Production of Vascular Endothelial Growth Factor by Murine Macrophages

MPMs and RAW264.7 Cells

Balb c mice (male, 6 to 8 weeks old, Taconic, Germantown, NY) were injected intraperitoneally with 2.5 ml sterile Brewer’s thioglycollate broth (3% w/v) (DIFCO Laboratories, Detroit, MI). Five days later, the mice were sacrificed and MPMs were harvested using phosphate-buffered saline containing 100 U/ml of heparin. Cells were centrifuged at 300 × g for 5 minutes at 4°C, washed twice with serum-free (DMEM), and resuspended in DMEM containing 10% fetal calf serum (FCS) and 50 μg/ml gentamycin (DMEM-10% FCS). Cells were seeded into 60-mm tissue culture dishes (Costar, Cambridge, MA) (4 × 10⁶ cells/dish) and incubated at 37°C in a humidified incubator in 95% air/5% CO₂ for 4 hours to allow the cells to adhere. In some experiments, cells were seeded in Contur Permanox gas-permeable dishes (Miles, Naperville, IL) rather than regular tissue culture dishes, to increase the availability of ambient gases to the cells on the base of the dishes. Nonadherent cells were removed by washing with serum-free DMEM, and the cells were refed with DMEM-1% FCS. MPMs were activated using 100 U/ml murine IFNγ (Sigma Chemical Co., St. Louis, MO) and 100 ng/ml of LPS (Escherichia coli serotype 055:B5, Sigma) either in the presence or absence of either of the iNOS inhibitors, N^g-nitro-l-arginine-methyl ester (l-NAME) (1.5 mmol/L) or aminoguanidine (AG) (1 mmol/L). To test the effects of lactate on MPMs, sodium lactate (25 mmol/L) was added to the cultures at the start of the incubation period. To test the effects of hypoxia, MPMs were incubated in Permanox dishes, either under normoxic conditions (95% air/5% CO₂) or under hypoxic conditions (95% N₂/5% CO₂). Media and cells were harvested at the indicated time points after addition of IFN-γ/LPS and/or lactate. Aliquots of media were sampled immediately after incubation and analyzed in a Blood Gas Analyzer (Instrumentation Laboratories, Lexington, MA). The remaining media were centrifuged at 4°C for 5 minutes at 15,000 × g to remove cellular debris and stored at −80°C before analysis.

RAW264.7 cells were obtained from American Type Culture Collection (Manassas, VA) and routinely maintained in DMEM-10% FCS. Cells were passaged by scraping and plated in either regular or Permanox dishes, with or without IFNγ/LPS, with or without sodium lactate, and under hypoxic conditions, as described above. The effects of l-NAME and AG on the production of VEGF and TNFα by these cells were also tested. Media and cells were harvested and treated as described above.

Isolation of Total Cellular RNA

Total cellular RNA was isolated from macrophage cell cultures using TRI reagent (Molecular Research Center, Inc., Cincinnati, OH). Medium was removed from the cells, TRI reagent was added directly to the culture dishes, and the cell lysate was passed several times through a 21-gauge syringe needle. Samples were stored at room temperature for 5 minutes, 0.2 ml chloroform was then added per milliliter lysis reagent, and the mixture was vortexed for 15 seconds and then incubated at room temperature for 10 minutes. The resultant mixture was centrifuged at 12,000 × g for 15 minutes at 4°C. The aqueous (upper) phase was transferred to a fresh microfuge tube, and RNA was precipitated by adding 0.5 ml isopropanol per 1 ml TRI reagent used for the original extraction. Samples were incubated at room temperature for 5 minutes and then centrifuged at 12,000 × g for 10 minutes at 4°C. The RNA pellets were washed with 75% ethanol, air dried for 5 minutes, and dissolved in RNase-free water.

Quantitative Reverse Transcription-Polymerase Chain Reaction (RT-PCR) Analysis of VEGF mRNA Levels

VEGF mRNA levels were determined by RT-PCR using an internal minigene RNA standard that is present through both the RT and the PCR reaction stages. The 293-bp VEGF minigene RNA standard, containing a 69-bp gene deletion, was prepared as follows. Total RNA from MPMs was subjected to RT and PCR through 35 cycles, using the following primers: sense minigene primer (18-mer) in exon 1 (positions 41 to 58), 5′-GGACCCTGGCTTTACTGC-3′, and antisense minigene primer (39-mer), starting in exon 5, spanning an intron, and continuing into exon 4 to position 387, deleting 69 bp of the gene to position 318, and continuing to position 300. The primer thus spans an intron, and contains a 69 bp deletion: 5′-TTGGTCTGCATTCACATCGGC-GTGATGTTGCTCTCTGAC-3′. The PCR band was purified from primers by ethanol precipitation and blunt end ligated into the pCR-Script AmpSK(+) vector (Stratagene, La Jolla, CA). The orientation of the minigene fragment in the vector was determined by dideoxy sequence analysis. A clone containing the minigene insert in an antisense orientation was used for subsequent in vitro transcription for the preparation of the RNA minigene. The vector was linearized with NotI, treated with proteinase-K (4 μg/ml) for 1 hour at 37°C, and purified by phenol extraction and ethanol precipitation. The linearized plasmid was then transcribed in vitro using a VTRAN-7 transcription kit (Sigma), with T7 RNA polymerase, yielding sense RNA. The reaction product was treated with RNase-free DNase-1 (10 U/mg DNA in the transcription reaction) (Promega, Madison WI) for 2 hours/at 37°C. The reaction mixture was then heated to 90°C for 5 minutes and cooled, and 10 × transcription stop solution (5 mmol/L ammonium acetate and 0.1 mmol/L ethylenediaminetetraacetic acid) was added, followed by phenol extraction and isopropanol precipitation. The RNA concentration was determined spectrophotometrically. VEGF RNA minigene (2.5 pg per reaction) was then incorporated into the RT-PCR reactions. Total RNA from macrophages treated under various conditions was added to the RT-PCR reactions in amounts ranging from 1 to 200 ng/reaction. The oligonucleotide primers used for the competitive RT-PCR reaction were 18-mers nested into the initial primers used to prepare the minigene: sense primer in exon 1, 5′-ACCCTGGCTTTACTGCTG-3′, and antisense primer (intron spanning), 5′-GGTCTGCATTCACATCGG-3′. The antisense primer was used for the initial RT reaction; the reverse transcriptase was inactivated at 99°C for 5 minutes, and added to a PCR mix containing an equivalent amount of sense primer. PCR was then carried out for 25 cycles. The reactions were analyzed by electrophoresis on 1.5% agarose gels in Tris-acetate-ethylenediamine tetraacetic acid (TAE) buffer, stained with ethidium bromide, and scanned using the Molecular Dynamics FluorImage Analyzer. The concentrations of input RNA that gave bands of equal intensity to that of the internal VEGF RNA minigene were then determined. Although intron-spanning primers were used throughout, controls for genomic DNA contamination of total RNA preparations were routinely carried out. These controls involved the performance of parallel reactions in the absence of reverse transcriptase.

As a control for a housekeeping gene that is not markedly modulated by the various culture conditions used, an RT-PCR procedure for the enzyme glyceraldehyde-3-phosphate dehydrogenase (G3PDH) was also developed (details not shown). Parallel reactions for G3PDH mRNA levels were performed on the various macrophage RNA samples, and the VEGF mRNA levels determined by RT-PCR were normalized to the G3PDH levels.

RT-PCR Analysis of VEGF mRNA Isoforms

For RT, 1.0 μg of total RNA was reverse transcribed using 100 ng of the reverse VEGF-specific primer indicated below, using 50 U murine leukemia virus reverse transcriptase with an RNA PCR Kit (Perkin Elmer, Foster City, CA), following the manufacturer’s protocol. After the initial RT reaction step, the 20-μl reaction volumes were boiled for 5 minutes to inactivate the reverse transcriptase. One hundred ng forward primer (see below) was added, together with 80 μl of a PCR master mix, to give a final concentration of 1 mmol/L MgCl₂, 1× PCR buffer II, and 2.5 U Taq polymerase (Perkin Elmer) per reaction. PCR primers were selected to enable the amplification of the three differentially spliced murine isoforms of VEGF mRNA formed from the VEGF gene. These VEGF mRNA isoforms are derived from a gene containing eight exons. 31 The largest, VEGF-1, is formed using all eight exons. VEGF-2 lacks exon 7, and VEGF-3 lacks exons 6 and 7. By using PCR primers in exons 3 and 8, the three different isoforms of VEGF generate PCR amplification products of different sizes, and because they amplify from the same primers, the ratio of intensities of the three bands gives an estimate of the relative abundance of the three differentially spliced mRNA isoforms. The primers selected for the PCR amplifications were: forward primer, located in exon 3, 5′-GATGAAGCCCTGGAGTGC-3′, and reverse primer, located in exon 8, 5′-TCCCAGAAACAACCCTAA-3′.

The following cycling program for PCR was used: Denaturation at 94°C for 1 minute, annealing at 54°C for 1 minute, and extension for 2 minutes at 72°C, for 25 cycles, with a final extension at 72°C for 15 minutes. PCR reactions were then analyzed by electrophoresis on 1.5% agarose gels using TAE buffer, and stained with ethidium bromide. Gels were scanned using a Molecular Dynamics FluorImage analyzer, and the staining intensities of the PCR-amplified VEGF isoform bands were analyzed using the ImageQuant image analysis software package (Molecular Dynamics).

Assay of VEGF Protein Levels by Enzyme-Linked Immunosorbent Assay (ELISA)

VEGF in conditioned media was assayed using a sandwich ELISA kit (Quantikine M, R&D Systems, Minneapolis, MN), following the manufacturer’s protocol. This assay detects murine VEGF with sensitivity in the range of 3 to 500 pg/ml. Samples with VEGF concentrations above this range were diluted with RPMI and reassayed. All samples were assayed in triplicate. Results are presented as means ± standard deviations of the mean (SD).

Assay of TNFα by ELISA

Murine TNFα was assayed using a sandwich ELISA kit (TNF-A Minikit, Endogen, Woburn, MA), following the procedure of the manufacturer. All samples were assayed in triplicate. Results are presented as means ± SD.

Assay of Nitrite

To determine the production of nitric oxide by the cells under the various conditions tested, the media were analyzed for nitrite using the Griess reaction, as described previously. Briefly, 50 μl of culture medium was placed in a 96-well plate, followed by 50 μl of cold 350 mmol/L ammonium chloride, pH 9.6. One hundred μl of a mixture of 1 part 5 mmol/L sulfanilic acid, 1 part 5 mmol/L N-(1-naphthyl)ethylenediamine, and 3 parts glacial acetic acid was added. After 10 minutes of incubation in the dark at room temperature, absorbance at 570 nm was determined using a microplate scanner (BioTek Instruments, Burlington, VT). The system was calibrated using freshly prepared standard nitrite solutions. A linear regression line was determined from the standards, and the experimental nitrite concentrations calculated. Results are means ± SD.

Assay of Angiogenic and Antiangiogenic Activity

Conditioned media from MPM cultures were concentrated 20-fold and diafiltered using Amicon centrifugal spin filters (3-kd cutoff) (Beverly, MA). Five μl of concentrated media was incorporated into equal volumes of slow-release Hydron (12% w/v in 95% ethanol) (Interferon Sciences, New Brunswick, NJ) and allowed to dry. Hydron pellets were implanted aseptically into pockets within rat corneal stromas 2 mm from the limbal vasculature, as described previously. 1,2,4,9 Corneas were examined daily for 7 days using a stereomicroscope and perfused with colloidal carbon at the end of the observation period to provide a permanent record of the angiogenic responses. Corneas were examined histologically for any evidence of nonspecific inflammation. Angiogenic responses were assessed on a graded scale as follows: 0, no response, or slight budding of the limbal vasculature that regresses rapidly; 1, formation of a few capillary buds and sprouts that progress less then 0.2 mm from the limbus and start to regress; 2, persistent growth of a network of capillary buds and sprouts that grow at least 1 mm toward the implant, but do not reach and invade the implant and 3, strong growth of a dense network of capillary buds and sprouts that reaches and surrounds the implant. Four corneal implants were prepared per test sample, and the responses were summed. A maximal response thus has a score of 12, whereas a minimal response has a score of 0. For the assay of antiangiogenic activity, test-conditioned media (20× concentrated) were combined with 20 ng recombinant human VEGF₁₆₅ (gift of Dr. Napoleone Ferrara, Genentech Inc., South San Francisco, CA). The effects of the test media on the angiogenic activity of the rVEGF were then determined using the corneal bioassay.

Effects of Anti-VEGF Antibodies on Macrophage Angiogenic Activity

To determine the contribution of VEGF to the angiogenic activity of the MPM-conditioned media, an affinity-purified neutralizing polyclonal antibody to VEGF (gift of Dr. Napoleone Ferrara) was used. Concentrated conditioned media prepared as described above were incubated with anti-VEGF antibody at a final concentration of 10 μg/ml at 37°C for 2 hours. Controls were incubated with preimmune immunoglobulin at the same concentration. These treated media were then assayed for angiogenic activity in the rat corneal bioassay.

ADP-Ribosylation of rVEGF

Initial attempts to metabolically label VEGF endogenously synthesized in MPMs, using ³²P-labeled nicotinamide adenine dinucleotide ([³²P]NAD⁺) were unsuccessful, as macrophages are impermeable to NAD⁺, which cannot enter the cells and provide a substrate for the cytoplasmic ADP-ribosyl transferases. 32-34 We therefore used either permeabilized MPMs (data not shown) or macrophage cytoplasmic extracts to determine whether exogenous rVEGF is a substrate for macrophage ADP-ribosyl transferases. Similarly, rVEGF was tested as a substrate for cholera toxin (an arginine-specific ADP-ribosyl transferase) and for pertussis toxin (a cysteine-specific ADP-ribosyl transferase). 35,36

Cytosolic extracts of MPMs were prepared as follows: MPMs were plated in 100-mm culture dishes (10 × 10⁶ cells per dish in 10 ml of medium) in RPMI 1640 containing 10% FCS and incubated at 37°C overnight. The medium was then removed, and the cells were washed two times with cold phosphate-buffered saline. The cells were then harvested by scraping into cold phosphate-buffered saline (1 ml/dish). The cells were spun down at 300 × g and resuspended on ice in 20 mmol/L Tris-HCl pH 7.5, 1 mmol/L ethylenediaminetetraacetic acid, 5 mmol/L MgCl₂, 1 mmol/L dithiothreitol DTT, 2 mmol/L mercaptoethanol, 1 mmol/L phenylmethylsulfonyl fluoride, 1 μg/ml leupeptin, 1 μg/ml aprotinin, and 0.25 mol/L sucrose (1 ml/50 × 10⁶ cells) and sonicated briefly. The extract was centrifuged in the cold for 15 minutes at 1100 × g to remove nuclei and insoluble debris. The protein content of the extracts was determined using the Bradford method (BioRad, Richmond, CA), and the extracts were stored at −80°C until use. To determine whether these extracts were able to ADP-ribosylate rVEGF₁₆₅, labeling reactions were set up containing: 500 ng VEGF₁₆₅, 10 μg macrophage protein extract, 20 mmol/L Tris-HCl pH 7.8, 20 mmol/L isoniazid, 120 mmol/L MgCl₂, 10 mmol/L NaF, 0.02% leupeptin, 0.54 mmol/L NAD phosphate, 0.4 mmol/L isobutyl-methylxanthine, 0.1% lubrol, 2 mmol/L dithiothreitol, 10 mmol/L thymidine, and 7 μCi ³²P-labeled NAD⁺ (800 Ci/mmol) (DuPont-NEN, Wilmington, DE). After 2 hours of incubation at 30°C the reaction mixture was placed on ice and precleared for 30 minutes with 10 μl of protein A/G-agarose (Santa Cruz Biotechnology, Santa Cruz, CA). Ten μg of a murine anti-VEGF monoclonal antibody (gift of Texas Biotechnology, Inc., Dallas, TX) was added to the supernatant, and the mixture was incubated on ice for 2 hours. Protein A/G-agarose beads (10 μl) were then added, and the mixture was further incubated for 2 hours at 4°C with gentle rocking. The beads were harvested by centrifugation and washed three times with cell lysis buffer. The beads were then incubated in an equal volume of 2× electrophoresis sample buffer (final concentration of 100 mmol/L dithiothreitol), and heated at 95°C for 10 minutes to elute bound VEGF from the beads. The samples were then separated using 0.1% sodium dodecyl sulfate (SDS) 15% polyacrylamide gel electrophoresis (PAGE), and the fractionated proteins were transferred to a nitrocellulose membrane by semidry electrophoretic transfer. The filters were then immunostained using anti-VEGF antibody, and the VEGF bands were detected using enhanced fluorecence detection reagents (Amersham Vistra reagents, Amersham, Arlington Heights, IL) and a Fluorimage Analyzer (Molecular Dynamics, Sunnyvale, CA). The nitrocellulose blots were then analyzed using a PhosphorImager analyzer (Molecular Dynamics), to determine the localization of ³²P-labeled bands.

rVEGF₁₆₅ was incubated for up to 2 hours at 30°C with cholera toxin as follows: 500 ng rVEGF and 250 μg cholera toxin (A-subunit, Sigma Chemical Co.), in the reaction buffer described above. The reaction was terminated by the addition of an equal volume of cold 10% trichloroacetic acid. The precipitated protein was washed three times with water-saturated chloroform and finally resuspended in an equal volume of 2× PAGE sample buffer, as above. The samples were separated by SDS-PAGE and transferred to a nitrocellulose membrane as described above.

Five hundred ng rVEGF₁₆₅ was incubated for up to 2 hours at 30°C with 25 μg pertussis toxin (Sigma Chemical Co., cat. no. P-0317) in the reaction mixture described above. Pertussis toxin was preactivated by incubation for 30 minutes with 10 mmol/L ATP and 20 mmol/L dithiothreitol before addition to the VEGF reaction mixture. The reaction was terminated and analyzed as described above.

Effects of ADP-Ribosylation on the Angiogenic Activity of VEGF

To determine whether ADP-ribosylation of VEGF modulates its bioactivity as an angiogenic factor rVEGF₁₆₅ was treated as described above with either cholera toxin or macrophage cytosolic extract, but in the presence of unlabeled NAD⁺. To facilitate the recovery of rVEGF from the reaction mixture, rather than using immunoprecipitation for the recovery of VEGF, which requires the use of harsh, denaturing conditions for the recovery of VEGF from the protein A/G-agarose beads, heparin-Sepharose binding was used to recover the VEGF. After the labeling reaction, 10 μl washed heparin-Sepharose beads were added, and the mixture was incubated at 4°C for 4 hours with gentle agitation. The beads were then washed three times with 100 μl 20 mmol/L Tris-HCl pH 7.8 containing 0.4 mmol/L NaCl. VEGF was eluted from the beads by incubation with 20 μl Tris-HCl containing 1.5 mol/L NaCl. Recovery of VEGF was determined by specific ELISA. Control reactions were carried out in the absence of bacterial toxins and macrophage extract. To ensure that antiangiogenic activity was not present in the macrophage extracts or the cholera toxin preparations, similar labeling reactions were carried out in the absence of VEGF, and the heparin-Sepharose eluates from these reactions were tested in the antiangiogenesis assay.

Production of Nitrite by RAW264.7 Cells and MPMs

Figure 1 ▶ shows the production of nitrite by MPMs. Nitrite was not produced by nonactivated cells, either with or without lactate. After challenge with IFNγ/LPS, nitrite production was strongly induced, with nitrite accumulating over the 48-hour incubation period. l-NAME (1.5 mmol/L) blocked nitrite production by about 70 to 80%; AG (1 mmol/L) blocked nitrite production by >95%. RAW264.7 cells produced nitrite in a similar manner, and l-NAME and AG blocked nitrite synthesis by RAW264.7 cells to a similar extent (data not shown).

Production of VEGF by RAW264.7 Cells and MPMs

The production of VEGF by RAW cells is shown in Figure 2A ▶ . Nonstimulated RAW cells produced VEGF in an apparently constitutive manner over the 48-hour incubation period. This spontaneous production of VEGF was similar in regular culture plates and in gas-permeable Permanox plates. Stimulation of cells with IFNγ and LPS increased the production of VEGF by RAW cells over the constitutive level produced by nonstimulated cells by about 3 to 4-fold by 18 hours. By 48 hours, the stimulated VEGF levels were only two-fold increased over the constitutive level. The iNOS inhibitors AG (1.0 mmol/L) and l-NAME (1.5 mmol/L) did not block the constitutive production of VEGF by nonstimulated RAW cells, but reduced the production of VEGF by IFNγ/LPS-activated RAW cells to a level markedly below that of the nonstimulated cells. Sodium lactate (25 mmol/L) did not alter the production of VEGF by these cells, either with or without IFNγ/LPS activation. RAW cells cultured under hypoxic conditions produced increased amounts of VEGF. After 18 hours, VEGF levels in the media of cells cultured under hypoxic conditions were about three-fold greater than those in the media of control, normoxic cells. This differential was less marked by 48 hours. Analyses of the dissolved oxygen levels in the conditioned media directly after harvesting indicated clearly that under normoxic conditions, oxygen levels were consistently high (partial pressure of oxygen > 145). After 24 and 48 hours of incubation under hypoxic conditions (95% N₂/5% CO₂), the partial pressures of oxygen were 71 mm and 46 mm, respectively.

The production of VEGF by MPMs was similar to that of RAW cells, with constitutive production occurring over 48 hours (Figure 2B) ▶ . Increased production was induced by IFNγ/LPS. However, in contrast to RAW cells, iNOS inhibitors did not significantly reduce the production of VEGF by IFNγ/LPS-activated MPMs. As was observed for RAW cells, sodium lactate did not modulate the production of VEGF by these cells. Culture of MPMs under hypoxic conditions resulted in an increase in VEGF production in the first 18 hours; after 48 hours, however, constitutive production of VEGF was only slightly higher than that of hypoxic cells. Oxygen levels determined in the conditioned media of MPMs were similar to those found in RAW cell media.

Quantitative RT-PCR Analysis of VEGF mRNA Levels

A typical example of a quantitative RT-PCR dilution series using the VEGF RNA minigene as internal standard is shown in Figure 3 ▶ . The PCR amplification product of the minigene is 293 bp in size. The native mRNA PCR amplification band is 362 bp in size. The point of equivalence for the amplified minigene and the amplified native mRNA is readily determined from the dilution series. The values determined from these analyses were normalized to the levels of G3PDH mRNA determined in parallel samples, although little variation in the G3PDH mRNA levels were in fact observed between samples. On this basis, the relative amounts of VEGF mRNA in the various macrophage preparations are shown in Table 1 ▶ . Both hypoxia and IFNγ/LPS activation up-regulated VEGF steady-state mRNA levels in MPMs at 4 and 10 hours. By 24 hours, however, the levels of VEGF mRNA were similar in all the groups. In RAW cells, VEGF mRNA levels remained elevated at 24 hours. AG treatment of IFNγ/LPS-treated MPMs did not significantly reduce their steady-state VEGF mRNA levels at any time point; in RAW cells, however, the VEGF mRNA levels were reduced by 70 to 80% at 4, 10, and 24 hours.

RT-PCR Analysis of VEGF mRNA Isoforms

Three isoforms of VEGF were found to be produced by both nonactivated and IFNγ/LPS-activated MPMs. These isoforms corresponded to VEGF-1 (652 bp), VEGF-2 (580 bp), and VEGF-3 (448 bp). 33 The relative proportions of the VEGF isoforms expressed by MPMs at each time point after IFNγ/LPS activation were only slightly modulated by IFNγ/LPS activation and by inhibition of iNOS with AG (Figure 4) ▶ . In RAW cells, VEGF mRNA isoforms were similarly unaffected by IFNγ/LPS activation and by AG treatment.

Production of TNFα by MPMs and RAW264.7 Cells

TNFα was not produced by either nonstimulated MPMs or by RAW264.7 cells over the 48-hour test period. Production of TNFα by MPMs is shown in Figure 5 ▶ . After stimulation with IFNγ/LPS, TNFα expression was strongly induced, with increased TNFα in the conditioned media being apparent by 8 hours after challenge. There was no significant difference in TNFα production in cells treated with or without sodium lactate. Similarly, culture of cells in Permanox dishes, under either normoxic or hypoxic conditions, did not modulate TNFα production. The iNOS inhibitors l-NAME and AG had no significant effect on the production of TNFα by MPMs. Production of TNFα by RAW cells was similar to that observed in MPMs (data not shown).

ADP-Ribosylation of VEGF by Bacterial Toxins and Macrophage Extracts

Labeling of rVEGF with ³²P-labeled NAD was observed using cholera toxin and macrophage cytosolic extracts (Figure 6) ▶ . Labeling with cholera toxin resulted in a single ³²P-labeled band corresponding to the size of the rVEGF₁₆₅ standard (Figure 6C) ▶ . Labeling with macrophage cytosolic extracts resulted in the ³²P labeling of a large number of bands, because of the endogenous labeling of macrophage cytosolic proteins (Figure 6A) ▶ . To clearly demonstrate labeling of rVEGF₁₆₅ in this mixture, immunoprecipitation of the macrophage cytosolic labeling mixture with anti-VEGF antibody was necessary. After immunoprecipitation, a prominent labeled band corresponding to rVEGF₁₆₅ was clearly visible (Figure 6B) ▶ . This band was not present in control reactions carried out in the absence of rVEGF₁₆₅. Labeling of VEGF using pertussis toxin was not observed (Figure 6E) ▶ .

Angiogenic and Antiangiogenic Responses in Rat Corneas

The angiogenic responses induced in rat corneas by the concentrated conditioned media from the MPMs cultured under various conditions are shown in Table 2 ▶ . Medium from nonactivated MPMs cultured under normoxic conditions did not induce angiogenesis. This medium did not contain antiangiogenic activity, as the angiogenic effects of VEGF (25 ng) were unaffected by this medium. Medium from IFNγ/LPS-activated MPMs was potently angiogenic, whereas medium from iNOS-inhibited IFNγ/LPS-activated MPMs showed markedly reduced angiogenic activity. In contrast to medium from normoxic, nonactivated MPMs, this medium was found to contain antiangiogenic activity, as we have reported previously. 37 Medium from normoxic, lactate-treated nonactivated MPMs showed significant angiogenic activity. Similarly, medium from nonactivated MPMs cultured under hypoxic conditions showed significant angiogenic activity. In both of these cases, a polyclonal antibody to VEGF neutralized the angiogenic activity in the conditioned media. Angiogenic responses induced by rVEGF₁₆₅ were neutralized by anti-VEGF antibody in control experiments, whereas those induced by basic fibroblast growth factor (20 ng/implant) and TNFα (20 ng/implant) were unaffected.

The angiogenic responses induced by rVEGF₁₆₅ that was ADP-ribosylated using cholera toxin or MPM cytosolic extract are shown in Table 3 ▶ . Whereas control VEGF (taken through a sham labeling procedure in the absence of cholera toxin and MPM cytoplasmic extracts) strongly induced angiogenesis, both cholera toxin-mediated and MPM cytoplasmic extract-mediated ADP-ribosylated VEGF showed greatly reduced angiogenic responses, indicating that the ADP-ribosylation abrogated the angiogenic activity of the VEGF. Because the VEGF was purified from the reaction mixtures using heparin-Sepharose binding and elution, we also tested eluates from control VEGF-free reactions prepared with cholera toxin or macrophage cytosolic extract, to determine, first, whether these extracts contained angiogenic activity in their own right, and second, whether any antiangiogenic activity might be enriched in the eluates through this procedure and interfere with the angiogenic activity of the VEGF. The eluates were therefore tested alone, and then with the postreaction addition of rVEGF₁₆₅. The sham eluates did not exhibit direct angiogenic activity, nor did they exhibit antiangiogenic activity when combined with VEGF.