Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Dendritic cells (DC) are bone marrow-derived cells (1, 2) that migrate via the bloodstream to peripheral tissues (3), where they adopt a sentinel function (4). After encountering antigen (5) or reacting with locally released inflammatory cytokines (8), DC resume their migration via lymphatics to local lymph nodes or via the bloodstream to the spleen (4), where they present processed antigenic peptides to local T cells. Unlike their mature counterpart, DC newly released from the bone marrow are monocyte-like in appearance, do not express major histocompatibility class II antigen (MHC II) or express it only weakly, and are capable of phagocytosis both in vitro (11) and in vivo (12). During their transit from peripheral nonlymphoid to lymphoid tissues, DC undergo maturation, their phagocytic capability is lost, and their accessory cell function is greatly augmented (12). That DC recently released from the bone marrow (3) and those emigrating from peripheral tissues (12) both may circulate via the bloodstream accounts, in part, for the phenotypically diverse, albeit trace, population of DC in the blood (13, 14).

The enormous capillary bed and vast alveolar surface area, though uniquely adapted to efficient gas exchange, make the lung a particularly vulnerable mucosal interface between the environment and internal tissue compartments (15). Guarding the airway mucosal surface is a population of rapidly turning-over sentinel DC (16), which, like their counterparts in the skin, are capable of limited phagocytosis (17). A second population of DC resides in the connective tissue cuffs around airways and vessels and in alveolar walls of the lung. These, in contrast to DC harvested from the airway epithelium, generally have a more mature phenotype and do not phagocytose particulates but readily activate naive T cells in a mixed leukocyte reaction (17). The turnover rate of this interstitial population of lung DC and the basis for their somewhat greater degree of differentiation have not been determined.

In response to intravascular and extravascular stimuli, both circulating and intravenously injected leukocytes are preferentially sequestered in the pulmonary capillary bed (18). By contrast, when splenic DC or cultured MHC II⁺, bone marrow-derived DC are injected intravenously into mice, there is a relatively low retention of these cells in the lung (19, 20). These studies, however, did not examine the possibility that a subpopulation of immature, MHC II DC precursors might be retained in the pulmonary capillary bed. Interestingly, treatment of mice with interferon- (IFN-) did not augment the sequestration of mature DC in the lung vasculature (20). Instead, MHC II expression by cells (including MHC II DC precursors in the interstitium) was increased, suggesting that a circulating, MHC II DC precursor population must colonize peripheral tissues. In the present study, experiments were devised specifically to examine the DC population harvested from the pulmonary vasculature and to determine whether MHC II DC precursors are preferentially sequestered in the pulmonary vascular compartment. To accomplish this, leukocytes from the pulmonary vascular bed were obtained by vascular perfusion. Blood draining into the inferior vena cava was harvested from the same animal as a source of leukocytes in pre-lung blood. The number, phenotype, proliferative activity, and accessory cell function of MHC II⁺ DC in the lung vasculature was compared with those harvested from the vena cava.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Animals

Pathogen-free, 6- to 8-wk-old female Lewis and Long Evans rats (160-180 g) were obtained from Charles River Breeding Laboratories (Kingston, NY). Animals, housed in restricted access research animal care facilities at Massachusetts General Hospital, were permitted free access to food and water and underwent monthly monitoring for viral infections.

Reagents and Antibodies

Recombinant mouse granulocyte macrophage colony-stimulating factor (GM-CSF) was a gift of Genetics Institute (Cambridge, MA). Antibodies and their specificities included: OX6 (MHC class II), OX8 (cytotoxic/suppressor T cells), OX19 (pan T cells), OX33 (B cells), OX52 (pan T cells), OX62 ( subunit of integrin on DC and T cells) (21), ED1 (macrophages, monocytes, and DC), ED2 (macrophages) (22) (Sera Lab Ltd., Crawley Down, Sussex, UK), 1A29 (intercellular adhesion molecule-1 [ICAM-1], CD54) (23), WT1 (lymphocyte function-associated antigen [LFA]-1, CD11a), and WT3 (2 integrin, CD18) (24). Other reagents included neuraminidase, RPMI-1640 medium, fetal bovine serum, Hanks' balanced salt solution, sodium nitrite, diaminobenzidine, penicillin/streptomycin, Na₂-ethylenediamenetetraacetic acid (EDTA) and Gill's hematoxylin (Sigma Chemical Co., St. Louis, MO), 3-amino-9-ethylcarbazole (Aldrich Chemical Co., Milwaukee, WI), and sheep red blood cells (Bio-Whittaker, Walkersville, MD).

Procedure for Obtaining Lung Vascular and Vena Cava Blood Leukocytes

Rats were anesthetized by intraperitoneal injection of chloral hydrate (0.04 g/100 g body weight). Heparin, 0.5 ml (1,000 U/ml), was injected intravenously. The abdominal cavity and thorax were opened and the inferior vena cava and aorta tied off above the diaphragm. To expose the cells to identical treatment, blood draining from the subdiaphragmatic organs was harvested through the vena cava and placed in four times the volume of the saline solution used to perfuse the lungs. The thymus was removed and the superior vena cava tied off. A 15-gauge needle, attached via tubing to a 12-ml syringe, was inserted into the left ventricle. After inserting a 19-gauge butterfly needle into the pulmonary artery, the lung vasculature was slowly perfused with 1.5 mM Na₂-EDTA and 0.075% sodium nitrite in phosphate-buffered saline (PBS), pH 7.3 (25). In some experiments sodium nitrite was omitted; this had no effect on the final result. The effluent was quantitatively retrieved by gentle aspiration from the left ventricle.

Retrieval of MHC II⁺ DC from the Lung Vasculature as a Function of Perfusate Volume

To determine the efficiency of MHC II⁺ DC retrieval from the lung vascular bed, lungs were perfused with graded volumes of perfusate (60-190 ml) as described previously. At the end of the perfusion the lungs were lavaged once with PBS and instilled with 1-2 ml optimal cutting temperature (OCT) compound (Miles, Inc., Elkhart, IN). Blocks of lung were embedded in OCT compound and frozen in liquid nitrogen in preparation for sectioning and immunostaining. Leukocytes in the perfusate and vena cava blood were processed in parallel as described below.

Cell Preparation for Culture and Immunophenotyping

The harvested blood cells were washed twice in PBS. Erythrocytes and polymorphonuclear leukocytes were removed by density gradient centrifugation on Ficoll. The leukocytes at the interface were harvested, washed in PBS, and counted. Cells, 5 × 10⁴/slide, were cytocentrifuged (Day 0), and the remaining cells were cultured at 1 × 10⁶ cells/ml in RPMI 1640, 10% fetal calf serum (FCS), and 500 U/ml GM-CSF. At daily intervals, up to 3 d, cells were harvested, cytocentrifuged, and immunostained as described below.

Isolation of Blood DC for Mixed Leukocyte Response

Blood from the vena cava and the lung vasculature was harvested and separated by Ficoll gradient centrifugation as described previously. Dendritic cells were enriched by negative, immunomagnetic cell separation using a Vario MACS device (Miltenyi Biotec, Auburn, CA). Briefly, the leukocyte fraction from each population was incubated for 30 min at 4°C in the following cocktail of mouse antirat monoclonal antibodies (mAbs), all diluted at 1:150 with 5 mM EDTA, 0.5% bovine serum albumin in PBS: OX19 (anti-T cell), OX52 (anti-pan T cell), OX8 (anticytotoxic/suppressor T cell), and OX33 (anti-B cell). This was followed by a 15-min incubation in secondary antibody conjugated to magnetic beads and finally separated by magnetic separation. The harvested cells from both sources, consisting of monocytes and 6 to 7% MHC II⁺ DC, were used immediately in a mixed leukocyte response (MLR) assay either with or without 500 U/ml GM-CSF. Alternatively, the cells were enriched further for MHC II⁺ DC by placing them in culture with 500 U/ml of GM-CSF for 1, 2, or 3 d. The nonadherent cells were then harvested. Their purity was assayed by immunostaining cytocentrifuged preparations with OX6 mAb and found by Day 3 to be 86 and 89% for the lung blood and vena cava blood, respectively. The accessory cell activity of DC, isolated from the two sources of blood after zero, 1, 2, or 3 d in culture with GM-CSF, was then compared in an MLR assay. Irradiated (1,000 rad) DC (1 × 10⁴ cells/well) were co-cultured with 2 × 10⁵ splenic T cells from Long Evans rats. The splenic T cells were purified by filtration through nylon wool and residual Ia⁺ cells were removed by OX6 immunopanning. After 7 d, the cells were pulsed with 1 µCi/well (5 µCi/ml) of tritiated thymidine ([³H]TdR) for 6 h at 37°C; they were harvested in a cell harvester (Skatron, Sterling, VA) and counted in a Tri-Carb liquid -scintillation spectrometer (Packard Instrument Co., Downers Grove, IL).

Immunoperoxidase

For immunophenotyping, cytocentrifuged leukocytes were stained by the avidin-biotin immunoperoxidase technique (26) for MHC II antigen (OX6),  subunit of integrin (OX62), monocyte/macrophage marker (ED1), macrophage marker (ED2), LFA-1 (WT1, CD11a), 2 subunit of integrin (WT3, CD18), ICAM-1 (1A29, CD54), or CD4 (W3/25), as previously described (27). Frozen sections of lung, 4 µm thick, were stained for MHC II antigen (OX6). Briefly, after blocking with normal horse serum, sections/ cells were incubated with optimal dilutions of mAbs for 60 min. Incubation with irrelevant mAbs of similar isotype or with PBS served as controls. Endogenous peroxidase was inhibited with 0.3% hydrogen peroxide in PBS for 30 min. This was followed by incubations with biotinylated horse antimouse IgG (diluted 1:100) for 30 min and with avidin-biotinylated peroxidase complex (diluted 1:50) from an ABC Elite kit (Vector Laboratories, Burlingame, CA) for 30 min. Each incubation was followed by a PBS rinse. Reaction product was generated by incubation with 0.03% H₂O₂, 0.03% 3-amino-9-ethylcarbazole, and 5% n-n-dimethylformamide in 0.1 M acetate buffer, pH 5.0. The sections were counterstained with Gill's hematoxylin.

Immunoelectron Microscopy

MHC II⁺ DC in alveolar walls were examined in nonperfused lungs from control Lewis rats that were gently instilled with Nakane's fixative (28) and submerged in the same fixative for 1 h at 4°C. After rinsing in 0.15 M cacodylate buffer, pH 7.3, slices of lung were embedded in 7% agar (Difco, Detroit, MI) and 100-µm-thick sections were cut on a vibratome (Lancer-Brunswick, St. Louis, MO). The immunocytochemical reaction was similar to that described previously, except that 0.05% diaminobenzidine in 0.05 M Tris buffer (pH 8.0) was used to generate reaction product. The sections were then processed for electron microscopy as previously described (29).

Morphometry of MHC II⁺ DC in Alveolar Walls of Perfused Lungs

MHC II⁺ cells in the alveolar walls of perfused and nonperfused control lungs were enumerated by counting the cells in 10 microscopic fields using a ×40 objective and a 1-cm² graticule containing 10 by 10 squares. Cell counts were corrected for the fraction of alveolar space included in the area counted. This was accomplished by estimating the fraction of each square devoid of tissue. The mean of this fraction was calculated and subtracted from 100. The mean number of positively stained cells was divided by the calculated tissue fraction and multiplied by 100. To obtain the number of cells per square centimeter of lung tissue, the corrected mean cell number was divided by 0.000625 (1 cm/40)². The Student's t test (Sigma Plot; SPSS Inc., Chicago, IL) was used for statistical analysis.

Cell Preparation for Autoradiography

The above procedure was slightly modified for pulse-labeling cells with [³H]TdR. Before placing the cells in culture, lymphocytes were partially depleted by incubation with neuraminidase-treated sheep red blood cells (30). The non-rosetted leukocytes were retrieved by density gradient centrifugation on Ficoll and cultured, as previously, with 500 U/ ml GM-CSF. Cells were pulse-labeled with 0.2 µCi/ml [³H]- TdR for 6 h. They were then washed three times in PBS and, after preparing a cytocentrifuged sample (Day 0), the remaining cells were placed in culture at 1 × 10⁶ cells/ml in RPMI 1640, 10% FCS, and 500 U/ml GM-CSF. Over the ensuing 3 d, cytocentrifuged preparations were made daily. All slides were processed for autoradiography at the end of the experiment. Cells were immunostained for MHC II antigen, except that diaminobenzidine was used as substrate (to avoid chemography) (31) and the color of the reaction product was intensified with 2% CuSO₄ · 5H₂O. The slides were dipped in NTB-2 Kodak nuclear track emulsion (Eastman Kodak Co., Rochester, NY), exposed for 1 wk at 4°C and developed in Kodak Dektol developer. The cells were counterstained with Gill's hematoxylin No. 2.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

The purpose of the present study was to characterize the DC population harvested from the vascular compartment of the lung. In the absence of specific immunologic markers for these cells in the rat, expression of MHC II antigen and several characteristic morphologic features served to identify mature DC. A similar lack of specific immunologic reagents also precluded the isolation and direct identification of circulating DC precursors that might be sequestered in the lung. Therefore, to retain putative DC precursors, the harvested leukocytes were partially enriched by removing neutrophils during the initial Ficoll separation used to eliminate erythrocytes. In preparing DC for immunophenotyping, the DC and their precursors were co-cultured with the T cells and monocytes present in the isolate. This strategy was selected to take advantage of T cell cytokines (e.g., interleukin-4) and monocyte factors which, in addition to exogenous GM-CSF, are known to promote DC maturation (32).

Localization of DC in Alveolar Walls

In the normal rat lung, the number of MHC II⁺ DC in alveolar walls is small. It was difficult to determine by light microscopy their precise anatomic location and to distinguish them from MHC II expressing alveolar type II cells (29) (Figure 1a). When examined by immunoelectron microscopy, however, MHC II⁺ cells were observed both in capillary lumens (Figure 2a) and in the adjacent connective tissue where the DC cell processes were often in close proximity to alveolar type I cells (Figure 2b). Their relatively abundant cytoplasm and paucity of lysosomes suggest that these cells are mature DC.

View larger version (128K): [in this window] [in a new window]   View larger version (134K): [in this window] [in a new window]   Figure 1.   (a) Frozen section of nonperfused Lewis rat lung immunostained for MHC II antigen with OX6 mAb. Several OX6+ cells are present in alveolar walls. Their precise location in the alveolar wall cannot be resolved at this magnification. Some of the positively stained cells may be alveolar type II cells expressing MHC II antigen (29). (b) Frozen section of Lewis rat lung perfused with 190 ml of saline and stained with OX6 mAb. Although there is a significant reduction in the number of OX6+ cells, a few still remain. Magnification: ×615.

View larger version (171K): [in this window] [in a new window]   View larger version (173K): [in this window] [in a new window]   Figure 2.   (a) Electron micrograph of an alveolar capillary from a nonperfused control Lewis rat. Sections of lungs were stained for MHC II antigen using OX6 mAb. A cell with abundant cytoplasm, virtually no lysosomes, and abundant OX6+ reaction product on its surface suggests that this is a circulating MHC II+ DC. The alveolar space is indicated with an asterisk. (b) An MHC II+ DC in the interstitium adjacent to an alveolar capillary (CL). Cell processes of the DC extend to close proximity of an alveolar type I cell lining an alveolar space (asterisk). Magnification (a and b): ×8,100.

Efficiency of DC Retrieval from the Lung Vascular Bed

In the majority of experiments, intravascular DC were harvested by perfusing with 60 or 120 ml of perfusate, a volume similar to that used by others to obtain leukocytes from the vasculature of the rat lung (33). Examination of immunostained sections of the perfused lung revealed, however, that a number of MHC II⁺ cells remained in alveolar walls (Figure 1b). To determine whether this was the result of an inefficient retrieval of these cells from the pulmonary vascular bed or, alternatively, whether a substantial fraction of MHC II⁺ DC is in fact located outside the capillary bed, lungs were perfused with graded volumes of perfusate (60-190 ml). Large volumes, up to 190 ml, were required to reduce the number of MHC II⁺ DC in alveolar walls by approximately 70% (Figure 3). These results indicate that DC, like other leukocytes (34), require prolonged perfusion to be quantitatively retrieved from the capillary bed of the lung. Furthermore, because not all MHC II⁺ cells were removed, even after prolonged perfusion, it indicates that a fraction of these cells resides in an extravascular location within the alveolar wall.

View larger version (15K): [in this window] [in a new window]   Figure 3.   To determine the efficiency of DC retrieval from the lung vasculature, Lewis rat lungs were perfused with either 0, 70, 140, or 190 ml of saline as described in MATERIALS AND METHODS. On completion of the perfusion, frozen sections of the five lobes of the lung were cut and stained for MHC II antigen using OX6 mAb. OX6+ cells were counted in sections from five different areas of the lung and the number of positive cells/cm2 of lung tissue calculated. Large volumes of perfusate were required to remove approximately 70% of OX6+ cells from alveolar walls. The data represent the means ± 1 SD.

Because of their large size, > 15 µm in diameter, it was assumed that mature MHC II⁺ DC would be firmly wedged in capillary lumens and therefore would be more difficult to dislodge by vascular perfusion than other leukocytes. However, when the number of MHC II⁺ DC was enumerated in the harvested perfusate, their fraction relative to the other leukocytes did not change in relation to the volume of perfusate used (Table 1). This suggests that mature intravascular DC are highly malleable cells that not only can be washed out of the lung vasculature, but also are apparently capable of traversing lung capillaries with diameters of only 7 µm.

Blood from the Lung Vasculature Is Enriched for DC Precursors

In view of the known retention of leukocytes in the lung vasculature (18), it was anticipated that the number of DC obtained from this anatomic compartment would be larger than from the blood entering the lung. In fact, the fraction of mature MHC II⁺ DC retrieved from both the vena cava and the lung vasculature was similar and was < 1% of total circulating leukocytes (Figure 4). However, when equal numbers of neutrophil-depleted leukocytes from both sources were cultured with GM-CSF and equal numbers of leukocytes were harvested from each well, there was a statistically significant increase in the number of mature MHC II⁺ DC in the lung blood sample by the second day. This number was increased further by the third day (Figure 4). In fact, by the third day, the number of MHC II⁺ DC was 76% higher in the population from the lung vasculature than that from the vena cava.

View larger version (30K): [in this window] [in a new window]   Figure 4.   Neutrophil-depleted leukocytes were prepared from the lung perfusate (lung blood) and vena cava blood (venous blood) and cultured at 1 × 106 cells/ml with 500 U/ml of GM-CSF. At each time point (0-3 d), equal numbers of cells of a single well from each category were harvested (black bars), counted, cytocentrifuged, and immunostained for MHC II antigen. On Day 0 there was no difference in the number of MHC II+ cells in the samples from either source. Beginning on Day 1, however, the number of OX6+ cells in the lung blood (gray bars) increased relative to those obtained from the vena cava (white bars). This difference became statistically significant by Day 2. Representative data from one of four separate experiments. The asterisk indicates a P value of < 0.005.

Mature MHC II⁺ DC Increase in Number by Differentiation and Not by Cell Division

To determine whether, following culture with GM-CSF, the observed increase in the number of mature DC was the result of cell division or differentiation, freshly harvested, neutrophil-depleted leukocytes from the lung vasculature and the vena cava were pulse-labeled for 6 h with [³H]TdR. The cells were then cultured with GM-CSF, harvested at daily intervals, and prepared for autoradiography. Fewer than 1% of MHC II⁺ DC, which themselves constituted < 4% of the leukocytes placed in culture, contained radioactive grains 24 h after pulse-labeling. This fraction declined further on the subsequent 2 d, during which time the number of MHC II⁺ DC grew in number (Table 2). These results indicate that the increase in the number of mature DC observed by the third day of culture with GM-CSF was the result of differentiation and not cell division.

Phenotype of DC in Venous and Lung Blood

In cytocentrifuged preparations, DC were identified as large cells, often with elongated cytoplasmic processes and with eccentrically placed ovoid or lobulated nuclei (Figures 5a-5d). Following culture with GM-CSF, many of the DC formed rosettes with T cells (Figures 5b-5d), a feature also used in their identification. On the day of isolation, DC from both sources constituted a trace population of the harvested leukocytes (< 3% of the neutrophil-depleted fraction); of these, approximately 39 and 58% of venous and lung blood DC, respectively, expressed MHC II (Table 3). The remaining cells in this category were large cells with eccentrically placed nuclei but lacked MHC II expression (Figure 5c). Their identity is uncertain, but because they did not immunostain with the macrophage marker ED2 and did display some of the phenotypic features of MHC II⁺ DC, including adherent lymphocytes, they were tentatively identified as MHC II DC precursors. As discussed subsequently, similar large cells were stained to varying degrees with the ED1 mAb that labels a lysosomal antigen, an observation consistent with the fact that immature DC, like monocytes and macrophages, contain a lysosomal compartment (35). The fraction of DC expressing MHC II antigen increased with time in culture with GM-CSF and was consistently greater in the population harvested from the lung vascular bed than from the vena cava (Figure 4).

View larger version (140K): [in this window] [in a new window]   View larger version (135K): [in this window] [in a new window]   View larger version (133K): [in this window] [in a new window]   View larger version (129K): [in this window] [in a new window]   View larger version (128K): [in this window] [in a new window]   View larger version (133K): [in this window] [in a new window]   Figure 5.   (a) Leukocyte preparation from vena cava blood, double immunostained with OX6 and OX62 mAbs. A rare OX62+ DC (red) is surrounded by a cluster of T cells similar to that surrounding an OX6+ DC (blue). (b) MHC II+ cells in a leukocyte preparation from the lung vascular blood cultured for 2 d with GM-CSF. MHC II+ cells vary in size and have one or more adherent T cells. MHC II+ DC from the vena cava (c) and lung blood (d) after culture for 3 d with 500 U/ml of GM-CSF; dendritic cells are more numerous in the leukocyte population from the lung than from the vena cava blood. The DC have eccentrically placed nuclei, abundant cytoplasm, and long cell processes, and they form rosettes with the surrounding T cells. Note in (c) the large MHC II cell with two adherent lymphocytes. ED1-stained leukocytes from lung blood on the day of isolation (Day 0) (e) and after 3 d in culture with GM-CSF (f ); in both preparations, ED1+ cells form clusters with large ED1 cells and T cells. Magnification: (a) ×200, (b-f ) ×256.

Very few of the circulating DC expressed the -like subunit of integrin that is recognized by the OX62 mAb (Table 3). This immunologic reagent has been used to identify DC precursors in fetal rat lungs (36) and in the airway epithelium (16). Dual immunostaining with OX6 and OX62 mAb showed that, when present, OX62 mAb labeled large cells that formed clusters with T cells and resembled OX6⁺ DC that formed similar clusters (Figure 5a). By contrast, ICAM-1, LFA-1, and the 2 subunit of integrin recognized by the WT3 mAb were expressed by most of the blood-borne MHC II⁺ DC even before cultivation with GM-CSF.

The mAb ED1 recognizes a lysosomal antigen present in monocytes, macrophages, and a subset of DC, whereas the ED2 mAb is specific for an antigen on the surface of macrophages (22). None of the leukocytes were immunolabeled by ED2, suggesting that the large cells immunolabeled by ED1 were immature DC. A spectrum of staining with ED1 was noted in these large cells (Figure 5f), which is consistent with the observation that the lysosomal compartment diminishes as the DC matures (35). A number of small mononuclear cells were also ED1⁺. Their fraction, as a percentage of total leukocytes counted, was consistently higher in the lung than in the venous blood (4.0, 2.7, and 1.6% in lung blood versus 3.2, 0.8, and 0.7% in the venous blood on Days 0, 2, and 3, respectively). ED1⁺ cells typically formed clusters with themselves, with T cells, and with large ED1 cells (Figures 5e and 5f), regardless of whether they had been cultured with GM-CSF.

Function of DC Isolated from the Venous and Lung Blood

The accessory cell function of DC may be modulated by the microenvironment in which they reside (17, 37). To determine whether the milieu of the lung vasculature modified their accessory cell function, DC were harvested from the lung vasculature and the vena cava and tested for their ability to stimulate T-cell proliferation in an allogeneic MLR. The harvested DC were enriched by negative immunomagnetic separation (T and B cells were removed, but monocytes were retained) and cultured with 500 U/ml of GM-CSF in 24-well trays. An aliquot of these cells was used immediately to set up an MLR assay (Day 0) with and without added GM-CSF. At 1, 2, and 3 d, loosely adherent cells, including DC, were harvested from the 24-well plates and MLR assays were similarly set up with and without GM-CSF. Allogeneic, splenic T cells from Long Evans rats were freshly isolated and purified for each assay. At all time periods studied, there was no difference in the ability of DC harvested from the two sources to stimulate T-cell multiplication (Figure 6). The radioactive counts were consistently higher when the MLR was conducted in the presence of GM-CSF than without (data not shown). The low level of accessory cell function observed on Days 0 and 1 was in part due to the small fraction of mature DC obtained after negative immunomagnetic separation. Furthermore, freshly isolated blood DC and the co-purified monocyte fraction from either compartment were incapable of stimulating T-cell multiplication in the absence of GM-CSF.

View larger version (32K): [in this window] [in a new window]   Figure 6.   Blood DC from either the lung vascular compartment or the vena cava of Lewis rats were enriched by negative immunomagnetic separation as described in MATERIALS AND METHODS. Except for an aliquot of cells used to set up the MLR on Day 0, the cells were placed in culture with 500 U/ml of GM-CSF. At 0, 1, 2, or 3 d the loosely adherent cells (1 × 104 cells/well) were added to nylon wool-purified splenic T cells (2 × 105 cells/ well) from Long Evans rats. The assay was set up either with or without 500 U/ml of GM-CSF. After 7 d, T-cell proliferation was assayed by the addition of 1 µCi/well of [3H]TdR for 6 h at 37°C. Incorporation of [3H]TdR by T cells was quantitated by liquid scintillation counting. Blood DC preparations obtained on the day of harvest (Day 0) or after 1 d in culture with GM-CSF (Day 1) required GM-CSF to maintain their accessory cell function during the assay. Assays conducted in the presence of GM-CSF yielded somewhat higher counts than in its absence (not shown). There was no difference in the ability of the blood DC from either source to stimulate T-cell proliferation. The data represent the means ± SEM.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

The key observations of the present study are the following. Circulating MHC II⁺ DC constitute a trace population of leukocytes that do not appear to be enriched in the lung vascular compartment. However, in contrast to the leukocytes in the vena cava blood, those sequestered in the lung vasculature are enriched in a population of nondividing mononuclear cells which, upon stimulation with GM-CSF, differentiate into DC. Freshly harvested, MHC II⁺ DC-enriched fractions obtained from either vascular compartment require the presence of GM-CSF to function as accessory cells in an MLR assay. After culture in vitro with GM-CSF for 1, 2, or 3 d, the accessory cell function of DC and/or their precursors is upregulated. No discernible difference was detected in the ability of DC from the two blood compartments to activate splenic T cells in an allogeneic MLR assay.

The impetus to examine the DC population in the lung vasculature came from a recent study from this laboratory (20). It was observed that intravenously injected bone marrow-derived MHC II⁺ DC neither were retained in the lung vasculature nor did they populate the lung interstitium, confirming previous observations made with spleen-derived DC (19). Because lung DC continuously turn over and those in the airway epithelium turn over particularly rapidly (16), it was somewhat surprising that cultured bone marrow-derived DC in various stages of differentiation did not enter the lung parenchyma. Furthermore, administration of IFN-, rather than augmenting retention of the injected DC in the lung vasculature, reduced their number even further while simultaneously upregulating the expression of MHC II antigen by cells in the lung interstitium. This suggested the possibility that, as in other tissues (4), a population of circulating MHC II precursor cells may be the source from which lung DC precursors are replenished.

The observation that there are MHC II mononuclear leukocytes sequestered in the lung vasculature that are capable of differentiating into DC upon stimulation with GM-CSF supports this notion. Their enrichment in the lung vascular compartment provides a readily available source from which to replenish tissue DC, both as part of the normal turnover of these cells and their recruitment during inflammatory reactions (38). Although a variety of chemokines and complement fragments are capable of recruiting DC into the airway epithelium (39), it is unclear whether these cells are derived from precursors residing in the interstitium or whether they are directly recruited from the lung vasculature. Further studies are required to determine the route of DC precursor migration from the vascular compartment into the lung interstitium, the mechanisms involved in the transmigration of these cells across the endothelium, and the factors that elicit DC precursor emigration from the lung vascular bed.

Previous studies from this laboratory indicated that DC isolated from the airway epithelium express Fc receptors, and that they endocytose, process soluble and particulate antigens, and present antigenic peptides to sensitized T cells, thereby resembling epidermal Langerhans cells (17). By contrast, DC isolated from the lung interstitium were less efficient in antigen uptake but were more competent in stimulating naive T cells in an MLR assay than those from the airway epithelium. These observations suggest that there are subsets of DC of differing phenotype and/or levels of differentiation in the lung. A difference in DC phenotype may be a response to locally released cytokines. Alternatively, recent studies have now delineated subsets of circulating DC precursors as well (40, 41). When CD34⁺ hematopoietic progenitors from human cord blood are cultured with GM-CSF plus tumor necrosis factor-, two subsets of DC precursors emerge. The CD1a⁺ precursors differentiate into E-cadherin-expressing Langerhans cells containing Birbeck granules, whereas the CD14⁺ subset expresses neither of these attributes. Both subsets are, however, equally effective in stimulating naive allogeneic T cells. It will be of interest to determine whether such subsets are present among the DC precursors sequestered in the lung vasculaturethose destined for the airway epithelium and those migrating primarily to the lung interstitium. Alternatively, it is possible that endothelial cells lining the bronchial vasculature might preferentially promote the sequestration of E-cadherin-bearing DC precursors, whereas those in the pulmonary vasculature might favor the retention of CD14⁺ precursors.

The ability to define phenotypically the blood DC precursors and their possible subsets in the rat is currently hampered by the lack of the necessary immunologic reagents. It is noteworthy that whereas MHC II⁺ DC harvested from both the lung vasculature and the vena cava express a variety of adhesion molecules (CD11a, CD54, and CD18) shown to be important in the adherence of DC to endothelium in vitro (42), these cells failed to populate the lung when they were injected intravenously (20). Similarly, few of the monocyte-like cells in these preparations expressed these adhesion molecules (data not shown), and culture for 3 d with GM-CSF did not appear to upregulate their expression in this cell population. It is conceivable that these precursor cells use a different set of adhesion molecules for adherence and migration across the endothelium into the lung interstitium. Alternatively, specific chemotactic signals may be required to induce these cells to express the necessary adhesion molecules and/or their receptors in order to facilitate their migration from the circulation into the various compartments of the lung.