Introduction

Top Abstract Introduction Materials & Methods Results Discussion References

Pulmonary alveolar macrophages (AMs) serve as the primary means of clearing inhaled particulates from the lung by engulfing a wide array of targets that vary in size, shape, and surface characteristics (1, 2). Given the diversity of particulates to which the lungs are exposed (3), AMs are capable of adapting to a wide variety of particulate characteristics, permitting successful phagocytosis of these particles. In this regard, specific proteins may serve as nonimmune opsonins to facilitate phagocytosis (4).

Vitronectin (Vn), a 75-kD adhesive glycoprotein found in the alveolar space, has recently been implicated as a nonimmune opsonin (5). Vn belongs to a diverse family of proteins that contain an arginine-glycine-aspartate (RGD) sequence recognized by integrins (6). Due to its multiple and diverse binding domains, Vn is multifunctional, with known roles in coagulation, fibrinolysis, complement- and lymphocyte-mediated cytolysis, and cell adhesion (7). The ability of Vn to adhere to a wide variety of different materials (7, 8) may be important in its role as a putative nonimmune opsonin. As evidence of this, Vn binds to sheep erythrocytes and enhances their phagocytosis by monocytes (9). In addition, a vitronectin receptor (VnR) mediates the phagocytosis of senescent neutrophils and lymphocytes by macrophages (4, 10). Vn is concentrated at sites of cellular debris where phagocytosis is required, such as keratin bodies (5) and senile plaques (6). In the lung, Vn is increased in granulomatous pulmonary disease (11). AMs secrete Vn (8), and cultured macrophages express at least two variants of VnR (12).

These findings suggest that in the lung, Vn may serve as a nonimmune opsonin that can bind to a wide variety of targets for clearance by AM. This report describes our work, which demonstrates that Vn enhances AM phagocytosis of liposomes and that this phagocytosis is mediated by a VnR.

	Materials and Methods

Top Abstract Introduction Materials & Methods Results Discussion References

Materials

Ammonium chloride, ammonium sulfate, bovine serum albumin (BSA), bovine Vn, calcium chloride, dibasic phosphate, dimethyl sulfoxide (DMSO), ethylenediamine tetraacetic acid (EDTA), heparin, N-2-hydroxyethylpiperazine-N'-2-ethane sulfonic acid (Hepes), monobasic phosphate, potassium bicarbonate, sodium chloride, tris(hydroxymethyl)aminomethane (Tris), trypsin, and urea were obtained from Sigma Chemical Co. (St. Louis, MO). Female rats with an average weight of 225 g were obtained from Harlan Sprague-Dawley, Inc. (Indianapolis, IN) and housed in the Laboratory Animal Resource Center at Indiana University Medical Center under the guidelines of the American Association for the Accreditation of Laboratory Animal Care. Diethyl ether was purchased from Mallinckrodt Specialty Chemicals Co. (Paris, KY). Beuthanasia-D (3.9 g/liter pentobarbital sodium) was supplied by Schering-Plough Animal Health Corp. (Kenilworth, NJ). RPMI 1640 culture medium (Life Technologies, Inc., Grand Island, NY) was fortified with 4 mM glutamine (Sigma) and penicillin-streptomycin (Bio-Whittaker Bioproducts, Inc., Walkersville, MD). Phosphate-buffered saline (PBS; 19 mM H₂PO₄, 81 mM HPO₄⁼, 150 mM NaCl, 1 mM CaCl₂), Hanks' balanced salt solution (HBSS), both titrated to pH 7.4, were obtained from Life Technologies. Cholesterol, dipalmitoylphosphatidylcholine (DPPC), and stearylamine were obtained from Avanti Lipids, Inc. (Alabaster, AL). Polystyrene microspheres (50-nm diameter) tagged with a proprietary yellow-green fluorophore were obtained from Polysciences Corp. (Warrington, PA). Antivitronectin IgG₁ (anti-Vn) and fluorescein-conjugated goat antimouse IgG (IgG-FITC) were purchased from Boehringer Mannheim Corp. (Indianapolis, IN). IODO-BEADS^®, anti-_v3 vitronectin receptor IgG_2A (anti-VnR), and bicinchoninic acid (BCA) protein assay kits were obtained from Pierce, Inc. (Rockford, IL). Oligopeptides gly-arg-gly-asp-ser (GRGDS), gly-pen-gly-arg-gly-asp-ser-pro-cys-ala (GPen), and gly-arg-gly-glu-ser-pro (GRGESP) were obtained from Life Technologies. A Pharmacia PD-10 column containing Sephadex G-25M gelatin sepharose and heparin sepharose were obtained from Pharmacia LKB Biotechnology, Inc. (Piscataway, NJ). 5(6)-Carboxy-X-rhodamin-N-hydroxysuccinimide ester (RHODOS) and phenylmethylsulfonyl fluoride (PMSF) were obtained from Boehringer Mannheim. Radiolabeled sodium iodide was obtained from Amersham Corp. (Arlington Heights, IL).

Isolation of Alveolar Macrophages

Rat AMs were obtained from female Sprague-Dawley rats by whole-lung lavage (13). The rats were rapidly anesthetized by placement in an ethyl ether chamber, then killed with an intraperitoneal injection of 0.5 ml of Beuthanasia-D. Five minutes after injection, the trachea was exposed and an 18-gauge angiocath inserted. Whole-lung lavage was performed with five 10-ml aliquots of HBSS containing 0.1 g EDTA and 5 ml penicillin-streptomycin per liter. For each rat, 50 ml of lavage fluid was collected in a sterile collection tube.

In the following steps, all centrifugation was at 1,200 × g for 5 min in a Beckman CPRK centrifuge (Beckman Instruments, Inc., Palo Alto, CA). The lavage fluid was centrifuged and the supernatant discarded. The pellet was resuspended in 11 mM KHCO₃ and 152 mM NH₄Cl to lyse red blood cells, and was then spun again. This was followed by two centrifugation washes in HBSS. The pellet was resuspended in 2 ml RPMI 1640 medium and the cells were counted with a grid cytometer. Typically, 4 to 5 × 10⁶ cells were obtained per rat, 95% of which were AMs (14).

Isolation of Bovine Vitronectin

Bovine Vn was purified by adapting two previously reported procedures (15, 16). Briefly, 1.5 liters of fibronectin-depleted serum was collected from a gelatin sepharose column (4.8 × 30 cm). To remove fibrinogen, the serum was precipitated by adding saturated ammonium sulfate solution to a final saturation of 30%. The precipitate was removed by centrifugation at 10,000 × g for 30 min at 4°C in a Beckman J2-21 centrifuge, using a Beckman JA-10 rotor. Saturated ammonium sulfate solution was added to the supernatant to a final saturation of 50%. The solution was incubated for 18 h at 4°C, centrifuged (10,000 × g for 30 min at 4°C), and the protein pellet was resuspended in 10 mM sodium phosphate (pH 7.7), 5.0 mM EDTA, 0.13 M NaCl, and 0.10 mM PMSF (resuspension buffer).

The resuspended protein was passed through a heparin sepharose column. The heparin column (2.5 × 24 cm) was first washed with 10 mM sodium phosphate (pH 7.7), 5.0 mM EDTA, and 2.0 M NaCl (500 ml), and was then equilibrated in resuspension buffer. The flow-through fraction was collected and urea was added to a final concentration of 8.0 M. The heparin sepharose column was washed with 10 mM sodium phosphate (pH 7.7), 5.0 mM EDTA, 8.0 M urea, and 10 mM -mercaptoethanol (500 ml), followed by equilibration in 10 mM sodium phosphate (pH 7.7), 5.0 mM EDTA, and 8.0 M urea (buffer A). The denatured flow-through fraction was then reapplied to the heparin sepharose column. The column was washed with buffer A plus 0.13 M NaCl (500 ml), and Vn was eluted with buffer A plus 0.50 M NaCl. The A₂₈₀-positive fractions were pooled and dialyzed with 3 × 4.0 liter of 50 mM Tris (pH 7.0) and 0.15 M NaCl. The purity of Vn was verified by polyacrylamide gel electrophoresis (PAGE [17]) and by the Western blot technique (18), using rabbit polyclonal antibody to Vn. The Vn concentration was determined by BCA protein assay (19).

Radioiodination of Vitronectin

Vn was radioiodinated by adding 1 mCi of Na¹²⁵I to prewashed Pierce IODO-BEADS^® and incubating for 5 min at room temperature. Vn (300 µg) in 50 mM Tris (pH 7.0) and 150 mM NaCl was added to the Na¹²⁵I/IODO-BEADS mixture. The mixture was incubated for 30 min at room temperature, after which the free ¹²⁵I was separated from the labeled protein using a 1 × 24 cm Sephadex G-25 desalting column equilibrated with 0.1% BSA/50 mM Hepes. Fractions containing labeled Vn were verified by repeat PAGE and autoradiography. The [¹²⁵I]Vn concentration was determined by BCA protein assay.

Liposome Production

Liposomes were prepared by aqueous reconstitution (20). DPPC (63 µmol), stearylamine (18 µmol), and cholesterol (9 µmol) were dissolved in 2 ml diethyl ether. The ether was evaporated with N₂ in a rotating glass vial. The vial was capped and heated to 60°C in a water bath (Forma Scientific, Inc., Marietta, OH). Swelling solution (150 mM NaCl, 20 mM Hepes, pH 7.4) was also heated to 60°C. One milliliter of swelling solution was added to the vial, followed by vigorous vortexing to create an emulsion. The emulsion was subjected to rapid and repetitive cycles of freezing and thawing by rolling the vial in a 80°C methanol bath (Virtis Corp., Gardiner, NY) and then quickly immersing it in the +60°C water bath, with this process being repeated five times. Thermal cycling enhances liposome formation and solute entrapment (20, 21). The liposome suspension was centrifuged for 5 min at 13,000 × g in a Sorvall 28C centrifuge equipped with a Sorvall SS-34 rotor (E.I. DuPont de Nemours and Co., Wilmington, DE). The pellet was resuspended in 5 ml swelling solution and recentrifuged. This rinsing centrifugation was repeated five times. The final pellet was resuspended in 2 ml RPMI 1640 medium at 37°C.

Variations in the foregoing method were used to make Vn-enriched, fluorescent liposomes that were used in the phagocytosis assays. For the Vn-enriched liposomes, the swelling solution contained 0.01 to 10 µM Vn, as the experiment required, and a 1:2 suspension of fluorescent polystyrene microspheres of 50 nm diameter. The liposome suspensions were centrifugally rinsed as described earlier to remove untrapped material.

Determination of Vn Entrapment in Liposomes

To determine whether Vn-enriched liposomes present Vn on their external surface in addition to entrapping Vn in their interior compartment, liposomes were prepared in the presence of rhodamine-conjugated Vn. To prepare rhodamine-Vn, RHODOS in DMSO (2.3 mg, 10 mg/ml) was added to 270 µg Vn (0.54 mg/ml), incubated with constant agitation for 5 h at room temperature, and dialyzed against PBS (19 mM H₂PO₄, 81 mM HPO₄⁼, 150 mM NaCl, 1 mM CaCl₂, pH 7.5) for 18 h at 4°C (22).

Liposomes were prepared as previously described in the presence of 500 µl rhodamine-Vn and 1.5 ml swelling solution. After vortexing, freezing-thawing, and rinsing as described above, the final liposome suspension was divided into three tubes at a volume of 1 ml/tube. The tube containing the control liposomes received BSA in PBS (final BSA concentration: 1%). The two tubes containing the sample liposomes each received 1:50 rabbit IgG antibody raised against bovine Vn (anti-Vn). All three liposome suspensions were incubated under constant agitation for 45 min at 37°C and then rinsed by centrifugation at 13,000 × g in PBS. The final pellets were resuspended in 2 ml of 1% BSA/PBS. One of the sample liposome suspensions was incubated with 0.1% trypsin in Tris pH 7.5 under constant agitation for 60 min at 37°C, then rinsed by centrifugation as described earlier. All three liposome suspensions were then incubated under constant agitation with 1:200 goat antimouse IgG-FITC for 60 min at 37°C, rinsed by centrifugation as described earlier, and viewed with differential interference contrast (DIC) and fluorescence microscopy, using a Zeiss Axioplan microscope fitted with Nomarski and epifluorescence optics (Carl Zeiss, Inc., Batavia, IL).

Determination of Vn-AM Binding

To assay Vn-AM binding, 2 × 10⁶ rat AMs were incubated with either 0.1, 1, or 2 µM Vn for 1 h at 37°C. Cells were harvested, washed in HBSS, and transferred to a 96-well filter plate, each well containing 100 µl of 50 mM Hepes (pH 7.4) and [¹²⁵I]Vn (2 × 10⁵ cpm/well, specific activity = 3 × 10⁶ cpm/µg). The cell suspension was then filtered with a 0.22-µm polyvinyldifluoride membrane, using the Millipore Multiscreen System (Millipore Corp., Bedford, MA). Unbound [¹²⁵I]Vn was washed away with excess HBSS containing 2 mM CaCl₂, leaving macrophage-bound [¹²⁵I]Vn on the well filters (23). The filter plate was dried, the filters were removed, and activity was measured with a Beckman 5500 gamma counter. AM binding of [¹²⁵I]Vn was expressed as cpm of [¹²⁵I]Vn retained on the filter, minus background counts. The background of [¹²⁵I]Vn binding to the polyvinyldifluoride filter averaged 5% of the total signal.

Immunolocalization of Exogenous Vitronectin on the Alveolar Macrophage Surface

To demonstrate that Vn does adhere to the AM surface even after rinsing, rat AMs were incubated with 1 µM Vn for 18 h at 37°C, rinsed three times with PBS, incubated with 1:40 anti-Vn for 30 min at 37°C, again rinsed three times with PBS, incubated with 1:100 goat antirabbit IgG- FITC for 30 min at 37°C, and then rinsed three more times with PBS and viewed under a fluorescence microscope. As a negative control, a separate group of cells was treated as described, except that no anti-Vn antibody was used. As a positive control to determine that anti-Vn antibody was recognizing the bovine Vn used in our experiments, the anti-Vn antibody was used in a Western blot assay to identify both bovine Vn isolated in our laboratory and commercially available rat Vn.

Determination of Phagocytic Activity

Phagocytosis was quantitated with a fluorometric assay (24). Briefly, rat AMs were placed in 24-well culture plates, each well containing 5 × 10³ cells in 300 µl RPMI 1640, and incubated for 18 h at 37°C under 5% CO₂. The wells were aspirated to remove nonadherent cells, and were filled with 150 µl of excess (> 10⁶) fluorescent liposomes suspended in RPMI 1640. The AMs were incubated for 1 h at 37°C. The wells were aspirated, vigorously rinsed three times with PBS at 4°C to remove nonphagocytosed liposomes, and then filled with 300 µl HBSS containing 685 µM EDTA. Adherent AMs were retrieved from each well with a rubber policeman and transferred into 12 × 75 mm cuvettes containing 1.8 ml HBSS. Fluorescence intensity was measured in a fluorometer (Sequoia-Turner Corp., Mountain View, CA), with _EX = 440 nm and _EM = 515 nm. The engulfed liposome concentration was determined by calibrated linear regression. AMs were counted with a grid cytometer. Phagocytosis was expressed as liposomes/cell.

Effect of Vn on AM Phagocytosis of Liposomes

To determine whether Vn enhances AM phagocytosis of liposomes, rat AMs (500 × 10³ cells/well × 12 wells) were pretreated with or without 1 µM Vn for 18 h at 37°C. AMs were subsequently incubated with excess fluorescent liposomes (> 10⁶ per well), made either with or without 10 µM Vn, for 1 h at 37°C. The cells were aspirated and rinsed three times with PBS to remove uningested liposomes, and phagocytosis was assayed as above.

To determine the effect of Vn concentration on phagocytosis, AMs (500 × 10³ cells/well × 12 wells) were pretreated for 18 h with 0, 0.01, 0.1, or 1 µM Vn. AM groups were subsequently incubated with excess fluorescent liposomes containing concomitant levels of Vn for 1 h at 37°C. After cell rinsing to remove uningested liposomes, phagocytic activity was determined with the fluorometric assay.

Mechanism of Vn Enhancement of Phagocytosis

To investigate the possibility that Vn-enhanced phagocytosis is VnR mediated, rat AMs (500 × 10³ cells/well × 21 wells) were pretreated for 18 h with 10 µM Vn and with 0, 0.02, 0.2, or 2 mM GRGDS (RGD). AMs were subsequently incubated with excess fluorescent liposomes containing 10 µM Vn for 1 h at 37°C. After cell rinsing to remove uningested liposomes, phagocytic activity was determined with the fluorometric assay.

In a related experiment, AMs (500 × 10³ cells/well × 21 wells) were pretreated for 18 h with 10 µM Vn and with 2 mM GPenGRGDSPCA (GPen), a cyclic RGD-containing polypeptide that is specific for VnR, and not only for the RGD-recognition site common to other integrins as well. Positive control AMs received 10 µM Vn but not GPen; negative control AMs received neither Vn nor GPen. All cells were incubated at 37°C for 18 h, and then incubated with fluorescent, Vn-enriched liposomes as described earlier. After 1 h, phagocytosis was assayed.

To establish that apparent RGD inhibition was due specifically to the RGD sequence in GRGDS, and not to a nonspecific, general inhibition by oligopeptides, rat AMs were pretreated with 10 µM Vn alone or with 10 µM Vn and either 20 µM GRGDS or 20 µM GRGESP (RGE) at 37°C for 18 h, and were then rinsed and incubated with the fluorescent, Vn-enriched liposomes and the same concentrations of Vn, RGD, and RGE. After 1 h, phagocytosis was assayed.

To determine the effect on phagocytosis of antibody blockade of VnR, rat AMs (500 × 10³ cells/well × nine wells) were pretreated with or without 10 µM Vn for 18 h at 37°C, then thoroughly rinsed three times in PBS. The AMs were then pretreated for 1 h at 37°C with or without a 1:200 dilution of anti-_v3-Vn-receptor IgG_2A (anti-VnR), an antibody to _v3 integrin, the major VnR expressed by AMs (12). Following the antibody pretreatment, AMs were incubated for 1 h at 37°C with excess fluorescent liposomes enriched with Vn; control AMs were incubated with excess fluorescent liposomes made without Vn. Phagocytosis was determined as described earlier.

Statistical Analysis

One-way analysis of variance (ANOVA) was applied to the concentration-related experiments (concentration dependence of Vn enhancement, RGD competition) and to the VnR blockade and RGE experiments (25). Two-way ANOVA was applied to the double-variable experiment (incubation/Vn). A significant difference between groups was declared for P  0.05.

	Results

Top Abstract Introduction Materials & Methods Results Discussion References

Determination of Vn Entrapment in Liposomes

When viewed with fluorescence microscopy, liposomes were strongly labeled with rhodamine-Vn (Figures 1B and 1E). The presence of Vn on the external surface of liposomes was detected with anti-Vn antibody (Figure 1C). Since liposome exposure to anti-Vn antibody occurred after the liposomes were formed, the antibody recognized epitopes of Vn that either extended through the membrane of the liposomes and/or adhered to the liposomes after they were formed. As confirmation of this, trypsinization, which also occurred after liposome formation, abolished binding of anti-Vn antibody (Figure 1F). In contrast, the Vn internalized within the liposomes was protected from trypsinization, and therefore gave an equally strong signal for rhodamine fluorescence in both the trypsinized and nontrypsinized liposomes (Figures 1B and 1E). The control liposomes were negative for nonspecific labeling with the second antibody (not shown). These results indicate that when entrapped during liposome formation, Vn is both internalized within the liposome and exposed on the external surface of the liposome membrane, presumably by random incorporation (26). Thus, Vn-enriched fluorescent liposomes are useful as probes for studying Vn-enhanced phagocytosis.

View larger version (106K): [in this window] [in a new window]   Figure 1.   Vn is internalized within liposomes and also spans the external membrane of the liposome. Liposomes were prepared in the presence of rhodamine-conjugated Vn, rinsed, incubated without (A, B, C) or with (D, E, F) 0.1% trypsin, rinsed, incubated with 1:50 mouse anti-Vn IgG, rinsed, incubated with 1:200 goat antimouse IgG-FITC, and rinsed again. The nontrypsinized liposome (A) is labeled both for rhodamine-conjugated Vn (B) and for anti-Vn antibody (C); the trypsinized liposome (D) is also labeled for rhodamine-conjugated Vn (E), but not for the externally applied anti-Vn antibody (F). A, D: differential interference contrast images; B, C, E, F  : fluorescence images. Magnification: ×125.

Determination of Vn-AM Binding

AMs that were pretreated with Vn subsequently demonstrated increased binding to [¹²⁵I]Vn (Figure 2). Moreover, this binding activity was Vn concentration-dependent: gamma counts increased from 23,622 ± 3,328 cpm, 40,847 ± 6,530 cpm, and 57,149 ± 2,789 cpm to 124,852 ± 42,930 cpm for AMs pretreated with 0, 0.1, 1, and 2 µM Vn, respectively (P < 0.05). These findings demonstrate that AM binding to Vn increases in direct proportion to increased prior exposure to Vn.

View larger version (9K): [in this window] [in a new window]   Figure 2.   Incubation of AMs with Vn enhances AM-Vn binding in a concentration-dependent manner. AMs were pretreated with 0, 0.1, 1, or 2 µM Vn at 37°C for 18 h, rinsed, incubated with 8.9 nM [125I]Vn at room temperature for 30 min, and then filtered on a 0.22-µm filter plate. The filter plate was rinsed to remove unbound [125I]Vn and then dried. The filters were removed and activity measured in a gamma counter. The amount of AM binding to [125I]Vn was directly proportional to the amount of AM pretreatment with Vn. *P < 0.05.

Immunolocalization of Vn on AM surface

AMs retained Vn on their surface even after being incubated with 1 µM Vn for 18 h and rinsed three times in PBS. AMs showed labeling for Vn over the cell surface (Figures 3A and3B); control cells were unlabeled (Figures 3C and 3D), indicating that the labeling seen in Figure 3B was not caused by nonspecific adherence of IgG-FITC to Vn. In a separate experiment done to assure that the labeling seen in Figure 3 was due to adherent Vn (positive control), anti-Vn antibody recognized both the bovine Vn isolated in our laboratory and bovine Vn that was obtained commercially, as demonstrated by Western blotting (data not shown). These findings demonstrate that AM pretreatment with Vn results in persistent Vn attachment to the cell surface, making Vn available for subsequent interaction with Vn-enriched liposomes.

View larger version (66K): [in this window] [in a new window]   Figure 3.   Incubation of AMs with Vn results in Vn attachment to AM surface. AMs were incubated with 1 µM Vn for 18 h at 37°C, rinsed, incubated with (A, B) or without (C, D) 1:40 rabbit anti-Vn polyclonal antibody, rinsed, incubated with 1:100 goat antirabbit IgG-FITC, and rinsed again. Pretreated cells (A) were well labeled for Vn even after rinsing three times in PBS (B). The fluorescent signal was not caused by nonspecific adherence of IgG- FITC to Vn, as evidenced by the control cells (C), which also were pretreated with Vn but were then exposed to IgG-FITC only; no fluorescence is seen (D). Magnification: ×50.

Effect of Vn on AM Phagocytosis of Liposomes

Eighteen hours of AM incubation with Vn significantly enhanced phagocytosis of Vn-enriched liposomes (Figure 4). AMs pretreated with Vn ingested significantly more Vn-enriched liposomes (20.7 ± 0.4 liposomes/cell) than AMs not pretreated with Vn but incubated with either empty or Vn-enriched liposomes (12.6 ± 0.7 and 11.5 ± 0.5 liposomes/cell, respectively; P < 0.01). This indicates that the presence of Vn on the surface of liposomes is necessary but not sufficient for enhancement of phagocytosis; AMs must also be exposed to Vn prior to the phagocytic challenge. In addition, pretreatment of AMs with Vn had no effect on AM phagocytosis of empty liposomes (12.6 ± 0.7 versus 12.1 ± 1.3 liposomes/cell) (Figure 4). This indicates that Vn enhancement of phagocytosis requires the precence of Vn during phagocytosis, which suggests that Vn does not simply stimulate phagocytosis through nonspecific activation of AMs.

View larger version (17K): [in this window] [in a new window]   Figure 4.   Pretreatment of AMs with Vn enhances the phagocytosis of Vn-enriched liposomes but not of control liposomes. AMs were pretreted with or without 1 µM Vn for 18 h, rinsed, and incubated with fluorescent liposomes made in the presence or absence of 10 µM Vn. No difference in phagocytic activity was found between untreated AMs that received either empty or Vn-enriched liposomes and pretreated AMs that received empty liposomes. In contrast, AMs pretreated with Vn and which also received Vn-enriched liposomes had a significantly higher level of phagocytic activity. *P < 0.05.

Pretreatment of AMs with increasing concentrations of Vn was associated with a significant increase in phagocytosis of Vn-enriched liposomes (Figure 5). In the presence of 0, 0.01, 0.1, and 1 µM Vn, phagocytosis increased from 3.7 ± 0.3 to 6.5 ± 0.2, 11.5 ± 0.5, and 16.5 ± 0.6 liposomes/ cell, respectively (P < 0.05 at each level). The concentration-dependent response of AMs to Vn strongly suggests that Vn itself is responsible for the enhanced phagocytosis observed for AMs.

View larger version (11K): [in this window] [in a new window]   Figure 5.   Phagocytosis of liposomes by AMs is enhanced by Vn. Fluorescent liposomes were prepared by aqueous reconstitution in the presence of Vn ranging in concentration from 0 to 1 µM and a 1:2 suspension of fluorescein-polystyrene microspheres (50 nm diameter). AMs were incubated with 0 to 1 µM Vn at 37°C for 18 h and then incubated with the fluorescent, Vn-enriched liposomes. Phagocytosis was assayed after 1 h. Phagocytic activity increased in direct proportion to the amount of Vn. *P < 0.05.

Mechanism of Vn Enhancement of Phagocytosis

Vn-enhanced phagocytosis was markedly reduced in the presence of RGD (Figure 6A). Phagocytosis decreased from 8.1 ± 0.4 to 3.4 ± 0.2, 2.4 ± 0.4, and 2.2 ± 0.2 liposomes/ cell in the presence of 0, 0.02, 0.2, and 2 mM RGD, respectively (P < 0.05). This reduction was concentration-dependent, suggesting that RGD interferes with Vn-enhanced phagocytosis through competitive inhibition. Furthermore, the inhibition of Vn-enhanced phagocytosis is specific for the VnR (Figure 6B). Positive control AMs demonstrated typical Vn-enhanced phagocytosis; (10.9 ± 1.5 liposomes/ cell); phagocytic activity of AMs treated with GPen (4.7 ± 0.8 liposomes/cell) was the same as that of negative control AMs (4.4 ± 0.2 liposomes/cell). The difference between the first and the latter two groups was significant at P < 0.05. Since GPen acts specifically to inhibit Vn-VnR binding (27), this result shows that Vn-enhanced phagocytosis depends on initial binding of Vn to VnR, and that Vn enhancement of phagocytosis is mediated exclusively by the AM VnR. To rule out the possibility that oligopeptides could cause a general, nonspecific inhibition of phagocytosis, GRGESP (RGE) was used as a control (Figure 6C). Cells exposed to Vn and RGE had a similar level of phagocytosis as control cells exposed to Vn only; in contrast, cells exposed to Vn and RGD had markedly attenuated phagocytosis (14.4 ± 3.3, 10.9 ± 1.0, and 4.6 ± 2.1 liposomes/cell, respectively; P < 0.05 for RGD cells compared with either control or RGE cells).

View larger version (20K): [in this window] [in a new window]   Figure 6.   RGD-containing oligopeptides competitively inhibit Vn-enhanced AM phagocytosis of liposomes. Fluorescent liposomes were prepared in the presence of 10 µM Vn. AMs were pretreated with 10 µM Vn and 0, 0.02, 0.2, or 2 mM RGD, 0.02 mM RGE, or 0.02 mM GPen at 37°C for 18 h, and then incubated with the fluorescent, Vn-enriched liposomes. After 1 h, phagocytosis was assayed. At all three concentrations of RGD, phagocytic activity was significantly less than that of control AMs not pretreated with RGD (A: *P < 0.05). In addition, the level of phagocytic activity was inversely proportional to the concentration of RGD. GPen, which is specific for Vn- VnR binding, similarly inhibited phagocytosis (B: *P < 0.05). Conversely, RGE, an oligopeptide that does not bind to the VnR recognition site, does not inhibit phagocytosis (C).

As further evidence of the role of VnR in Vn-enhanced phagocytosis, AM pretreatment with an antibody directed against the _v3 VnR (anti-VnR) markedly attenuated phagocytosis of Vn-enriched liposomes (Figure 7). Phagocytosis decreased from 8.9 ± 1.7 to 3.3 ± 0.4 liposomes/cell in the presence of anti-_v3 antibody (P < 0.01). These results strongly suggest that Vn-enhanced phagocytosis is receptor mediated, and that the receptor is either _v3 or a VnR closely related to it.

View larger version (15K): [in this window] [in a new window]   Figure 7.   Anti-Vn receptor antibody blocks Vn-enhanced phagocytosis. AMs were pretreated with or without 10 µM Vn for 18 h at 37°C and then thoroughly rinsed three times with PBS. AMs then were or were not pretreated for 1 h at 37°C with 1:200 polyclonal antibody directed against the v3-Vn receptor. After this pretreatment, AMs were incubated for 1 h at 37°C with excess fluorescent liposomes made in the presence or absence of 10 µM Vn. AMs that were pretreated with Vn and received Vn-enriched liposomes had significantly greater phagocytic activity than control AMs that were not pretreated and received empty liposomes. *P < 0.05. AMs that were pretreated with Vn and received Vn-enriched liposomes, but were also exposed to anti-Vn receptor antibody, had a level of phagocytic activity similar to that of the untreated control AMs.

	Discussion

Top Abstract Introduction Materials & Methods Results Discussion References

The experiments done in the study demonstrate that: (1) Vn pretreatment of AMs enhances Vn-AM binding; (2) Vn adheres to the AM surface even after rinsing; (3) Vn enhances AM phagocytosis of Vn-enriched liposomes; and (4) this enhancement requires AM pretreatment with Vn, is directly proportional to Vn concentration, is competitively inhibited by RGD-containing oligopeptides, including an oligopeptide specific for the Vn recognition site on VnR, and is blocked by anti-VnR antibody.

The Vn-AM binding data show that the amount of AM binding to Vn is directly proportional to exposure to Vn in a dose-dependent manner. Similarly, the concentration dependence of Vn on phagocytic enhancement suggests that this enhancement is directly mediated by Vn. In addition, the adherence of Vn to AMs even after rinsing suggests that Vn mediates phagocytosis through interaction at the cell surface. Since pretreatment of AMs with Vn did not increase the phagocytosis of empty liposomes, the enhancement cannot be due to a nonspecific activation of AMs by Vn. Moreover, since the amount of liposomes presented to the AMs remained constant at all levels of Vn, the enhancement cannot be due to nonspecific AM activation by a factor intrinsic to the liposomes themselves. When purifying plasma proteins such as Vn, the possibility of endotoxin (lipopolysaccharide [LPS]) contamination exists. LPS is a well-documented activator of AMs that can increase AM phagocytic activity (28). To rule out the possibility that endotoxin contamination of the Vn, rather than Vn itself, was increasing phagocytosis, we repeated the Vn dose-response experiment in the presence of polymyxin B. The results were the same as those seen without polymyxin B (data not shown).

The predominant VnR expressed by AMs is a ₃ integrin, _v3 (12). Integrins are surface receptors that recognize a number of proteins involved in hemostasis, cell adhesion, and cell migration (29). Although structurally and genetically diverse, all known ligands that bind ₃ integrins (including _v3) contain sequences of Arg-Gly-Asp (RGD) (32, 33). The results of the RGD competition experiment demonstrate that Vn enhancement of phagocytosis is mediated through a receptor that has an RGD binding domain. In addition, the cyclic polypeptide GPen is specific for Vn-VnR binding (27), and this inhibits Vn-induced enhancement of phagocytosis to the control level. These findings strongly suggest that Vn-enhanced phagocytosis is VnR mediated. The blockade of Vn enhancement of phagocytosis by an anti-VnR antibody further supports this hypothesis.

The enhancement of phagocytosis of Vn-enriched liposomes requires pretreatment of AMs with Vn prior to the phagocytic challenge. This finding could suggest that VnR is being upregulated after exposure to Vn, in the manner of positive feedback. However, immunoprecipitation with anti-VnR antibody did not show an increase in the surface expression of VnR after AM incubation with increasing concentrations of Vn (data not shown). Alternatively, if VnR is not being upregulated, preincubation might be necessary to allow time for Vn crosslinking to occur. Vn can self-associate into multimers (34). The multimeric form of Vn, which is found in the extracellular matrix, may represent the "activated" form of Vn that is involved in the scavenging of extracellular debris. The exact mechanism of this self-association in not understood. Vn readily adheres to a wide variety of surfaces, including glass, agarose, and a number of plastics (7, 8). This nonfastidious adhesion, probably produced by separate domains of differing charge and hydrophobicity (35), may induce conformational changes in Vn that allow it to undergo self-association. Adhesion-triggered self-assembly would allow Vn to function as a nonimmune opsonin. In keeping with this possibility, there is much circumstantial evidence that Vn participates in nonimmune phagocytosis: Vn is localized at scavenging sites such as the keratin bodies created by senescent keratinocytes (5) and the senile plaques of brain tissue in Alzheimer's disease, which are subject to phagocytosis by microglia (6). Vn enhances the internalization of asbestos by pleural mesothelial cells (36). Macrophages use _v3-VnR in the scavenging of senescent thymocytes, lymphocytes, and neutrophils undergoing apoptosis (10, 37, 38). As is characteristic of VnR binding, the adhesion of neutrophils to macrophages is divalent-cation dependent, and is competitively inhibited by free Vn and by RGD (10). In addition, macrophage phagocytosis of neutrophils is blocked by monoclonal anti-VnR antibodies (4). Fibroblasts in culture have been shown to ingest apoptotic neutrophils via the VnR (39). In light of these findings, the findings in the current study support the concept of Vn acting as a molecular bridge that can link a phagocyte with a nonspecific target for engulfment.