INTRODUCTION

Top Abstract Introduction Methods Results Discussion References

SINCE METCHNIKOFF (17) described the ordered inflammatory response to wounding in the 19th century, wound cells have been assigned specific functions. The classical role of the macrophage, present in the early inflammatory phase of wound healing, has been the removal of debris and the phagocytosis of microorganisms (22). However, the role of the macrophage in wound healing continues to broaden as new subpopulations of macrophages with differing functions are described (4, 22). The fibroblast, present in the later phases of wound healing, has been thought to be uniquely responsible for collagen production and remodeling (3). Recent data have blurred these classical roles, expanded the role of the macrophage in wound healing and fibrosis, and suggested a phenotypic continuum between macrophages and fibroblasts (1, 2, 11, 26, 27). These studies suggest that cells displaying hematopoetic markers can differentiate into a collagen-producing phenotype. To date, however, no investigations have shown coexpression of macrophage-specific markers and fibroblast-specific markers by a wound cell population.

Because the availability of antibodies that distinguish intermediate or transitional cell types may be limiting, phagocytosis was chosen as a means to assess the possible continuity between the macrophage and fibroblast phenotype. Phagocytosis is a primary function used to classify cells as macrophages. In vitro and in vivo bead phagocytosis by macrophages has been well characterized. Cultured monocytes and macrophages acquire an opsonin-independent ability to engulf beads in culture after adherence, whereas alveolar macrophages can engulf nonopsonized particles or beads when in solution (10). Alveolar macrophages ingesting beads coated with IgG release tumor necrosis factor- (TNF-) both in vitro and in vivo (9, 10). This TNF- release does not occur for nonopsonized beads, suggesting substrate-specific macrophage activation. This specific phagocytosis-mediated activation in macrophages has also been measured by flow cytometry using hydroethidium (HE; see Ref. 10). Alveolar macrophages show oxidation of HE after phagocytosis of IgG-coated beads and no oxidation after ingesting uncoated beads, consistent with receptor-mediated activation (9). When the scavenger receptor is blocked, uncoated bead phagocytosis by macrophages is markedly decreased, whereas opsonin-dependent (IgG coated) phagocytosis is unaltered (8).

Fibroblasts are known to demonstrate phagocytosis (21). However, fibroblasts, which are classified as "nonprofessional" phagocytes, differ from "professional" phagocytes, like macrophages, primarily based on the limited range of particles they ingest (21). Because fibroblasts lack Fc receptors, they are unable to phagocytize immunoglobulins and complement as efficiently as macrophages. Unlike macrophages, fibroblasts are not thought to produce reactive oxygen or nitrogen products in response to particle uptake. Apoptotic neutrophils are phagocytosed by fibroblasts by a mechanism involving the fibroblast vitronectin receptor and a mannose/fucose-specific lectin (6). Fibroblasts also ingest beads coated with collagen or fibronectin in culture (7, 12-15). This process appears to be receptor specific and can be blocked by competitive inhibition. Gingival fibroblasts have been shown to engulf type I, III, and IV collagen-coated beads, and this effect is blocked by a synthetic peptide, which is a competitive inhibitor demonstrating receptor specificity (7). Similarly, peptides and specific antibodies also block the phagocytosis of fibronectin-coated beads by fibroblasts (15).

We examined phagocytosis and HE oxidation by wound fibroblasts both in vivo and in vitro. Phagocytosis of fluorescent beads was chosen since this method is easily quantifiable by flow cytometry and is a well-established method for both macrophage and fibroblast phenotypes.

	METHODS

Top Abstract Introduction Methods Results Discussion References

Preparation of fluorescent beads. Carboxylate-modified 1-µm yellow green fluorospheres (excitation 490 nm, emission 515 nm; Molecular Probes, Eugene, OR) were used in all experiments. Before use in the phagocytosis assays, 100 × 10⁹ beads were coated by incubation with 1 ml phosphate-buffered saline (PBS; GIBCO-BRL, Grand Island, NY) containing either 0.1 mg/ml mouse IgG (Accurate Chemical, Westbury, NY) or 0.1 mg/ml rat tail collagen (Sigma, St. Louis, MO) for 1 h at 4°C.

In vivo phagocytosis. Male Fisher rats (n = 3) were anesthetized with intraperitoneal pentobarbital sodium (50 mg/kg; Abbott Laboratories, N. Chicago, IL) and wounded by subcutaneous implantation of four large (1 × 1 × 0.3 cm) polyvinyl alcohol (PVA; Unipoint Industries, Thomasville, NC) sponges into their shaved dorsal side. Nine days after implantation, the rats were again anesthetized with pentobarbital sodium, and each sponge was injected with 50 µl of PBS containing either 100 million uncoated, IgG-coated, or collagen-coated fluorescent beads. One sponge in each rat was injected with 50 µl of PBS as a control. Sponges were harvested 24 h after injection of the beads. Sponges were then incubated with constant shaking at 37°C with PBS containing protease [0.2% DNAse, 0.5% collagenase, 0.2% protease (Sigma)] for 15 min, which maximized cell recovery from the sponge (20). The resulting cell suspension was centrifuged (500 g for 5 min), and contaminating erthyrocytes were removed by hypotonic lysis in sterile distilled water for 15 s. Lysis was stopped by the addition of an equal volume of 2× concentrated Hanks' balanced salt solution. Cells were then treated with 2.5% trypsin (GIBCO-BRL) for 10 min at 37°C to remove any bound but not engulfed beads. Cells were then centrifuged at 500 g for 5 min through a solution of PBS containing 10% bovine serum albumin (BSA) to mechanically cleave loosely bound beads and then were resuspended in PBS. The cells were then counted with a hemocytometer, and viability was assessed by trypan blue exclusion.

Flow cytometry. Wound-derived cells (1 × 10⁶) were fixed with 70% ethanol in PBS at 4°C, centrifuged (500 g, 5 min), and resuspended in PBS with 0.1% BSA. This wash step was repeated for all subsequent steps in staining. Polyclonal rabbit anti-rat procollagen I (a gift from Dr. Joseph Madri) was added to the cells at 1:125 dilution in PBS with 0.1% BSA for 1 h at 4°C. Rabbit serum at an equivalent dilution was used as a nonimmune control. After washing with 0.1% BSA/PBS, the secondary antibody, phycoerythrin-tagged donkey anti-rabbit IgG (Accurate) at 1:100 dilution in PBS, was added to the cells for 30 min at 4°C. Cells were analyzed by flow cytometry (FACSCAN; Becton-Dickinson) simultaneously for fluorescent bead phagocytosis in the FL1 channel and procollagen I positivity in the FL2 channel. Controls were performed with either FL1 staining alone or FL2 staining alone to determine appropriate compensation settings.

Expression of procollagen I was used to characterize cells isolated from PVA sponge wounds as fibroblasts (collagen-producing cells). Two cell types, rat peritoneal exudate cells (RPEC) and wound fibroblasts (WFIB), were used as controls to establish the specificity of the polyclonal antibody for fibroblasts (Fig. 1). The RPEC showed no fluorescence for procollagen I while the majority (90%) of WFIBs showed positive fluorescence.

View larger version (26K): [in this window] [in a new window]   Fig. 1.   Cellular specificity of procollagen I antibody. Rat peritoneal exudate cells (RPEC) and wound fibroblasts (WFIB) are used as controls for rabbit anti-rat procollagen I. Labeling and flow cytometry were performed as described in METHODS. RPEC are negative, whereas WFIB are strongly collagen positive compared with nonimmune control as indicated by the increased percentage of events in marker region (M1). These controls document the specificity of this procollagen antibody and allow for detection of collagen-producing cells in vivo. A: RPEC control; B: RPEC procollagen I; C: WFIB control; D: WFIB procollagen I.

An example of an individual analysis is shown in Fig. 2. In this analysis, uncoated beads were injected into a sponge, and wound cells were harvested 1 day later and then simultaneously analyzed for bead phagocytosis (FL1) and procollagen I (FL2). In the dot plots shown, the nonimmune control is compared with the procollagen I, and a horizontal marker is used for procollagen I positivity. Controls for phagocytosis (PBS-injected sponges) showed no events in the right two quadrants. The cells that have associated beads were separated by a vertical marker in the FL1 channel. Cells detected in the two left quadrants were considered positive for phagocytosis. Quadrant analysis was used to compare the procollagen I profile of phagocytic and nonphagocytic cells.

View larger version (31K): [in this window] [in a new window]   Fig. 2.   Two-color analysis of sponge inflammatory cells. These dot plots are examples of fluorescence-activated cell sorter (FACS) analysis of the sponge inflammatory cells removed from the wound 10 days after sponge implantation and 1 day after uncoated fluorescent bead injection. In the dot plots shown, the nonimmune control (normal rabbit serum instead of procollagen I antiserum) was compared with the procollagen I, and a horizontal marker was set to establish procollagen I positivity in the FL2 channel. Cells from PBS-injected sponges (instead of fluorescent beads) were analyzed as controls for phagocytosis, and a vertical marker was set to establish phagocytosis positivity in the FL1 channel. Cells in the lower left quadrant are considered phagocytosis and procollagen I negative; cells in the upper left quadrant are phagocytosis negative and procollagen I positive; cells in the lower right quadrant are phagocytosis positive and procollagen I negative; cells in the upper right quadrant are phagocytosis and procollagen I positive. A: nonimmune control; B : procollagen I.

WFIB cultures. Male Fisher rats were anesthetized with intraperitoneal pentobarbital sodium (50 mg/kg) and wounded by implantation of ten 0.5-cm³ PVA sponges into their shaved dorsal side. Sponges were removed by sharp dissection 7 days after implantation and quartered. Under sterile conditions, quartered sponges were placed into DMEM (GIBCO-BRL) containing 10% fetal bovine serum (FBS; HyClone, Logan, UT), 10 mM 3-(N-morpholino)propanesulfonic acid (MOPS), 50 µg/ml penicillin, and 50 U/ml streptomycin (Penstrep; complete medium). This medium favors the growth of fibroblasts, producing a monolayer of cells with typical fibroblast morphology after 7 days. These cells are uniformly positive for intracellular procollagen I, as detected by a rabbit anti-rat polyclonal antibody to procollagen I by immunohistology (not shown) and flow cytometry (Fig. 1). The fibroblasts were passed using 2.5% trypsin. Cells were used for phagocytosis studies at passages 3-7. Resident RPEC were collected by sterile peritoneal lavage using medium consisting of RPMI 1640 supplemented with 1% FBS, 10 mM MOPS, and Penstrep.

HE dose response. HE is taken up by the cell and then oxidized by both O₂ and H₂O₂ to ethidium (24). Ethidium binds to cellular DNA and emits a red fluorescence that can be detected by the flow cytometer in conjunction with phagocytosis of yellow-green fluorescent spheres. HE (Molecular Probes) was added to WFIBs in complete medium at 0.25-16 µg/ml. Cells were then incubated at 37°C for 30 min. Subsequent analysis was by flow cytometry in the FL2 channel. Doses of HE at a concentration of 1 µg/ml or higher produced a bimodal distribution of FL2 fluorescence (Fig. 3). More fluorescence was seen in fewer cells with increasing concentrations of HE and is consistent with the known toxicity of high doses of HE. Thus 1 µg/ml was chosen as the dose for the phagocytosis experiments since it was the lowest concentration of HE that resulted in optimal fluorescence with minimal toxicity.

View larger version (29K): [in this window] [in a new window]   Fig. 3.   Dose response for hydroethidium (HE). WFIBs incubated with increasing concentrations of HE at 37°C show a bimodal distribution when analyzed by flow cytometry. As the concentration of HE increases, the mean fluorescence (FL2) of the more fluorescent peak increases, but the number of cells in this peak decreases. A concentration of 1 µg/ml was chosen for subsequent experiments, since it produced optimal fluorescence with minimal cell toxicity. A: no HE; B: 0.25 µg/ml HE; C: 1 µg/ml HE; D: 4 µg/ml HE; E: 16 µg/ml HE.

In vitro measurement of phagocytosis and activation state. Stock cultures of WFIBs were trypsinized, and the resulting cell suspension was pelleted by centrifugation. The cell pellet was resuspended in complete medium, and cell aliquots were seeded into wells of 24-well plates at a density of 10,000 cells/ml. The WFIBs were allowed to recover for 2 days before determination of phagocytic activity. Before assaying for phagocytosis, old culture medium was removed, and 1 ml of fresh complete medium was added at 4°C. The plates were then divided into cold (4°C) and warm (37°C), and all further manipulations were done in parallel. The cold plates served as a phagocytosis control. Phagocytosis is an energy-requiring process, and at this low temperature only the binding of beads to the cell surface occurs (13).

HE was added to individual wells of WFIBs in complete medium at 1 µg/ml in both cold and warm for 30 min. Controls were performed without HE. A total of 50 million beads (either uncoated, IgG coated, or collagen coated) was then added to each well. After 24 h, medium containing unbound beads was removed, and the cells were incubated in 2.5% trypsin for 10 min at 37°C (for both cold and warm plates) to cleave any bound but not engulfed beads as well as to remove the cells from the plates. Cells were then centrifuged at 500 g for 5 min through a 10% BSA/PBS solution to mechanically cleave loosely bound beads. The cells were finally resuspended in 0.1% BSA in PBS.

Flourescence-activated cell sorter analyses were performed (n = 2) to evaluate oxidation state (HE) and bead phagocytosis in both warm and cold environments, with and without HE, and in the absence of beads and for coated beads (none, IgG, and collagen). For two-color flow cytometry, controls were performed with either FL1 staining alone or FL2 staining alone. These controls were compared with the two-color analysis to determine accurate compensation. Quadrant analysis was used to compare the oxidation state of phagocytic and nonphagocytic cells. An example of an individual analysis is shown in Fig. 4. The percentage of cells in each quadrant was determined and classified as follows: lower left, HE and phagocytosis negative; upper left, HE positive and phagocytosis negative; lower right, HE negative and phagocytosis positive; upper right, HE and phagocytosis positive.

View larger version (22K): [in this window] [in a new window]   Fig. 4.   Two-color analysis of phagocytosis and activation state (HE). These dot plots are examples of FACS analysis of WFIBs incubated with HE and then IgG-coated beads in both cold (4°C; A) and warm (37°C; B) conditions. WFIBs incubated without HE or beads were used as a control to set a horizontal marker in the FL2 channel and a vertical marker in the FL1 channel. Cells in each quadrant were considered to express the following characteristics: lower left, HE and phagocytosis negative; upper left, HE positive and phagocytosis negative; lower right, HE negative and phagocytosis positive; upper right, HE and phagocytosis positive.

Scavenger receptor phagocytosis. The role of scavenger receptors in WFIB phagocytosis was assessed using the selective inhibitor polyinosinic acid [poly(I)]. WFIB or adherent RPEC was incubated with either complete medium or complete medium containing 500 µg/ml of poly(I) or chondroitin sulfate for 30 min before incubation with beads (8). Uncoated beads were incubated with RPEC for 2 h and WFIB overnight. Flow cytometric detection of bead fluorescence in the FL1 channel was used to determine phagocytosis.

	RESULTS

Top Abstract Introduction Methods Results Discussion References

In vivo phagocytosis. The percentage of cells that engulf beads is shown in Fig. 5A. Uncoated, IgG-coated, or collagen-coated beads were engulfed by ~12, 18, and 2% of the wound cells, respectively. Approximately 35% of the cells isolated from the day 10 PVA sponge wounds were procollagen I positive. The ability of procollagen I-positive cells to phagocytize beads with different coatings was also assessed. Approximately 4% of procollagen I-positive cells engulfed collagen-coated beads, whereas 16% engulfed uncoated and 28% engulfed IgG-coated beads (Fig. 5B). Approximately 50% of the phagocytic cells that engulfed uncoated or IgG-coated beads and 65% of the phagocytic cells that engulfed collagen-coated beads were procollagen I positive (Fig. 5C).

View larger version (12K): [in this window] [in a new window]   View larger version (14K): [in this window] [in a new window]   View larger version (15K): [in this window] [in a new window]   Fig. 5.   In vivo phagocytosis of fluorescent beads. Data shown for each set of bars are means ± SE from FACS analysis cells collected from sponges in 3 different rats. Analysis of phagocytosis and procollagen I expression was performed as described in METHODS and the example analysis (Fig. 2). * P < 0.05 vs. none and IgG-coated beads by unpaired Student's t-test. A: total phagocytic cells. Total percentage of cells that engulfed each bead type was determined by summing the positive cells in both the upper and lower right quadrants of the FACS analysis. B: phagocytic, procollagen I positive cells. The percentage of procollagen I positive cells that were also phagocytic (upper right quadrant divided by the sum of upper left and right quadrants). C: procollagen I positive, phagocytic cells. The percentage of phagocytic cells that are procollagen I positive (upper right quadrant divided by the sum of upper and lower right quadrants).

In vitro phagocytosis. Phagocytosis by a purified population of WFIB was assessed to determine similarities to macrophage phagocytosis. Table 1 shows the results of two-color analyses of WFIB oxidation state (using HE) and phagocytosis. The analysis of fluorescent beads with various coatings at 4°C shows the degree of bead binding to the surface of these cells (right quadrants) and is used to distinguish beads bound to the cell surface from internalized beads. At 37°C there was a larger percentage of cells with bead-associated fluorescence, indicating more phagocytosis than bead binding. WFIBs engulfed beads regardless of bead coating (uncoated, IgG, or collagen; Fig. 6). The majority of cells that had engulfed beads were positioned in the upper right quadrant, indicating a correlation between elevated oxidation state and phagocytosis (Table 1). In the absence of beads, ~57% of the WFIBs were in an elevated oxidation state at 37°C.

View larger version (15K): [in this window] [in a new window]   Fig. 6.   Phagocytosis by WFIBs in vitro. The percentage of WFIBs in an enhanced oxidation state that engulfed beads with the indicated coatings (upper right quadrant). Values are means ± SD of duplicate analyses.

WFIBs incubated with poly(I), a well-described inhibitor of macrophage bead phagocytosis by the scavenger receptor, showed a stimulation of uncoated bead phagocytosis compared with chondroitin sulfate-inhibited control (Fig. 7). The RPECs showed an expected decrease in phagocytosis when incubated with poly(I) compared with control.

View larger version (12K): [in this window] [in a new window]   Fig. 7.   Analysis of macrophage scavenger receptor-mediated phagocytosis in WFIBs. WFIB (A) and RPEC (B) were incubated with polyinosinic acid [poly(I)] or chondroitin sulfate (CS) as described in METHODS. Values are means ± SD of duplicate analyses.

	DISCUSSION

Top Abstract Introduction Methods Results Discussion References

Macrophages are a heterogeneous population of cells capable of many functions (4, 22). Maturational stage, state of activation, and the local microenvironment are key factors that create this heterogeneity. Of particular interest is the possibility that macrophages contribute to collagen production in the wound. Numerous reports support this possibility by suggesting that cells of myelomonocytic origin give rise to collagen-producing cells at sites of fibrosis (1, 2, 11, 26, 27). A typical finding of these studies is the development of a fibroblast-like morphology by these cells. The antigenic markers used in these studies were not macrophage specific, and the possibility that these cells were a subpopulation of macrophages arising in the wound environment has not been explored. Although existing markers identify specific populations of macrophages, they do not identify all macrophages. A marker that exclusively recognizes all members of macrophage phylogeny is not available.

A primary approach used to identify macrophages is to evaluate phagocytosis (4). The mechanism macrophages use to internalize particles (usually >0.5 µm) is well characterized (18). In general, phagocytosis by so-called "professional phagocytes" such as neutrophils, monocytes, and macrophages is a receptor- and actin-dependent and clathrin-independent process. Particles may be recognized by receptors on the phagocyte surface or by specific recognition of opsonins, which coat the particles before ingestion. The major opsonins are complement components in serum and immunoglobulins.

However, cells other then the professional phagocytes also are capable of varying degrees of phagocytosis (21). Fibroblasts are among the cell types classified as "nonprofessional phagocytes." Mechanisms similar to those used by macrophages are also used by fibroblasts to engulf particles and effete neutrophils (21). These mechanisms involve fibronectin, laminin receptors, or heparan sulfates displayed by the cell. Nonetheless, differences also exist in particle uptake by macrophages and fibroblasts. Fibroblasts lack Fc receptors, and thus antibodies cannot act as opsonins to enhance particle uptake by fibroblasts (21). In addition, fibroblasts do not produce reactive oxygen or nitrogen species in response to a phagocytic stimulus (21).

On the basis of our hypothesis of the continuum between macrophage and fibroblast function in the wound, we predicted that WFIBs may display features of macrophage phagocytosis. We tested this by injecting uncoated beads, IgG-coated beads, and collagen-coated beads into a rat sponge wound, harvesting the wound cells and examining cells that had engulfed beads. Opsonization with IgG did not enhance phagocytosis by the total phagocytizing population in the wound or in the subpopulation of procollagen I-positive cells. Equivalent uptake of unopsonized and opsonized latex beads in vivo by macrophages has been reported (8). The finding of a similar result for the procollagen I population suggests that these cells engulf the beads through a similar mechanism to the other phagocytes in the wound, which are presumably macrophages at this time.

Fibroblasts lack Fc receptors and thus are thought to not produce an Fc receptor-activated respiratory burst (21). The intracellular oxidation state after phagocytosis secondary to respiratory burst has been well characterized in macrophages but not in fibroblasts (19, 24). The HE assay is well established in the macrophage literature as a measure of intracellular oxidative burst (19, 24). When HE was added to fibroblasts, a bimodal distribution of fluorescence was obtained, defining two activation states. Our results suggest that either cell activation favors ingestion of beads or that WFIBs, like macrophages, produce a respiratory burst upon ingestion of beads. The methods used do not allow us to discern between these two possibilities.

The role that fibroblast phagocytosis of collagen may play in collagen turnover and connective tissue remodeling could be important to wound repair. In vivo studies of fibroblast phagocytosis have primarily shown ultrastructural identification of ingested collagen fibrils. Melcher and Chan (16) demonstrated collagen fibrils that had been digested within fibroblasts in rat gingiva. Hall and Squier (5) examined the effect of phenytoin on gingival hyperplasia and suggested that the hyperplasia was a result of impaired fibroblast phagocytosis of collagen. Sakai and colleagues (25) investigated the increase in collagenase activity in lacerated rat corneas. They cultured rat corneas in the presence of leupeptin, a protease inhibitor, and found that the previous collagen seen in fibroblast lysosomes was now absent, implying an important in vivo role for fibroblast collagen phagocytosis and degradation. A deficiency in the phagocytosis of collagen-coated beads has been demonstrated in human fibroblasts cultured from fibrotic lesions, suggesting a potential mechanism for fibrosis in these patients (14).

In vivo, the number of cells that engulf collagen-coated beads is significantly less than those that engulf uncoated or IgG-coated beads. The ingestion of collagen beads appears to be more selective than other bead coatings. It is possible that a specific cell population (present in relatively low numbers in a 10-day wound) has increased affinity for ingestion of collagen beads. This possibility is supported by the finding that procollagen I-positive phagocytic cells ingest collagen-coated beads more readily than other bead types.

Also of interest is the finding that ingestion of beads in vitro is accomplished by WFIB in an enhanced activation state. To our knowledge, this is the first demonstration of a correlation of increased oxidation reactions (as measured by HE) and phagocytosis in WFIBs. The possibility that ingestion of beads elicits a respiratory burst in a manner similar to that observed in Fc-mediated respiratory burst in macrophages requires further investigation.

Some similarities may exist in the phagocytic mechanisms of macrophages and WFIBs. However, the evaluation of scavenger receptor-mediated phagocytosis suggests that there are distinct differences. The macrophage scavenger receptor has an unusually broad binding specificity. Ligands include modified low-density lipoproteins and some polyanions [for example, poly(I) but not chondroitin sulfate; see Ref. 23]. Our findings suggest that the receptor for bead phagocytosis by fibroblasts in vitro is either an altered scavenger receptor with paradoxical stimulation by poly(I) or an entirely different receptor altogether.

In conclusion, fibroblasts appear to have a significant contribution to phagocytic activity in the healing wound. The results continue to support the possibility of a connection between macrophage and fibroblast phenotype. However, a direct connection is still elusive. Studies combining immunophenotyping and functional analyses of the multiple macrophage and fibroblast wound phenotypes may provide further resolution of this question.

Perspectives