Abstract
Vascular endothelial growth factor (VEGF) increases vascular permeability and is important in pleural effusion formation. In studies using transforming growth factor β (TGF- β ) to produce pleurodesis, we observed that although TGF- β was more effective than talc or doxycycline, it induced transient production of large pleural effusions. We hypothesized that TGF- β stimulates VEGF production in pleural tissues in vivo, and by mesothelial cells in vitro. New Zealand White rabbits (n = 33) were given TGF- β2 (1.7 or 5.0 μ g), talc (400 mg/kg), doxycycline (10 mg/kg), or buffer intrapleurally. Pleural fluid was collected at 24 h. Intrapleural injection of TGF- β2 induced a dose-dependent increase in VEGF production. The pleural fluid VEGF level was 2-fold higher in rabbits given 5.0 μ g of TGF- β2 than in those given 1.7 μ g of TGF- β2 and 3-fold higher than in those given buffer. VEGF levels were higher after the injection of TGF- β2 than after administration of talc or doxycycline. The pleural fluid VEGF correlated significantly with the volume of pleural effusions (r = 0.79, p < 0.00001). In vitro, TGF- β2 stimulated a dose-dependent increase in VEGF production from murine pleural mesothelial cells. At 4 and 24 h, TGF- β2, but not talc or doxycycline, induced a significant increase in VEGF, when compared with controls. The mesothelial cell VEGF production was significantly reduced by anti-TGF- β antibody in the TGF- β2, talc, and control (buffer and medium) groups. In conclusion, mesothelial cells are an important source of VEGF. TGF- β increases the VEGF production by mesothelial cells in vivo and in vitro.
Keywords: doxycycline; pleural effusion; talc; transforming growth factor β; vascular endothelial growth factor
Pleural effusion is common in clinical practice and affects more than 3,000 people per million population each year (1). Pleural fluid accumulates as a result of increased pleural fluid formation and/or reduced drainage (2). The pathophysiology of increased pleural fluid formation remains poorly understood.
Current evidence suggests that vascular endothelial growth factor (VEGF) plays an important role in effusion formation (3). VEGF is a cytokine known for its potent ability to induce vascular leakage, and hence formation of effusion and ascites. It has been shown that tumor cells implanted in the pleural (4) or peritoneal cavity (5) in mice secrete VEGF, which increases the permeability of microvessels and results in development of pleural effusion and ascites, respectively. Conversely, blockade of VEGF receptor phosphorylation inhibited the formation of malignant pleural effusions from lung adenocarcinoma in a murine model (6). In humans, pleural fluid VEGF levels were significantly higher in exudates than in transudates, and VEGF receptors were present at high density in the pleura (7, 8), suggesting that VEGF plays a functional role in the development of effusions. However, the factors responsible for the accumulation of VEGF in the pleural space remain largely unknown.
We investigated the effectiveness of transforming growth factor β (TGF-β), a profibrotic cytokine, in producing pleurodesis (9-11). Interestingly, although TGF-β stimulated more effective pleurodesis than talc or doxycycline, the intrapleural administration of TGF-β induced a large volume of pleural fluid in the first few days (9). There is evidence that TGF-β can stimulate VEGF release in some epithelial and cancer cell lines (12-14), although its effect on mesothelial cells has not been studied. Also, we have shown that the levels of both TGF-β1 and -β2 concentrations correlated with the VEGF levels in all common types of pleural effusions in humans (15).
Hence, we hypothesized that TGF-β induces VEGF pleural fluid in vivo and in vitro. The aim of this study is (1) to compare VEGF production in the pleural space in vivo after intrapleural injection of TGF-β as compared with other profibrotic agents, and (2) to assess the effect of TGF-β on VEGF production of pleural mesothelial cells in vitro.
For further detail, see online data supplement.
TGF-β2 and its buffer were obtained from Genzyme (Framingham, MA). Doxycycline (Fujisawa, Deerfield, IL) and sterilized asbestos-free talc (Sigma, St. Louis, MO) were diluted with 0.9% NaCl for the animal studies and with Dulbecco's modified Eagle's medium (DMEM) for cell culture experiments. A murine monoclonal anti-TGF-β antibody (1D11) was used to bind the active forms of TGF-β1, -β2, and -β3 (16).
Intrapleural administration of agents in rabbitsChest tubes were inserted into the right pleural cavities of 33 New Zealand White rabbits (1.5– 2.0 kg), using a method previously described (9, 10). TGF-β2 at 5.0 μg (n = 9 rabbits) or 1.7 μg (n = 9), talc at 400 mg/kg (n = 9), doxycycline at 10 mg/kg (n = 3), or TGF-β2 buffer (n = 3) in 2.5 ml was injected intrapleurally via the chest tube. The chest tube was aspirated after 24 h. Pleural fluid was collected on ice, and the supernatant was stored immediately at −70° C. The volume and the biochemical analysis of the pleural fluid in these rabbits have been included in previous reports (9, 10, 17).
To compare the systemic and pleural fluid VEGF levels, plasma (at baseline and 24 h) and pleural fluid (at 24 h) were collected after intrapleural TGF-β2 (5.0 μg) administration in another five rabbits.
Culture of mesothelial cellsPleural mesothelial cells were obtained from C57BL/6 mice according to the methodology described for rabbits (18). The cells were cultured in 75-cm2 flasks with DMEM, l-glu-tamine, penicillin–streptomycin, and 10% fetal bovine serum (FBS) at 37° C and 5% CO2. The medium was changed the following day, and thereafter every 3–5 d. Cells of Passage 2 were used for experiments. The cells demonstrated typical cobblestone morphology of mesothelial cells and > 96% were cytokeratin positive. Cell culture materials were purchased from Life Technologies (Grand Island, NY) unless otherwise stated.
Pilot study to establish the toxic and optimal doses for VEGF production.TGF-β2 (0.01–100 ng/cm2), talc (1–100 μg/cm2), doxycycline (0.1– 1,000 μg/cm2) in half-log increments, the TGF-β buffer, and saline were applied to mesothelial cells on six-well plates (n = 2 for each dose) for 4, 8, and 24 h. The culture media were then aspirated for VEGF quantification. Cell viability was tested by trypan blue exclusion and the number of cells in each well was measured with a hemacytometer.
Full study.TGF-β2 (0.1 ng/cm2), talc (10 μg/cm2), doxycycline (3 μg/ cm2), buffer, or medium alone was applied to mesothelial cells plated in 24-well plates. Doses were determined from the pilot study. Before the study, the medium was changed to serum-free DMEM to remove any TGF-β in the FBS. The supernatant was collected at 4 h, and stored at −70° C. The cells in each well were then lysed with 500 μl of a lysis agent containing 0.5% sodium dodecyl sulfate (SDS). The protein in the lysate was measured by a bicinchoninic acid (BCA) assay (Pierce Chemical, Rockford, IL), and represented the amount of cells in each well. Cytokine measurements were normalized to the protein concentration to ensure that the differences were not due to variation of cell numbers in individual wells. In the in vitro study, VEGF levels were expressed as the fold increase over the medium control group.
The experiment was performed similarly for 8-h and 24-h time points. The full study was repeated for a total of 8 wells (for the 4- and 8-h studies) and 10 wells (for the 24-h studies) for each agent at each time point.
Dose–response of TGF-β2 –induced VEGF production.To establish the dose response of VEGF production from TGF-β2, mesothelial cells (five wells in each group) were stimulated with TGF-β2 at 0.04, 0.1, and 0.4 ng/cm2 for 24 h and the supernatant was collected for VEGF measurement. The cell experiment was performed in the same manner as in the full study.
Anti-TGF-β antibody study.TGF-β2, talc, doxycycline, buffer, and medium alone were each applied to 10 wells of mesothelial cells, using the same protocol and doses as described above. Anti-TGF-β antibodies (47 μg/ml) were added to half the wells of each group 5 min before the different reagents were applied. The supernatants were collected at 24 h for cytokine measurements.
To ensure that the inhibitory effect seen was not a result of nonspecific inhibition, mouse gammaglobulin (Sigma) was used as a control antibody for comparison. Mesothelial cells were treated with TGF-β2, TGF-β2 with anti-TGF-β antibodies (47 μg/ml), TGF-β2 with mouse IgG (47 μg/ml), medium with mouse IgG, and medium only (five wells in each group) for 24 h, and VEGF was measured in the supernatant.
Cytokine concentrations were determined with enzyme-linked immunosorbent assay kits (R&D, Minneapolis, MN) and normalized to protein. In the case of TGF-β, all samples were acidified to convert all TGF-β present to the immunoreactive form for measurement, according to the manufacturer's instructions.
Data are expressed as means ± standard error. The differences among groups were compared by one-way ANOVA (parametric) or one-way ANOVA on-ranks (nonparametric). Multiple comparisons were performed by the Tukey (pairwise comparison) or Dunnett (versus control) method. The Student t test was used to compare the VEGF levels with or without anti-TGF-β antibodies. The Pearson correlation was used for linear regression analysis. A value of p < 0.05 was considered significant. Data were analyzed with the SigmaStat version 2.03 program (Jandel Scientific, San Rafael, CA).
Intrapleural injection of TGF-β2 induced a dose-dependent increase in pleural fluid VEGF concentrations (Figure 1A). The pleural fluid VEGF level was significantly higher in rabbits that received 5.0 μg of TGF-β2 than in those given 1.7 μg, which in turn had higher pleural fluid VEGF levels than rabbits receiving buffer alone (3,195 ± 275 versus 1,512 ± 129 versus 1,135 ± 215 pg/ml, respectively), p < 0.05. The pleural fluid VEGF level induced after intrapleural injection of 5.0 μg of TGF-β2 (3,195 ± 275 pg/ml) was also significantly higher than the VEGF levels after the injection of talc at 400 mg/kg (748 ± 98 pg/ml), and doxycycline at 10 mg/kg (1,127 ± 77 pg/ml) (Figure 1A).
Fig. 1. (A) The VEGF levels in pleural fluids induced after the intrapleural injection of different pleurodesing agents and the control buffer in rabbits. The VEGF levels were higher in both the TGF-β2 5.0 and 1.7 μg groups when compared with its buffer controls (p < 0.05 for both, one-way ANOVA). *p < 0.05 when compared with TGF-β2 5 μg. (B) Plasma and pleural fluid VEGF levels in rabbits given intrapleural TGF-β2 (5.0 μg) injections. *p < 0.01 versus pleural fluid VEGF.
[More] [Minimize]To establish whether the VEGF was produced locally within the pleural space or was a result of filtration from the systemic circulation, plasma (at baseline and at 24 h) and pleural fluid (at 24 h) were collected after intrapleural injection of 5.0 μg of TGF-β2 in another five rabbits. There was no difference in the plasma VEGF levels at baseline and at 24 h (37.2 ± 1.6 and 38.0 ± 2.1 pg/ml, respectively) after intrapleural TGF-β2 injections. Pleural fluid VEGF (2,795 ± 272 pg/ml) was about 70-fold higher than the plasma VEGF level at either baseline or 24 h (p < 0.01 for both comparisons) (Figure 1B).
The higher VEGF production induced by TGF-β2 relative to other agents paralleled the development of a significantly larger amount of pleural fluid induced 24 h after the injection of TGF-β2 (Figure 2). The pleural fluid VEGF concentration was strongly correlated with the volume of pleural fluid induced after the injection of different agents (r = 0.79, p < 0.000001) (Figure 3). The pleural fluid VEGF levels correlated inversely with lactate dehydrogenase (LDH) or leukocyte concentrations in the effusion (p < 0.00001 for both). There was no significant correlation between the pleural fluid VEGF concentration and the effusion protein levels (Table 1).
Fig. 2. The volume of pleural fluids induced in rabbits 24 h after the intrapleural injection of different pleurodesing agents and the control buffer. The volume of the effusions were higher in both the TGF-β2 5.0 and 1.7 μg groups than the buffer controls (p < 0.05, one-way ANOVA), when the TGF-β2 treatment groups were compared with the buffer group. *p < 0.05 when compared with TGF-β2 5 μg.
[More] [Minimize]
Fig. 3. Correlation of the pleural fluid VEGF concentrations and the volume of pleural fluid induced in rabbits at 24 h. Triangles, TGF-β2; circles, other agents.
[More] [Minimize]| Correlation with Pleural Fluid VEGF Level* | p Value | |||
|---|---|---|---|---|
| Protein | −0.03 | NS | ||
| LDH | −0.84 | < 0.00001 | ||
| Total leukocyte count | −0.71 | < 0.00001 |
To assess the ability of mesothelial cells to produce VEGF, primary cultures of murine pleural mesothelial cells were harvested. The cells were stimulated with TGF-β2 and compared with the other profibrotic agents. The pilot study demonstrated that pleural mesothelial cells were capable of producing VEGF. The lethal doses of these agents on the pleural mesothelial cells are shown in Figures 4A–4C. Interestingly, even at extremely high doses of TGF-β2, none of the cells took up the trypan blue dye. However, at concentrations > 1 ng/cm2, the cells demonstrated obvious morphologic changes with severe frag-mentation and in most cells only a residual nucleus was seen. This morphologic change corresponded with a reduction in VEGF protein production, confirming cell injury. Cells cultured with 0.9% saline, TGF-β buffer, or culture medium alone all had ⩾ 95% survival at 24 h.
Fig. 4. Percentage survival of murine pleural mesothelial cells at 24 h after exposure to (A) TGF-β2, (B) talc, and (C ) doxycycline as measured by trypan blue dye exclusion method. The arrow in A indicates the dosage at which significant morphologic changes were observed.
On the basis of the results of the survival studies, the following doses were chosen for each reagent: TGF-β2, 0.1 ng/cm2; talc, 10 μg/cm2; doxycycline, 3 μg/cm2 for the full study of VEGF production. In general, the optimal dose was about 1-log dose lower than the lethal concentration. TGF-β2 stimulated significantly higher VEGF production from pleural mesothelial cells in vitro than did talc, doxycycline, or the controls (buffer or culture medium only) as early as 4 h (Figure 5A). The levels of VEGF at each group increased over time, but the same pattern was observed at 8 and 24 h (Figure 5B), with the TGF-β2 group inducing the highest amount of VEGF. Results of the preliminary study and the two full studies were highly consistent.
Fig. 5. VEGF production by murine pleural mesothelial cells at (A) 4 h and (B) 24 h. The results represent data from two experiments of a combined total of 8–10 wells in each group. At both 4 and 24 h, TGF-β2 induced significantly higher VEGF production than its buffer controls (p < 0.05, Student t test). VEGF levels in the talc and doxycycline groups were lower than their medium control (p < 0.05, one-way ANOVA) at 4 h, but no differences were observed at 24 h (p = NS, one-way ANOVA). (A) *p < 0.05 when compared with TGF-β2; (B ) *p < 0.05 versus TGF-β2.
[More] [Minimize]To further confirm the effect of TGF-β2 on VEGF production, a dose–response study was performed by applying TGF-β2 at half log increments (0.04, 0.1, 0.4, 1, and 4 ng/cm2) to murine pleural mesothelial cells. The concentration of VEGF in the supernatant increased in a dose-dependent manner (Figure 6). At concentrations higher than 0.4 ng/cm2, VEGF production reached a plateau, and then decreased with further increases in TGF-β2 concentrations (not shown), mirroring the results observed in the survival study.
Fig. 6. Dose-dependent increase in VEGF following TGF-β2 stimulation in murine pleural mesothelial cells. *p < 0.001 versus medium.
[More] [Minimize]The concentration of TGF-β2 was below the detectable range (50 pg/ml) in all groups at all time points, including the group that received TGF-β2 (0.1 ng/cm2 or 2 ng/ml) at 4 and 24 h. TGF-β1 levels were measurable in all groups, but did not differ significantly at 4 h. At 24 h, a significantly higher TGF-β1 level was present in the TGF-β2 group when compared with the talc or doxycycline group, p < 0.05 (Table 2).
| TGF-β1(fold increase over the TGF-β1 levels of medium group at 4 h) | ||
|---|---|---|
| Cells treated for 4 h with: | ||
| TGF-β2 | 1.42 ± 0.14 | |
| Talc | 1.05 ± 0.13 | |
| Doxycycline | 1.28 ± 0.15 | |
| Buffer | 1.55 ± 0.08 | |
| Culture medium | 1.00 ± 0.06 | |
| Cells treated for 24 h with: | ||
| TGF-β2 | 2.44 ± 0.15 | |
| Talc | *1.72 ± 0.14 | |
| Doxycycline | *1.75 ± 0.11 | |
| Buffer | 2.20 ± 0.15 | |
| Culture medium | 2.15 ± 0.17 |
VEGF production from the mesothelial cells was significantly reduced when TGF-β2 was administered with anti-TGF-β antibodies, but not with murine gammaglobulin (Figure 7). The VEGF concentrations were reduced in all groups with the coapplication of anti-TGF-β antibody (Figure 8).
Fig. 7. Murine pleural mesothelial cell production of VEGF in the presence of control antibodies (murine IgG) and anti-TGF-β antibodies. *p < 0.001 versus medium; #p < 0.001 versus TGF-β2.
[More] [Minimize]
Fig. 8. VEGF production at 24 h by murine pleural mesothelial cells in the presence of anti-TGF-β antibodies. Solid bars, without anti-TFG Ab; open bars, with anti-TGF Ab. *p < 0.05; **p < 0.005; ***p < 0.001.
[More] [Minimize]The present study demonstrated that direct intrapleural administration of TGF-β2 induced the accumulation of VEGF in pleural fluids in a dose-dependent fashion in vivo. This increase in VEGF production was associated with the development of large pleural effusions. This study also demonstrated that pleural mesothelial cells are capable of producing VEGF, and mesothelial cell production of VEGF increases in a dose-dependent fashion after stimulation by TGF-β2. To our knowledge, this is the first study to demonstrate the in vivo induction of VEGF by TGF-β, and the first to show that mesothelial cells plays a role in VEGF production in the pleural space.
Pleural effusion is a common clinical problem and affects 0.32% of the general population each year (1). Pleural effusion develops when the rate of pleural fluid formation exceeds its rate of absorption (2). Evidence suggests that VEGF plays a critical role in the formation of pleural effusion and ascites (3, 19). VEGF is a potent inducer of vascular hyperpermeability, and can also stimulate vasodilatation and induce fenestrations in endothelial cells in vitro and in vivo (3). These properties contribute to a leakage of plasma and proteins from the vascular space. In animal studies, tumor cells implanted by intrathoracic (4) or intraperitoneal (5) injections produced VEGF and resulted in fluid accumulation. Deletion of the VEGF gene in tumor cells before implantation significantly reduced its expression and effusion formation (4). Likewise, inhibition of VEGF receptor phosphorylation decreased the formation of malignant pleural effusion in mice (6). In humans, VEGF was present in significantly higher concentrations in exudative effusions. Also, VEGF (Fms-like tyrosine kinase 1) receptors are present at high densities in human pleural mesothelial cells, implying an active biologic role for VEGF in the pleural space (7).
Little, however, is known about the factors that induce VEGF production or the principal source of VEGF in the pleural space. The present study showed that TGF-β can stimulate VEGF accumulation in the pleura, which is associated with the development of large pleural effusions in rabbits. TGF-β is a ubiquitous multifunctional cytokine with potent profibrotic activities (20) and, given intrapleurally, could induce more effective pleurodesis than talc and doxycycline (9). However, we previously observed that whereas TGF-β induces more pleural fibrosis, initially it is associated with the production of larger amounts of pleural fluids compared with talc or doxycycline (9). TGF-β itself does not affect vascular permeability (21). The present study shows that TGF-β2 stimulates significantly more VEGF production than does talc or doxycycline, which can explain our previous observation.
It is interesting that whereas TGF-β stimulated the accumulation of VEGF and large pleural effusions, it was also effective in inducing pleurodesis. TGF-β is a potent profibrotic agent and we have shown that its intrapleural administration can stimulate significant pleural adhesions even within 24 h (10). By Day 4, most rabbits have developed partial symphysis between the visceral and parietal pleurae. We believe that this rapid induction of pleural fibrosis allows early obliteration of the pleural space that inhibits continual accumulation of pleural fluids.
What is the source of the pleural fluid VEGF induced after intrapleural injections of TGF-β? The accumulation of VEGF in the pleural or peritoneal space is believed to be a result of local production rather than passive diffusion from the systemic circulation (22, 23). This is further supported by the observation that in humans the VEGF levels in malignant effusions were up to 10 times higher than in the corresponding serum. In our study, the VEGF concentration in the pleural fluids was about 70-fold higher than the corresponding plasma VEGF levels, strongly suggesting that the large amount of VEGF in the effusion originated predominantly from the pleural space. Previous investigations, however, have focused on the role of infiltrating cells, such as malignant cells and inflammatory cells (5), as the source of VEGF in the pleural fluids. Our study demonstrated that resident pleural mesothelial cells express VEGF basally and, more importantly, that this expression is upregulated by TGF-β. In humans, the pleural space is lined by an extensive monolayer of mesothelial cells with an estimated area of 2,000 cm2 (24). Although infiltrating cells (such as monocytes [5] and macrophages [25]) may contribute to the production of VEGF, we believe that the mesothelial cells represent an important source of VEGF in the pleura.
In this study, we used primary cultures of murine pleural mesothelial cells and showed that they can be harvested and cultured to a high degree of purity. Most studies of murine mesothelial cells employed peritoneal mesothelial cells (26– 28), but it is unknown whether results with peritoneal mesothelial cells can be extrapolated to pleural mesothelial cells. We also established the toxic doses for TGF-β, talc, and doxycycline in mesothelial cells. The lethal dose of talc to murine mesothelial cells was similar to the published data for human pleural mesothelial cells and mesothelioma cells (29), suggesting that the behavior of murine pleural mesothelial cells is similar to that of human mesothelial cells.
This is the first study to show VEGF production from mesothelial cells. A significant rise in VEGF was seen by 4 h after application of TGF-β2, and the VEGF levels in the TGF-β2 group remained higher than in the other groups at 24 h. This is consistent with the results of Chua and coworkers (30) who demonstrated that in osteoblasts, TGF-β1 stimulates an increase in VEGF mRNA expression within 1 h, which peaks by 2 h. Likewise, TGF-β–induced VEGF mRNA expression peaked at 4–8 h and subsided to background levels afterward in AKR-2B (fibroblast) and A549 (lung adenocarcinoma) cell lines (31).
In humans, there are three isoforms of TGF-β: β1, β2, and β3. Although occasional functional differences have been reported, it is generally believed that TGF-β1 and -β2 have similar biologic functions (32, 33). TGF-β1 have been shown to induce VEGF in cultured osteoblasts (30), fibroblasts (34), and tumor cell lines (13, 14). In the only other study using TGF-β2, it was shown to exert the same effect as TGF-β1 on the induction of VEGF production from human glioma cells (13). In our previous study of human effusions, TGF-β1 and -β2 were both correlated with VEGF levels (15). Hence, we believe that the effect of TGF-β1 on mesothelial VEGF production would be similar to what we have shown with TGF-β2.
Although we have shown that the administration of TGF-β to mesothelial cells increased VEGF concentrations, the exact mechanism of induction and the intracellular signaling pathways will require further investigations. On the basis of the effect of TGF-β on other cell lines, we speculate that TGF-β upregulates VEGF gene transcription (30, 31), although other mechanisms that can lead to an increase in VEGF levels, such as release of intracellular stores, cannot be excluded. Also, VEGF consists of several isoforms with biologic differences (35). The differential expression of these isoforms by mesothelial cells after stimulation by TGF-β warrants further studies.
The TGF-β2 (2,000 pg/ml) applied to pleural mesothelial cells was cleared rapidly. By 4 h, the TGF-β2 levels were below the detectable range of 50 pg/ml. In vivo, TGF-β is known to have an extremely short half-life of < 5 min, and is removed by binding to α2-macroglobulins and by hepatic clearance (33). On the other hand, in mesothelial cells the application of TGF-β2 stimulated the production of TGF-β1, which increased with time. This would be in keeping with the previously described autoinduction characteristics—the ability to stimulate its own production—of TGF-β (36, 37).
TGF-β exists predominantly in an inactive form through binding to a latent protein, and the active molecule is released only on activation (32). The active form is usually present in minute quantities and this, together with the short half-life, makes it difficult to measure. Hence, most studies, including ours, measure the total amount of TGF-β. The total amount of TGF-β1 was only slightly higher in the TGF-β group, when compared with other treatment and control groups (Table 2). Such data are difficult to interpret, as the portion of activated TGF-β in each group is not known. For that reason, we applied an anti-TGF antibody, which binds to all active TGF-β isoforms, but not the latent protein (16), in the cell cultures and showed that VEGF production was significantly reduced in all groups. This suggests that the basal VEGF production from mesothelial cells is in part due to endogenous TGF-β.
Factors other than TGF-β, such as platelet-derived growth factor (PDGF), epidermal growth factor (EGF), keratinocyte growth factor (KGF), interleukin 1 (IL-1), IL-6, and IL-8, have been shown to induce VEGF production in other cell lines (3). However, these factors do not consistently induce VEGF, as does TGF-β1. For example, whereas PDGF stimulated VEGF production in vascular smooth muscle cells (38) and fibroblasts (39), it did not stimulate VEGF production in human airway epithelial cells. TGF-β, but not PDGF, EGF, or KGF, was capable of inducing VEGF synthesis in transformed airway epithelial cells (12). Furthermore, TGF-β is synergistic with PDGF (38) and IL-1 (34) in enhancing VEGF production.
We believe the results of this study have clinical implications. Management of recurrent pleural effusions is a common clinical problem. The treatments frequently employed are that of repeated aspiration, or the application of a pleurodesing agent to induce fibrous obliteration of the pleural space. No strategy actually targets the control of increased pleural vascular permeability, which underlies the accumulation of most cases of exudative pleural effusions. Anti-VEGF antibodies have been safely used in a Phase I study (40), and its use against malignant pleural effusions is currently under clinical trial (41). Our results provide further understanding of the mechanism of VEGF induction in the pleural space, and may provide more target options in the pathway of increased pleural fluid formation.
In conclusion, this study showed that TGF-β is a potent stimulator of VEGF production in the pleural space in vivo. The amount of pleural fluids induced after the intrapleural administration of TGF-β correlated significantly with the pleural fluid VEGF levels. In the in vitro study, we demonstrated that pleural mesothelial cells are capable of producing VEGF, and that this production was upregulated by TGF-β. Conversely, anti-TGF-β antibodies inhibited VEGF production by mesothelial cells. Targeting the TGF-β and VEGF pathway may provide novel treatment strategies for management of pleural effusions.
The authors thank Genzyme Corporation for providing the TGF-β2, and Dr. Steve Ledbetter for providing the anti-TGF-β antibodies used in the experiments.
Supported by the St. Thomas Foundation (Nashville, TN). Dr. Lee is the recipient of a United States–New Zealand Fulbright Graduate Award.
| 1. | Marel M, Zrustova M, Stasny B, Light RWThe incidence of pleural effusion in a well-defined region: epidemiologic study in central Bohemia. Chest104199314861489 |
| 2. | Light RW. Pleural diseases. 4th edition. Baltimore, MD: Lippincott, Williams & Wilkins; 2001. |
| 3. | Ferrara NMolecular and biological properties of vascular endothelial growth factor. J Mol Med771999527543 |
| 4. | Yano S, Shinohara H, Herbst RS, Kuniyasu H, Bucana CD, Ellis LM, Fidler IJProduction of experimental malignant pleural effusions is dependent on invasion of the pleura and expression of vascular endothelial growth factor/vascular permeability factor by human lung cancer cells. Am J Pathol157200018931903 |
| 5. | Yeo K-T, Wang HH, Nagy JA, Sioussat TM, Ledbetter SR, Hoogewerf AJ, Zhou Y, Masse EM, Senger DR, Dvorak HF, et al.Vascular permeability factor (vascular endothelial growth factor) in guinea pig and human tumor and inflammatory effusions. Cancer Res53199329122918 |
| 6. | Yano S, Herbst RS, Shinohara H, Knighton B, Bucana CD, Killion JJ, Wood J, Fidler IJTreatment for malignant pleural effusion of human lung adenocarcinoma by inhibition of vascular endothelial growth factor receptor tyrosine kinase phosphorylation. Clin Cancer Res62000957965 |
| 7. | Thickett DR, Armstrong L, Millar ABVascular endothelial growth factor (VEGF) in inflammatory and malignant pleural effusions. Thorax541999707710 |
| 8. | Cheng D, Rodriguez RM, Perkett EA, Rogers J, Bienvenu G, Lappalainen U, Light RWVascular endothelial growth factor in pleural fluid. Chest1161999760765 |
| 9. | Light RW, Cheng DS, Lee YCG, Rogers J, Davidson J, Lane KBA single intrapleural injection of transforming growth factor beta-2 produces an excellent pleurodesis in rabbits. Am J Respir Crit Care Med162200098104 |
| 10. | Lee YCG, Teixeira LR, Devin CJ, Vaz MAC, Vargas FS, Thompson PJ, Lane KB, Light RWTransforming growth factor-β2 induces pleurodesis significantly faster than talc. Am J Respir Crit Care Med1632001640644 |
| 11. | Lee YCG, Lane KB, Parker RE, Ayo DS, Rogers JT, Diters RW, Thompson PJ, Light RWTransforming growth factor beta-2 (TGFβ2) produces effective pleurodesis in sheep with no systemic complications. Thorax55200010581062 |
| 12. | Boussat S, Eddahibi S, Coste A, Fataccioli V, Gouge M, Housset B, Adnot S, Maitre BExpression and regulation of vascular endothelial growth factor in human pulmonary epithelial cells. Am J Physiol Lung Cell Mol Physiol2792000L371L378 |
| 13. | Koochekpour S, Merzak A, Pilkington GJVascular endothelial growth factor production is stimulated by gangliosides and TGF-β isoforms in human glioma cells in vitro. Cancer Let1021996209215 |
| 14. | Donovan D, Harmey JH, Toomey D, Osborne DH, Redmond HP, Bouchier-Hayes DJTGF beta-1 regulation of VEGF production by breast cancer cells. Ann Surg Oncol41997621627 |
| 15. | Cheng DS, Lee YC, Rogers JT, Perkett EA, Lappalainen U, Rodriguez RM, Light RWVascular endothelial growth factor level correlates with transforming growth factor beta isoform levels in pleural effusions. Chest118200017471753 |
| 16. | Dasch JR, Pace DR, Waegell W, Inenaga D, Ellingsworth LMonoclonal antibodies recognizing transforming growth factor-beta: bioactivity neutralization and transforming growth factor beta 2 affinity purification. J Immunol142198915361541 |
| 17. | Lee YCG, Devin CJ, Teixiera LR, Rogers JT, Thompson PJ, Lane KB, Light RWTransforming growth factor beta-2 induced pleurodesis is not inhibited by corticosteroids. Thorax562001643648 |
| 18. | Antony VB, Owen CL, Hadley KJPleural mesothelial cells stimulated by asbestos release chemotactic activity for neutrophils in vitro. Am Rev Respir Dis1391989199206 |
| 19. | Neufeld G, Cohen T, Gengrinovitch S, Poltorak ZVascular endothelial growth factor and its receptors. FASEB J131999922 |
| 20. | Border WA, Okuda S, Languino LR, Sporn MB, Ruoslahti ESuppression of experimental glomerulonephritis by antiserum against transforming growth factor β1. Nature3461990371374 |
| 21. | Murohara T, Horowitz JR, Silver M, Tsurumi Y, Chen D, Sullivan A, Isner JMVascular endothelial growth factor/vascular permeability factor enhances vascular permeability via nitric oxide and prostacyclin. Circulation97199899107 |
| 22. | Kraft A, Weindel K, Ochs A, Marth C, Zmija J, Schumacher P, Unger C, Marme D, Gastl GVascular endothelial growth factor in the sera and effusions of patients with malignant and nonmalignant disease. Cancer851999178187 |
| 23. | Zweers MM, de Waart DR, Smit W, Strujik DG, Krediet RTGrowth factors VEGF and TGF-beta 1 in peritoneal dialysis. J Lab Clin Med1341999124132 |
| 24. | Sahn SAThe pleura. Am J Respir Crit Care Med1381988184234 |
| 25. | Harmey JH, Dimitriadis E, Kay E, Redmond HP, Bouchier-Hayes DRegulation of macrophage production of vascular endothelial growth factor (VEGF) by hypoxia and transforming growth factor beta-1. Ann Surg Oncol51998271278 |
| 26. | Muller J, Yoshida TInteraction of murine peritoneal leukocytes and mesothelial cells: in vitro model system to survey cellular events on serosal membranes during inflammation. Clin Immunol Immunopathol751995231238 |
| 27. | Moalli PA, MacDonald JL, Goodglick LA, Kane ABAcute injury and regeneration of the mesothelium in response to asbestos fibers. Am J Pathol1281987426445 |
| 28. | Gibson FC, Onderdonk AB, Kasper DL, Tzianabos AOCellular mechanism of intraabdominal abscess formation by Bacteroides fragilis. J Immunol160199850005006 |
| 29. | Nasreen N, Mohammed KA, Dowling PA, Ward MJ, Galffy G, Antony VBTalc induces apoptosis in human malignant mesothelial cells in vitro. Am J Respir Crit Care Med1612000595600 |
| 30. | Chua CC, Hamdy RC, Chua BHMechanism of transforming growth factor-beta-1-induced expression of vascular endothelial growth factor in murine osteoblastic MC3T3-E1 cells. Biochim Biophys Acta149720006976 |
| 31. | Pertovaara L, Kaipainen A, Mustonen T, Orpana A, Ferrara N, Saksela O, Alitalo KVascular endothelial growth factor is induced in response to transforming growth factor-beta in fibroblastic and epithelial cells. J Biol Chem269199462716274 |
| 32. | Border W, Noble NATransforming growth factor-beta in tissue fibrosis. N Engl J Med331199412861292 |
| 33. | Kelley J. Transforming growth factor-beta. In: Kelley J, editor. Cytokines of the Lung. New York: Marcel Dekker; 1993. p. 101–137. |
| 34. | Berse B, Hunt JA, Diegel RJ, Morganelli P, Yeo K, Brown F, Fava RAHypoxia augments cytokine (transforming growth factor-beta (TGF-beta) and IL-1)-induced vascular endothelial growth factor secretion by human synovial fibroblasts. Clin Exp Immunol1151999176182 |
| 35. | Ferrara N, Keyt BVascular endothelial growth factor: basic biology and clinical implications. EXS791997209232 |
| 36. | Van Obberghen-Schilling E, Roche NS, Flanders KC, Sporn MB, Roberts ABTransforming growth factor beta 1 positively regulates its own expression in normal and transformed cells. J Biol Chem263198877417746 |
| 37. | Bascom CC, Wolfshohl JR, Coffey RJ, Madisen L, Webb NR, Purchio AR, Derynck R, Moses HLComplex regulation of transforming growth factor beta 1, beta 2, and beta 3 mRNA expression in mouse fibroblasts and keratinocytes by transforming growth factor beta 1 and beta 2. Mol Cell Biol9198955085515 |
| 38. | Kronemann N, Bouloumi A, Bassus S, Kirchmaier CM, Busse R, Schini-Kerth VBAggregating human platelets stimulate expression of vascular endothelial growth factor in cultured vascular smooth muscle cells through a synergistic effect of transforming growth factor-beta 1 and platelet-derived growth factor AB. Circulation1001999855860 |
| 39. | Enholm B, Paavonen K, Ristimaki A, Kumar V, Gunji Y, Klefstrom J, Kivinen L, Laiho M, Olofsson B, Joukov V, Eriksson U, Alitalo KComparison of VEGF, VEGF-B, VEGF-C and Ang-1 mRNA regulation by serum, growth factors, oncoproteins and hypoxia. Oncogene14199724752483 |
| 40. | Gordon MS, Talpaz M, Margolin K, Holmgren E, Sledge GW, Benjamin R, Stalter S, Shak S, Adelman D. Phase I trial of recombinant humanized monoclonal anti-vascular endothelial growth factor in patients with metastatic cancer. In: 34th annual meeting of the American Society of Clinical Oncologists; 1998, May 16–19, Los Angeles, 1998. |
| 41. | Verheul HMW, Hoekman K, Jorna AS, Smit EF, Pinedo HMTargeting vascular endothelial growth factor blockade: ascites and pleural effusion formation. Oncologist520004550 |
This article has an online data supplement, which is accessible from this issue's table of contents online at www.atsjournals.org
