Cytokine 69 (2014) 284–288
Contents lists available at ScienceDirect
Cytokine journal homepage: www.journals.elsevier.com/cytokine
VEGF increases the permeability of sheep pleura ex vivo through VEGFR2 stimulation Vasiliki I. Peppa a, Zoi V. Arsenopoulou a, Sotirios G. Zarogiannis a,b,⇑, Triantafyllia Deligiorgi a, Rajesh Jagirdar a, Ioannis Makantasis a, Ioannis Stefanidis c, Vassilios Liakopoulos c, Paschalis-Adam Molyvdas a, Konstantinos I. Gourgoulianis b, Chrissi Hatzoglou a,b a
Department of Physiology, Medical School, University of Thessaly, Biopolis, Larissa 41110, Greece Department of Respiratory Medicine, Medical School, University of Thessaly, Biopolis, Larissa 41110, Greece c Department of Nephrology, Medical School, University of Thessaly, Biopolis, Larissa 41110, Greece b
a r t i c l e
i n f o
Article history: Received 14 November 2013 Received in revised form 30 May 2014 Accepted 8 June 2014 Available online 28 June 2014 Keywords: Mesothelium Permeability Pleura VEGF VEGF receptors
a b s t r a c t Vascular endothelial growth factor (VEGF), a cytokine that increases vascular permeability to water and proteins and induces angiogenesis, has been implicated in the development of pleural effusions. Inﬂammatory and malignant pleural effusions are rich in VEGF content while mesothelial cells produce and excrete VEGF. In this report we aimed at investigating by means of electrophysiology the direct effects of VEGF on the parietal and visceral sheep pleura as well as the type of receptors that mediate this effect. Our ﬁndings show that VEGF has a direct effect on the pleural mesothelium rendering it more permeable and this effect is mediated through the stimulation of VEGF receptor 2. Our ﬁndings shed more light to the role of VEGF in the pathogenesis of pleural effusions and provide functional evidence for a role of VEGFR2 on the pleural mesothelium that has never been studied before. Ó 2014 Elsevier Ltd. All rights reserved.
1. Introduction Vascular endothelial growth factor (VEGF) has been implicated in the development of pleural effusions . It has been reported to be signiﬁcantly elevated in exudative pleural effusions and mesothelial cells are a known source of VEGF production [2,3]. In exudative pleural effusions the levels of VEGF are very high in empyemas and complicated parapneumonic effusions being also a signiﬁcant predictor of clinically signiﬁcant residual pleural thickening and it is also highly elevated in malignant effusions [2–6]. Human pleural mesothelial cells in culture produce VEGF as a response to bacterial insult that leads to the reduction of the electrical resistance and thus to the increase of the mesothelial monolayer permeability . Moreover, mesothelial cells have been shown to express VEGFR1/Flt-1 receptors in their plasma membrane while the soluble form of VEGFR1 has been found elevated in malignant pleural effusions [6,7]. The biological activity of VEGF is exerted though its ligation to the VEGF receptors, VEGFR-1/Flt-1 and VEGFR-2/Flk-1 KDR ⇑ Corresponding author. Address: Environmental Pleural & Lung Diseases Group, Membrane Permeability Group, Department of Physiology, University of Thessaly, School of Medicine, Biopolis, Larissa 41110, Greece. Tel.: +30 2410685558; fax: +30 2410685555. E-mail address: [email protected]
(S.G. Zarogiannis). http://dx.doi.org/10.1016/j.cyto.2014.06.014 1043-4666/Ó 2014 Elsevier Ltd. All rights reserved.
[1,7,8]. In health, activation of VEGFR-2 mediated pathway promotes angiogenesis, increased vascular permeability as well as endothelial cell proliferation and migration [7,8]. VEGFR-1 acts as a decoy receptor but its role is more prominent in pathological conditions such as malignancies, where it becomes the main mediator of angiogenesis induction . In contrast to VEGFR1 being described in the pleura no data exist regarding VEGFR2. Electrophysiological studies with the Ussing system testing the effects of various agents on sheep serosal membrane permeability have been shown to correspond well with human samples as well as with in vivo models [9–12]. To this end, the aim of our study was to assess whether VEGF has direct effects on the permeability of healthy sheep parietal and visceral pleura and whether this effect is mediated through VEGFR1 or VEGFR2 receptors. To our knowledge this is the ﬁrst study to assess the effects of VEGF in more complex pleural structures than pleural cell lines.
2. Materials and methods 2.1. Specimen collection and preparation Intact sheets of visceral and parietal pleura were obtained from 60 adult sheep (males and females). The samples were collected from the slaughterhouse immediately after the death of the animal
V.I. Peppa et al. / Cytokine 69 (2014) 284–288
(time of warm ischemia close to 0) and transferred to the laboratory in oxygenated Dulbecco Modiﬁed Eagle’s Medium (DMEM) at 4 °C within 30 min of the death of the animal. Visceral pleura was carefully stripped from the underlying lung, while parietal pleura was carefully stripped from the chest wall and then examined for evidence of holes or adherent tissue by visual inspection. Pieces of parietal pleura not likely to contain stomas were used, as suggested from anatomical studies in sheep . 2.2. Transmesothelial electrical resistance measurements Pleura specimens were carefully mounted in Ussing chambers (Dipl.-Ing. K. Mussler Scientiﬁc Instruments, Aachen, Germany) with an opening surface area of 1 cm2. Tissues were bathed with 4 ml of Krebs–Ringer bicarbonate (KRB) solution on each side of the membrane, continuously oxygenated with 95% O2/5% CO2 circulated by gas lift. KRB solution was balanced at pH 7.4 and contained (in mM) 117.5 NaCl, 1.15 NaH2PO4, 24.99 NaHCO3, 5.65 KCl, 1.18 MgSO4, 2.52 CaCl2, and 5.55 glucose. Two pairs of Ag/AgCl electrodes monitored the transmesothelial electrical resistance (RTM; X cm2) under open-circuit conditions every 60 s. Experiments were conducted in computer-controlled chambers (Clamp version 2.14 software: AC Micro-Clamp, Aachen, Germany). RTM was measured in the basal state (end of equilibration time of 10–40 min) and after the addition of different substances. Active ion transport is temperature dependent so the Ussing chambers were held at 37 °C. The voltage responses to applied current pulses of given amplitude (50 lA) and duration (200 ms) were measured. RTM was calculated by automatically deducing the initially measured resistance of the solution. Changes in RTM after the addition of the chemicals were determined as percent (%) changes (DRTM). 2.3. Experimental procedure The mesothelial cell surface facing in vivo the pleural ﬂuid is referred to as apical, while the one facing the blood supply is referred to as basolateral. Measurements of RTM were made before and after exposure to substances at time points (0, 1, 10, 20 and 30 min). Values before the addition of substances served as controls. In the initial set of experiments, VEGF165 (ﬁnal concentration of 1 ng/mL) was added apically on the parietal (n = 6) and visceral (n = 6) pleura, as well as basolaterally on the parietal (n = 6) and visceral (n = 6) pleura, in different pleura specimens. In another set of experiments the VEGFR1/R2 (VEGFR-1/Flt-1 & 2) inhibitor SU5416 (10 6 M) or the VEGFR2/Flt-KDR selective inhibitor SU1498 (10 6 M) plus VEGF (1 ng/mL) were added apically on the parietal (n = 6) and visceral (n = 6) pleura, as well as basolaterally on the parietal (n = 6) and visceral (n = 6) pleura. Equal numbers of experiments with the inhibitors alone were also performed in order to detect potential non-speciﬁc agonistic effects. Solutions were freshly prepared before each experiment, heated to 37 °C, and bubbled continuously with a 95% O2/5% CO2 gas mixture. All chemicals were purchased from Sigma–Aldrich Chemie GmbH, Munich, Germany. 2.4. Protein isolation from pleural membranes and Western blotting Parietal and visceral pleural specimens were rinsed with PBS and soaked in RIPA buffer containing protease inhibitor cocktail (Roche Diagnostics, Mannheim, Germany). The lysates were centrifuged at 15,000 rpm at 4 °C for 10 min and the protein content of the supernatant was measured by the MicroBCA method (Thermoscientiﬁc, Rockford, IL). Equal amounts of protein (300 lg) were separated by 4–20% SDS–PAGE Bis-Tris polyacrylamide gel
and then transferred to nitrocellulose membranes. Membranes were probed with rabbit anti-VEGFR2 and anti-b-actin (Cat.#: sc-505 and sc-130656, Santa Cruz Biotechnology, Germany) followed by incubation with goat anti-rabbit IgG-HRP (Cat.#: sc-2004). Protein bands were revealed using chemiluminescence HRP substrate (Pierce Biotechnology, Rockford, IL) and X-ray ﬁlms. The experiment was performed three times. 2.5. Statistical analysis Statistical analyses were performed with GraphPad Prism v4 for Mac OSX (GraphPad Software Inc., San Diego, CA). All data are expressed as mean ± SEM. The probability of error for comparison of the mean values was calculated using two-way ANOVA with Bonferroni post-test. Values of p < 0.05 were regarded as signiﬁcant. 3. Results 3.1. Parietal pleura The control RTM of parietal pleura (i.e. mean value from the total number of experiments before the addition of any substance, n = 60) was 18.35 ± 0.93 X cm2. VEGF induced a signiﬁcant decrease from baseline in the RTM of the parietal pleura when added both apically being signiﬁcant from the 20th min ( 18.14 ± 9.82%, p < 0.01; Fig. 1A) and basolaterally being signiﬁcant from the 10th min ( 14.89 ± 3.79%, p < 0.01; Fig. 1B). This effect was sustained throughout the experiment and the maximal effect of VEGF from baseline in both apical and basolateral additions was observed on the 30th min as shown in Fig. 1A and B (Apical: 25.48 ± 5.79%, p < 0.01 compared to baseline; Basolateral: 28.87 ± 6.36%, p < 0.01 compared to baseline). At this time point, the effect of VEGF was inhibited apically by both SU5416 ( 7.23 ± 4.8% from baseline, p < 0.01 compared to VEGF; Fig. 1A) and SU1498 (0.16 ± 6.53% from baseline, p < 0.001 compared to VEGF; Fig. 1A), The same effect was also observed basolaterally [SU5416 ( 7.18 ± 3.02% from baseline, p < 0.05 compared to VEGF; Fig. 1B) and SU1498 (1.99 ± 3.81% from baseline, p < 0.001 compared to VEGF; Fig. 1B)]. The inhibitors did not have any effect when applied to both surfaces alone (p > 0.05 compared to baseline). 3.2. Visceral pleura The control RTM of visceral pleura (n = 60) was 19.35 ± 0.70 X cm2. VEGF induced a signiﬁcant decrease from baseline in the RTM of the parietal pleura when added both apically ( 16.39 ± 4.93%, p < 0.01 compared to baseline; Fig. 2A) and basolaterally ( 21.59 ± 4.97%, p < 0.001 compared to baseline; Fig. 2B) being signiﬁcant from the 10th min in both cases. This effect was sustained throughout the experiment (30 min). At this time point the maximal effect of VEGF from baseline regarding the apical addition of VEGF was 30.83 ± 6.99% (p < 0.01 compared to baseline; Fig. 2A) and regarding the basolateral addition of VEGF was 32.29 ± 4.88% (p < 0.01 compared to baseline; Fig. 2B). At this time point the effect was inhibited apically only by SU1498 ( 5.57 ± 2.78% from baseline, p < 0.01 compared to VEGF; Fig. 2A). SU5416 partially inhibited the VEGF effect ( 13.82 ± 6.43% from baseline, p > 0.05 compared to VEGF and p < 0.05 compared to baseline; Fig. 2A) and when it was added alone it exhibited an agonist effect ( 15.51 ± 5.95% from baseline, p < 0.01 compared to baseline; Fig. 2A). Moreover, there was a signiﬁcant difference when SU5416 and SU1498 effects were compared (p < 0.05). As far as the basolateral application of VEGF is concerned, its effect was abolished only by SU1498 ( 7.3 ± 3.67% from baseline,
V.I. Peppa et al. / Cytokine 69 (2014) 284–288
Fig. 1. Temporal effects of VEGF and its inhibitors on the RTM of the parietal sheep pleura. (A) The percentage change in transmesothelial electrical resistance (% DRTM) as a function of time in given time points (0, 1, 10, 20 and 30 min), after the addition of VEGF (1 ng/mL), VEGF (1 ng/mL) plus SU1498 (10 6 M), VEGF (1 ng/mL) plus SU5416 (10 6 M), SU1498 (10 6 M), SU5416 (10 6 M) apically in the parietal pleura. (B) The percentage change in transmesothelial electrical resistance (% DRTM) as a function of time in given time points (0, 1, 10, 20 and 30 min), after the addition of VEGF (1 ng/mL), VEGF (1 ng/mL) plus SU1498 (10 6 M), VEGF (1 ng/mL) plus SU5416 (10 6 M), SU1498 (10 6 M), SU5416 (10 6 M) basolaterally in the parietal pleura. Values are mean and standard error of 6 experiments in each case of substances addition. *P < 0.05 comparison vs. baseline, @P < 0.05 comparison vs. VEGF.
p < 0.001 compared to VEGF; Fig. 2B). Like in the case of apical side related experiments, SU5416 partially inhibited the VEGF effect ( 11.82 ± 5.04% from baseline, p < 0.001 compared to VEGF and p < 0.05 compared to baseline; Fig. 2B) and when it was added alone it exhibited an agonist effect ( 20.45 ± 3.35% from baseline, p < 0.001 compared to baseline; Fig. 2B). Moreover, there was a signiﬁcant difference when SU5416 and SU1498 effects were compared (p < 0.001). 3.3. Western blot analysis for VEGFR2 detection Specimens from both the parietal and visceral pleura were probed in order to determine the expression of VEGFR2 in both our tissue types. As shown in Fig. 3, VEGFR2 can be detected in both types of pleura in equal amounts. 4. Discussion VEGF levels in the pleural ﬂuid of patients due to inﬂammation and malignancy are elevated and have been implicated in the
pathophysiology of pleural effusions [1–7]. In various models it was shown that pleural mesothelial cells produce VEGF, although the distribution of VEGF receptors is not well characterized since only the VEGFR-1 and the soluble VEGFR-1 have been evaluated histologically in the pleura and by ELISA measurements in the pleural ﬂuid respectively [6,7]. In the current study we investigated the effects of VEGF on the sheep pleural permeability by assessing the changes in RTM, a surrogate marker of membrane ionic permeability, those two being inversely correlated [9,10,13,14]. Results from such studies have been shown to correlate well with in vivo animal models of induced hydrothoraces, therefore provide a representative means of projecting to in vivo conditions [11,12]. Our principal ﬁnding was that VEGF induces a rapid decrease on the RTM, both when added apically and basolaterally, indicative of a potent direct effect on the pleural membrane. These results are in good agreement with the unique study that has investigated the direct effects of VEGF on human pleural mesothelial cell monolayers (Met5A cell line), a structure more simple than our tissue specimen, where it was also shown that mesothelial cells produce
V.I. Peppa et al. / Cytokine 69 (2014) 284–288
Fig. 2. Temporal effects of VEGF and its inhibitors on the RTM of the visceral sheep pleura. (A) The percentage change in transmesothelial electrical resistance (% DRTM) as a function of time in given time points (0, 1, 10, 20 and 30 min), after the addition of VEGF (1 ng/mL), VEGF (1 ng/mL) plus SU1498 (10 6 M), VEGF (1 ng/mL) plus SU5416 (10 6 M), SU1498 (10 6 M), SU5416 (10 6 M) apically in the visceral pleura. (B) The percentage change in transmesothelial electrical resistance (% DRTM) as a function of time in given time points (0, 1, 10, 20 and 30 min), after the addition of VEGF (1 ng/mL), VEGF (1 ng/mL) plus SU1498 (10 6 M), VEGF (1 ng/mL) plus SU5416 (10 6 M), SU1498 (10 6 M), SU5416 (10 6 M) basolaterally in the visceral pleura. Values are mean and standard error of 6 experiments in each case of substances addition. *P < 0.05 comparison vs. baseline, @P < 0.05 comparison vs. VEGF, #P < 0.05 comparison of SU5416 vs. SU1498.
PP ~ 150 kDa ~ 42 kDa
VP VEGFR2 ACTB1
Fig. 3. VEGFR-2 is expressed in both the parietal and visceral sheep pleura. Western blot for VEGFR-2 detection on specimens of sheep parietal pleura (PP) and visceral pleura (VP). Parietal and visceral pleura membrane lysates were probed for the expression of VEGFR2 protein. The immuno-blot shows the expression of VEGFR2 protein in both parietal and visceral pleura at predicted molecular weights (in accordance with data from Ensembl.org).
VEGF that decreases the electrical resistance and therefore increases the permeability of the mesothelial monolayer to ﬂuid and proteins . In the same study use of VEGF neutralizing antibodies abolished the effect of mesothelial cell barrier dysfunction. Furthermore, we demonstrated that the effect of VEGF occurs through the stimulation of VEGFR-2 since the application of
SU1498, a selective inhibitor of VEGFR2, abolished the effect of VEGF on both parietal and visceral pleura. There are no studies either in humans or in animals reporting the existence of such a receptor in the visceral and parietal pleura. Given the fact that our specimens were derived from healthy sheep the detection of VEGFR-2 as the mediator of VEGF effect is in agreement with the notion that in health, the activation of VEGFR-2 mediated pathway mainly regulates the effects of VEGF . However, the only available data on human pleura show that both in healthy and inﬂamed pleura the VEGFR-1 receptors are present in mesothelial cells, although very few samples were included in this study . Nevertheless, the mechanism of action for VEGFR-2 to increase the permeability of the pleura might be similar to the one described in microvascular endothelial cells . More speciﬁcally, VEGFR2 activation increases paracellular permeability by up-regulating EphA2 via its associated downstream signaling cascades, PI3K/ Akt and ERK1/2 . On the other hand, the use of SU5416, an inhibitor of both VEGFR1 and VEGFR2, abolished the effect of VEGF on the parietal
V.I. Peppa et al. / Cytokine 69 (2014) 284–288
pleura. This effect was expected since the SU1498 was able to fully inhibit the decrease of the RTM caused by VEGF. As far as the visceral pleura is concerned, we found an agonist effect of SU5416 since it partially inhibited the effect of VEGF when added together – due to VEGFR2 inhibition – while when added alone it exhibited an equal decrease in the RTM suggestive of non-speciﬁc effects on the visceral pleura. SU5416 has recently being shown to stimulate the intracellular Aryl Hydrocarbon Receptors (AHR) in AHR-mutant hepatoma cells . The existence of such receptors has not been described in mesothelial cells. The effect of SU5416 was similar when added both apically and basolaterally in the visceral pleura and suggests that AHR receptors maybe present only in the visceral mesothelial cells. Still, this hypothesis requires further experimental testing. In conclusion we found that VEGF has direct effects on the pleural membrane increasing the pleural permeability in healthy sheep parietal and visceral pleura through stimulation of VEGFR2. VEGFR-1 were either not present in our specimens or were inactive in the presence of VEGFR-2. Finally, the presence of VEGFR-2 in our samples was demonstrated at the protein level. The clinical implication of our study need to be veriﬁed in human pleural specimens and in mouse animal models of induced hydrothoraces in the context of potential intra-pleural administration for the inhibition of VEGFR2 as a means of reduction of pleural effusion production. Acknowledgments This research was supported by a postgraduate Grand of the Hellenic General Secretariat of Research and Technology (PENED 2003-03ED782) and by ARITI S.A. by an unrestricted grand. References  Papaioannou A, Kostikas K, Kollia P, Gourgoulianis KI. Clinical implications for vascular endothelial growth factor in the lung: friend or foe? Respir Res 2006;7:128.
 Thickett DR, Armstrong L, Millar AB. Vascular endothelial growth factor (VEGF) in inﬂammatory and malignant pleural effusions. Thorax 1999;54:707–10.  Marchi E, Vargas FS, Acencio MM, Antonangelo L, Genofre EH, Teixeira LR. Evidence that mesothelial cells regulate the acute inﬂammatory response in talc pleurodesis. Eur Respir J 2006;28:929–32.  Mohammed KA, Nasreen N, Hardwick J, Logie CS, Patterson CE, Antony VB. Bacterial induction of pleural mesothelial monolayer barrier dysfunction. Am J Physiol Lung Cell Mol Physiol 2001;281:119–25.  Papaioannou AI, Kostikas K, Tsopa P, Kiropoulos T, Tsilioni I, Oikonomidi S, et al. Residual pleural thickening is related to vascular endothelial growth factor levels in parapneumonic pleural effusions. Respiration 2010;80:472–9.  Tomimoto H, Yano, Muguruma H, Kakiuchi S, Sone S. Levels of soluble vascular endothelial growth factor receptor 1 are elevated in exudative pleural effusion. J Med Invest 2007;54:146–53.  Hooper CE, Elvers KT, Welsh GI, Millar AB, Maskell NA. VEGF and sVEGFR-1 in malignant pleural effusions: association with survival and pleurodesis outcomes. Lung Cancer 2012;77:443–9.  Froudarakis M. Pleural effusion in lung cancer: more questions than answers. Respiration 2012;83:367–76.  Stefanidis I, Zarogiannis S, Hatzoglou C, Liakopoulos V, Kourti P, Poultsidi A, et al. Enhancement of the transmesothelial resistance of the parietal sheep peritoneum by epinephrine in vitro: using-type chamber experiments. Artif Organs 2005;29:919–22.  Stefanidis I, Liakopoulos V, Kourti P, Zarogiannis S, Poultsidi A, Mertems PR, et al. Amiloride sensitive sodium channels on the parietal human peritoneum: evidence by Ussing-type chamber experiments. ASAIO J 2007;53:335–8.  Zarogiannis S, Deligiorgi T, Stefanidis I, Liakopoulos V, Gourgoulianis K, Molyvdas PA, et al. Dexamethasone decreases the transmesothelial electrical resistance of the parietal and visceral pleura. J Physiol Sci 2009;59:335–9.  Zarogiannis S, Psallidas I, Hatzoglou C, Gourgoulianis K, Kalomenidis I. Dexamethasone acutely accelerates pleural ﬂuid absorption in mice hydrothoraces. J Physiol Sci 2010;60:299–302.  Zarogiannis S, Hatzoglou C, Stefanidis I, Ioannou M, Paraskeva E, Gourgoulianis K, et al. Comparison of the electrophysiological properties of the sheep isolated costal and diaphragmatic parietal pleura. Clin Exp Pharmacol Physiol 2007;34:129–31.  Zarogiannis S, Hatzoglou C, Stefanidis I, Liakopoulos V, Gourgoulianis K, Molyvdas PA. Adrenergic inﬂuence on the permeability of sheep diaphragmatic parietal pleura. Respiration 2007;74:118–20.  Miao Z, Dong Y, Fang W, Shang D, Liu D, Zhang K, et al. VEGF increases paracellular permeability in brain endothelial cells via upregulation of EphA2. Anat Rec 2014;297:964–72.  Mezrich JD, Nguyen LP, Kennedy G, Nukaya M, Fechner JH, Zhang X, et al. SU5416, a VEGF receptor inhibitor and ligand of the AHR, represents a new alternative for immunomodulation. PLoS ONE 2012;7:e 44547.