INTRODUCTION

Top Abstract Introduction Materials & Methods Results Discussion References

PARTICLES, such as aged red blood cells, bacteria, and immune complexes, are rapidly removed from the blood by macrophages with access to the circulation. This system, classically called the reticuloendothelial system and now termed the mononuclear phagocyte system (MPS), is a collection of mononuclear cells that originate from the bone marrow. They enter the circulation and localize in various organs where they mature into tissue macrophages. Macrophages with direct access to the circulating blood have historically been reported in the liver, spleen, and bone marrow (23). However, recent studies (2, 3, 15, 24-26, 28) showed that the lung is also a significant site for intravascular macrophages.

It has been demonstrated that sheep, calves, horses, and pigs have significant pulmonary uptake of intravenously injected particles, bacteria, or endotoxin (3, 15, 24, 26, 28). These studies revealed that distinct populations of resident phagocytic cells within the pulmonary capillaries are responsible for particle uptake. These pulmonary intravascular macrophages (PIMs) are large (20- to 80-µm-diameter), mature macrophages that are bound to the pulmonary capillary endothelium. PIMs have characteristic morphological features of differentiated macrophages, such as irregular shape, an indented nucleus, lysosomal granules, pseudopods, phagosomes and phagolysosomes, tubular micropinocytosis vermiformis structures, and a fuzzy glycocalyx (24, 31). Atwal et al. (1) described in detail the electron-dense surface coat of PIMs that appears to be primarily lipoprotein in nature.

PIMs also form membrane-adhesive complexes with underlying endothelial cells. These adhesions have an intercellular separation of 12-15 nm and electron-dense material present both in the intercellular space and subjacent to the plasma membrane of both cells (24, 30). These adhesion complexes most closely resemble adherent or intermediate junctions of epithelia, which have a 20-nm interspace and some associated cytoplasmic filaments. These adhesive complexes are believed to be important in anchoring macrophages to the microvasculature. Their presence is a useful criterion for distinguishing PIMs from adherent monocytes. PIMs attach preferentially to the thick portion of the air-blood barrier (29, 31), thus perhaps minimizing any potential interference with gas exchange at the air-blood barrier. This may be important because ~15 m² of the sheep lung capillary endothelial surface is covered with PIMs (24).

There is functional evidence consistent with the existence of PIMs from particle uptake studies (9, 17) carried out more than 70 years ago. More recently, other investigators (2, 16, 31) have demonstrated the existence of PIMs. However, inevitably, different investigators have used a variety of tracer particles and protocols to study removal of particles from the circulation and to demonstrate the existence of PIMs. In this paper, we describe a systematic comparative survey of 13 species with regard to the clearance kinetics and organ distribution of intravenously injected ¹⁹⁸Au colloid and magnetic iron oxide particles. Radioactive ¹⁹⁸Au colloid was used to determine clearance kinetics and organ uptake. Iron oxide particles were also used to identify the cell types responsible for organ uptake of particles because these particles are visible by light and electron microscopy.

	MATERIALS AND METHODS

Top Abstract Introduction Materials & Methods Results Discussion References

Animals. The characteristics of the animals used in this study are summarized in Table 1. Four healthy adult animals of each of the 13 species (two for goats) were obtained from various sources, kept in conventional housing, and given commercial food and water ad libitum. The calves and sheep were studied as part of previously published reports (24, 25). All animals were confirmed free from pneumonia or other significant pulmonary pathology at postmortem examination.

Radiolabeled ¹⁹⁸Au colloid: clearance and organ uptake. Four animals of each species except goat and hamster were injected intravenously with 0.01 mg/kg of ¹⁹⁸Au colloid (0.001 mg/kg for sheep and calves due to their large size; Amersham, Arlington Heights, IL). The average particle diameter was 20 nm. The colloid concentration was 0.01 mg ¹⁹⁸Au colloid/ml. It was given through the jugular (sheep, calves, hyrax, guinea pigs, and pigs), tail (rats and mice), wing (chicken), ear (rabbits), or cephalic (cats and macaque monkeys) vein. Duplicate aliquots of blood were taken before and at 1, 3, 5, 10, 15, 20, 25, 30, 40, 50, and 60 min after injection of ¹⁹⁸Au. The volume of blood collected was adjusted depending on the estimated total blood volume of each species (from <1% of blood volume in calves to 10% in mice). The rates of vascular clearance were determined by measuring radioactivity in sequential blood samples. The animals were humanely killed, and the heart, left lung (right lungs were used for morphology), liver, spleen, kidney, and samples of bone marrow and skeletal muscle were removed. Whole organs were weighed, and the radioactivity in up to 10 randomly selected samples of each tissue was determined in a Packard model 2001 gamma counter. Radioactivity, expressed as counts per minute per gram of tissue, was determined for each organ or tissue, and the percentage of the total recovered dose localized in each organ was computed. The organ and tissue weights not measured in all species were estimated (as a percentage of total body weight) as follows: bone marrow, 3%; skeletal muscle, 45%; and peripheral blood, 7% (8, 11, 12). The total lung weight was calculated using the estimate that the left lung is 67% of the right lung based on measurements made in normal animals of each species.

	RESULTS

Top Abstract Introduction Materials & Methods Results Discussion References

Clearance kinetics of intravenously injected radioactive ¹⁹⁸Au colloid. The clearance kinetics of intravenously injected ¹⁹⁸Au colloid particles from the blood in the 11 species studied were similar (Fig. 1A). The fact that the individual curves are indistinguishable demonstrates the similarity in rates of vascular clearance of ¹⁹⁸Au colloid. In every species studied, the blood was cleared of gold particles very rapidly during the first 5 min. The mean clearance half-times did not differ significantly and ranged from 1.04 (calves) to 2.14 (hyraxes) min (Fig. 1B). After 10 min, >90% of the ¹⁹⁸Au had been cleared in all species. The small remaining fraction of ¹⁹⁸Au was cleared more slowly over the next hour. At 60 min, essentially 100% of the injected ¹⁹⁸Au had disappeared from the circulating blood.

View larger version (26K): [in this window] [in a new window]   View larger version (19K): [in this window] [in a new window]   Fig. 1.   A: vascular clearance of radioactive 198Au colloid in various species. Data are percentages of extrapolated initial value. Clearance curves in 11 species studied are almost indistinguishable. B: clearance half-times of radioactive 198Au colloid. Data are means ± SE; n, no. of animals. Consistent with A, half-times are not significantly different (P > 0.05 by ANOVA).

Organ distribution of recovered ¹⁹⁸Au colloid. The distribution of recovered ¹⁹⁸Au from the lungs and liver of 11 species 1 h postinjection is shown in Fig. 2A. In all species, there was detectable pulmonary uptake of ¹⁹⁸Au colloid particles. Ninety percent or more of the recovered radioactivity was found in the livers of the monkey, hyrax, rabbit, guinea pig, rat, mouse, and chicken, and minimal gold uptake was detected in the lungs. On the other hand, moderate to substantial pulmonary uptake was observed in the cat, pig, calf, and sheep (17-60%). In all species studied, the lungs and the liver together accounted for 89.1 (rat) to 96.6% (chicken) of the total radioactivity recovered from the carcass. Nearly all of the injected dose of ¹⁹⁸Au colloid was recovered from each animal. The percentage of injected dose recovered from the entire carcass ranged from 94 to 100%.

View larger version (27K): [in this window] [in a new window]   View larger version (30K): [in this window] [in a new window]   View larger version (21K): [in this window] [in a new window]   View larger version (20K): [in this window] [in a new window]   View larger version (20K): [in this window] [in a new window]   Fig. 2.   Distribution of recovered intravenously injected radioactive 198Au colloid and magnetic iron oxide particles in various species. Data are expressed as percentage of total recovered radioactive 198Au colloid (A) or magnetic iron oxide (B) that was found in lungs or liver. C-E: organ and tissue distribution of intravenously injected 198Au and iron oxide particles in cats (C), monkeys (D), and guinea pigs (E). Data are expressed as percentage of total recovered dose that was found in each organ or tissue 1 h after 198Au colloid and 30 min after iron oxide injections. Data are means ± SE; n = 4 animals/species.

View larger version (66K): [in this window] [in a new window]   Fig. 3.   Overview of lungs after injection with iron oxide particles. A: lung from a rat injected with iron oxide particles. Lung appears normal and devoid of detectable iron oxide particles (glycol methacrylate; magnification, ×4,500). B: sheep lung after injection of iron oxide particles. In contrast to rat lung, abundant iron oxide particles (arrow) are visible (glycol methacrylate; magnification, ×4,500).

View larger version (202K): [in this window] [in a new window]   Fig. 4.   Cells responsible for pulmonary uptake of iron oxide particles. A: typical cat pulmonary intravascular macrophage (PIM), with irregular shape and extensive areas of interaction with capillary endothelial cells (E). Note electron-dense particles within phagosomes (arrow). Box magnified in inset to show characteristic adhesion complexes (arrowheads), 2 plasma membranes ~12-15 nm from each other, and segments of electron-dense regions in both PIM and endothelial cell. A, alveolar space; C, capillary lumen. B: monocyte (M) in monkey lung with engulfed iron oxide particles (arrow). No adhesion complexes were seen. C: neutrophil (N) in a hyrax lung with engulfed iron oxide particles (arrow). D: monkey lung. Occasionally in non-PIM species, endothelial cells were seen with particles within vacuoles (arrow). E: PIM with characteristic morphology from a calf not injected with any tracer particle. Box magnified in inset to show adhesion complex (arrowheads), PIM, and endothelial cells. Bars, 1 µm.

View larger version (87K): [in this window] [in a new window]   Fig. 5.   Overview of liver after injection of iron oxide particles. A: section of rat liver showing numerous dense iron oxide particles (arrows; glycol methacrylate; magnification, ×7,440). B: in sheep, no iron oxide particles were observed (glycol methacrylate; magnification, ×7,200).

View larger version (154K): [in this window] [in a new window]   Fig. 6.   Hepatic macrophage responsible for liver uptake. Guinea pig liver shows a large Kupffer cell (K) with crystalline electron-dense iron oxide particles (arrows) within phagosomes. S, sinusoid; H, hepatocyte. Bar, 1 µm.

	DISCUSSION

Top Abstract Introduction Materials & Methods Results Discussion References

Clearance of particles from the blood. The half-times for clearance of a nonsaturating dose of very small (20-nm) ¹⁹⁸Au colloid particles from the circulation were not different among the species tested (sheep, calf, pig, cat, monkey, hyrax, rabbit, guinea pig, rat, mouse, and chicken). Thus the presence or absence of a dominant pulmonary compartment in the MPS does not alter the clearance rates of injected ¹⁹⁸Au colloid, only the organ distribution of cleared particles.

Attempts to quantitate vascular clearance of iron oxide particles were abandoned because the clearance was too fast to accurately measure. In the sheep and calves, it occurred during the first pass through the lungs, and no particles were detected in blood samples at 1 min postinjection. The more rapid clearance and the greater likelihood of first-pass clearance of iron oxide particles compared with ¹⁹⁸Au colloid is likely due to their larger particle size, which increases the likelihood of particle-phagocyte receptor interaction as particles pass through the pulmonary circulation. A second possible reason is the larger number of gold particles due to their much smaller size. The relatively longer circulation time (i.e., less complete first-pass clearance) of smaller and more plentiful ¹⁹⁸Au colloid particles may also explain the greater liver uptake of ¹⁹⁸Au colloid than of iron oxide in calves, pigs, and cats despite the presence of PIMs (Fig. 2A). In these species, iron oxide particles never reach the liver when given intravenously because of pulmonary uptake.

Pulmonary versus hepatic uptake. The quantitative measurement of organ uptake of intravenously injected particles was used as a functional assay for the presence of phagocytic cells lining endothelial cells in various organs. More than 80% of the iron oxide particles rapidly removed from the blood were found within the lungs of sheep, calves, pigs, goats, and cats. It has been previously shown (5) that if iron oxide particles are presented first to the liver in the sheep, they are efficiently removed by Kupffer cells. Thus the lack of iron oxide in the liver of these species reflects the effectiveness of PIMs, not a dysfunction of hepatic macrophages. The hepatic uptake of radioactive ¹⁹⁸Au colloid in species with PIMs likely reflects the fact that many of these smaller particles escape initial clearance during pulmonary circulation. For both particles, the lungs and liver accounted for the bulk of their removal from the blood in all species. On the basis of other studies in our laboratory, the distribution patterns of the same dose of intravenously injected iron oxide were similar regardless of prior ¹⁹⁸Au colloid administration. This was true in rats (without PIMs) (27) and sheep (with PIMs) (unpublished data). The differences in pulmonary uptake of both ¹⁹⁸Au colloid and iron oxide particles among animal species with PIMs may be due to variation in their PIM volume densities or in the phagocytic activities of PIMs. A comparative morphometric estimation of PIM numbers or volumes or rates of phagocytosis by isolated PIMs from different species is necessary to elucidate this observation.

Function of PIMs. It can be assumed that natural endogenous targets similar in size and number to ¹⁹⁸Au colloid and iron oxide would be removed from the blood by PIMs in a pattern resembling that seen here. During an inflammatory response, porcine PIMs were shown to engulf neutrophils and fibrin after Haemophilus pleuropneumoniae-induced pneumonia (3). Several investigators have noted erythrophagocytosis by PIMs in, for example, normal goat (2), sheep (28), and cat (21). These findings suggest that PIMs may play a role in normal removal of effete erythrocytes or fibrin and cell debris during inflammation.

PIMs have a role in the clearance of circulating bacteria, an important host defense function of the MPS. This function is usually ascribed to hepatic macrophages. Intravascular infusion of bacteria into pigs has been used as a model of septicemia, and investigators measured pulmonary clearance of bacteria by quantitative culture of blood samples drawn simultaneously from the aorta and pulmonary artery. Efficient clearance of Escherichia coli, Pseudomonas aeruginosa, and Staphylococcus aureus during pulmonary circulation was demonstrated (6). In sheep, Warner et al. (28) showed >90% uptake of recovered intravenously injected P. aeruginosa in the lungs. A subsequent study (26) documented similar pulmonary uptake of injected radiolabeled endotoxin in sheep.

Thus the results seen here would be anticipated in a general way for a wide variety of targets given experimentally by intravenous injection or entering naturally via lymphatics or venous drainage. However, PIM-containing species can have effective hepatic clearance. Thus, when bacteria enter through the gut and are first presented to the liver through the portal circulation, liver uptake will be far greater (5). Interestingly, endotoxin is still retained to a greater extent in the lungs in the sheep even when given through the portal vein. Perhaps this is due to its slower clearance and smaller macromolecular aggregate size.

Is pulmonary uptake beneficial to the host? There are circumstances when PIMs may benefit the host. Their capacity to remove cell debris, bacteria, immune complexes, and endotoxin from the circulation may protect critical downstream organs such as the brain or heart. However, pathogen phagocytosis by PIMs can also produce an inflammatory response in the microvasculature and elicit acute lung injury. Once activated, PIMs may influence events in the pulmonary microvasculature through release of inflammatory mediators.

Acute respiratory distress syndrome, a clinical entity seen frequently in humans after major trauma, burns, or sepsis (14), is frequently associated with gram-negative sepsis or endotoxemia. Many animal species, such as rats and rabbits, are quite resistant to endotoxin-induced acute lung injury, but three species that have been commonly used as animal models for this process are sheep, goats, and pigs (10, 18, 20). Their increased sensitivity to endotoxin is likely due to the presence of abundant PIMs.

Dehring and Wismar (7) reported large mononuclear cells with phagocytic vacuoles in clinical human lung biopsy specimens, but their number was not estimated. A morphometric study of human lung (32) did not show macrophages or macrophage-like cells in the pulmonary capillaries. Results of particle uptake studies in humans suggest that they do not have PIMs. Under normal circumstances, ^99mTc-sulfur colloid injected into the human venous system is largely taken up by hepatic and splenic macrophages, which is the basis of the liver-spleen scans used clinically (13). However, there are reported instances of detectable lung uptake, usually in cases of severe liver damage (13). Perhaps pulmonary uptake in humans is more likely when Kupffer cells are compromised (19) or when organ injury results in monocyte margination in lung capillaries and the differentiation of these monocytes into mature macrophages (4, 27). When phagocytic uptake of circulating pathogens in human lungs is enhanced, the risk of lung injury and acute respiratory distress syndrome is likely increased.

In conclusion, our results show marked differences in the distribution of uptake of intravenously injected ¹⁹⁸Au colloid and iron oxide particles among various species. These differences are largely attributable to the anatomic distribution of mononuclear phagocytes that have access to the circulating blood in these different species. Ruminants, pigs, and cats have an extensive resident pulmonary intravascular component (PIMs) that avidly removes circulating particles from the blood. Rodents, monkeys, and chickens do not. In these latter species, the modest localization of particles observed in the lungs was primarily by marginated monocytes and neutrophils. Thus, in selected species, there is functional and morphological evidence that the lungs are an important component of the MPS.