would be liable to stress failure. However, this likelihood is reduced because the blood-water barrier is much thicker
than the blood-gas barrier in mammals. The reason for the thicker
barrier is that fishes are homeothermic (cold-blooded) with relatively low levels of maximal oxygen consumption, and therefore
the diffusion characteristics of the blood-gas barrier are not as
stringent as they are in mammals. Because the barrier is much
thicker, it can be much stronger and thus stress failure is not an
issue. Interestingly, there are some highly aerobic fishes such as
the albacore tuna Thunnus alalunga, where the mean dorsal aortic
pressure is as high as 79 mmHg. Such a high pressure would
break the pulmonary capillaries of mammals, but because the
blood-gas barrier is much thicker, this does not apparently occur.
It would be impossible to have the arrangement shown for
fishes in Fig. 5 in the amphibians, reptiles, mammals, and birds.
This is because as evolution proceeds, there is an increase in
maximal oxygen consumption, reaching its peak in the mammals and birds, and this necessitates a thinner blood-gas barrier
to allow the adequate diffusive gas exchange to occur (Fig. 4).
Therefore, a gradual separation of the pulmonary circulation
from the systemic circulation is critically important. This is the
only way that it is possible to maintain a relatively low pressure
in the lung capillaries while maintaining a high systemic
arterial pressure to perfuse the tissues including the limbs that
require a high blood flow.
Figure 5 shows that in modern amphibians, the beginnings of
the separation between the pulmonary and systemic circulations
occurs. Now there are two atria, one of which receives oxygenated
blood from the lung, and the other the mixed venous blood from
the tissues. However, the ventricle is undivided and so there is
some mixing of blood from the two atria. Nevertheless, there is
some streaming of blood within the heart with the result that much
of the oxygenated blood from the lungs finds its way into the
Review
R176
EVOLUTION OF THE PULMONARY CIRCULATION
Potential for Stress Failure in the Developing Fetus
REFERENCES
1. Bensley RD, Bensley SH. Studies on the lining of the alveolus of normal
lungs of adult animals. Anat Rec 64: 41–49, 1935.
2. Bhattacharya J, Nanjo S, Staub NC. Micropuncture measurement of
lung microvascular pressure during 5-HT infusion. J Appl Physiol 52:
634 –637, 1982.
AJP-Regul Integr Comp Physiol • doi:10.1152/ajpregu.00444.2012 • www.ajpregu.org
Downloaded from http://ajpregu.physiology.org/ by 10.220.33.1 on June 17, 2017
It is fascinating that as the fetus develops in utero and finally
emerges into the environment, the pulmonary and systemic circulations follow a pattern somewhat similar to that shown in Fig.
5. In the mammalian fetus, such as a human, the pulmonary artery
and aorta are connected by a large patent ductus arteriosus. This
anatomy is similar in some ways to that in amphibians and
noncrocodilian reptiles in that in the human fetus, both ventricles
discharge into what is essentially a common conduit. In fact in
crocodilian reptiles, there is a shunt between the pulmonary artery
and aorta similar to the ductus arteriosus. However, as in amphibians and reptiles, there is streaming in the fetal heart so that much
of the well-oxygenated blood from the placenta crosses the right
atrium to the left atrium through the patent foramen ovale and is
distributed to the brain.
A feature of the shunt between the pulmonary artery and the
aorta means that the pulmonary capillaries are potentially exposed
to the relatively high pressure in the systemic circulation. In
principle, this makes the pulmonary capillaries vulnerable to stress
failure. Two factors lessen this outcome. One is that the fetal lung
contains a large amount of vascular smooth muscle in the small
pulmonary arteries that contracts as a result of hypoxic pulmonary
vasoconstriction and thus decreases the pressure in the capillaries.
In addition, the developing fetal lung has no need of thin-walled
pulmonary capillaries because it is not involved in gas exchange.
In fact, only about 15% of the cardiac output in the human fetus
reaches the lung. This is sufficient for the development of the
organ and, of course, a large blood flow through the lung is
unnecessary because the placenta, not the lung, is responsible for
gas exchange.
Dramatic changes occur at birth when the gas exchange has
to change in a very short time from placental to pulmonary.
Blood flow through the lung of the newborn increases greatly
mainly as a result of the release of hypoxic pulmonary vasoconstriction, which is very effective because of the abundance
of vascular smooth muscle. In addition, the patent ductus
arteriosus gradually closes protecting the pulmonary circulation from the high systemic pressures. One of the triumphs of
evolution is that the fall in pulmonary vascular resistance and
the closure of the ductus arteriosus are so well synchronized in
the majority of births. Occasionally the ductus is slow to close
and then the pulmonary capillaries are potentially exposed to a
high pressure with resulting damage to their walls.
In summary, the fact that gas exchange across the blood-gas
barrier occurs through passive diffusion necessitates a very thin
barrier in highly aerobic vertebrates such as mammals and
birds. This very thin barrier makes it vulnerable to stress failure
especially at high levels of exercise when the capillary pressure
rises. The gradual separation of the pulmonary and systemic
circulations as evolution progresses from fishes to amphibia,
reptiles, and finally birds and mammals is essential to protect
the pulmonary capillaries from stress failure.
3. Burggren W. Ventricular haemodynamics in the monitor lizard Varanus
exanthematicus: pulmonary and systemic pressure separation. J Exp Biol
96: 343–354, 1982.
4. Erickson BK, Erickson HH, Coffman JR. Pulmonary artery, aortic and
oesophageal pressure changes during high intensity treadmill exercise in
the horse: a possible relation to exercise-induced pulmonary haemorrhage.
Equine Vet J Suppl 9: 47–52, 1990.
5. Groves BM, Reeves JT, Sutton Wagner PD, Cymerman A Jr, Malconian MK, Rock PB, Young PM, Houston CS. Operation Everest II:
elevated high-altitude pulmonary resistance unresponsive to oxygen. J
Appl Physiol 63: 521–530, 1987.
6. Hackett PH, Bertman J, Rodriguez G, Tenney J. Pulmonary edema
fluid protein in high-altitude pulmonary edema. JAMA 256: 36 –36, 1986.
7. Hopkins SR, Schoene RB, Martin TR, Henderson WR, Spragg RG,
West JB. Intense exercise impairs the integrity of the pulmonary bloodgas barrier in elite athletes. Am J Respir Crit Care Med 155: 1090 –1094,
1997.
8. Hultgren HN. High altitude pulmonary edema. In: Biomedicine Problems
of High Terrestrial Altitude, editd by A. H. Hegnauer. New York:
Springer-Verlag, 1969, p. 131–141.
9. Jones JH, Smith BL, Birks EK, Pascoe JR, Hughes TR. Left atrial and
pulmonary artery pressures in exercising horses. FASEB J 6: A2020, 1992.
10. Lendrum AC. Pulmonary haemosiderosis of cardiac origin. J Pathol
Bacteriol 62: 555–561, 1950.
11. Loosli CG, Adams WE, Thornton TM. The histology of the dog’s lung
following experimental collapse. With special reference to the nature of
the alveolar lining. Anat Rec 105: 697–721, 1949.
12. Low FN. Electron microscopy of the rat lung. Anat Rec 113: 437–44, 1952.
13. Low FN. The pulmonary alveolar epithelium of laboratory mammals, and
man. Anat Rec 117: 241–263, 1953.
14. Macklin CC. Pulmonic alveolar epithelium. J Thorac Surg 6: 82–88, 1936.
15. Maina JN, West JB. Thin and strong! The bioengineering dilemma in the
structural and functional design of the blood-gas barrier: comparative and
evolutionary perspectives. Physiol Rev 85: 811–844, 2005.
16. Miller WS. The epithelium of the lower respiratory tract. In: Section V.
Special Cytology (2nd ed.), edited by Cowry LV. New York: Hoeber,
1932, P. 133–150.
17. Policard A. Les nouvelles ideés sur la disposition de la surface respiratoire pulmonaire. Presse Med 37: 1293–1295, 1929.
18. Poole DC, Erickson HH. Highly athletic terrestrial mammals: horses and
dogs. Comp Physiol 1: 1–37, 2011.
19. Potter EL, Loosli CG. Prenatal development of the human lung. AMA Am
J Dis Child 82: 226 –228, 1951.
20. Reeves JT, Groves BM, Cymerman A, Sutton Wagner PD, Turkevich
D Jr, Houston CS. Operation Everest II: cardiac filling pressures during
cycle exercise at sea level. Respir Physiol 80: 147–154, 1990.
21. Schoene RB, Hackett PH, Henderson WR, Sage EH, Chow M, Roach
RC, Mills WJ, Martin TR. High-altitude pulmonary edema. Characteristics of lung lavage fluid. J Am Med Assoc 256: 63–69, 1986.
22. Starling EH. On the absorption of fluids from the connective tissue
spaces. J Physiol 19: 312–326, 1896.
23. Tsukimoto K, Mathieu-Costello O, Prediletto R, Elliot AR, West JB.
Ultrastructural appearances of pulmonary capillaries at high transmural
pressures. J Appl Physiol 71: 573–582, 1991.
24. Wagner PD, Gale GE, Moon RE, Torre-Bueno JR, Stolp BW, Saltzman HA. Pulmonary gas exchange in humans exercising at sea level and
simulated altitude. J Appl Physiol 61: 260 –270, 1986.
25. Weibel ER. Morphological basis of alveolar-capillary gas exchange.
Physiol Rev 53: 419 –495, 1973.
26. West JB. Thoughts on the pulmonary blood-gas barrier. Am J Physiol
Lung Cell Mol Physiol 285: L501–L513, 2003.
27. West JB. Comparative physiology of the pulmonary circulation. Comp
Physiol 1: 1525–1539, 2011.
28. West JB, Tsukimoto K, Mathieu-Costello O, Prediletto R. Stress
failure in pulmonary capillaries. J Appl Physiol 70: 1731–1742, 1991.
29. West JB, Mathieu-Costello O, Jones JH, Birks EK, Logerman RB,
Pascoe JR, Tyler WS. Stress failure of pulmonary capillaries in racehorses with exercise-induced pulmonary hemorrhage. J Appl Physiol 75:
1097–1109, 1993.
30. West JB, Fu Z, Deerinck TJ, Mackey MR, Obayashi JT, Ellisman
MH. Structure-function studies of blood and air capillaries in chicken lung
using 3D electron microscopy. Respir Physiol Neurobiol 170: 202–209,
2010.