Macrophage developmentIII. Transformation of pulmonary macrophages from precursors in fetal lungs and their later maturation in organ cultureкод для вставкиСкачать
THE ANATOMICAL RECORD 232551-571 11992) Macrophage Development: 111. Transformation of Pulmonary Macrophages From Precursors in Fetal Lungs and Their Later Maturation in Organ Culture SERGE1 P. SOROKIN, NANCY A. McNELLY, DANA G. BLUNT, AND RICHARD F. HOYT, JR. Laboratory of Pulmonary Cell Biology, Department of Anatomy and Neurobiology, Boston University School of Medicine, Boston, Massachusetts ABSTRACT The fate of macrophage precursors residing in 14-day prenatal rat lungs was followed in organ cultures to obtain a detailed, ultrastructurally resolved picture of the sequence and timing of events accompanying their transformation into typical pulmonary macrophages. Cultures were examined at close intervals during the first day (1, 2, 3, 4, 6, 9, 12, 15, 18, and 24 hr) and at wider intervals thereafter (2, 4,5, 7, 9, and 13 days) to yield a developmental series of cells identified as in the macrophage line based on binding of peroxidase-coupled isolectin B, of Griffonia simplicifolia (GSA I-B4) to cell membranes and on negligible content of peroxidase-positive granules in the cytoplasm. Organ culturing stimulated virtually all precursors to develop into macrophages. GSA-positive cells in explants occurred outside vessels in pulmonary connective tissue, and a t the outset none were typical macrophages: 71% were angular cells, resembling unlabeled mesenchymal cells around them, 16% were undifferentiated leukocytes, and the remainder were irregularly shaped cells with few vacuoles intermediate between the preceding and the macrophages. During the first 12 hr in culture the proportion of angular cells and leukocytes fell to zero, and that of intermediate cells first rose, then receded. In the same interval the proportion of macrophages rose to 87.5%, and by 24 hr all GSA-positive cells were typical macrophages generally engorged with phagocytosed material; about 8 hr appear necessary for converting half the population. Notable ultrastructural changes during this period of transformation involved the centrioles and cytoskeleton, reflecting enhanced cell mobility and phagocytosis. A period of maturation followed, marked by disappearance of cellular debris from phagosomes and an increased prevalence of cells with elaborate lamellipodia. This accords with earlier work showing that macrophage Fc receptor density increases sharply during the first 24 hr, but elevated levels of histochemically demonstrable acid phosphatase appear only later. Mitotic activity was conspicuous in GSA-positive cells throughout both periods. 3H-thymidine labeling indices for precursors and macrophages, determined at six intervals between 1hr and 24 hr, remained steady at -34%, whereas indices of other categories of lung cells (GSA-negative stromal cells, pleural cells, and airway epithelium) began at this level but rapidly declined, indicating that the GSA-positive cells constitute a single population distinct from others in the lungs. Macrophages found outside the lung cultures after 4-5 days qualify as a mature population, but having migrated away from direct contact with the lung stroma, they survive only a week or two and no longer divide. It is now known that precursors of pulmonary mac- nary tissues and from mitotic activity, prominent rophages exist in mammalian lungs from earliest times among cells already recognizable as macrophages. and that cells with morphological characteristics of Judged by many criteria, these cells are indistinguishmacrophages gradually accumulate in the organ dur- able from typical alveolar macrophages of adult lungs ing much of fetal life, to build up more rapidly just prior and subsequent to birth (Sorokin et al., 1984; Received June 19, 1991; Accepted October 2, 1991. Honda et al., 1989). Such cells appear on an accelerated Address reprint requests to Professor Sergei P. Sorokin, Pulmonary schedule after lungs from late embryonic rats or ham- Cell Biology, Department of Anatomy and Neurobiology, Boston Unisters are explanted to organ culture. This results from versity School of Medicine, 80 East Concord Street, Boston, MA rapid transformation of precursors present in pulmo- 02118. 0 1992 WILEY-LISS, INC 552 S.P. SOROKIN ET AL. (Sorokin and Hoyt, 1987; Sorokin et al., 1989). The origin of some alveolar macrophages is therefore attributable to early resident precursors, although the relative importance of this source has not yet been determined, as against seeding of the lungs by circulating stem cells or more specialized monocytes at later times in life. This paper reconstructs the process of transformation from precursors into typical pulmonary macrophages as it occurs in organ cultured embryonic lungs. It is based on electron microscopy, which we consider essential for accurate staging of events. Two major objectives of the study were to determine the time scale of transformation and to depict morphological stages along the pathway of transformation in some detail. In addition, an effort was made to select out the cell type or types of origin for this population, considered from candidates present in the lungs at explantation. As shown here, an outstanding difference between the developmental pathways of these “premedullary” pulmonary macrophages and typical bone marrow-derived macrophages is that none of the candidate precursors is a typical monocyte. This also seems true for kindred “primitive” or “fetal” macrophages located outside the lungs. In mice and rats this line of cells appears to originate directly from precursor cells in blood islands of the yolk sac (Takahashi et al., 1989, 1990) and we have seen that a considerable buildup of the population occurs in the embryo itself prior to the time macrophages first appear in the organs (Sorokin et al., 1992a). Such early macrophages are later supplemented by monocyte-derived macrophages, but as in the lungs, the extent to which their direct descendents persist in adults is still uncertain. In recent years acceptance has grown that macrophage populations of adults are also heterogeneous, and that diversity may even exist within the population of a single organ, with subsets sometimes differing greatly in renewal kinetics (van Rooijen et al., 1989). Although the idea that blood leukocytes give rise to macrophages is well over 60 years old (Lewis, 19251, and ultrastructural aspects of monocyte transformation have been known for about 25 years (Sutton and Weiss, 1966), many important questions remain unanswered about the transition between precursor and macrophage, or require reinvestigation in the light of current concepts of the heterogeneity of these cells. Observations made in relation to specific macrophage subsets, like those that follow in relation to the primitive/ fetal macrophage line, may well discourage uncritical application of the term, “monocyte,”to any immediate precursor. Indeed, “exact description andlor identification of the transitional stages” has been considered essential to understanding the differentiation pathway of macrophages (Kreipe et al., 1987). more, with the assistance of a marker that adequately selects developing macrophages and their precursors from other cells inside these cultures, morphological stages of transformation can be reconstructed simply by harvesting the cultures at suitable intervals, noting the form of labeled cells predominating at each interval, and arranging these forms in the correct temporal sequence. The marker chosen for our work was the peroxidase-coupled isolectin B, of Griffonia simplicifolia (GSA I-B4), which recognizes terminal a1pha-Dgalactose on cell coats. It is selective for cell membranes of macrophages and antecedents in the macrophage line, although not to the exclusion of all other leukocytes, since it tags heterophils and eosinophils equally well (Sorokin and Hoyt, 1992). Two special circumstances, however, made GSA I-B, an entirely satisfactory marker for the macrophage line in our organ culture system: 1)Comparatively few cells intrinsic to embryonic or fetal rat lungs bind this lectin. 2) The growing macrophage population in the cultures is remarkably pure; virtually no granulocytes and few lymphocytes have ever been seen in Giemsa-stained mature cultures from embryonic lungs despite exhaustive examination of serial sections (Sorokin and Hoyt, 1987; Sorokin et al., 1989). Consequently, possible confusion from GSA binding to other leukocytes was not of serious concern. General Procedure Eleven litters of 13.5 t o 14 day prenatal rats [crownrump (CR) length 8.5-10 mml furnished the main specimens used for this study-either the right lung or the left, but most frequently the paired lungs. Of these, ten specimens were dissected out and preserved as records of starting conditions; all were examined. The remaining lungs were placed in organ culture and subsequently harvested for processing at the following hourly or daily intervals (in parentheses, number of specimens examined): l ( n = 31, 2(n = 61, 3(n = lo), 4(n = 9), 6(n = 12),9(n = 12), 12(n = l l ) , 15(n = 13), 18(n = lo), and 24 hours (n = 121, and at 2(n = 1),4(n = 8),5(n = 2),7(n = 5),9(n = 51, and 13 days (n = 1). The lungs and lung cultures were fixed in 1%glutaraldehyde in 0.1 M phosphate-buffered saline (pH 7.2), incubated in a solution containing peroxidase-coupled GSA I-B4, followed by development in a diaminobenzidine substrate, osmication, and embedding in Epon. Details of organ culturing, fixation and staining procedures, and controls used to ascertain the specificity of GSA I-B, binding were given previously (Sorokin and Hoyt, 1992). Semi-thin (0.5 pm) sections of each lung and culture were counterstained with 0.05% toluidine blue and examined a t 1-8 generally horizontal planes spaced about 10 pm apart, to maximize surveillance of macrophages within each block and minimize excessive encounters with the same cell. Subsequent ultrastructural examination of thin sections at these MATERIALS AND METHODS levels proved that most cells of interest were identifiable by light microscopy, although electron microscopy Rationale was required to resolve cellular detail needed for stagGiven that embryonic rat lungs are free from mor- ing the transition to macrophages. Sufficient thin secphologically recognizable macrophages until after the tions a t each sampling plane were examined to provide lungs are placed in organ culture, an opportunity serial views of certain cells, useful for determining arises for determining the length of time needed for their shape and relationship to neighboring cells in the precursors to transform into macrophages. Further- lungs. DIFFERENTIATION O F LUNG MACROPHAGES H- Thymidine Labeling Indices 3H-thymidine labeling indices were obtained in organ cultures of 14-day prenatal lungs at 1,4, 7,13, 19, and 24 hr after explantation, for cells in the following categories: macrophages and precursors (GSA I-B, stained), stromal cells, pleural cells, airway epithelium, and fetal erythroblasts. Explants were placed on medium containing 4 pCi/ml 3H-thymidine for 1hr immediately before fixation. After processing with peroxidase-coupled lectin and embedding in glycol methacrylate, 2 pm sections were cut, mounted on glass slides, dipped in Kodak NTB-2 emulsion mixed 1:l with distilled water, and exposed for 20 days at 4°C. Slides were developed in Kodak D-19 developer, fixed in 24% sodium thiosulfate, and stained with 0.05% toluidine blue. Cells with five or more grains (above background) over or touching the nucleus were considered labeled. RESULTS Morphological Form of GSA I-B,-Positive Cells at Explantation In lungs examined from the 14th day of gestation, GSA I-B,-positive cells occurred only in the pulmonary connective tissue, a t this time constituted as a loose mesenchyme somewhat condensed about developing bronchi. There were no labeled cells in the epithelium nor among mesothelial cells of the visceral pleura. Within the connective tissue positive cells were thinly scattered throughout the extravascular compartment. Blood vessels contained peroxidase-positive erythroblasts and a leukocytic population consisting almost entirely of immature megakaryocytes rounded out by a very few undifferentiated cells of uncertain lineage. None of these intravascular leukocytes stained by the lectin marker; as exceptions, however, two GSA-positive leukocytes were found inside vessels of newly cultured lungs (see below), and so may have been there at explantation. The plasmalemma of vascular endothelial cells occasionally appeared stained. When this occurred, however, it was usually on only one surface, more frequently the luminal. Instances with both luminal and basal surface staining occurred too infrequently to certify pulmonary endothelial cells as GSAIB,-positive, and by this criterion to admit them for consideration as macrophage precursors. Cells with the form and ultrastructural appearance of macrophages were not found in any of our lung specimens at 14 days of gestation. Those qualifying for consideration as macrophage precursors by the criterion of GSA I-B, binding were primarily 1) rather angular, fusiform cells sandwiched in among the mesenchymal fibroblasts, having a moderately euchromatic nucleus, a large nucleolus, and a cytoplasm drawn out into two or more major processes each bearing a varying number of short secondary processes (Fig. 1). As seen in thin sections, typical GSA-positive cells of this kind rather resembled surrounding fibroblasts in size and shape, with the bulk of cytoplasm in the processes rather than centered about the nucleus. Such cells usually appeared to be about twice as long as their nuclear diameter, but those fortuitously sectioned along their major axis measured as much as 6 diameters long. The presence of many GSA-positive, nucleus-free cell pro- 553 files in the connective tissue testified further to the elongate, branching form of these cells. Their cytoplasm appeared undifferentiated, characterized by abundant polysomes, moderate numbers of mitochondria, little development of cytoplasmic membrane systems, and rarely a small cluster of peroxidase-positive granules, too meager in diameter (ca. 100 nm) and too few in the aggregate to register by light microscopy. Alongside equally undifferentiated fibroblasts, these cells at times stood out by their greater electron density in addition to their GSA binding. They were often observed brushing their cell membranes against those of the fibroblasts but never appeared to form specific junctions with them. In counts made from the electron micrographic record of 14-day prenatal lungs, these angular cells made up 71% of all GSA-positive cells present. 2) The second largest population of GSA I-B,-positive cells in the extravascular stroma of 14-day lungs consisted of mixed undifferentiated leukocytes. Rarely exceeding three to four cells per specimen, they made up only a fraction of the combined total of 60-70 intraand extra-vascular leukocytes present in each lung pair at 14 days (Sorokin and Hoyt, 1987). Some of these labeled cells resembled medium sized lymphocytes with an electron-dense cytoplasm rich in polysomes, and few exhibited more than an occasional profile of granular endoplasmic reticulum. Exceptionally, one lung specimen contained three cells characterized by a distinctly notched nucleus and coliform, peroxidasepositive granules within a dense, polysome-rich cytoplasm. The granules were 0.1-0.4 pm long, radially arrayed about the cytocentrum, and visible by light microscopy, causing the cells to stand out in semi-thin sections. They resembled lysosomal granules present in bone marrow-derived neutrophilic promyelocytes or promonocytes and afforded some basis for affiliating the cells with the granulocyte-macrophage line (GM cells). A third type of leukocyte contained conspicuous, peroxidase-positive cisternae of endoplasmic reticulum and large, peroxidase-positive and distinctly more rounded granules than in the progranulocyte-macrophages. These cells closely resembled eosinophilic promyelocytes of adult rat bone marrow (Bainton and Farquhar, 1970). None was found in any of the uncultured lungs we examined; however, two were found in a culture a t 3 hr and another a t 4 hr. Consequently, they likely had been present a t explantation. Cells with any of the foregoing descriptions (Fig. 2) were simply grouped together as leukocytes and totaled 16% of the GSA-positive cells present. 3) The third category of GSA I-B,-positive cells in 14-day lungs was intermediate in appearance between the angular cells and the more agranular of the leukocytes and so equally could be thought to derive from them. Cell processes were less extended and more cytoplasm surrounded the nucleus than in the angular cells; on the other hand, cell outlines were more irregular than among the leukocytes. Prior to explantation or immediately afterwards, intermediate cells resembled those of the other two classes in possessing an undifferentiated cytoplasm and very few vacuoles (Fig. 3). Peroxidase-positive granules were either absent or inconspicuous, as in angular cells. The intermediate cells made up 14% of the total. 554 S.P. SOROKIN ET AL. All figures are electron micrographs of 14-day prenatal rat lungs or of organ cultures of 14-day explants, incubated with peroxidase-coupled GSA I-B, to reveal precursors and maturing pulmonary macrophages by the binding of the isolectin to the cell membrane. All have been counterstained with uranyl acetate and lead citrate. In the designation, 14 +(n), n equals the time in culture. Beginning at zerotime (Fig. 1)and in the time-resolved sequence ending at 7 days (Figs. 7-15), illustrations have been selected to portray the most characteristic cell o f macrophage lineage present in the culture at the stated time. In addition, low-power views are included to illustrate GSApositive cells in relation to other pulmonary tissues as well as changed conditions in the cultures between + 6 and + 12 hr (Figs. 3, 5, respectively). Additional illustrations (Figs. 2, 6, 16, 17) are not in chronological order but document special points raised in the text. TRANSFORMATION OF PRECURSORS INTO MACROPHAGES climbed to 65% of the population, when a new category of GSA-positive nascent (or early) macrophages first appeared. These cells were larger and more rounded than the intermediates, having phagocytic vacuoles with variable, electron-dense contents. Intermediate cells continued to hold to around 65% of the GSA-positive population through the 6th hr. The remainder consisted almost exclusively of nascent macrophages, but by 9 h r these cells, augmented now by a comparatively small number of more mature-appearing macrophages, had risen to 67%of the total GSA-positive population. By 12 hr, the macrophage pool (87.5%)mainly consisted of the more mature cells, and by 15 h r the GSA-positive population comprised 90% macrophages and only 10% intermediate cells. After 24 hr, immature lectin-positive cells occasionally were present in the cultures, although they never made up more than a small fraction of the total. Comparison of low-magnification electron micrographs of prenatal lungs after 6 h r and after 12 h r of culturing underscores the dramatic shift that the GSApositive population undergoes over the interval, from a predominance of intermediate cells to outright macrophages (Figs. 3, 5). Immediately after the 14-day prenatal lungs were placed in organ culture, GSA I-B,-labeled precursors began to t u r n into macrophages. In a n initial phase of this process, GSA I-B,-labeled cells shifted from a mixed population of angular cells, leukocytes, and intermediate cells to one consisting exclusively of overtly phagocytic cells having ultrastructural characteristics of macrophages. A second phase continued past 24 hr and was characterized by maturation changes affecting cells already classifiable as macrophages. Only ultrastructurally visible aspects of maturation are addressed here, but some of these can be correlated with functional aspects of maturation examined in earlier studies on these cells. For present purposes, transformation will be defined as equivalent to the first of these two phases and maturation to the second. During the first hours of organ culturing, the proportion of intermediate cells in the total GSA-positive population grew rapidly while that of angular cells and leukocytes declined (Table 1, Fig. 4). By 2 hr, they had Fig. 1. 14-day prenatal lung a t explantation. GSA-positive angular cell in the mesenchymatous connective tissue. x 11,140. DIFFERENTIATION OF LUNG MACROPHAGES 555 Fig. 2. a: “Primitive” macrophage present in a pulmonary blood vessel at 14 + 6 hr. The cell has an enlarged nucleolus and remains a n undifferentiated leukocyte showing little response to the trauma of lung explantation, compared to GSA-positive cells in the connective tissue at this time. x 11,530.b Intermediate cell in pulmonary connective tissue at 14 + 4 hr. It is one of few found differing from typical intermediate cells in possessing coliform peroxidase-positive granules of appreciable size, possible grounds for association with GM lineage. x 6,680. c: GSA-positive cell with round peroxidase-positive granules and reaction product in the endoplasmic reticulum, present a t 14 + 3 hr. This cell type was not observed in cultures beyond + 4 hr. x 8,610. The angular cells appeared to be mainstream precursors of the macrophages because they were the commonest of GSA-positive cells in 14-day explants and underwent a steady, proportionate decline upon cultur- ing, in tandem with a rise in intermediate cells initially and later with macrophages (Fig. 4). Leukocyte-like GSA-positive cells also underwent early, proportionate decline, but their initial small 556 S.P. SOROKIN ET AL. Fig. 3. Low-power view of a lung culture at 14 +6 hr. Three intermediate cells (arrows) are present outside blood vessels in the connective tissue, which has the texture of mesenchyme. The epithelium is seen at the upper right, and the blood vessels contain dark-staining erythroblasts. Compare with Figure 5. ~2,340. number and comparatively irregular occurrence in the cultures disqualified them as prime candidate-precursors for the macrophages. We found no reason to deny that labeled primitive leukocytes (Fig. 2a) were capable of changing into “intermediate cells,” although direct evidence was limited. Two GM cells observed in a culture a t 4 hr were interpretable as intermediate forms (Fig. 2b), exhibiting similarly to agranular GSA- positive cells in their initial reaction to organ culturing. In contrast, all three promyelocytes with large peroxidase-positive granules observed in the stroma of 3 and 4 hr cultures appeared to be holding to an eosinophilic line of development, without extension of pseudopods or other visible response to their surroundings (Fig. 2c). Similar cells were never found in older cultures. 557 DIFFERENTIATION OF LUNG MACROPHAGES TABLE 1. Changing composition of GSA I-B,+ cell types in lung cultures over time Angular cells (%) 70.5 44.0 32.3 20.8 14.1 1.8 1.2 0.0 0.0 11.1 0.0 0.0 1.1 0.0 0.0 0.0 6.2 N Age 14 + 0 hr. 14 + 1 hr. 14 + 2 hr. 14 + 3 hr. 14 + 4 hr. 14 + 6 hr. 14 + 9 hr. 14 + 12 hr. 14 + 15 hr. 14 + 18 hr. 14 + 24 hr. 14 + 2 d. 14 + 4 d. 14 + 5 d. 14 + 7 d. 14 + 9 d. 14 + 13 d. 46 25 31 24 71 53 86 40 51 9 8 5 94 18 21 3 16 loo Leukocytelike cells (%) 15.9 12.0 0.0 Intermediate cells (%) 13.6 44.0 64.5 66.7 59.2 64.2 30.2 12.5 9.8 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 2.8 0.0 1.2 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 I Macrophages (%) Early Mature 0.0 0.0 0.0 0.0 3.2 0.0 12.5 0.0 23.9 0.0 34.0 0.0 52.3 15.1 22.5 65.0 29.4 60.8 22.2 66.7 0.0 100.0 0.0 100.0 14.9 84.0 0.0 100.0 0.0 100.0 33.3 66.7 0.0 93.8 1 % of cells 0 1 2 3 4 6 9 15 12 18 24 hours in culture Fig. 4. The changing composition of GSA-positive cell types in lung cultures over time, graphic representation of data for the first 24 hr from Table 1. Half of the population of transforming cells becomes recognizable as macrophages around + 8 hr. 0, angular cells; (01, leukocyte-like cells; (w), intermediate cells; (o), macrophages. Extravascular vs. lntravascular Formation of Macrophages During the first 6 hours of organ culturing, all GSA I-B,-labeled transitional forms and nascent macrophages were localized to the extravascular connective tissue. In contrast, neither of the two intravascular GSA-labeled cells found during this period provided more telling morphological evidence of entry into transformation than a large nucleolus (Fig. 2a). Nascent macrophages were first found within vessels at 9 hr. They became more numerous at 12 hr and, together with more mature forms, common at 15 hr and thereafter, particularly after tissue breakdown had left gaps in the endothelium for cells to pass through (Fig. 6 ) . Inside the vessels, macrophages frequently stationed themselves beside stagnant blood pools and by 24 hr were engorged with fragments or entire protoplasts of red cells and expanded to many times their original cell volume. Ultrastructural Changes During Transformation and Maturation A staged and time-resolved morphological reconstruction of the transformation process was obtained by selecting an electron micrograph to depict the predominant GSA I-B,-positive cell in the cultures a t each specific interval after explantation and arranging these images in chronological order; the resulting series covers close intervals in the first 15 hr, and wider intervals from there to 7 days (Figs. 7-15). Ultrastructurally, cells in the “intermediate” category underwent subtle changes as they gradually took on typical morphological characteristics of macrophages. Throughout, one or more nucleoli were con- 558 S.P. SOROKIN ET AL. Fig. 5. Low power view of a lung culture at 14 + 12 hr. By this time the prevailing GSA-positive cell contains phagocytosed cell debris and is recognizable as a macrophage. Several are shown above the blood vessel in the middle and upper part of the field. Compare with Figure 3. x 2,950. spicuous in the nucleus (Fig. 7,8). The cytoplasm only perceptibly increased its content of granular endoplasmic reticulum and gradually developed a few clear vacuoles (Figs. 9, lo), so that after 4 hr, few nonvacuolated cells remained. The Golgi apparatus was rarely prominent in these cells. Early-stage macrophages graded away from this appearance as they increased in size and phagocytic activity (Fig. 11). The Golgi apparatus began to expand in certain young macrophages around 9 hr, forming a small but distinct region of cisternae and vesicles on one side of the nucleus. This had grown into a complex of many stacks by 15 hr, so that five or six might be included in a section passing through the cytocentrum (Fig. 12b). Frequently they were turned with their long axis extending radially outward, in an “open” position 559 DIFFERENTIATION OF LUNG MACROPHAGES TABLE 2. 3H-thymidinelabeling indices for lung explants in culture GSA-IB, Age 14 1 hr + N 14 + 4 hr N + 7 hr 14 + 13 hr N 14 + 19 hr N 14 + 24 hr 14 N N cells + 33.63% 113 33.98% 103 32.65% 98 34.03% 238 35.61% 264 34.05% 279 Stromal cells Pleural cells Epithelial 32.20% 441 26.43% 401 26.40% 428 12.56%** 422 9.88%** 476 13.58%** 508 34.54% 414 36.43% 409 26.12% 402 8.26%** 436 6.11%** 409 9.39%** 479 cells 37.67% 438 34.70% 415 29.90% 418 18.26%** 449 15.35%** 404 30.69% 613 Fetal erythroblasts 25.62% 441 20.33%* 861 17.68%** 1103 13.75%** 1084 0.84%** 830 1.31%** 917 Comparisons with GSA-IB,+ cells by X2 tests: *P<0.005 and **P<O.OOl. often seen among actively phagocytic cells (Sorokin, 1983b), the stacks being separated by radial arrays of microtubules coming from centrioles within the Golgi ring. Most of the ribosomes of transforming cells were arranged in free polysomal clusters. The granular endoplasmic reticulum of macrophages appeared only moderately extended beyond that seen in older intermediate cells (Figs. 10a, 12a), not a t all like the expansion sometimes present among mature alveolar macrophages provoked into production of lysosomes (Sorokin, 1983a). Indeed, the most singular aspect of cytoplasmic differentiation among transitional cells and macrophages was the widespread occurrence of satellites and astral fibers about the centrioles (Figs. 11, 12b, 16a), manifestly contrasting to stromal fibroblasts which instead sprouted a primary cilium from the diplosome (not illustrated). Like angular cells, but unlike cells resembling blood leukocytes, intermediate cells and all gradations of macrophages characteristically maintained extensive surface contact with adjacent stromal mesenchymal cells or fibroblasts, as if interacting at a cell membrane level (Fig. 16b), and like them they formed no specialized junctions with the fibroblasts. Macrophages seen in 24 hr cultures were generally large and replete with phagocytosed matter of considerable, though varied electron density (Fig. 13). By 48 hr these dense phagosomes had greatly receded from the cytoplasm, being supplanted in typical stromal macrophages by many small (ca. 0.2-0.5 pm) electronlucent (lipid-containing) vacuoles. These vacuoles remained as a characteristic of macrophages for the duration of a week or more that cultures were grown (Figs. 14a, 14b, 15,16b). During the same period, macrophages within lung tissue appeared to increase their complement of mitochondria, giant cells began to be seen, and many cells, giant and smaller alike, displayed well ruffled membranes (Figs. 14, 15). These ultrastructural changes were the most consistently found among maturing cells and remained characteristic of macrophages as many began a centrifugal migration that resulted in a buildup of these cells at the pleural surface and gradual clearing of the lungs (Fig. 15). Mitotic Activity and Labeling Indices of GSA-Positive Cells Mitotic figures were observed among GSA I-B,-labeled cells in both intermediate and macrophage categories throughout the period the lungs were cultured. They were first documented in electron micrographs of intermediate cells at 4 hr, and in intermediate cells and nascent macrophages a t 6 hr (Fig. 10b,c). They were also seen at 9 hr (intermediate cells), 12, 15, and 18 hr (maturing macrophages), and 24 hr (maturing and well-developed macrophages, Fig. 13).Thereafter, all dividing GSA-positive cells observed were mature macrophages; they were especially numerous in 4 day cultures. In our electron microscopic survey, labeled mitotic cells were found only in the extravascular stroma, although dividing macrophages were sometimes observed by light microscopy in vessels of 2-day and older cultures. 3H-thymidine labeling indices for macrophages and precursors, stromal cells, pleural cells, airway epithelium, and fetal erythroblasts in lung cultures are given in Table 2. At explantation ( + 1 hr), grain counts indicated a high labeling index for cells in all categories, as expected of rapidly growing embryonic tissues. Lung stromal and pleural cells and those in the macrophage line (GSA I-B,-positive) labeled similarly with a third of the cells in S-phase. This figure slightly led erythroblasts in the blood vessels (26% labeled) but trailed epithelial cells of the terminal buds (38%labeled). During the next 18 hr, however, the labeling index of cells in all categories except the GSA-positive pool declined, followed by some indication of recovery in the airway epithelium at 24 hr. The fall was steepest among the erythroblasts but was sufficiently similar in each of the three remaining categories as to make them appear a single population, perhaps reflecting equal susceptibility of lung cells to an “explantation shock” that requires some hours to overcome. In marked contrast, GSA-positive cells maintained their initial labeling index throughout the entire 24 hr period, exhibiting no explantation shock and objectively behaving as a population distinct from other cells in the lungs, including their immediate neighbors in the connective tissue. Figs. 6-7. DIFFERENTIATION O F LUNG MACROPHAGES 561 Fig. 8.14 + 2 hr. Intermediate cells are now the predominant GSA-positive cells in the lung stroma: resembling those present a n hour earlier, they as yet contain few vacuoles. x 11,180. Immature Cells in Older Cultures A small number of incompletely transformed cells may be present in lung organ cultures long past the time when virtually all other GSA-positive cells have become transformed into macrophages (Fig. 17). Among the younger cultures to contain nearly 100% macrophages (12-24 hr), the immature forms more frequently were classifiable as intermediate cells, similar Fig. 6.Intravascular vs. extravascular origin of macrophages: 14 + 15 hr. Macrophages (*) present in the vascular lumen (above and left) as well as in the pulmonary stroma (center). Intravascular cells may have migrated into the blood vessel through a rent in the endothelial lining (arrow). x 3,320. Fig. 7.14 + 1 hr. Compared to an angular cell, this intermediate cell appears to have shorter processes and more cytoplasm about the nucleus. Two nucleoli are seen; the cytoplasm remains rather undifferentiated, although a few tiny granules can be seen near the centriole below the nuclear notch. x 9,750. to intermediate cells seen earlier except for an apparently delayed response to explantation (Fig. 17b). In contrast, among older cultures of 4 days (Fig. 17a) or as many as 13 days (Fig. 17~1,immature GSA-positive cells more nearly resembled angular cells and appeared further separated from mature macrophages because transitional forms between them were not seen. DISCUSSION This paper presents a time-resolved, ultrastructural reconstruction of events taking place within a distinct, GSA I-B,-labeled population of cells in organ cultures of 14-day prenatal rat lungs as it undergoes a shift in composition from precursors to macrophages. There is no doubt that macrophages are produced in large number by living cultures over a comparable interval because the process can be continuously followed under the light microscope. In the present instance, sequential passage of cells from one stage t o another has been inferred from study of fixed specimens which are required for electron microscopy. We are confident of its 562 S.P. SOROKIN ET AL. Fig. 9. 14 + 3 hr. The intermediate cell has begun to exhibit a few profiles of granular endoplasmic reticulum and to contain a few, still mostly clear vacuoles. x 11,760. validity for a number of reasons: 1)The cultures were collected at close, in initial phases hourly, intervals. 2) The GSA-positive population presented a changing appearance with increasing time from explantation but was the same for all specimens fixed at a given interval. 3) In contrast to other cell types, this population showed no pyknosis or other visible evidence of cellular degeneration. 4) On the contrary, it maintained a uniformly high (34%)3H-thymidine labeling index during the transformation period. 5) Many fine gradations were seen between the main morphological classifications. Replication vs. Transformation of Macrophage Precursors Our study indicates that in the organ cultures, macrophages initially arise in extravascular connective tissue rather than the intravascular spaces. For the most part they appear to develop from angular cell precursors situated among stromal mesenchymal cells or fibroblasts which to a degree they resemble morphologically; angular cells differ markedly from their neighbors in binding GSA I-B,. Seemingly, macro- phages can also arise from undifferentiated GSA-positive leukocytes present in prenatal rat lungs, particularly if the cells first migrate into the extravascular connective tissue. In 14-day explants, however, such leukocytes alone could not account for more than a small part of the macrophages existing in 12 hr cultures because too few were present to have made up the macrophage number by direct transformation, and little time was available for these candidate precursors to increase through mitotic activity. From what we have learned from study of GSA-positive cells in intact rat embryos (Sorokin et al., 1992a), this distinction may be of small account, since GSA-labeled angular cells and GSA-labeled agranular leukocytes equally represent the primitive/fetal macrophage line, the angular form probably reflecting greater length of time spent residing in the connective tissues of the lungs or other organs, compared to the more rounded, leukocyte-like forms. As they occurred in 14-day prenatal rat lungs, angular and other GSA-positive cells constituted an actively growing population with a high division rate. The 3Hthymidine labeling index (34%) remained unchanged DIFFERENTIATION O F LUNG MACROPHAGES Fig. 10. 14 + 6 hr. a: An intermediate cell has continued trends first noted at + 3 hr: further expansion of the granular reticulum and increased vacuolization. GSA-positive staining of a vacuolar profile along the lower surface (arrow) may represent infolding of cell mem- 563 brane, hence early formation of a phagosome. x 9,800. b Mitosis in a n intermediate cell. x 8,375. c: Mitosis in a n early macrophage. x 5,800. 564 S.P. SOROKIN ET AL. Fig. 11. 14 + 9 hr. Typical cells at this time are nascent (early) macrophages like this one with numerous phagocytic inclusions but as yet little differentiation of other cytoplasmic membrane systems, and centrioles with prominent satellites. x 9,180. for 24 hr of organ culturing while the population transformed into one made up entirely of macrophages. From this it appears on the one hand that macrophage precursors can replicate rapidly, as in keeping pace with lung growth, without being stimulated to transform into macrophages. On the other hand, transformation need not be accompanied by slowing in the rate of DNA synthesis by affected cells, at least for the first 24 hr. Significance of Macrophage Precursors in Embryonic Lungs A number of our findings indicate that the main purpose of the early appearance of macrophage-like cells in the lungs and several other organs is to establish a resident cell population of potential macrophages in those places, rather than primarily to meet immediate local challenges. Such a process differs fundamentally from the dispatching of leukocytes to an organ in response to an infectious or immunological cell. If for argument’s sake we put aside the question of de novo formation of angular cells in the lungs (Sorokin et al., 1992a) and assume that the original consignment of macrophage precursors is entirely by seeding from the blood stream, then the events that follow seeding are still markedly different from what is generally understood to follow the arrival of monocytes in older lungs: In the first case the cells extend processes out among the mesenchymal fibroblasts and blend into the tissue as angular cells, whereas in the second monocytes begin to transform into phagocytic macrophages as soon as they cross out of the blood vessels, much as activated precursors behave in lung cultures, or chicken monocytes in tissue culture (Sutton and Weiss, 1966). The angular cells neither increase greatly in volume, round up, engage in phagocytosis, or increase production of lysosomes but remain phagocytically quiescent, evidently responding to quite different signals from those initiating transformation. Given the high rate of thymidine incorporation and the close matching of labeling indices between GSA-positive and other constituent cells of intact lungs, what more appropriate behavior could be expected from cells establishing a permanent residency? Transformation and Maturation of Macrophages In developing lungs resident precursors normally transform into macrophages irregularly and a few at a time. Extensive trauma like that brought about by dissection and organ culturing brings about transfor- DIFFERENTIATION OF LUNG MACROPHAGES Fig. 12. 14 + 15 hr. a: Typical macrophages are now more matureappearing than at + 9 hr. They are larger and often exhibit greater expansion of the granular reticulum (left side and top of cell) and prominence of mitochondria, although development of these organelles remains moderate in comparison to the full potential of alveolar macrophages. x 5,620. b: The Golgi apparatus generally is 565 larger than in nascent macrophages and may have several stacks of lamellae. These are sometimes arrayed with their long axes raidal to the diplosome in a stellate or “open” pattern characteristic of phagocytizing cells. The centrioles characteristically are surrounded by many satellites. x 17,600. 566 S.P. SOROKIN ET AL. Fig. 13.14 + 24 hr. Dividing macrophage in the lung stroma. This cell is typical of those in the cultures a t the conclusion of the transformation process, large, filled with phagocytosed but not yet well digested material. x 6,890. mation of a great many cells and demonstrates the developmental potential inherent in the normally inconspicuous resident population. As our material tends to show, initial actions of precursor cells during the transformation period are related to becoming mobile and actively engaged in phagocytosis, a t once mirrored in the cytoplasm by increased prominence of microtubular arrays about the centrioles. Earlier work has shown by rosetting assay that few cells in 14-day lung explants have sufficient Fc receptors to bind IgG-opsonized erythrocytes effectively, but the proportion of cells binding them increases sharply during the first 24 hr in culture (Sorokin et al., 1989).Consequently, macrophage precursors become more efficient phagocytes during transformation. Notable expansion of the granular endoplasmic reticulum is held off for a time, however, and so is large-scale synthesis of lysosomal enzymes such as acid phosphatase, which is not histochemically conspicuous in macrophages before 24 hr and attains levels of activity comparable to activated macrophages of postnatal lungs only a day or two later (Sorokin and Hoyt, 1987). From these observations one can perceive an ordering of priorities that produces efficient, motile phagocytes before they are fully capable of digesting materials ingested; hence typical transformed macrophages in the lung cultures at 24 hr are enlarged and their phagosomes filled with recognizable cellular debris, and digestively competent macro- DIFFERENTIATION OF LUNG MACROPHAGES Fig. 14. 14 + 4 days. a: Macrophage typical of those occurring well within the maturation period. The plasmalemma has developed elaborate lamellipodia (ruffled membranes), shown here at the advancing edge of the cell (right), and the vacuoles are homogeneous and filled with nearly electron-lucent contents. Ingested cellular debris has been hydrolyzed. It is uncertain whether one or two nuclei are 567 present. x 10,100. b Binucleate macrophage in the lung stroma. The cytoplasm appears similar to that in the macrophage above, but the cell diameter is greater. A detail of the cytoplasm is shown to the right. X 3,280. c: Cytocentrum of the cell a t left. More than two centrioles are present, making it likely that the cell has more than one nucleus-an early example of giant cell formation. x 32,500.. 568 S.P. SOROKIN ET AL. Fig. 15. 14 + 7 days. Macrophages on the outer surface of a lung culture. Pleural cells form an arc sweeping down from the middle of the left side of the figure and are themselves GSA-negative, but they are enmeshed in lectin-positive processes from the macrophages. These possess electron lucent vacuoles and are typical of mature cells produced by the organ culture system. X 4,175. 569 DIFFERENTIATION O F LUNG MACROPHAGES Fig. 16. a: Microtubules radiating from satellites about a centriole in a pulmonary macrophage a t 14 + 24 hr. This configuration is typical of intermediate cells and young macrophages and unlike that seen in stromal fibroblasts during the same period. x 25,780. b: Mac- + rophage within lung tissue a t 14 4 days. Cell processes are extensively spread over the surfaces of two neighboring fibroblasts (arrows). x 7,700. 570 S.P. SOROKIN ET AL. Fig. 17. a: Elongate GSA-positive cell (*) beneath the pleural lining of a lung culture at 14 + 4 days. The cell has accommodated itself to the layering of the tissue, does not resemble typical macrophages present at this time, and may represent conversion to a dendritic, or monitor cell. x 6,840. b: Late occurrence of a transforming GSA-positive cell in a culture aged 14 +15 hr. Located just beneath the endothelium, it resembles a blast form more than most intermediate cells, and possibly reflects delayed activation of an originally intravascular leukocyte (Fig. 2a). x 13,060. c: Undifferentiated GSA-positive cell present in pulmonary stroma at 14 + 13 days. It is at the borderline between an angular and intermediate cell. x 8,530. DIFFERENTIATION OF LUNG MACROPHAGES phages seen a day or two later have clear vacuoles. In earlier work we had also found that these vacuoles stain with Oil Red 0 and therefore contain lipids. Such alteration in vacuolar content should not be taken to imply disorder in macrophage function, as it is known that mouse peritoneal macrophages subjected to lipidloading regimens in culture exhibit no loss in Fc receptor-mediated endocytosis and continue to produce a variety of products, although the balance secreted may be subtly altered (Montgomery and Cohn, 1989). “Quiescence”in Older Cultures The normal course in the organ culture system used here is for matured macrophages with clear vacuoles and often well-developed lamellipodia to migrate within lung tissue a few days but eventually to be drawn out onto the pleural surface. There they lose contact with stromal cells and remain largely unchanged for a week or two with the population increasing from outward migration of other macrophages and not by local mitotic activity. A few giant cells may appear among them, and these have many mitochondria and an outwardly more metabolically active appearance than the remaining mononuclear cells; but as a population the cells become quiescent, scavenging activity gradually ceases, and the population eventually dies, as if from lack of stimulation. Relatively few macrophages are left inside the lung culture, and among those remaining, most find their way into fluid filled airways where they become caught up in the currents set up by action of ciliated cells of the bronchi and bronchioles and never return to the tissues. A very few GSA-positive cells were found not conforming to the life pattern summarized above, retaining a habit nearer angular cells than macrophages long after most other GSA-positive cells had matured. It is as yet unknown whether these simply required a higher threshold of stimulation to transform or, having once transformed, subsequently converted into dendritic cells. With respect to the vast majority of GSA-positive precursors that become macrophages, however, it is of interest to know whether their fate can be altered experimentally by administration of hematological colony stimulating factors or other cytokines. Macrophages on the pleural surface of the culture are present in sufficient number for certain types of quantitative analyses, and being easily exposed to these agents, make good experimental subjects. In particular, they may provide information useful for resolving uncertainty about the longevity of early pulmonary phagocytes and the possibility for continuance of this macrophage line into postnatal life. These topics are further 571 explored in the next paper of the series (Sorokin et al., 199213). ACKNOWLEDGMENTS This work was supported by USPH research grant HL-33070. LITERATURE CITED Bainton, D.F., and M.G. Farquhar 1970 Segregation and packaging of granule enzymes in eosinophilic leucocytes. J . Cell Biol., 45:5473. Honda, T., B.A. Schulte, and S.S. Spicer 1989 Glycoconjugate with terminal galactose. A selective property of macrophages in developing rat lung. Histochemistry, 91t61-67. Kreipe, H., H.J. Radzun, M.R. Parwaresch, A. Haislip, and M.L. Hansmann 1987 Ki-M7 monoclonal antibody specific for myelomonocytic cell lineage and macrophages in human. J . Histochem. Cytochem., 35~1117-1126. Lewis, M.R. 1925 The formation of macrophages, epitheloid cells and giant cells from leucocytes in incubated blood. Am. J . Pathol., lt91-100. Montgomery, R.R., and Z.A. Cohn 1989 Endocytic and secretory repertoire of the lipid-loaded macrophage. J . Leuk. Biol., 45t129138. Sorokin, S.P. 1983a Dynamics of lysosomal elements in pulmonary alveolar macrophages. I. The postactivation lysosomal cycle. Anat. Rec., 206t117-143. Sorokin, S.P. 1983b Dynamics of lysosomal elements in pulmonary alveolar macrophages. 11. Transitory intracytoplasmic events in activated cells. Anat. Rec., 206:145-170. Sorokin, S.P., R.F. Hoyt, Jr., and M.M. Grant 1984 Development of macrophages in the lungs of fetal rabbits, rats, and hamsters. Anat. Rec., 208:103-121. Sorokin, S.P., and R.F. Hoyt, J r . 1987 Pure population of nonmonocyte derived macrophages arising in organ- cultures of embryonic rat lungs. Anat. Rec., 21 7t35-52. Sorokin, S.P., L. Kobzik, R.F. Hovt. Jr.. and J.J. Godleski 1989 Development of surface membrane characteristics of “premedullary” macrophages in organ cultures of embryonic rat and hamster lungs. J . Histochem. Cytochem., 37t365-376. Sorokin, S.P., and R.F. Hoyt, J r . 1992 Macrophage development. I. Rationale for using G r i o n i a simplicifolia isolectin B, as a marker for the line. Anat. Rec. 232520-526, Sorokin, S.P., R.F. Hoyt, Jr., D.G. Blunt, and N.A. McNelly 1992a Macrophage development. 11. Early ontogeny of macrophage populations in brain, liver, and lungs of rat embryos as revealed by a lectin marker. Anat. Rec. 232527-550. Sorokin, S.P., N.A. McNelly, and R.F. Hoyt, Jr. 199213 Macrophage development. IV. Effects of blood factors on macrophages from prenatal lung cultures. Anat. Rec. (in press). Sutton, J.S., and L. Weiss 1966 Transformation of monocytes in tissue culture into macrophages, epithelioid cells, and multinucleated giant cells. An electron microscope study. J . Cell Biol., 28t303332. Takahashi, K., F. Yamamura, and M. Naito 1989 Differentiation, maturation, and proliferation of macrophages in the mouse yolk sac. J. Leuk. Biol., 45:87-96. Takahashi, K., M. Naito, and M. Takeya 1990 Development, differentiation, and maturation of primitivelfetal macrophage population in rat ontogeny. J . Leuk. Biol., Suppl. 1.57 (abstr.) van Rooijen, N., N. Kors, and G. Kraal 1989 Macrophage subset repopulation in the spleen. Differential kinetics after liposome-mediated elimination. J . Leuk. Biol., 45t97-104.