JOURNALOF
Vol.
A~pmm
38, No. 1,January
PHYSIOLOGY
2975.
Distribution
Printed
in U.S.A.
of red and white
cells in alveolar
blood
walls
S. PERLO, A. A. J’ALOWAYSKI, C. M. DURAND, AND J. B. WEST De~artmenf of Medicine, School of Medicine, Universiy of California San Diego, La Jolla, Gulifurnia
PERLO, &A. A. JALOWAYSKI, C. M. DURAND, ~z’stributim of red and white blood cells in alveolar
J. B. WEST. walls. J- Appl.
AND
Yhysiol. 38( 1) : 117-124. 1975.~Dog lungs were perfused with blood and rapidly frozen with liquid Freon gas at various pulmonary artery and venous pressures. The numbers of red and white blood cells per mm2 of alveolar wall were counted in lung sections and, in addition, the proportion (by area) of the wall occupied by the cells was measured by point counting. The number and proportional area of the red blood cells rapidly increased as perfusing pressure was raised. These findings are consistent with earlier observations of capillary recruitment and distension. An unexpected observation was the large number of leukocytes in the capillaries especially at low perfusing pressures. For example when arterial exceeded alveolar pressure by 5 cmHn0 (as occurs near the apex of the upright human lung), there were about 5,000 red cells and 4,000 white cells per mm2 of alveolar wall* As perfusing pressure was increased, the number of leukocytes paradoxically decreased in zone 3 but remained constant in zone 2. Most of the white cells were mononuclear cells. These results suggest: that the lung behaves as a mechanical sieve for large cells and that the number of trapped cells depends on the capillary pressure. recruitment; Aow behavior
distension;
sequestration;
zones
of
lung;
pressure-
THERE HAVE BEEN several recent studies of the morphology of the pulmonary microcirculation including the concentration of red blood cells in the capillaries, the number of open capillaries, and their dimensions in relation to the pulmonary vascular pressures (13, 28-30) examining rapidly frozen sections of dog and cat lung. Sobin and his colleagues (26) have emphasized the large proportion of the alveolar wall which can be covered by open capillaries as viewed en face but there is uncertainty about how this area is related to capillary pressure. The present study was designed to determine the morphology of the pulmonary capillaries in the dog viewed en face in relation to the pulmonary arterial, venous, and alveolar pressures. An unexpected finding was the large proportion of white blood cells in the capillaries especially at low perfusing pressures. METHODS
Lung Pre$mation The preparation has been described in detail elsewhere (13, 30) and will only be summarized here. Healthy mongrel dogs weighing between 16 and 30 kg were anesthetized
9.2037
with Nembutal sodium, intubated, and ventilated with a Harvard respirator. A tracheotomy was performed, a largebore catheter was advanced into the inferior vena cava from a femoral vein cutdown, and the heart was exposed through a left thoracotomy. The animal was then heparinized (2,000 units/kg) and 400-500 ml of blood were removed over a 3- to 5-min period. The heart was then fibrillated with an electrocautery, the chest wall removed, and the pulmonary artery and left atrium cannulated in a rapid sequence that averaged less than 10 min. After cannulation the lungs were gently inflated with 100 % 02 until all areas were fully expanded and the dog secured upright. The dog was ventilated with 6 % CO2 in air and the lungs were perfused with blood from the venous reservoir by means of a roller pump. The arterial venous and tracheal pressures were measured by saline manometers. Either the right or left lung was selected for rapid freezing based on its appearance and amount of tissue available. The pleural surface was marked with methylene blue. The arterial pressure was adjusted by altering the pump speed, and the venous pressure by raising or lowering the reservoir. A brass tube with fine holes 5 mm apart from top to bottom was connected to a funnel and set next to the lung. Freon 22 cooled to - 155°C was poured into the funnel so that it sprayed a vertical strip of lung for 35-40 s, The lung was rapidly cut down (10 s) and plunged into a tub of liquid nitrogen. Care was taken to keep the lung submerged in liquid nitrogen during all subsequent handling. A specially designed drill gave plugs of lung measuring 1 cm in diameter. The lung was sampled at 5-cm intervals from apex to base and all drilling was done with the lung submerged in liquid nitrogen. The plugs were freeze dried in either a Thermovac freeze drier (model FD lV7DG) or Edwards-Pearse freeze drier (model EPDZ) at -5O”C, at a pressure between 0.01 and 0.03 mmHg for 3 days. Both driers were fitted with PzO5 traps. Histological
Preparation
The tissue plugs were trimmed to a depth of less than 5 mm and immersed in fixative solution of ether-alcohol (1: 1) for 3 h. The pl u g s were then advanced into flexible collodion (Mallinckrodt no. 4580) from 2 to 10 % in strength for 12-l 5 h. This was followed by three changes in chloroform. Embedding in paraffin was done under vacuum at 10, 15, and 25 mmHg pressure for 1 h each. The plugs were cut on a rotary microtome and the ribbons were floated out on an ethanol/water bath at room temperature and then transferred to a pure water bath warmed to 45°C. The
Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (132.174.254.157) on January 15, 2019.
118
PERLO,
sections were mounted with Biebrich’s scarlet
on glass slides and stained either with aniline blue or Giemsa stain
(8, 18). Physiological
Conditions During
Freezing
Special attention was given to whether venous pressure was less than alveolar pressure (zone II) or exceeded it (zone III). Eight dogs were prepared in zone II conditions with the pulmonary arterial exceeding alveolar pressure at the top of the lung. Thirteen dogs were studied under zone III conditions. For these animals, the venous pressure was set at least 10 cm above the apex of the lung so that the whole lung was in zone III. Nine of these dogs were studied with narrow arterial-venous pressure differences (3-5 cmHzO) and the other four dogs were studied with wide a-v differences (15-17 cmHe0). All lungs were frozen at end inspiration (alveolar pressure 10 cmHi0). One dog was heparinized, exsanguinated and had its pulmonary vasculature rinsed out with 1,500 ml of heparinized saline and 1,500 ml of Dextran 70 while the lungs were being tidally ventilated. It was then rapidly frozen with liquid nitrogen. One intact dog was placed in the head-up position with the left lung exposed through a midline thoracotomy. The lung was rapidly frozen without cannulation and freezedried. Airway pressure and pulmonary arterial and left atria1 pressures were measured via catheters. Another open-chested dog was prepared by fixing the lungs with isotonic cacodylate-buffered glutaraldehyde (3.5 %) infused into the trachea. At the time of fixation the arterial supply to the left lung was rapidly ligated (30 s) but the lobar arterial supply to the right lung was left untouched. Histolgoical
Measurements
and Quantitative
JALOWAYSKI,
DURAND,
AND
WEST
them onto a counting grid provided with black dots 5 mm apart. A Leitz Ortholux microscope fitted with an xenon light source was used. For the thick (50+ pm sections) a X40 planar objective and X 10 eyepiece was used. The total magnification for the thick lung sections was X 1,000. Thus 1 pm of alveolar wall equaled 1 mm on the grid. For the measurements made by point-counting at least 3,000 points were sampled. Visualization of the superimposed image on the grid was excellent. Further resolution was achieved by focusing up and down the image with the fine focus attachment of the microscope. Proportion of nucleated cells that were intracapillary. This measurement was made directly through the microscope on thin (2 pm) histological sections stained with Biebrich’s scarlet and aniline blue. The microscope was fitted with a conventional light source, oil immersion objective (X 100) and X 10 oculars. This measurement determined the pro-
Techniques
Distribution of red blood cells and nucleated cells in alveolar walls viewed en face. Four measurements were made on alveolar walls viewed en face in thick (50+ pm) lung sections stained with Biebrich’s scarlet and aniline blue. They were : I) the number of red blood cells per square millimeter of alveolar wall; 2) the proportion by area of the alveolar wall occupied by red blood cells; 3) the number of nucleated cells per square millimeter of alveolar wall; 4) the proportion of the alveolar wall occupied by nucleated cells. Ten different alveolar wall areas divided between at least two sections 50f pm thick were studied from each plug. Six plugs were studied at each 5-cm interval of height. The wall areas were selected on the basis of wall area, horizontal orientation to the field of viewing, clarity of structures, and freedom from artifacts. If one wall area failed to provide a sufhciently large enough area for study, adjacent areas were examined until an area of 7,500 pm2 was studied. Within any given wall area examined no measurements were made within 5 pm of the wall edge. Walls that were immediately adjacent to the pleural surface, arteries, or veins were excluded. All walls selected were within a depth of 2 mm of the pleural surface where freezing was rapid and the histological appearance was uniformly good. Examples of the appearance of alveolar walls face on are shown in Fig. 1. The histological sections were examined by projecting
FIG. 1. A; low-power view of alveolar walls seen face on. Large clear areas are sections through alveoli. Section was stained with Giemsa to demonstrate the population of nucleated cells in the alveolar walls (chiefly mononuclear and polymorphonuclear leucocytes). Less conspicuous gray cells are red blood cells. Zone II preparation where arterial pressure exceeded aheolar pressure by 5 cinHr0. B: highpower view of one alveolar wall seen face on. Two polymorphonuclear and several mononuclear leucocytes are visible. Red blood cells are pale gray. In original section with the advantage of focusing up and down, alveolar pores of Kohn can be seen. In this projection, it is not possible to determine whether nucleated cells are intraor extracapillary in location.
Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (132.174.254.157) on January 15, 2019.
RED
AND
WHITE
BLOOD
CELLS
IN
LUNG
CAPILLARIES
portion of nucleated cells that were intracapillary in location. Twenty-five septa were selected from each plug. The entire septum was examined including corner vessels at points of septal intersectionThe number and position (intracapillary vs. extracapillary) of the nucleated cell was recorded for each septum. An intracapillary nucleated cell was defined as one whose outer border fell within the capillary wall as outlined by basement membrane. These two measurements were made on plugs from three zone II lungs and two zone III lungs over a pulmonary arterial pressure range of 10-20 crnHn0. Pru~ortion of j.wlymor@s in aholar walls. Measurements were made on thick (SO+ Pm> section stained with Giemsa. Alveolar wa “11sen face were exa mined directly through the fitted with aX 40 planar microscope objective and X 10 ocular. Differential counts were made on all cells in a wall area. Wall areas were randomly sampled until 200 nucleated cells had been counted. Nucleated cells were counted either as polymorphonuclear cells or mononuclear cells. Endothclial nuclei were not included when identifiable. Two zone II dogs and two zone III dogs were examined over a pulmonary artery pressure range of 5-20 cmHz0. Volumeof aluaolarseptum occu@ed by ofEn capillaries. The proportion of the volume of alveolar septum occupied by open capillaries was determined by examining cross sections of alveolar septa in thin (2 pm) histological sections. Three or four serial sections were taken from each plug and stained with Biebrich’s scarlet and aniline blue. The slides were projected onto the grid. An oil immersion (X 100) objective and X 10 ocular were used yielding a total magnification of X2,500. The determination of the proportion of the volume of alveolar septum occupied by open capillaries was made by point counting the number of dots that fell on open capillaries in the septum and those that fell on other parts of the septum. The lateral 10 pm of septum adjoining corner vessels were excluded from the study. Twenty-five septa were examined from each plug- Septa were selected only if they were fully in tact and provided at least 20 pm of septal length exclusive of the 10 pm lengths that adjoined the corner vessel, that is the vessel in the junction between adjoining alveolar walls. An open capillary was defined as one whose walls were more than 2 purn apart. A distinction was made between open capillaries that contained red blood cells or nucleated cells and closed capillaries containing these cellular structures. A capillary containing a red blood cell was considered open only if a space representing a margin of plasma was visible around any border of the cell. If the cell appeared trapped in the capillary, i.e., the capillary wall was hugging the cell border and no clear space could be discerned then the capillary was considered closed. Any capillary that did not have at least 90 % of its wall visible was rejected and the entire septum excluded. RESULTS
In Fig;. L. 2 and subsequent figures, the results are related to the difference between pulmonary arterial and alveolar pressures both when venous pressure was less than alveolar pressure (zone II) and also when it exceeded alveolar pressure (zone III). For zone II, it is reasonable to ignore
22
20
&
-
4
9L L
0’
I
I
I
I
I
I
1
5
IO
15
20
25
30
35
ARTERIAL-ALVEOLAR
PRESSURE (cm water)
2. Number of red blood cells per mm2 of alveolar against arterial minus alveolar pressure. Cells were counted the alveolar walls face on. In zone II venous pressure but greater than alveolar pressure alveolar pressure, Vertical bars are SD of mean. FIG.
wall plotted by viewing is less than in zone III.
venous pressure since this has been shown not to influence blood flow. For zone I II, venous pressure presumably influences capillary pressure but the relations between arterial, venous, and capillary pressure are not yet known. The choice of arterial pressure is somewhat arbitrary although it is known that the pulmonary blood volume, diffusing capacity for carbon monoxide, and number of red blood cells in the capillaries are closely related to arterial but not to venous pressure (13, 19, 22). There are also theoretical studies which suggest that capillary pressure is close to arterial pressure (9). Number
of Red Blood
Cells per mm2
of Alveolar
Wdl
Figure 2 shows the number of red blood cells per mm* of alveolar wall as determined by the analysis of six plugs at each 5-cm level. The number of red blood cells increased linearly in both zones II and III with increases in pulmonary arterial pressure. The rate of increase in zone II was greater than that for zone III. However, the absolute number of red blood cells in zone III is greater than that in zone II for any given perfusing pressure. This presumably reflects the higher capillary pressure in zone III and the results are consistent with those of Glazier et al. (13). For example, at a 5 cmHz0 pulmonary artery pressure the number of red blood cells per mm2 of alveolar wall was about 5,000. In zone III for the same upstream pressure the value was about double this. An arterial pressure 5 cmHz0
Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (132.174.254.157) on January 15, 2019.
120
PERLO,
above alveolar corresponds to the conditions near the apex of the upright human lung. On the other hand, at 20 cmH20 pulmonary arterial pressure the number of red blood cells per mm2 of alveolar wall was about 13,000 in zone II which is about three-fourths that found in zone III at an equivalent arterial pressure using the line of best fit shown in the figure. At 35 cmHz0 pressure there were about 21,000 red blood cells per mm2 of alveolar wall. This number probably underestimates the number of red blood cells because it was difficult to distinguish overlapping cells when they were very tightly packed. There was no indication that the number of red blood cells reached a maximum as the pulmonary arterial pressure was raised over the range of pressures studied. However Glazier et al. reported that the number of red blood cells per 10 pm of alveolar septum as measured in thin sections did reach a maximum above a capillary pressure of 50 cm Hz0 in zone III.
This was obtained by recording the number of cChits” from a counting grid on red blood cells and the rest of the alveolar wall. Figure 3 shows that in both zones II and 111 the proportion of alveolar wall occupied by red blood cells increased linearly over the range of pulmonary arterial pressures studied. The rate of increase in zone II was greater than that in zone III. At a pulmonary arterial
I
I
I
t
I
I
I
1
5
IO
15
20
25
30
35
40
ARTERIAL-ALVEOLAR
PRESSURE
(cm water)
FIG. 3. Proportion of the area of alveolar wall covered by red blood cells. Notice that where the arterial pressure is only 5 cm&O above alveolar pressure as would be the case near the apex of the upright human lung, only some 12% of the wall is covered. However this increases to over 45% when the arterial pressure is 30 cmHzO as it is near the base of the upright human lung. Vertical bars are SD of mean.
01
JALOWAYSKI,
tained
4, by
AND
I
I
I
1
I
I
1
5
IO
15
20
25
30
35
ARTERIAL-ALVEOLAR FIG.
DURAND,
Comparison of the data Glazier et al. (f 3).
PRESSURE from
the
present
WEST
1 40
(cm woted study
with
that
ob-
pressure of 5 cmHg0, only about 12 % of the wall area was occupied by red blood cells while in zone III about 30 % of the wall area was occupied. At a higher perfusing pressure of 20 cmHz0, 33 % of the wall area was occupied by red blood cells in zone II compared with 40% in zone III. At 35 cmH20 arterial pressure, about 50 % of alveolar wall area was occupied by red blood cells in zone III. It appears that the regression line calculated for zone II (R = 0.80, slope = 1.42) would join that calculated for zone III = 0,68, slope = 0.67) between 25 and 30 cmHy0 upCR stream pressure. There is no evidence that the proportional area of the wall occupied by red blood cells reached a maximum over the range of pulmonary arterial pressures studied. Glazier et al. found that the proportion of alveolar septum occupied by red blood cells in thin sections reached a maximum above a capillary pressure of 50 cmHz0 in zone III. A comparison of the regression lines in Figs. 2 and 3 shows that the number of red blood cells per square millimeter of alveolar wall increased more than the proportional area of the wall which they occupied. For example, in zone III, the number of red cells doubled as the pressure was increased from 0 to 30 cmHz0 pressure while the proportional area took 45 cm pressure for the same increase. In zone II, the difference was in the same direction though less marked. This is consistent with the overlapping of red celfs which could be seen in cross sections as the pressures increased. It means that fewer red cells have their surfaces in contact with the alveolar wall and this will increase the diffusion distance for gases. The proportion (by area) of the wall occupied by red blocd cells determined by point counting alveolar walls en face from thick (50 pm) lung sections is similar to the measurement by Glazier et al. of the percent septum occupied by red blood cells made on cross sections of alveolar septa in thin (2 pm) lung sections. These investigators superimposed an eyepiece graticule calibrated every 5 pm along the septum so that the graduation lines cut the septum perpendicularly. They counted the number of intersections of the lines with red blood cells and thus determined the proportion by length of the septum occupied by red cells. This measurement is equivalent to that obtained in the present study by superimposing a grid of points on the alveolar wall seen en face. Figure 4 compares the present data with those of Glazier
Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (132.174.254.157) on January 15, 2019.
RED -
AND
WHITE
BLOOD
CELLS
IN
6r
“I
E2
LUNG
121
CAPILLARIES
-3-
IO]
1 5
1
t 15
10 ARTERIAL-ALVEOLAR
I 20
I 25
PRESSURE
1 30
t 35
(cm water)
FIG. 5. Number of nucleated cells per mm2 of alveolar wall plotted against arterial minus alveolar pressure. Note that in zone II there was no consistent change but there was a fall in zone III. Contrast this behavior with that of the number of red blood cells shown in Fig. 1, Vertical bars are SD of mean.
This measurement is made by point counting nucleated cells in alveolar walls en face. Figure 6 shows that the proportion of the wall occupied by nucleated cells is similar to that expected from the data presented in Fig. 5, In zone II there is no consistent difference (slope = 0.00, R = 0.01) over the range of perfusing pressures studied while in zone III there is a decrease (slope = -0.09, R = -0.27). At a low perfusing pressure of 5 cmH20 in zone II the relative wall area occupied by nucleated cells was about 16 % which is slightly greater than the relative wall area occupied by red blood cells (Fig. 3) for the same conditions (11 %). This d’ff 1 erence can be explained by the larger size of the nucleated cells compared with the red blood cells. Distribution htracupihry
0’
I
5
I IO
I
1
I
I
I
I
15
20
25
30
35
40
ARTERIAL-ALVEOLAR 6. Proportion
cleated
cells. Compare
by area of the Fig. 4. Vertical
PRESSURE
h
woted
alveolar wall occupied bars are SD of mean.
by nu-
et al. In that study, the zone III results were related to venous minus alveolar pressure, rather than arterial minus alveolar pressure as here. The regression line should therefore be slightly displaced to the right but the error is small because the arterial-venous pressure difference was so low. It can be seen that there is good agreement between the two sets of measurements. Number
of Nucleated
Cells per mm2 of Alveolar
of Nucleated Cells. Vmus Extracapillary
Figure 7 shows that the nucleated cells are predominantly intracapillary both in zones II and III. For this measurement which was made on thin (2 pm) sections, endothelial cell nuclei could be recognized and were not included as either intracapillary or extracapillary. In zone II over the range of pulmonary artery pressures of 10 and 20 cmHz0 about 78 % of the nucleated cells were located within capillaries. In zone III at an upstream pressure of 10 cmHz0, 72 % of the nucleated cells were intracapillary while at a higher perfusing pressure of 20 cmHz0 only 62 % of the cells were located within the capillaries. The differences between zones II and III were significant at the 1 % level. Thus it appears from Figs. 5 and 7 that in zone III there is a decrease in the number of nucleated cells found in alveolar walls as the perfusing pressure is raised and that
WuIl
Figure 5 shows that in zone II there was no consistent change in the number of nucleated cells per mm2 of alveolar wall for increases in pulmonary arterial pressure. However, in zone III there was a decrease, the number reducing from 5,000 at an arterial pressure of 5 cmH20 to 3,000 when the pressure was 35 cmH20. Comparison of the number of red and nucleated cells shows that there was a considerably greater number of nucleated cells in alveolar walls than would be expected from the ratio of nucleated cells to red blood cells & peripheral blood. For example, in zone II at 5 cmHs0 perfusing pressure there were about 5,000 nucleated cells per mm2 of alveolar wall and about the same number of red blood cells (Fig. 2). In zone III at this same upstream pressure the number of nucleated cells were about half the number of red blood cells. It should be emphasized that not all these nucleated cells are white blood cells, some being alveolar, epi thelial, endothelial cells, and macrophages. However, as shown below, the majority of these nucleated cells were white blood cells.
IQ
I5
ARTERIAL-ALVEOLAR FIG.
capillaries.
7. Number Notice
of nucleated that between
20 PRESSURE
(cm weted
cells which are inside and outside 60% and 80% are intracapillary.
the
Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (132.174.254.157) on January 15, 2019.
122
PERLO,
this decrease can be attributed to a fall in the number of nucleated cells within capillaries. We thought that it might be possible to determine the proportions of white cells which were intra- and extracapillary by washing the vascular bed out with saline or dextran. This was tried on one lung but we found that although all the red blood cells could be removed, some intracapillary white cells remained. They were presumably either jammed tightly in the small vessels or Intere adhering to the walls. The bindings presented thus far indicate that the nucleated cells behave differently in zones II and 111 when the arterial pressure is raised. The number of nucleated cells in zone II appears to be uninfluenced by changes in perfusing pressure while the nucleated cells in zone III appear to decrease with increases in pulmonary artery pressure. In addition, most of the nucleated cells are intracapillary. A reasonable mechanism for these differences is that the capilla ries are more d istended in zone III than in zone II for the same perfusin g pressures and that under both conditions distension increases as the perfusing pressure is raised. They are thus less likely to trap the relatively large nucleated cells as they flow through the bed. Glazier et al. showed that the mean width of the capillaries was less in zone II than zone II I for the same perfusing pressure, and also that the width increased with perfusing pressure down zone III. Prqtmrdion of nucleated cells which were @ymor~honuclaar cells. This study was directed at the differential morphology of the cells. Observations of nucleated cells in alveolar walls stained with Giemsa and viewed en face revealed that they could only be confidently divided into polymorphonuclear and mononuclear white blood cells. Figure 8 shows the proporti .on 01 polymorphs ai nongst nucleated cells with arteri al pressu re. It can be seen that changes in pulmonary the proportion is generally low, in the range of about 825 % of all nucleated cells. In zone II the proportion stays fairly constant at 8 %. In zone III at perfusing pressures of 5 and 10 c1nH~0 the proportion of polymorphs amongst nucleated cells is about 11 % but increases to 25 % at an
ARTERIAL-ALVEOLAR 8. Percentage of nucleated cells polymorphonuclear. This was generally pressures in zone III where it was about predominately mononuclear. FIG.
PRESSURE
(cm water)
in the alveolar less than 127, 25oj’. Therefore
walls which are except at higher the cells were
8 -
JALOWAYSKI,
DURAND,
AND
WEST
60
0 ARTERIAL-ALVEOLAR PRESSURE (cm water) 9. I’ercentage by volume of the alveolar septa taken up by open capillaries. Marked increase with the arterial pressure, especially in zone II, is consistent with the mechanisms of recruitment and distension. FIG.
upstream pressure of 20 cmHz0. The differences between zones II and III were significant at the 2 % level These findings are consistent with the observation that the population of intracapillary nucleated cells in zone II remains approximately constant in the face of changes in The increase in proportion of polyperfusion pressure. morphs in zone 111 can perhaps be explained by the preferential loss of smaller nucleated cells through the capillaries as they distend. Observahs on lungs from intact animals. We thought that the unexpectedly high concentration of white blood cells in the pulmonary capillaries might be caused by the perfusing circuit. To test this, we made additional measurements on two dogs in which operative interference was kept to a minimum. In one dog, the lung was rapidly frozen irnmcdiately after opening the chest, inserting a catheter into the pulmonary artery to measure pressure, and placing the animal in the head-up position. In another anesthetized dog, the lungs were flooded with glutaraldehyde introduced via the trachea. In the rapidly frozen lung, the average numbers of nucleated cells and red blood cells in the alveolar walls near the apex of the lung were 3,666 and 4,217 per mm2, respectively. These values are comparable with those found in the excised lungs as shown in Figs. 2 and 5. Toward the base of the lung, the average numbers of nucleated and red cells were 2,994 and 16,373 square mm2, respectively, which again Iit with the previ ous data. I n the glu ta raldehyde-fixed lung, the nu mber of nu cleated cells was approximately 5,500 per mm2 although in this preparation the time taken for fixation to occur was unknown. We conclude from these additional observations that the large concentration of white cells in th e capillaries was not caused by the procedures connected with perfusi .ng the lu ngs. Vulume of Alveolar Septum Occufied by Upen Capillaries This measurement was made by point-counting random cross sections of alveolar septa in thin (2 pm) sections. The
Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (132.174.254.157) on January 15, 2019.
RED
AND
WHITE
BLOOD
CELLS
IN
LUNG
CAPILLAKIES
percentage volume of the septum occupied by open capillaries was determined by the proportional number of hits on open capillaries. Figure 9 shows that in zone II the proportional volume of the septum occupied by open capillaries increased from about 10 % at 5 cmHp0 pressure to about 62 % at 20 cmHk0 pressure (R = 0.80, slope = 3.53). In zone III the proportional volume of the septum increased from about 43 % at 5 crnHk0 pressure to 73 Yo at 20 cmHz0 pressure (R = 0.73, slope = 2.05). The proportional volume of the septum occupied by open capillaries in zone II increased rapidly over the range of pressures studied (O-20 cmHz0). This finding is consistent with that of Glazier et al. (13) and Warrell et al. (30) that the number of open capillaries and/or capillary width increased up to a capillary pressure of 15 cmHa0 in zone II. The sixfold increase in the volume of the septum occupied by open capillaries in zone II is consistent with the data of Warrell et al. (30) which demonstrated that recruitment is the predominant mechanism by which the puhnonary capillary bed responds to increases in upstream pressure in zone II in this pressure range. The smaller increase in the volume of the septum occupied by open capillaries in zone III is consistent with the data of Glazier et al. which showed distension to be the predominant mode of response to increasing upstream pressure. The different patterns exhibited in zones II and III can be explained by an increased capillary pressure in zone III when compared with zone II for equivalent upstream pressures. Note that the data presented here do not of themselves provide evidence in favor of recruitment or distension of capillaries, DISCUSSION
Red Blood GeIl I)istribution in AIuednr Beds: Sign$cant for Gns Exchange
Cu/dlary
Several studies have shown that the regional diffusing capacity increases down the upright human lung (7, 9, 3 1). Glazier et al. (13) f ound that the number of red blood cells per 10 pm of alveolar septum increased linearly with increases in capillary pressure to 50 cmHl0 and suggested that the maximum diffusing capacity could probably not be attained in man because of the very large pulmonary vascular pressures which would be necessary. The present study similarly demonstrates a linear increase in -lilling of the alveolar wall area with red blood cells as pulmonary arterial pressure rises. Over the pressure range studied the percentage of wall area occupied by red blood cells did not reach a maximum although the silicone injected preparation of Sobin et al. (26) and the frozen lung preparation of Warrell et al. (30) would suggest that about 90 ‘% of the wall area is available for capillary blood. These findings further suggest that regional differences in diffusing capacity, DL, could be explained by regional differences in filling of the alveolar wall in the vertical lung. Recently Michaelson et al. have compared diffusing capacity per liter of lung volume (DIJVA) with pulmonary capillary blood flow per liter of lung volume (Qc/VA) in the upright lung using a bolus gas inhalation technique (20). They found that both DL/VA and Qc/VA increased down the lung but that the rate of increase of capillary blood Aow was greater.
123
White cell sequestration by the lung has long been recognized (11, 24). B ierman and colleagues (3, 4) have shown that the lung selectively removes leukemic leukocytes infused into circulating blood whereas a peripheral capillary bed will not. They also reported that histamine induces white cell sequestration and epinephrine stimulates white cell release from the lung. Ambrus et al. (2) postulated that the lung acts as a Yeukostat” in regulating the white count of circulating blood. They infused white cell-rich blood into Starling heart-lung preparations and found the circulating white cell level to be constant and indenendent of the concentration in infused blood despite repeated infusions. In a subsequent study they found that this depletion was primarily due to sequestration of granulocytes by the lung (1). The present study shows that there is a large n umber of n uclea ted cells with in alveolar walls and that thev a re chiefly intercapillary in location. The predominant cell type is mononuclear while a minority are polymorphonuclear leukocytes. Our study further suggests that the puL monary capillary bed contributes to the sequestration of white blood cells on a mechanical basis. In zone II the capillary bed serves as an efficient filter as evidenced by the large number of nucleated cells in relation to the number of red blood cells per mm2 of alveolar wall and the constant number and proportion of the wall occupied by nucleated cells. The capillary bed in zone I II serves as a less eficient filter on the basis of these criteria. This can be explained by the relatively greater distention of the bed in zone III as shown by Glazier et al. (13) which is presumably a reflection of the higher capillary pressure. It should be noted that our observations were confined to the extreme periphery of the lung and the possibility remains that other patterns may apply to other regions. The large concentration of leukocytes in the alveolar wails may affect the pressure-flow relations of the capillary bed. Warrell et al. (30) demonstrated uneven filling of the capillary bed within arteriolar domains and concluded that recruitment occurs on a capillary rather than arteriolar level. White cell sequestration by the capillary bed could certainly influence flow through a capillary segment by altering the pressure drop across a plugged capillary. Preferential flow to a neighboring unobstructed capillary segment could then result. The predominance of the mononuclear cell in the bed is of interest. Recent evidence suggests that one precursor of the alveolar rnacrophage is the blood monocyte derived from a bone marrow stern cell (15, 23, 27). Morphological evidence demonstrating the transformation of the intracapillary monocyte into the alveolar macrophage is lacking although studies of enzyme content and phagocytic abilities of the two cells support this notion (27)* Bowden and Adarnson suggested that the pulmonary interstitial cell is the essential link between the circulating monocyte and the alveolar macrophage (5, 6). A second source of alveolar rnacrophages may be migrating hepatic Kupffer cells caught up in the pulmonary capillary bed (16, 17, 2 1, 27). There is evidence that these cells then move into the alveolar spaces and are su bsequently lost via the airways.
Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (132.174.254.157) on January 15, 2019.
124
PERLO,
We are Gaines This
indebted to Myra for assistance. work was supported
Bailey, by
Richard
National
Matthews, Institutes
and of Health
Richard Grants
HL-13687-03 Administration Received
JALOWAYSKI,
and HL-05931-03 Grant NGL-05-009for
publication
and
DURAND, National 109.
28 November
AND
Aeronautics
WEST
and
Space
1973.
REFERENCES
1, AMBRUS,
C. M., AND J- L. AMBRUS. Regulation of the leukocyte level. Ann. iV. Y. Acad. Sci. 77 : 445-486, 1959. 2. AMBRUS, C. M., J. L. AMBRUS, G. C. JOHNSON, E. W. PACKMAN, W. S. CHERNICK, N. BACK, AND J. W. E,. HARRISON. Role of the lungs in regulation of the leukocyte level. Am. J. Physiol. 178: 3344, lY54. H. R. The hematologic role of the lung in man. Am. J. 3. BIERMAN, Surg. 8Y: 130-140, lY55. H. R., K. H. KELLEY, AND F. 1;. CORDES. The sequestra4. BIERMAN, tion and visceral circulation of leukocytes in man. Ann. NY. Acad.
Sci. 59 : 850-862, 5. BOWDEN, D. H.,
6.
7.
8.
9.
10. 11. 12.
13.
14,
15.
16.
1955.
AND I. Y. IX. ADAMSON. Pulmonary interstitial cell as immediate precursor of alveolar macrophage. Am. J. Pathol. 68: 521-536, 1972. BOWDEN, D. H., I. Y. R. ADAMSON, W. G. GRANTHAM, AND J. P. WYATT. Origin of the lung macrophage. Arch. Puthol. 88: 540-546, 196Y. BURROWS B., H. NIDEN, C. MITTMAN, R. C. TALLEY, AND W. R. BARCLAY. Non-uniform pulmonary diffusion as demonstrated by the carbon monoxide equilibration technique : experimental results in man. J. Clin. Invest. 39 : 943-949, 1960. CONN, I-3. J., M. A. DARROW, AND V. M. EMMEL. Staining Procedures Used by the Biological Stain Commission (2nd ed.). Baltimore, Md. : Williams & Wilkins, 1960. FORSTER, R. E., W. S. FOWLER, D. V. BATES, AND B. VAN LINGEN. The absorption of carbon monoxide by the lungs during breath holding. J. Clin. Invest. 33 : 1135-l 145, 1954. FUNG, Y. C., AND S. S. SOBIN. Elasticity of the pulmonary alveolar sheet. Circulation Res. 30: 45 1469, 1972. GARREY, W. E., AND W. R, BRYAN. Variations in white blood cell counts. physiol. Rev. 15 : 597-638, 1935. GLAZIER, J. B., J. M. B. HUGHES, J. E. MALONEY, AND J. B. WEST. Vertical gradient of alveolar size in lungs of dogs frozen intact. J. A@Z. Physiol. 23 : 694-705, 1967. GLAZIER, J. B., J. M. B. HUGHES, J. E. MALONEY, AND J- B. WEST, Measurement of capillary dimensions and blood volume in rapidly frozen lungs. J. A$$. Physiol. 26: 65-76, 1969. GLAZIER, J. B., AND J. F. MURRAY. Sites of pulmonary vasomotor reactivity in the dog during alveolar hypoxia and serotonin and histamine infusion. J. Clin. Invest. 50 : 2550-2558, 197 1. GODELSKI, J. J., AND J. D. BRAIN. The origin of alveolar macrophages in mouse radiation chimeras. J. Exptl. Med. 136: 630-643, 1972. IRWIN, D. A. Kupffer cell migration. Can. Med. Assoc. J. 27 : 353356, 1932.
17. 18. 19.
20.
21.
22.
23.
24.
KLINGENSMZTH, W. C. III, AND T. W. RYERSON. Lung uptake of 9grnTc-sulfur colloid. J. i2rzlcl. Med. 14: 201-204, 1973, Manual of Histologic Staining Methods of theArmed Forces Institute of Pathology (3rd ed.). New York: McGraw, 1960. MASERI, A., P. CALDINI, P. HARWARD, R, C. JOSHI, S. PERMUT’I-, AND K L. ZIERLER. Determinants of pulmonary vascular volume: recruitment vs. distensibility. Circulation Res. 3 1 : 218-228, 1972. MICHAELSON, E. D., M. A. SACKNER, AND R. L. JOHNSON, JR. Vertical distributions of pulmonary diffusing capacity and capillary blood flow in man. J. Clin. Invest. 52: 359-369, 1973. NICOL, T., AND D. L. J. DILBEY. Elimination of macrophage cells of the reticuloendothelilial system by way of the bronchial tree. Nature 182: 192-193, 1958. PERMUTT, S., P. CALDINI, A. MASERI, W. H. PALER, T. SASAMORI, AND K. ZIERLER. Recruitment versus distensibility in the pulmonary vascular bed. In : The Pulmonary Circulation and Interstitial Space, edited by A. P. Fishman and H. H. Hecht. Chicago, Ill.: Univ. of Chicago Press, 1969. PINKETT, M., C. K. COWDREY, AND P. C. NOWELI,. Mixed hematopoietic and pulmonary origin of alveolar macrophages as demonstrated by chromosome markers. Am. J. Pathol. 48: 859-867, 1966. RATLIFF, N. B., J. W. WILSON, E. MTKAT, D. B. HACKEL, AND T. C. GRAHAM. The lung in hemorrhagic shock. Am. J. Pathol. 65: 325-
331, 1971. 25. SIMPSON, M.
29.
E. The experimental production of macrophages in the circulating blood. J. Med. Res. 43: 77-144, 1922. SOBIN, S. S., H. M. TREMER, AND Y. C. FUNG. Morphometric basis of the sheet-flow concept of the pulmonary alveolar microcirculation in the cat. Circulation Res. 26 : 397-414, 1970. VIJEYARATNAM, G. S., AND B. CORRIN. Origin of the pulmonary alveolar macrophage studied in the iprindole-treated rat. J. Pathol. 108: 115-118, 1972. VREIM, C. E., AND N. C. STAUB. Indirect and direct pulmonary capillary blood volume in anesthetized open-thorax cats. J. A@. Physiol. 34: 452459, 1973. VREIM, C. E., AND N. C. STAUB. Pulmonary vascular pressures and
30.
capillary Physiol. WARRELL,
26.
27.
28.
blood
volume
36: 275-279, D.
A.,
changes 1974. J. W. EVANS,
AND J. B, WEST.
Pattern
J. A#.
32 : 346-356,
31. WEST,
Physiol. J. B., R.
of filling
A. B. HOLLAND,
MATTHEWS. Interpretation lung, J. Appl. Ph+ysiol.
in R.
anesthetized 0.
in the
CLARKE,
cats. G.
pulmonary
J.
Af$Z.
P. KINGABY, capillary
bed.
1972. C.
of radioactive 17 : 14-20, 1962.
T. DOLLERY, gas clearance
AND
C. rates
M.
E.
in the
Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (132.174.254.157) on January 15, 2019.
Vol.
A~pmm
38, No. 1,January
PHYSIOLOGY
2975.
Distribution
Printed
in U.S.A.
of red and white
cells in alveolar
blood
walls
S. PERLO, A. A. J’ALOWAYSKI, C. M. DURAND, AND J. B. WEST De~artmenf of Medicine, School of Medicine, Universiy of California San Diego, La Jolla, Gulifurnia
PERLO, &A. A. JALOWAYSKI, C. M. DURAND, ~z’stributim of red and white blood cells in alveolar
J. B. WEST. walls. J- Appl.
AND
Yhysiol. 38( 1) : 117-124. 1975.~Dog lungs were perfused with blood and rapidly frozen with liquid Freon gas at various pulmonary artery and venous pressures. The numbers of red and white blood cells per mm2 of alveolar wall were counted in lung sections and, in addition, the proportion (by area) of the wall occupied by the cells was measured by point counting. The number and proportional area of the red blood cells rapidly increased as perfusing pressure was raised. These findings are consistent with earlier observations of capillary recruitment and distension. An unexpected observation was the large number of leukocytes in the capillaries especially at low perfusing pressures. For example when arterial exceeded alveolar pressure by 5 cmHn0 (as occurs near the apex of the upright human lung), there were about 5,000 red cells and 4,000 white cells per mm2 of alveolar wall* As perfusing pressure was increased, the number of leukocytes paradoxically decreased in zone 3 but remained constant in zone 2. Most of the white cells were mononuclear cells. These results suggest: that the lung behaves as a mechanical sieve for large cells and that the number of trapped cells depends on the capillary pressure. recruitment; Aow behavior
distension;
sequestration;
zones
of
lung;
pressure-
THERE HAVE BEEN several recent studies of the morphology of the pulmonary microcirculation including the concentration of red blood cells in the capillaries, the number of open capillaries, and their dimensions in relation to the pulmonary vascular pressures (13, 28-30) examining rapidly frozen sections of dog and cat lung. Sobin and his colleagues (26) have emphasized the large proportion of the alveolar wall which can be covered by open capillaries as viewed en face but there is uncertainty about how this area is related to capillary pressure. The present study was designed to determine the morphology of the pulmonary capillaries in the dog viewed en face in relation to the pulmonary arterial, venous, and alveolar pressures. An unexpected finding was the large proportion of white blood cells in the capillaries especially at low perfusing pressures. METHODS
Lung Pre$mation The preparation has been described in detail elsewhere (13, 30) and will only be summarized here. Healthy mongrel dogs weighing between 16 and 30 kg were anesthetized
9.2037
with Nembutal sodium, intubated, and ventilated with a Harvard respirator. A tracheotomy was performed, a largebore catheter was advanced into the inferior vena cava from a femoral vein cutdown, and the heart was exposed through a left thoracotomy. The animal was then heparinized (2,000 units/kg) and 400-500 ml of blood were removed over a 3- to 5-min period. The heart was then fibrillated with an electrocautery, the chest wall removed, and the pulmonary artery and left atrium cannulated in a rapid sequence that averaged less than 10 min. After cannulation the lungs were gently inflated with 100 % 02 until all areas were fully expanded and the dog secured upright. The dog was ventilated with 6 % CO2 in air and the lungs were perfused with blood from the venous reservoir by means of a roller pump. The arterial venous and tracheal pressures were measured by saline manometers. Either the right or left lung was selected for rapid freezing based on its appearance and amount of tissue available. The pleural surface was marked with methylene blue. The arterial pressure was adjusted by altering the pump speed, and the venous pressure by raising or lowering the reservoir. A brass tube with fine holes 5 mm apart from top to bottom was connected to a funnel and set next to the lung. Freon 22 cooled to - 155°C was poured into the funnel so that it sprayed a vertical strip of lung for 35-40 s, The lung was rapidly cut down (10 s) and plunged into a tub of liquid nitrogen. Care was taken to keep the lung submerged in liquid nitrogen during all subsequent handling. A specially designed drill gave plugs of lung measuring 1 cm in diameter. The lung was sampled at 5-cm intervals from apex to base and all drilling was done with the lung submerged in liquid nitrogen. The plugs were freeze dried in either a Thermovac freeze drier (model FD lV7DG) or Edwards-Pearse freeze drier (model EPDZ) at -5O”C, at a pressure between 0.01 and 0.03 mmHg for 3 days. Both driers were fitted with PzO5 traps. Histological
Preparation
The tissue plugs were trimmed to a depth of less than 5 mm and immersed in fixative solution of ether-alcohol (1: 1) for 3 h. The pl u g s were then advanced into flexible collodion (Mallinckrodt no. 4580) from 2 to 10 % in strength for 12-l 5 h. This was followed by three changes in chloroform. Embedding in paraffin was done under vacuum at 10, 15, and 25 mmHg pressure for 1 h each. The plugs were cut on a rotary microtome and the ribbons were floated out on an ethanol/water bath at room temperature and then transferred to a pure water bath warmed to 45°C. The
Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (132.174.254.157) on January 15, 2019.
118
PERLO,
sections were mounted with Biebrich’s scarlet
on glass slides and stained either with aniline blue or Giemsa stain
(8, 18). Physiological
Conditions During
Freezing
Special attention was given to whether venous pressure was less than alveolar pressure (zone II) or exceeded it (zone III). Eight dogs were prepared in zone II conditions with the pulmonary arterial exceeding alveolar pressure at the top of the lung. Thirteen dogs were studied under zone III conditions. For these animals, the venous pressure was set at least 10 cm above the apex of the lung so that the whole lung was in zone III. Nine of these dogs were studied with narrow arterial-venous pressure differences (3-5 cmHzO) and the other four dogs were studied with wide a-v differences (15-17 cmHe0). All lungs were frozen at end inspiration (alveolar pressure 10 cmHi0). One dog was heparinized, exsanguinated and had its pulmonary vasculature rinsed out with 1,500 ml of heparinized saline and 1,500 ml of Dextran 70 while the lungs were being tidally ventilated. It was then rapidly frozen with liquid nitrogen. One intact dog was placed in the head-up position with the left lung exposed through a midline thoracotomy. The lung was rapidly frozen without cannulation and freezedried. Airway pressure and pulmonary arterial and left atria1 pressures were measured via catheters. Another open-chested dog was prepared by fixing the lungs with isotonic cacodylate-buffered glutaraldehyde (3.5 %) infused into the trachea. At the time of fixation the arterial supply to the left lung was rapidly ligated (30 s) but the lobar arterial supply to the right lung was left untouched. Histolgoical
Measurements
and Quantitative
JALOWAYSKI,
DURAND,
AND
WEST
them onto a counting grid provided with black dots 5 mm apart. A Leitz Ortholux microscope fitted with an xenon light source was used. For the thick (50+ pm sections) a X40 planar objective and X 10 eyepiece was used. The total magnification for the thick lung sections was X 1,000. Thus 1 pm of alveolar wall equaled 1 mm on the grid. For the measurements made by point-counting at least 3,000 points were sampled. Visualization of the superimposed image on the grid was excellent. Further resolution was achieved by focusing up and down the image with the fine focus attachment of the microscope. Proportion of nucleated cells that were intracapillary. This measurement was made directly through the microscope on thin (2 pm) histological sections stained with Biebrich’s scarlet and aniline blue. The microscope was fitted with a conventional light source, oil immersion objective (X 100) and X 10 oculars. This measurement determined the pro-
Techniques
Distribution of red blood cells and nucleated cells in alveolar walls viewed en face. Four measurements were made on alveolar walls viewed en face in thick (50+ pm) lung sections stained with Biebrich’s scarlet and aniline blue. They were : I) the number of red blood cells per square millimeter of alveolar wall; 2) the proportion by area of the alveolar wall occupied by red blood cells; 3) the number of nucleated cells per square millimeter of alveolar wall; 4) the proportion of the alveolar wall occupied by nucleated cells. Ten different alveolar wall areas divided between at least two sections 50f pm thick were studied from each plug. Six plugs were studied at each 5-cm interval of height. The wall areas were selected on the basis of wall area, horizontal orientation to the field of viewing, clarity of structures, and freedom from artifacts. If one wall area failed to provide a sufhciently large enough area for study, adjacent areas were examined until an area of 7,500 pm2 was studied. Within any given wall area examined no measurements were made within 5 pm of the wall edge. Walls that were immediately adjacent to the pleural surface, arteries, or veins were excluded. All walls selected were within a depth of 2 mm of the pleural surface where freezing was rapid and the histological appearance was uniformly good. Examples of the appearance of alveolar walls face on are shown in Fig. 1. The histological sections were examined by projecting
FIG. 1. A; low-power view of alveolar walls seen face on. Large clear areas are sections through alveoli. Section was stained with Giemsa to demonstrate the population of nucleated cells in the alveolar walls (chiefly mononuclear and polymorphonuclear leucocytes). Less conspicuous gray cells are red blood cells. Zone II preparation where arterial pressure exceeded aheolar pressure by 5 cinHr0. B: highpower view of one alveolar wall seen face on. Two polymorphonuclear and several mononuclear leucocytes are visible. Red blood cells are pale gray. In original section with the advantage of focusing up and down, alveolar pores of Kohn can be seen. In this projection, it is not possible to determine whether nucleated cells are intraor extracapillary in location.
Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (132.174.254.157) on January 15, 2019.
RED
AND
WHITE
BLOOD
CELLS
IN
LUNG
CAPILLARIES
portion of nucleated cells that were intracapillary in location. Twenty-five septa were selected from each plug. The entire septum was examined including corner vessels at points of septal intersectionThe number and position (intracapillary vs. extracapillary) of the nucleated cell was recorded for each septum. An intracapillary nucleated cell was defined as one whose outer border fell within the capillary wall as outlined by basement membrane. These two measurements were made on plugs from three zone II lungs and two zone III lungs over a pulmonary arterial pressure range of 10-20 crnHn0. Pru~ortion of j.wlymor@s in aholar walls. Measurements were made on thick (SO+ Pm> section stained with Giemsa. Alveolar wa “11sen face were exa mined directly through the fitted with aX 40 planar microscope objective and X 10 ocular. Differential counts were made on all cells in a wall area. Wall areas were randomly sampled until 200 nucleated cells had been counted. Nucleated cells were counted either as polymorphonuclear cells or mononuclear cells. Endothclial nuclei were not included when identifiable. Two zone II dogs and two zone III dogs were examined over a pulmonary artery pressure range of 5-20 cmHz0. Volumeof aluaolarseptum occu@ed by ofEn capillaries. The proportion of the volume of alveolar septum occupied by open capillaries was determined by examining cross sections of alveolar septa in thin (2 pm) histological sections. Three or four serial sections were taken from each plug and stained with Biebrich’s scarlet and aniline blue. The slides were projected onto the grid. An oil immersion (X 100) objective and X 10 ocular were used yielding a total magnification of X2,500. The determination of the proportion of the volume of alveolar septum occupied by open capillaries was made by point counting the number of dots that fell on open capillaries in the septum and those that fell on other parts of the septum. The lateral 10 pm of septum adjoining corner vessels were excluded from the study. Twenty-five septa were examined from each plug- Septa were selected only if they were fully in tact and provided at least 20 pm of septal length exclusive of the 10 pm lengths that adjoined the corner vessel, that is the vessel in the junction between adjoining alveolar walls. An open capillary was defined as one whose walls were more than 2 purn apart. A distinction was made between open capillaries that contained red blood cells or nucleated cells and closed capillaries containing these cellular structures. A capillary containing a red blood cell was considered open only if a space representing a margin of plasma was visible around any border of the cell. If the cell appeared trapped in the capillary, i.e., the capillary wall was hugging the cell border and no clear space could be discerned then the capillary was considered closed. Any capillary that did not have at least 90 % of its wall visible was rejected and the entire septum excluded. RESULTS
In Fig;. L. 2 and subsequent figures, the results are related to the difference between pulmonary arterial and alveolar pressures both when venous pressure was less than alveolar pressure (zone II) and also when it exceeded alveolar pressure (zone III). For zone II, it is reasonable to ignore
22
20
&
-
4
9L L
0’
I
I
I
I
I
I
1
5
IO
15
20
25
30
35
ARTERIAL-ALVEOLAR
PRESSURE (cm water)
2. Number of red blood cells per mm2 of alveolar against arterial minus alveolar pressure. Cells were counted the alveolar walls face on. In zone II venous pressure but greater than alveolar pressure alveolar pressure, Vertical bars are SD of mean. FIG.
wall plotted by viewing is less than in zone III.
venous pressure since this has been shown not to influence blood flow. For zone I II, venous pressure presumably influences capillary pressure but the relations between arterial, venous, and capillary pressure are not yet known. The choice of arterial pressure is somewhat arbitrary although it is known that the pulmonary blood volume, diffusing capacity for carbon monoxide, and number of red blood cells in the capillaries are closely related to arterial but not to venous pressure (13, 19, 22). There are also theoretical studies which suggest that capillary pressure is close to arterial pressure (9). Number
of Red Blood
Cells per mm2
of Alveolar
Wdl
Figure 2 shows the number of red blood cells per mm* of alveolar wall as determined by the analysis of six plugs at each 5-cm level. The number of red blood cells increased linearly in both zones II and III with increases in pulmonary arterial pressure. The rate of increase in zone II was greater than that for zone III. However, the absolute number of red blood cells in zone III is greater than that in zone II for any given perfusing pressure. This presumably reflects the higher capillary pressure in zone III and the results are consistent with those of Glazier et al. (13). For example, at a 5 cmHz0 pulmonary artery pressure the number of red blood cells per mm2 of alveolar wall was about 5,000. In zone III for the same upstream pressure the value was about double this. An arterial pressure 5 cmHz0
Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (132.174.254.157) on January 15, 2019.
120
PERLO,
above alveolar corresponds to the conditions near the apex of the upright human lung. On the other hand, at 20 cmH20 pulmonary arterial pressure the number of red blood cells per mm2 of alveolar wall was about 13,000 in zone II which is about three-fourths that found in zone III at an equivalent arterial pressure using the line of best fit shown in the figure. At 35 cmHz0 pressure there were about 21,000 red blood cells per mm2 of alveolar wall. This number probably underestimates the number of red blood cells because it was difficult to distinguish overlapping cells when they were very tightly packed. There was no indication that the number of red blood cells reached a maximum as the pulmonary arterial pressure was raised over the range of pressures studied. However Glazier et al. reported that the number of red blood cells per 10 pm of alveolar septum as measured in thin sections did reach a maximum above a capillary pressure of 50 cm Hz0 in zone III.
This was obtained by recording the number of cChits” from a counting grid on red blood cells and the rest of the alveolar wall. Figure 3 shows that in both zones II and 111 the proportion of alveolar wall occupied by red blood cells increased linearly over the range of pulmonary arterial pressures studied. The rate of increase in zone II was greater than that in zone III. At a pulmonary arterial
I
I
I
t
I
I
I
1
5
IO
15
20
25
30
35
40
ARTERIAL-ALVEOLAR
PRESSURE
(cm water)
FIG. 3. Proportion of the area of alveolar wall covered by red blood cells. Notice that where the arterial pressure is only 5 cm&O above alveolar pressure as would be the case near the apex of the upright human lung, only some 12% of the wall is covered. However this increases to over 45% when the arterial pressure is 30 cmHzO as it is near the base of the upright human lung. Vertical bars are SD of mean.
01
JALOWAYSKI,
tained
4, by
AND
I
I
I
1
I
I
1
5
IO
15
20
25
30
35
ARTERIAL-ALVEOLAR FIG.
DURAND,
Comparison of the data Glazier et al. (f 3).
PRESSURE from
the
present
WEST
1 40
(cm woted study
with
that
ob-
pressure of 5 cmHg0, only about 12 % of the wall area was occupied by red blood cells while in zone III about 30 % of the wall area was occupied. At a higher perfusing pressure of 20 cmHz0, 33 % of the wall area was occupied by red blood cells in zone II compared with 40% in zone III. At 35 cmH20 arterial pressure, about 50 % of alveolar wall area was occupied by red blood cells in zone III. It appears that the regression line calculated for zone II (R = 0.80, slope = 1.42) would join that calculated for zone III = 0,68, slope = 0.67) between 25 and 30 cmHy0 upCR stream pressure. There is no evidence that the proportional area of the wall occupied by red blood cells reached a maximum over the range of pulmonary arterial pressures studied. Glazier et al. found that the proportion of alveolar septum occupied by red blood cells in thin sections reached a maximum above a capillary pressure of 50 cmHz0 in zone III. A comparison of the regression lines in Figs. 2 and 3 shows that the number of red blood cells per square millimeter of alveolar wall increased more than the proportional area of the wall which they occupied. For example, in zone III, the number of red cells doubled as the pressure was increased from 0 to 30 cmHz0 pressure while the proportional area took 45 cm pressure for the same increase. In zone II, the difference was in the same direction though less marked. This is consistent with the overlapping of red celfs which could be seen in cross sections as the pressures increased. It means that fewer red cells have their surfaces in contact with the alveolar wall and this will increase the diffusion distance for gases. The proportion (by area) of the wall occupied by red blocd cells determined by point counting alveolar walls en face from thick (50 pm) lung sections is similar to the measurement by Glazier et al. of the percent septum occupied by red blood cells made on cross sections of alveolar septa in thin (2 pm) lung sections. These investigators superimposed an eyepiece graticule calibrated every 5 pm along the septum so that the graduation lines cut the septum perpendicularly. They counted the number of intersections of the lines with red blood cells and thus determined the proportion by length of the septum occupied by red cells. This measurement is equivalent to that obtained in the present study by superimposing a grid of points on the alveolar wall seen en face. Figure 4 compares the present data with those of Glazier
Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (132.174.254.157) on January 15, 2019.
RED -
AND
WHITE
BLOOD
CELLS
IN
6r
“I
E2
LUNG
121
CAPILLARIES
-3-
IO]
1 5
1
t 15
10 ARTERIAL-ALVEOLAR
I 20
I 25
PRESSURE
1 30
t 35
(cm water)
FIG. 5. Number of nucleated cells per mm2 of alveolar wall plotted against arterial minus alveolar pressure. Note that in zone II there was no consistent change but there was a fall in zone III. Contrast this behavior with that of the number of red blood cells shown in Fig. 1, Vertical bars are SD of mean.
This measurement is made by point counting nucleated cells in alveolar walls en face. Figure 6 shows that the proportion of the wall occupied by nucleated cells is similar to that expected from the data presented in Fig. 5, In zone II there is no consistent difference (slope = 0.00, R = 0.01) over the range of perfusing pressures studied while in zone III there is a decrease (slope = -0.09, R = -0.27). At a low perfusing pressure of 5 cmH20 in zone II the relative wall area occupied by nucleated cells was about 16 % which is slightly greater than the relative wall area occupied by red blood cells (Fig. 3) for the same conditions (11 %). This d’ff 1 erence can be explained by the larger size of the nucleated cells compared with the red blood cells. Distribution htracupihry
0’
I
5
I IO
I
1
I
I
I
I
15
20
25
30
35
40
ARTERIAL-ALVEOLAR 6. Proportion
cleated
cells. Compare
by area of the Fig. 4. Vertical
PRESSURE
h
woted
alveolar wall occupied bars are SD of mean.
by nu-
et al. In that study, the zone III results were related to venous minus alveolar pressure, rather than arterial minus alveolar pressure as here. The regression line should therefore be slightly displaced to the right but the error is small because the arterial-venous pressure difference was so low. It can be seen that there is good agreement between the two sets of measurements. Number
of Nucleated
Cells per mm2 of Alveolar
of Nucleated Cells. Vmus Extracapillary
Figure 7 shows that the nucleated cells are predominantly intracapillary both in zones II and III. For this measurement which was made on thin (2 pm) sections, endothelial cell nuclei could be recognized and were not included as either intracapillary or extracapillary. In zone II over the range of pulmonary artery pressures of 10 and 20 cmHz0 about 78 % of the nucleated cells were located within capillaries. In zone III at an upstream pressure of 10 cmHz0, 72 % of the nucleated cells were intracapillary while at a higher perfusing pressure of 20 cmHz0 only 62 % of the cells were located within the capillaries. The differences between zones II and III were significant at the 1 % level. Thus it appears from Figs. 5 and 7 that in zone III there is a decrease in the number of nucleated cells found in alveolar walls as the perfusing pressure is raised and that
WuIl
Figure 5 shows that in zone II there was no consistent change in the number of nucleated cells per mm2 of alveolar wall for increases in pulmonary arterial pressure. However, in zone III there was a decrease, the number reducing from 5,000 at an arterial pressure of 5 cmH20 to 3,000 when the pressure was 35 cmH20. Comparison of the number of red and nucleated cells shows that there was a considerably greater number of nucleated cells in alveolar walls than would be expected from the ratio of nucleated cells to red blood cells & peripheral blood. For example, in zone II at 5 cmHs0 perfusing pressure there were about 5,000 nucleated cells per mm2 of alveolar wall and about the same number of red blood cells (Fig. 2). In zone III at this same upstream pressure the number of nucleated cells were about half the number of red blood cells. It should be emphasized that not all these nucleated cells are white blood cells, some being alveolar, epi thelial, endothelial cells, and macrophages. However, as shown below, the majority of these nucleated cells were white blood cells.
IQ
I5
ARTERIAL-ALVEOLAR FIG.
capillaries.
7. Number Notice
of nucleated that between
20 PRESSURE
(cm weted
cells which are inside and outside 60% and 80% are intracapillary.
the
Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (132.174.254.157) on January 15, 2019.
122
PERLO,
this decrease can be attributed to a fall in the number of nucleated cells within capillaries. We thought that it might be possible to determine the proportions of white cells which were intra- and extracapillary by washing the vascular bed out with saline or dextran. This was tried on one lung but we found that although all the red blood cells could be removed, some intracapillary white cells remained. They were presumably either jammed tightly in the small vessels or Intere adhering to the walls. The bindings presented thus far indicate that the nucleated cells behave differently in zones II and 111 when the arterial pressure is raised. The number of nucleated cells in zone II appears to be uninfluenced by changes in perfusing pressure while the nucleated cells in zone III appear to decrease with increases in pulmonary artery pressure. In addition, most of the nucleated cells are intracapillary. A reasonable mechanism for these differences is that the capilla ries are more d istended in zone III than in zone II for the same perfusin g pressures and that under both conditions distension increases as the perfusing pressure is raised. They are thus less likely to trap the relatively large nucleated cells as they flow through the bed. Glazier et al. showed that the mean width of the capillaries was less in zone II than zone II I for the same perfusing pressure, and also that the width increased with perfusing pressure down zone III. Prqtmrdion of nucleated cells which were @ymor~honuclaar cells. This study was directed at the differential morphology of the cells. Observations of nucleated cells in alveolar walls stained with Giemsa and viewed en face revealed that they could only be confidently divided into polymorphonuclear and mononuclear white blood cells. Figure 8 shows the proporti .on 01 polymorphs ai nongst nucleated cells with arteri al pressu re. It can be seen that changes in pulmonary the proportion is generally low, in the range of about 825 % of all nucleated cells. In zone II the proportion stays fairly constant at 8 %. In zone III at perfusing pressures of 5 and 10 c1nH~0 the proportion of polymorphs amongst nucleated cells is about 11 % but increases to 25 % at an
ARTERIAL-ALVEOLAR 8. Percentage of nucleated cells polymorphonuclear. This was generally pressures in zone III where it was about predominately mononuclear. FIG.
PRESSURE
(cm water)
in the alveolar less than 127, 25oj’. Therefore
walls which are except at higher the cells were
8 -
JALOWAYSKI,
DURAND,
AND
WEST
60
0 ARTERIAL-ALVEOLAR PRESSURE (cm water) 9. I’ercentage by volume of the alveolar septa taken up by open capillaries. Marked increase with the arterial pressure, especially in zone II, is consistent with the mechanisms of recruitment and distension. FIG.
upstream pressure of 20 cmHz0. The differences between zones II and III were significant at the 2 % level These findings are consistent with the observation that the population of intracapillary nucleated cells in zone II remains approximately constant in the face of changes in The increase in proportion of polyperfusion pressure. morphs in zone 111 can perhaps be explained by the preferential loss of smaller nucleated cells through the capillaries as they distend. Observahs on lungs from intact animals. We thought that the unexpectedly high concentration of white blood cells in the pulmonary capillaries might be caused by the perfusing circuit. To test this, we made additional measurements on two dogs in which operative interference was kept to a minimum. In one dog, the lung was rapidly frozen irnmcdiately after opening the chest, inserting a catheter into the pulmonary artery to measure pressure, and placing the animal in the head-up position. In another anesthetized dog, the lungs were flooded with glutaraldehyde introduced via the trachea. In the rapidly frozen lung, the average numbers of nucleated cells and red blood cells in the alveolar walls near the apex of the lung were 3,666 and 4,217 per mm2, respectively. These values are comparable with those found in the excised lungs as shown in Figs. 2 and 5. Toward the base of the lung, the average numbers of nucleated and red cells were 2,994 and 16,373 square mm2, respectively, which again Iit with the previ ous data. I n the glu ta raldehyde-fixed lung, the nu mber of nu cleated cells was approximately 5,500 per mm2 although in this preparation the time taken for fixation to occur was unknown. We conclude from these additional observations that the large concentration of white cells in th e capillaries was not caused by the procedures connected with perfusi .ng the lu ngs. Vulume of Alveolar Septum Occufied by Upen Capillaries This measurement was made by point-counting random cross sections of alveolar septa in thin (2 pm) sections. The
Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (132.174.254.157) on January 15, 2019.
RED
AND
WHITE
BLOOD
CELLS
IN
LUNG
CAPILLAKIES
percentage volume of the septum occupied by open capillaries was determined by the proportional number of hits on open capillaries. Figure 9 shows that in zone II the proportional volume of the septum occupied by open capillaries increased from about 10 % at 5 cmHp0 pressure to about 62 % at 20 cmHk0 pressure (R = 0.80, slope = 3.53). In zone III the proportional volume of the septum increased from about 43 % at 5 crnHk0 pressure to 73 Yo at 20 cmHz0 pressure (R = 0.73, slope = 2.05). The proportional volume of the septum occupied by open capillaries in zone II increased rapidly over the range of pressures studied (O-20 cmHz0). This finding is consistent with that of Glazier et al. (13) and Warrell et al. (30) that the number of open capillaries and/or capillary width increased up to a capillary pressure of 15 cmHa0 in zone II. The sixfold increase in the volume of the septum occupied by open capillaries in zone II is consistent with the data of Warrell et al. (30) which demonstrated that recruitment is the predominant mechanism by which the puhnonary capillary bed responds to increases in upstream pressure in zone II in this pressure range. The smaller increase in the volume of the septum occupied by open capillaries in zone III is consistent with the data of Glazier et al. which showed distension to be the predominant mode of response to increasing upstream pressure. The different patterns exhibited in zones II and III can be explained by an increased capillary pressure in zone III when compared with zone II for equivalent upstream pressures. Note that the data presented here do not of themselves provide evidence in favor of recruitment or distension of capillaries, DISCUSSION
Red Blood GeIl I)istribution in AIuednr Beds: Sign$cant for Gns Exchange
Cu/dlary
Several studies have shown that the regional diffusing capacity increases down the upright human lung (7, 9, 3 1). Glazier et al. (13) f ound that the number of red blood cells per 10 pm of alveolar septum increased linearly with increases in capillary pressure to 50 cmHl0 and suggested that the maximum diffusing capacity could probably not be attained in man because of the very large pulmonary vascular pressures which would be necessary. The present study similarly demonstrates a linear increase in -lilling of the alveolar wall area with red blood cells as pulmonary arterial pressure rises. Over the pressure range studied the percentage of wall area occupied by red blood cells did not reach a maximum although the silicone injected preparation of Sobin et al. (26) and the frozen lung preparation of Warrell et al. (30) would suggest that about 90 ‘% of the wall area is available for capillary blood. These findings further suggest that regional differences in diffusing capacity, DL, could be explained by regional differences in filling of the alveolar wall in the vertical lung. Recently Michaelson et al. have compared diffusing capacity per liter of lung volume (DIJVA) with pulmonary capillary blood flow per liter of lung volume (Qc/VA) in the upright lung using a bolus gas inhalation technique (20). They found that both DL/VA and Qc/VA increased down the lung but that the rate of increase of capillary blood Aow was greater.
123
White cell sequestration by the lung has long been recognized (11, 24). B ierman and colleagues (3, 4) have shown that the lung selectively removes leukemic leukocytes infused into circulating blood whereas a peripheral capillary bed will not. They also reported that histamine induces white cell sequestration and epinephrine stimulates white cell release from the lung. Ambrus et al. (2) postulated that the lung acts as a Yeukostat” in regulating the white count of circulating blood. They infused white cell-rich blood into Starling heart-lung preparations and found the circulating white cell level to be constant and indenendent of the concentration in infused blood despite repeated infusions. In a subsequent study they found that this depletion was primarily due to sequestration of granulocytes by the lung (1). The present study shows that there is a large n umber of n uclea ted cells with in alveolar walls and that thev a re chiefly intercapillary in location. The predominant cell type is mononuclear while a minority are polymorphonuclear leukocytes. Our study further suggests that the puL monary capillary bed contributes to the sequestration of white blood cells on a mechanical basis. In zone II the capillary bed serves as an efficient filter as evidenced by the large number of nucleated cells in relation to the number of red blood cells per mm2 of alveolar wall and the constant number and proportion of the wall occupied by nucleated cells. The capillary bed in zone I II serves as a less eficient filter on the basis of these criteria. This can be explained by the relatively greater distention of the bed in zone III as shown by Glazier et al. (13) which is presumably a reflection of the higher capillary pressure. It should be noted that our observations were confined to the extreme periphery of the lung and the possibility remains that other patterns may apply to other regions. The large concentration of leukocytes in the alveolar wails may affect the pressure-flow relations of the capillary bed. Warrell et al. (30) demonstrated uneven filling of the capillary bed within arteriolar domains and concluded that recruitment occurs on a capillary rather than arteriolar level. White cell sequestration by the capillary bed could certainly influence flow through a capillary segment by altering the pressure drop across a plugged capillary. Preferential flow to a neighboring unobstructed capillary segment could then result. The predominance of the mononuclear cell in the bed is of interest. Recent evidence suggests that one precursor of the alveolar rnacrophage is the blood monocyte derived from a bone marrow stern cell (15, 23, 27). Morphological evidence demonstrating the transformation of the intracapillary monocyte into the alveolar macrophage is lacking although studies of enzyme content and phagocytic abilities of the two cells support this notion (27)* Bowden and Adarnson suggested that the pulmonary interstitial cell is the essential link between the circulating monocyte and the alveolar macrophage (5, 6). A second source of alveolar rnacrophages may be migrating hepatic Kupffer cells caught up in the pulmonary capillary bed (16, 17, 2 1, 27). There is evidence that these cells then move into the alveolar spaces and are su bsequently lost via the airways.
Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (132.174.254.157) on January 15, 2019.
124
PERLO,
We are Gaines This
indebted to Myra for assistance. work was supported
Bailey, by
Richard
National
Matthews, Institutes
and of Health
Richard Grants
HL-13687-03 Administration Received
JALOWAYSKI,
and HL-05931-03 Grant NGL-05-009for
publication
and
DURAND, National 109.
28 November
AND
Aeronautics
WEST
and
Space
1973.
REFERENCES
1, AMBRUS,
C. M., AND J- L. AMBRUS. Regulation of the leukocyte level. Ann. iV. Y. Acad. Sci. 77 : 445-486, 1959. 2. AMBRUS, C. M., J. L. AMBRUS, G. C. JOHNSON, E. W. PACKMAN, W. S. CHERNICK, N. BACK, AND J. W. E,. HARRISON. Role of the lungs in regulation of the leukocyte level. Am. J. Physiol. 178: 3344, lY54. H. R. The hematologic role of the lung in man. Am. J. 3. BIERMAN, Surg. 8Y: 130-140, lY55. H. R., K. H. KELLEY, AND F. 1;. CORDES. The sequestra4. BIERMAN, tion and visceral circulation of leukocytes in man. Ann. NY. Acad.
Sci. 59 : 850-862, 5. BOWDEN, D. H.,
6.
7.
8.
9.
10. 11. 12.
13.
14,
15.
16.
1955.
AND I. Y. IX. ADAMSON. Pulmonary interstitial cell as immediate precursor of alveolar macrophage. Am. J. Pathol. 68: 521-536, 1972. BOWDEN, D. H., I. Y. R. ADAMSON, W. G. GRANTHAM, AND J. P. WYATT. Origin of the lung macrophage. Arch. Puthol. 88: 540-546, 196Y. BURROWS B., H. NIDEN, C. MITTMAN, R. C. TALLEY, AND W. R. BARCLAY. Non-uniform pulmonary diffusion as demonstrated by the carbon monoxide equilibration technique : experimental results in man. J. Clin. Invest. 39 : 943-949, 1960. CONN, I-3. J., M. A. DARROW, AND V. M. EMMEL. Staining Procedures Used by the Biological Stain Commission (2nd ed.). Baltimore, Md. : Williams & Wilkins, 1960. FORSTER, R. E., W. S. FOWLER, D. V. BATES, AND B. VAN LINGEN. The absorption of carbon monoxide by the lungs during breath holding. J. Clin. Invest. 33 : 1135-l 145, 1954. FUNG, Y. C., AND S. S. SOBIN. Elasticity of the pulmonary alveolar sheet. Circulation Res. 30: 45 1469, 1972. GARREY, W. E., AND W. R, BRYAN. Variations in white blood cell counts. physiol. Rev. 15 : 597-638, 1935. GLAZIER, J. B., J. M. B. HUGHES, J. E. MALONEY, AND J. B. WEST. Vertical gradient of alveolar size in lungs of dogs frozen intact. J. A@Z. Physiol. 23 : 694-705, 1967. GLAZIER, J. B., J. M. B. HUGHES, J. E. MALONEY, AND J- B. WEST, Measurement of capillary dimensions and blood volume in rapidly frozen lungs. J. A$$. Physiol. 26: 65-76, 1969. GLAZIER, J. B., AND J. F. MURRAY. Sites of pulmonary vasomotor reactivity in the dog during alveolar hypoxia and serotonin and histamine infusion. J. Clin. Invest. 50 : 2550-2558, 197 1. GODELSKI, J. J., AND J. D. BRAIN. The origin of alveolar macrophages in mouse radiation chimeras. J. Exptl. Med. 136: 630-643, 1972. IRWIN, D. A. Kupffer cell migration. Can. Med. Assoc. J. 27 : 353356, 1932.
17. 18. 19.
20.
21.
22.
23.
24.
KLINGENSMZTH, W. C. III, AND T. W. RYERSON. Lung uptake of 9grnTc-sulfur colloid. J. i2rzlcl. Med. 14: 201-204, 1973, Manual of Histologic Staining Methods of theArmed Forces Institute of Pathology (3rd ed.). New York: McGraw, 1960. MASERI, A., P. CALDINI, P. HARWARD, R, C. JOSHI, S. PERMUT’I-, AND K L. ZIERLER. Determinants of pulmonary vascular volume: recruitment vs. distensibility. Circulation Res. 3 1 : 218-228, 1972. MICHAELSON, E. D., M. A. SACKNER, AND R. L. JOHNSON, JR. Vertical distributions of pulmonary diffusing capacity and capillary blood flow in man. J. Clin. Invest. 52: 359-369, 1973. NICOL, T., AND D. L. J. DILBEY. Elimination of macrophage cells of the reticuloendothelilial system by way of the bronchial tree. Nature 182: 192-193, 1958. PERMUTT, S., P. CALDINI, A. MASERI, W. H. PALER, T. SASAMORI, AND K. ZIERLER. Recruitment versus distensibility in the pulmonary vascular bed. In : The Pulmonary Circulation and Interstitial Space, edited by A. P. Fishman and H. H. Hecht. Chicago, Ill.: Univ. of Chicago Press, 1969. PINKETT, M., C. K. COWDREY, AND P. C. NOWELI,. Mixed hematopoietic and pulmonary origin of alveolar macrophages as demonstrated by chromosome markers. Am. J. Pathol. 48: 859-867, 1966. RATLIFF, N. B., J. W. WILSON, E. MTKAT, D. B. HACKEL, AND T. C. GRAHAM. The lung in hemorrhagic shock. Am. J. Pathol. 65: 325-
331, 1971. 25. SIMPSON, M.
29.
E. The experimental production of macrophages in the circulating blood. J. Med. Res. 43: 77-144, 1922. SOBIN, S. S., H. M. TREMER, AND Y. C. FUNG. Morphometric basis of the sheet-flow concept of the pulmonary alveolar microcirculation in the cat. Circulation Res. 26 : 397-414, 1970. VIJEYARATNAM, G. S., AND B. CORRIN. Origin of the pulmonary alveolar macrophage studied in the iprindole-treated rat. J. Pathol. 108: 115-118, 1972. VREIM, C. E., AND N. C. STAUB. Indirect and direct pulmonary capillary blood volume in anesthetized open-thorax cats. J. A@. Physiol. 34: 452459, 1973. VREIM, C. E., AND N. C. STAUB. Pulmonary vascular pressures and
30.
capillary Physiol. WARRELL,
26.
27.
28.
blood
volume
36: 275-279, D.
A.,
changes 1974. J. W. EVANS,
AND J. B, WEST.
Pattern
J. A#.
32 : 346-356,
31. WEST,
Physiol. J. B., R.
of filling
A. B. HOLLAND,
MATTHEWS. Interpretation lung, J. Appl. Ph+ysiol.
in R.
anesthetized 0.
in the
CLARKE,
cats. G.
pulmonary
J.
Af$Z.
P. KINGABY, capillary
bed.
1972. C.
of radioactive 17 : 14-20, 1962.
T. DOLLERY, gas clearance
AND
C. rates
M.
E.
in the
Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (132.174.254.157) on January 15, 2019.
