Functional diameters of alveolar microvessels
at high lung volume in zone II
ROBERT L. CONHAIM AND LANCE A. RODENKIRCH
Department of Surgery, University of Wisconsin-Madison, Madison, Wisconsin 53705
isolated rat lungs; confocal microscopy; rapid freezing; pulmonary microcirculation; latex particle perfusion
are subjected to a range of
perfusion and inflation pressures that affect their functional diameters. In a recent report (2), our laboratory
showed in zone I, where inflation pressure exceeds
perfusion pressure, that these microvessels were compressed but not obstructed. In a subsequent report (3),
we estimated the functional diameter of these vessels
in zone I to be 1.7 µm. When we raised the pulmonary
arterial pressure (Ppa) to equal the inflation pressure
(zone I/II border), we estimated that the microvessel
diameter would increase to 6–8 µm (4).
In those previous studies, all lungs were inflated to
near-total lung capacity (25 cmH2O). Ppa was set to
either 20 cmH2O (zone I) or 25 cmH2O (zone I/II border)
(14). Results of those studies showed that a rise in Ppa
from 5 cmH2O below inflation pressure to a value equal
to inflation pressure caused the functional diameter to
rise by 4–6 µm.
We also found in those studies that the fraction of
alveoli containing particles rose with the Ppa. We found
that 13% of alveoli contained 1-µm-diameter particles
in zone I, but this increased to 27% at the zone I/II
border. Thus both the fraction of perfused alveoli and
the microvessel diameter rose with Ppa.
In this report, we describe results of perfusing lungs
in zone II. As in our previous studies (2–4), the lungs
were inflated to 25 cmH2O. However, in these studies,
Ppa was set to 30 cmH2O so that it exceeded inflation
pressure by 5 cmH2O. In addition, we also measured for
the first time the particle concentrations in the venous
ALVEOLAR MICROVESSELS
http://www.jap.org
effluent. We used these data as an index of the size of
the particles trapped within specific microvascular
segments.
METHODS
Animal preparation. Eighteen male rats (450–550 g) were
anesthetized with intraperitoneal ketamine (40 mg/kg), xylazine (6 mg/kg), and acepromazine (1 mg/kg) and placed in a
supine position. A polyethylene cannula (PE-60) was inserted
into the femoral vein, and heparin (750 U/kg) was injected
through it. After 10 min to allow the heparin to circulate, we
severed the femoral artery and let the animal exsanguinate.
The shed blood volume was recorded minute by minute
during exsanguination, and an equal volume of Ringer lactate
was infused to facilitate the washout of red blood cells from
the circulation. This improved subsequent perfusion of the
isolated lungs. The volume of Ringer lactate infused equaled
the volume of blood shed and averaged 25–35 ml (5–6% of the
animal’s body wt). We collected blood and infused Ringer
lactate until the animal’s spontaneous breathing ceased.
After tying a polyethylene cannula (PE-200) into the
trachea, we inflated the lungs to 5–10 cmH2O using a small
air compressor. We then opened the chest with a sternumsplitting incision, cut the thoracic walls along the ribs above
the diaphragm, and tied back the walls to fully expose the
heart and lungs.
A liquid-filled cannula (PE-200) was inserted into the
pulmonary artery through an incision in the right ventricle
and was secured with no. 000 silk. We placed another
liquid-filled cannula into the left atrium through an incision
in the left ventricle and anchored it with a loop of suture
around the base of the heart. The cannulated heart and lungs
were then removed from the chest and placed dorsal side
down into a styrofoam perfusion chamber (vol 300 ml). After
passing the cannulas through the chamber walls, we set the
outlet of the atrial cannula level with the surface on which the
lungs rested [left atrial pressure (Pla) 5 0 cmH2O]. A drop
counter placed directly below the cannula outlet recorded
perfusate outflow. The arterial cannula was connected through
a Y to two reservoirs placed on a bench jack. One of the
reservoirs contained Ringer lactate, and the other contained
the latex perfusate.
The latex particle perfusate consisted of 0.1% latex by
weight (Molecular Probes) in phosphate-buffered saline.
Therefore, the particle concentration was inversely proportional to the particle diameter, although all lungs received
equal latex concentrations (0.1%). We added 0.5% bovine
serum albumin to the perfusate to coat the particles and
neutralize their surface charge. This prevented the particles
from clumping in the presence of saline ions and minimized
hindrance by the vascular endothelium. We also added a
fluorescent membrane stain [5-(N-dodecanoyl) aminofluorescein (Molecular Probes, Eugene, OR)] for subsequent visualization of the tissue in the fluorescence microscope. Latex
particles of one specific diameter (1.0, 2.0, 4.0, 5.0, 6.0, or 10.0
µm) were infused into each lung. A total of three lungs each
was prepared with 1.0-, 2.0-, 4.0-, or 5.0-µm-diameter particles for confocal microscopic analysis. We prepared addi-
8750-7587/98 $5.00 Copyright r 1998 the American Physiological Society
47
Downloaded from http://jap.physiology.org/ by 10.220.32.246 on April 2, 2017
Conhaim, Robert L., and Lance A. Rodenkirch. Functional diameters of alveolar microvessels at high lung volume
in zone II. J. Appl. Physiol. 85(1): 47–52, 1998.—To estimate
the functional diameter of alveolar microvessels, we perfused
isolated rat lungs with fluorescent latex particles (1 diameter/
lung) at inflation, pulmonary arterial, and left atrial pressures of 25, 30, and 0 cmH2O, respectively. We used confocal
microscopy to count latex particles within septal microvessels
and flow cytometry to count particle concentrations in venous
outflow. We found 1-, 2-, and 4-µm-diameter particles within
septal vessels of 45 6 12, 31 6 12, and 25 6 9%, respectively,
of examined alveoli. Particles of 5-µm diameter were absent
from septal vessels but were present within a small percentage of corner vessels. Particle concentrations in the venous
outflow for 1-, 2-, 4-, and 5-µm-diameter particles were 54 6
28, 67 6 32, 2.2 6 0.3, and 0.4 6 0.3%, respectively, of the
arterial inflow. Particles with diameters of 6 or 10 µm were
absent from venous outflow. Our results suggest that, under
these conditions, the functional diameter of the septal microvessels is ,4 µm and that the diameter of the adjacent
corner vessels is slightly larger but ,6 µm.
48
ALVEOLAR MICROVESSEL DIAMETERS
the fact that digital images are composed of discrete picture
elements (pixels). Our images measured 768 pixels wide by
512 pixels high (393,216 pixels/image); each pixel had a
gray-scale value that ranged from 0 (white) to 255 (black).
The gray-scale value was proportional to fluorescence intensity.
The image analyses, performed by computer, consisted of
simply counting the pixels caused by latex particle fluorescence within a defined portion of the septal plane (version
1.52, Image, National Institutes of Health Research Services
Branch). The defined portion consisted of the central ,50% of
the septal image. This ensured that peripheral septal areas
that might be considered to be corner vessels were excluded.
The area of the counted portion was measured from the
Downloaded from http://jap.physiology.org/ by 10.220.32.246 on April 2, 2017
tional lungs for measurement of venous effluent concentrations (Table 1). A total of 20 lungs were prepared.
Lung perfusate flows were recorded with a drop counter,
and pressures were recorded with transducers (Statham,
Hato Rey, Puerto Rico) connected to an oscillograph (Grass
Instruments, Quincy, MA). Ppa and Pla were referenced to
the surface on which the lungs lay dorsal side down. Because
the dorsal-to-ventral dimension of the lungs averaged 2–3
cm, vascular pressures at the ventral (upward) surface were
2–3 cmH2O less than those at the dorsal (downward) surface.
After the inflation pressure [alveolar pressure (PA )] was set to
25 cmH2O, perfusion was begun by raising the Ringer lactate
reservoir to set the Ppa to 30 cmH2O. With the Pla at 0 cmH2O
(bottom of the lung), this produced a double-occlusion pressure of 19 cmH2O (8). Thus inflation pressure exceeded
estimated microvascular pressure by 6 cmH2O (zone II) (14).
After 10 min of perfusion with Ringer lactate to allow the flow
to stabilize, we began perfusion with the latex solution. The
latex reservoir was set at equal height to the Ringer lactate
reservoir so that no change in Ppa would occur as latex
perfusion began.
During latex particle perfusion, we collected samples of the
venous effluent for subsequent measurement of latex particle
concentrations. Five samples were collected at evenly spaced
intervals throughout the 10-min perfusion period.
After 10 min of latex perfusion, the lungs were rapidly
frozen by pouring chilled (2150°C) isopentane into the styrofoam reservoir.
Histology. The lungs were removed from the isopentane
and placed into liquid nitrogen where they were cut, by using
a scalpel, into blocks of ,73737 mm. Ten blocks, representing ,30% of the total lung volume, were randomly selected
and placed into dehydrated, chilled (270°C) ethanol. The
blocks were stored in a freezer (270°C) for 3–5 days to
dehydrate the tissue by freeze substitution. After being
warmed (4°C), the ethanol was replaced with unpolymerized
embedding resin (JB-4; Polysciences), and the blocks were
placed under vacuum to remove residual air. The blocks were
then embedded in polymerized resin. We cut three to four
sections (75-µm-thick) from each block using a heavy-duty
microtome (Reichert Ultracut) and mounted the sections
unstained on glass slides.
The sections were examined by using a confocal laser
scanning fluorescence microscope (MRC 600, Bio-Rad)
equipped with a krypton-argon laser, two sets of excitationemission filters, and two solid-state photodetectors. With this
instrument, we were able to view fluorescence simultaneously from tissue fluorescein and latex rhodamine on the
microscope’s computer monitor. We were also able to save the
viewed images on disk. With the use of a confocal microscope,
only the tissue within the plane of focus is visible (1, 15). This
allowed us to scan the tissue by focusing up and down in the
vertical (z) axis (optical sectioning), as well as by moving the
stage (x- and y-axes). We scanned the tissue until we located a
latex-containing alveolar septum seen face on (Fig. 1). We
used the following criteria (3). In the confocal microscope,
only tissue within the plane of focus was visible. Therefore,
we required that the entire septum appear and disappear
within a focusing range of ,6–8 µm. Furthermore, only
alveolar air space could be present above and below the septal
plane. This technique ensured that we were not viewing
corner vessels, which had a smaller overall cross-sectional
area than septa had and which appeared and disappeared
from the focal plane over a focusing range of 20–40 µm.
We saved the image of the septum on disk and then used
image-analysis techniques to measure the latex particle
density within it (particles/µm2 ). The image analysis utilized
Fig. 1. Confocal optical sections from lungs perfused with 4- (A) and
5-µm-diameter (B) fluorescent latex particles. The 4-µm-diameter
particles (arrows, A) were able to enter septal vessels under zone II
conditions, but 5-µm-diameter particles (arrow, B) were not. These
results are consistent with our hypothesis that septal microvessel
diameter in these conditions lies between 4 and 5 µm. Optical section
thickness (A and B) is 0.3 µm; scale is 25 µm. A lies within alveolar
septal plane; B lies within air space above septal plane. See Fig. 2 for
three-dimensional reconstructions of A.
49
ALVEOLAR MICROVESSEL DIAMETERS
RESULTS
Perfusate flows. During 10 min of perfusion with
Ringer lactate before latex perfusion, flows averaged
7.2 6 3.0 ml/min. During the 10 min of subsequent
latex perfusion before freezing, flows averaged 6.3 6
3.4 ml/min. Flows did not differ significantly among
lungs perfused with particles of each diameter.
Particle densities. Average particle densities measured within septa ranged from 0.068 6 0.024 to
0.015 6 0.007 particles/µm2 for 1.0- and 4.0-µmdiameter particles, respectively (Table 2). We found no
examples of 5-µm-diameter particles within septa, although we did see examples of these particles within
corner vessels (Fig. 1). Densities for particles of any one
diameter differed significantly from densities of the
others (P # 0.05).
Fraction of septa with latex. In lungs perfused with
1.0-µm-diameter particles, the fraction of septa found
to contain latex was 0.45 6 0.12 (Table 2). This fraction
fell to 0.31 6 0.12 and 0.25 6 0.09 for the 2.0- and
4.0-µm-diameter particles, respectively. The 5.0-µmdiameter particles were absent from septa.
Table 1. Latex particle concentrations
in venous effluent
Latex Particle
Diameter, µm
n
Venous Concentration
(Fraction of Arterial
Concentration)
1.0
2.0
4.0
5.0
6.0
10.0
2
3
4
3
2
2
0.54 6 0.28
0.67 6 0.32
0.022 6 0.026*†
0.004 6 0.003*†
0
0
Values for venous concentration are averages of 5 measurements/
lung (means 6 SD). n, No. of lungs. * Significantly different from
1.0-µm-diameter particles, P # 0.5; † significantly different from
2.0-µm-diameter particles, P # 0.5.
Venous particle concentrations. Particle concentrations in venous effluent, expressed as fractions of the
concentrations in the arterial inflow, were 0.54 6 0.28
and 0.67 6 0.32 (not significant) for lungs perfused with
1.0- and 2.0-µm-diameter particles, respectively (Table
1). The venous concentration ratios dropped to 0.022 6
0.026 and 0.004 6 0.003 for lungs perfused with 4.0and 5.0-µm-diameter particles, respectively. Particles
of 6 and 10 µm diameter were absent from the venous
effluent.
DISCUSSION
Our data show that, in the zone II conditions under
which the lungs were perfused, particles with diameters of 1.0, 2.0, and 4.0 µm entered septa, whereas
5.0-µm-diameter particles did not. However, 5.0-µmdiameter particles entered corner vessels adjacent to
septa, whereas 6.0-µm-diameter particles did not. These
results suggest that, under these conditions, the largest alveolar septal microvessels had a functional diameter .4.0 but ,5.0 µm and further suggest that corner
vessels had diameters between 5.0 and 6.0 µm.
This conclusion is supported by the venous effluent
concentrations, which were near zero for the 5.0-µmdiameter particles and at zero for the 6.0-µm-diameter
particles. Concentrations were only 0.022 6 0.026 for
the 4.0-µm-diameter particles, suggesting that 4.0-µmdiameter particles were trapped within septal vessels,
whereas 5.0-µm-diameter particles were trapped within
corner vessels. The 1.0- and 2.0-µm-diameter particles
were partially trapped by septal vessels, based on their
Table 2. Latex particle densities within septa and the
fraction of alveolar septa containing latex particles
Latex Particle
Diameter, µm
1.0
2.0
4.0
5.0
n
Particle Densities
within Septa,
particles/µm2
Fraction of
Septa Containing
Particles
3
3
3
3
0.068 6 0.024
0.032 6 0.013*
0.015 6 0.007*†
0
0.45 6 0.12
0.31 6 0.12
0.25 6 0.09*
0
Values for particle densities within septa are averages from 9
alveoli/lung (means 6 SD). Values for fraction of septa containing
particles are averages of 110 alveoli examined/lung. n, No. of lungs.
* Significantly different from 1.0-µm-diameter particles, P # 0.5;
† significantly different from 2.0-µm-diameter particles, P # 0.5.
Downloaded from http://jap.physiology.org/ by 10.220.32.246 on April 2, 2017
product of the number of pixels within it and the area per
pixel (determined from the magnification). Pixels caused by
latex fluorescence (gray-scale values .128) within the portion
were also counted and divided by the number of pixels per
particle to yield the number of particles (3). Dividing the
number of particles by the area produced the particle density
(particles/µm2 ). We measured densities in 9 alveoli/lung.
We also counted the fraction of septa that contained latex,
because we found that the number of alveoli containing latex
increased with the Ppa. Ten fields averaging 10 alveoli each
were examined in each lung (,100 alveoli/lung). The number
of alveoli in each field that contained latex was counted and
divided by the number of alveoli present in the field to express
the fraction of septa that contained latex.
We prepared three-dimensional reconstructions of latexcontaining tissue from individual confocal images. The reconstructions were prepared from individual images (optical
sections) that were obtained by focusing sequentially downward through the tissue in 0.2-µm steps. The stack of sections
was assembled digitally into a three-dimensional image that
could be viewed by using a computer (Macintosh Quadra
8500). We used Voxblast software (version 1.0; Vaytek, Fairfield, IA) designed specifically for this purpose.
Venous particle concentrations. Latex particle concentrations in venous effluent were measured by using a flow
cytometer (FACScan, Becton Dickinson). The cytometer was
calibrated so that each latex perfusate produced 2,000–4,000
forward and side-scatter events per sample, corresponding to
one event per particle. Particle events in venous samples
were then measured and expressed as fractions of those in the
perfusate. Five venous samples were measured from each
perfused lung. Results were averaged among all lungs perfused with particles of each diameter.
Statistics. Data obtained for particles of each diameter
were pooled and expressed as means 6 SD. The number of
measurements made and the number of alveoli studied are
shown in Tables 1 and 2, respectively. Results among particle
diameters were compared by using one-way analysis of
variance. Post hoc comparisons were performed by using
Fisher’s paired least-significant difference test (version 1.03,
StatView; Abacus Concepts, Berkeley, CA). Differences were
considered to be significant at P # 0.05.
50
ALVEOLAR MICROVESSEL DIAMETERS
Table 3. Latex particle densities in alveolar septa of
isolated rat lungs at the zone I/II border and in zone I
Particle
Diameter, µm
Particle Densities
Within Septa,
particles/µm2
Fraction of Septa
Containing
Particles
Zone I/II border
0.5
1.0
2.0
4.0
0.062 6 0.019
0.038 6 0.013
0.019 6 0.008
0.007 6 0.004
0.31 6 0.08
0.27 6 0.12
0.16 6 0.04
0.09 6 0.03
5 cmH2O into zone I
0.24
0.5
1.0
0.077 6 0.020
0.041 6 0.025*
0.022 6 0.009*†
0.32 6 0.06
0.25 6 0.07
0.13 6 0.04
Values are means 6 SD. * Significantly different from 0.24-µmdiameter particles, P # 0.05; † significantly different from 0.5-µmdiameter particles, P # 0.05.
ticles, particle densities within septa increased from
0.019 6 0.008 particles/µm2 at the zone I/II border to
0.032 6 0.013 particles/µm2 in zone II. The fraction of
septa perfused with these particles increased from
0.16 6 0.04 to 0.31 6 0.12, respectively. A similar trend
existed for the 4.0-µm-diameter particles. Particle densities increased from 0.007 6 0.004 particles/µm2 at the
zone I/II border to 0.015 6 0.007 particles/µm2 in zone
II. The fraction of septa perfused with 4.0-µm-diameter
particles increased from 0.09 6 0.03 to 0.25 6 0.09,
respectively. Thus the trend was toward higher densities within septa and greater numbers of perfused
septa as PA was raised from below to above inflation
pressure.
In our previous studies (2–4), we estimated functional diameters of septal microvessels by fitting curves
produced by the Verniory equation to the particle
density data. The Verniory equation was originally
developed to estimate diameters of transendothelial
pores, based on interstitium-to-plasma concentration
ratios for solutes of specific molecular radii (12). We
hypothesized that exclusion of latex particles by narrowed septal microvessels could be likened to the
reflection of plasma solutes by small transendothelial
pores. Using this technique, we estimated diameters of
septal microvessels to be 1.7 µm in zone I and 6–8 µm
at the zone I/II border. In these previous studies, we did
not measure venous effluent particle concentrations
but assumed that particles that did not enter septa
were reflected away into corner vessels, much as large
plasma solutes were reflected by small endothelial
pores.
The venous concentration measurements in our present study show that this analogy is inaccurate. Particles that could not enter septal vessels, such as the
5.0- and 6.0-µm-diameter particles, had venous concentrations near zero, indicating that they could not flow
through the lung. Thus they were not reflected away
from the septal microvessels but were trapped within
adjacent, larger vessels.
This finding means that the septal diameter estimates of our previous studies should be revised. In zone
I, on the basis of Verniory curves, we estimated the
functional diameter of septal microvessels to be 1.7 µm.
However, we found that 1.0-µm-diameter particles had
very low particle densities within septa (0.022 6 0.009
particles/µm2, Table 3) and found that 4.0-µm-diameter
particles were excluded from septa. We did not perfuse
with 2.0-µm-diameter particles, which were not available at that time. Based solely on the particle density
data, the 1.7-µm-diameter estimate seems reasonable.
At the zone I/II border, on the basis of Verniory
curves, we estimated the septal microvessel diameter
to be 6–8 µm. However, under these conditions, we
found that 4.0-µm-diameter particles had very low
particle densities (0.007 6 0.004 particles/µm2, Table
3). We did not perfuse with 5.0-µm-diameter particles,
but, on the basis of the low particle density of the
4.0-µm-diameter particles and the fact that 5.0-µmdiameter particles were excluded from septa in zone II
(Table 2), it seems likely that these larger particles
Downloaded from http://jap.physiology.org/ by 10.220.32.246 on April 2, 2017
venous concentrations, but were able to flow through
the lungs with much less restriction than the larger
particles could.
Our results also suggest that no significant shunt
pathway around the corner vessel-septal vessel network was present within these lungs. If a shunt pathway existed, we would have expected higher venous
concentrations of particles with diameters of 5.0 µm
and larger.
Our results can be better appreciated by comparing
them with the results of our previous studies conducted
in zone I and at the zone I/II border (3, 4). In those
studies, the lungs were inflated to the same pressures
as those described here (25 cmH2O) but were perfused
at Ppa of 20 or 25 cmH2O, respectively. The latex
concentrations in the perfusates and the number of
minutes the lungs were perfused are identical to the
values in our present study. Therefore, it is reasonable
to compare results of those studies with our present
one. Particle densities and the fraction of septa that
contained latex in the previous studies are shown in
Table 3.
Comparing septal densities for particles of common
diameter among the studies shows that densities were
higher in zone II than either at the zone I/II border or in
zone I. Because all lungs were inflated to the same
pressure, these results suggest that the septal capillary
diameter increased with the vascular pressure. For
example, particles with diameters of 1.0 µm had septal
densities of 0.068 6 0.024 particles/µm2 in zone II
(Table 2), 0.038 6 0.013 particles/µm2 at the zone I/II
border (Table 3), and 0.022 6 0.009 particles/µm2 in
zone I (Table 3). The fraction of septa containing
particles of 1.0-µm diameter decreased from 0.45 6
0.12 in zone II to 0.27 6 0.12 at the zone I/II border and
to 0.13 6 0.04 in zone I. Thus both the particle densities
and the fraction of perfused septa rose as Ppa was
raised from below PA (zone I) to PA (zone I/II border)
and above PA, suggesting that the functional diameter
of the septal microvessels rose.
This conclusion is also supported by data of the 2.0and 4.0-µm-diameter particles used in the zone II and
zone I/II border studies. For the 2-µm-diameter par-
ALVEOLAR MICROVESSEL DIAMETERS
ments in the casts had diameters of 27.4 6 10.1 µm
when vascular pressure was 2 cmH2O lower than
inflation pressure and 23.6 6 9.4 µm when vascular
pressure was 17 cmH2O lower than inflation pressure.
They concluded that the diameters of these smallest
patent segments represented choke points in the venous circulation, upstream from which smaller veins
and capillaries were collapsed. They concluded that
these choke points were responsible for flow limitation
in zone II and suggested that the collapsible segments
operated in an all-or-none fashion, remaining collapsed
Downloaded from http://jap.physiology.org/ by 10.220.32.246 on April 2, 2017
would have been excluded from septa at the zone I/II
border. A more likely microvessel diameter, based solely
on the density data, would be near but slightly larger
than 4.0 µm.
This is similar to our estimate of the microvessel
diameters in zone II. However, we found that the
densities of 4.0-µm-diameter particles in zone II were
greater than those at the zone I/II border (0.015 6
0.007 vs. 0.007 6 0.004 particles/µm2; P # 0.05). This
suggests that the microvessel diameter was larger in
zone II, but the difference must be ,1.0 µm, because
5.0-µm-diameter particles were excluded from septa in
zone II. Thus the septal microvessel diameters must be
4.0 µm , zone I/II border , zone II , 5.0 µm. This
conclusion is supported by our finding that 25 6 9% of
alveoli contained 4.0-µm-diameter particles in zone II
(Table 2), whereas only 9 6 3% of alveoli contained
these particles at the zone I/II border (Table 3).
Lamm and colleagues (10), describing red blood cell
flow through septa at the pleural surface of intact dog
lungs in zone I, reported that red blood cells could be
seen flowing through corner vessels but not septal
vessels. They reported that, as the lung was placed into
zone II, red blood cell flow through septal vessels began.
Based on our diameter estimates for these vessels and
considering the fact that our studies were not conducted in dogs, the results of Lamm and colleagues
suggest that entry of red blood cells into septa in zone II
may require some red blood cell deformation. If red
blood cells have an undeformed diameter of 5.0 µm and
if the septal microvessel diameter is ,5.0 µm, as our
results suggest, then some red blood cell deformation
would be necessary for entry into septa in zone II.
Our septal microvessel diameter estimates are similar to those of others. Guntheroth et al. (7) measured
alveolar microvessel diameters from scanning-electron
micrographs of latex casts made from rat lungs perfused in zone III. They estimated the microvessel
diameter to be 5.8 µm. Glazier and colleagues (6)
estimated the diameters to be 2.0–2.8 µm near the zone
I/II border. In zone II, they reported that microvessel
diameters ranged from 2.8–5.0 µm from the top of the
lung to the bottom. Weibel (13) and Pump (11) reported
diameters of 6.0 µm for fixed human lungs.
On the other hand, our results are markedly different
from those of Fung and colleagues (5), who investigated
the patency of pulmonary veins under conditions in
which inflation pressure exceeded venous pressure.
Their findings are relevant to our present study, because we also set inflation pressure above venous
pressure. These investigators filled the vasculature of
cat lungs with casting resin and allowed the resin to
polymerize while the inflation pressure was set at 10
cmH2O; the vascular pressure was set 2–17 cmH2O
lower than the inflation pressure (5). Arterial and
venous pressures were equal. Examination of the casts
revealed that capillary segments were absent. They
concluded that, with inflation pressure set above vascular pressure, the capillaries must have been compressed to the point that the resin was excluded. They
also found that the smallest pulmonary venous seg-
51
Fig. 2. Three-dimensional reconstructions of optical sections from a
lung perfused with 4-µm-diameter fluorescent latex particles. A: in
low-power view, particles (arrows) appear to be confined to a corner
vessel pathway that lies between adjacent alveoli. Particles have
entered septa of one alveolus (*), which is shown enlarged in B. A total
of 49 optical sections were assembled digitally to produce A; 109
sections were assembled to produce B (Voxblast, version 1.0). One
section appears in Fig. 1A. Reconstructions are displayed obliquely
(right upper corner tipped back) to enhance 3-dimensional effect.
Scale is 75 µm (A) and 25 µm (B).
52
ALVEOLAR MICROVESSEL DIAMETERS
Furthermore, capillary opening pressures and functional diameters undoubtedly vary among septa.
We conclude that the largest functional diameter of
alveolar septal microvessels at high lung volumes in
zone II lies between 4.0 and 5.0 µm. This is similar to
our revised estimated diameter for these vessels at the
zone I/II border, although the functional diameter in
zone II appears to be larger by ,1.0 µm. These estimates apply only to the specific zone II pressure
conditions under which the lungs were perfused.
This study was supported by National Heart, Lung, and Blood
Institute Grant HL-49985. The confocal laser-scanning microscope
was provided by the Integrated Microscopy Resource (IMR) of the
University of Wisconsin-Madison. The IMR is a Biomedical Research
Technology Resource of the National Institutes of Health.
Address for reprint requests: R. L. Conhaim, Dept. of Surgery,
B-7055A, Veterans’ Hospital, 2500 Overlook Terrace, Madison, WI
53705 (E-mail: [email protected]).
Received 23 October 1997; accepted in final form 4 March 1998.
REFERENCES
1. Brankenhoff, G. J., P. Blom, and P. Barends. Scanning light
microscopy with high aperature lenses. J. Microsc. 117: 219–232,
1979.
2. Conhaim, R. L., and B. A. Harms. Perfusion of alveolar septa
in isolated rat lungs in zone 1. J. Appl. Physiol. 75: 704–711,
1993.
3. Conhaim, R. L., and L. A. Rodenkirch. Functional diameter of
alveolar septal microvessels 5 and 10 cmH2O into zone 1. Am. J.
Physiol. 271 (Heart Circ. Physiol. 40): H996–H1003, 1996.
4. Conhaim, R. L., and L. A. Rodenkirch. Estimated functional
diameter of alveolar septal microvessels at the zone I-II border.
Microcirculation 4: 51–59, 1997.
5. Fung, Y.-C., S. S. Sobin, H. Tremer, M. R. T. Yen, and H. H.
Ho. Patency and compliance of pulmonary veins when airway
pressure exceeds blood pressure. J. Appl. Physiol. 54: 1538–
1549, 1983.
6. Glazier, J. B., J. M. B. Hughes, J. E. Maloney, and J. B. West.
Measurements of capillary dimensions and blood volume in
rapidly frozen lungs. J. Appl. Physiol. 26: 65–76, 1969.
7. Guntheroth, W. G., D. L. Luchtel, and I. Kawabori. Pulmonary microcirculation: tubules rather than sheet and post. J.
Appl. Physiol. 53: 510–515, 1982.
8. Hakim, T. S., R. P. Michel, and H. K. Chang. Partitioning of
pulmonary vascular resistance in dogs by arterial and venous
occlusion. J. Appl. Physiol. 52: 710–715, 1982.
9. Hanger, C. C., R. G. Presson, Jr., O. Okada, S. J. Janke, J. J.
Watkins, W. W. Wagner, Jr., and R. L. Capen. Computer
determination of perfusion patterns in pulmonary capillary
networks. J. Appl. Physiol. 82: 1283–1289, 1997.
10. Lamm, W. J. E., K. R. Kirk, W. L. Hanson, W. W. Wagner, and
R. K. Albert. Flow through zone 1 lungs utilizes alveolar corner
vessels. J. Appl. Physiol. 70: 1518–1523, 1991.
11. Pump, K. K. The circulation in the peripheral parts of the
human lung. Dis. Chest 49: 119–129, 1966.
12. Verniory, A., R. DuBois, P. Decoodt, J. P. Gassee, and P. P.
Lambert. Measurement of the permeability of biological membranes. J. Gen. Physiol. 62: 489–507, 1973.
13. Weibel, E. R. Morphometry of the Human Lung. Berlin: SpringerVerlag, 1963, p. 78–82.
14. West, J. B., C. T. Dollery, and A. Naimark. Distribution of
blood flow in isolated lung; relation to vascular and alveolar
pressures. J. Appl. Physiol. 19: 713–724, 1964.
15. White, J. G., W. B. Amos, and M. Fordham. An evaluation of
confocal versus conventional imaging of biological structures by
fluorescence light microscopy. J. Cell Biol. 138: 41–48, 1987.
Downloaded from http://jap.physiology.org/ by 10.220.32.246 on April 2, 2017
until upstream pressure rose enough to open them,
then collapsing again as the pressure was relieved.
Our results suggest that flow limitation in zone II
occurs because capillaries are narrowed but not collapsed. However, our experiments were conducted under different conditions: we set inflation pressure 15
cmH2O higher than did Fung and colleagues (5) and
allowed the lungs to perfuse continuously. At the high
inflation pressures we used, capillaries may be stretched
and less collapsible than at the lower pressures used by
Fung et al. They suggested that pulmonary veins with
diameters larger than the choke point remained open
because of tethering by surrounding alveoli. However,
at high lung volumes, vascular tension may be great
enough to maintain patency of all vessels, including
septal microvessels.
These differences between our results and those of
Fung and colleagues (5) emphasize the point that zone
II is not a specific set of inflation and perfusion pressures but is any condition in which inflation pressure
lies between arterial and venous pressures. Septal
microvessel diameters in zone II at high lung volumes
may be different from those at low or intermediate
volumes, which is a possibility that deserves to be
addressed by future studies.
A disadvantage of our analytic approach is that we
were unable to view particle motion within capillaries
as the lung was perfused. Therefore, we could not
evaluate the dynamic properties of capillary particle
deposition. However, some aspects of these properties
can be inferred from the figures. In Fig. 2A, an interseptal corner vessel filled with 4-µm-diameter particles is
oriented diagonally across the view. Few of the alveoli
supplied by it contain particles, suggesting that few of
their septal capillaries were large enough to admit
these particles. The implication is that the majority of
capillary entrance diameters under these conditions
was ,4 µm. This would explain why we found that only
25% of the examined alveoli contained particles of this
size (Table 2). This suggests that capillary diameters
are not uniform; only the largest ones may have been
able to admit 4-µm-diameter particles.
Within the capillary network of individual septa, it
has been shown that flow can follow nonrandom pathways through the multiple segments that make up the
network (9). Subtle differences in pressure distribution
within the network may be responsible for this phenomenon. Similarly, subtle pressure differences at the
network entrance might determine whether or not
4-µm-diameter particles can enter. We did not raise
Ppa above 30 cmH2O, but, if we had, more septa might
have contained these particles. We also did not ventilate the lungs. Varying the difference between inflation
and vascular pressures while maintaining the lung in
zone II might also have caused 4-µm-diameter particles
to enter more septa.
Our point is that subtle pressure differences might
affect the number of septa that particles can enter.