Screening of Anti-Hypoxia/Reoxygenation Agents by an in vitro Method. Part 2: Inhibition of Tyrosine Kinase Activation Prevented Hypoxia/Reoxygenation-Induced Injury in Endothelial GapJunctional Intercellular Communication

• Abstract
• Introduction
• Materials and Methods
• Materials
• Test compounds
• Preparations of test samples
• Cell culture
• In vitro H/R model
• GJIC analysis
• Western blotting of connexin 43
• Results
• GJIC assay by the FRAP technique
• H/R-induced GJIC injury
• Relationship between this GJIC injury and PTK activation
• Effects of natural compounds on GJIC
• Interference with the gap junctional protein connexin 43
• Discussion
• Acknowledgements
• References

Abstract

In this study, we demonstrated that hypoxia/reoxygenation (H/R) induced an injury in gap junctional intercellular communication (GJIC) after 2 h of reoxygenation in cultured HUVEC. Free radical scavenger (DMSO) and antioxidant (SOD) did not prevent this GJIC injury at all. Protein kinase C inhibitor (calphostin C) partly blocked this injury. However, the protein tyrosine kinase (PTK) inhibitor genistein completely inhibited this GJIC injury. Compounds 1 [laxogenin-3-O-α-L-arabinosyl-(1¡ú 6)-β-D-glucopyranoside], 2 (macrostemososide A), 3 [laxogenin-3-O-β-D-xylopyranosyl-(1¡ú 4)-α-L-arabinopyranosyl-(1¡ú  6)-β-D-glucopyranoside], 4 (chinenoside II), 5 (β-sitosterol), 6 (daucosterine), 7 (ginsenoside-Rd), 29 (isocumarine), 52 (icariin), 53 (icariside), and 54 (icaritin), which showed obvious influence on H/R-induced PTK activation as stated in Part 1 (except 1), were explored for their effects on GJIC. The results showed that compounds 2 šC 7 and 52 šC 57 partly protected H/R-induced GJIC injury. Compounds 5 and 6 (especially 5), which showed the strongest inhibitory effects on PTK activation, completely blocked H/R-provoked GJIC injury. Compound 1, which did not influence PTK activation, failed to prevent this GJIC injury. In contrast, compound 29, which significantly promoted PTK activation, enhanced this H/R-induced GJIC injury further. Western blotting of connexin 43, an important gap junctional protein for modulating GJIC in HUVEC, revealed that interference with the gap junctional protein might be the most direct mechanism for compounds 2, 5, 29, and 53 to affect H/R-injured GJIC.

Introduction

It is known that an important characteristic of I/R- or H/R-induced endothelial injury is the damage of the tissue homeostasis [1], [2]. The endothelium is composed of only one single layer of endothelial cells (EC); the tissue homeostasis requires quite coordinated communications within the monolayer [3]. The cytosolic continuity of cells as a result of gap junction-mediated intercellular communication (GJIC) integrates connecting individual cells into a functional syncytia and has been postulated to be a principal means of maintaining the tissue homeostasis [4], including that in HUVEC.

Gap junctions are hydrophilic membrane channels connecting two adjacent cells [4], [5]. These channels are composed of hexameric hemichannels, or connexons, that are attached to other connexons in the plasma membranes of neighboring cells. Connexons are composed of highly conserved proteins called connexins. In vertebrates, no other channel can provide an enclosed conduit for direct diffusional exchange of ions and small molecules between adjacent cells, and few other membrane channels have pore diameters large enough to accommodate passage of metabolites and signaling molecules with molecular weights as high as 1000 - 1500 dalton [4], [6]. GJIC plays an important role in modulating the endothelial homeostasis [7].

In light of the results in Part 1 (see preceeding paper [14]), we have shown that H/R induced PTK activation in cultured HUVEC, and several compounds (2 - 7 and 52 - 54) purified from traditional Chinese herbs were found to be able to inhibit this PTK activation. In the present report, we investigated whether H/R could induce endothelial injury and PTK activation would be involved in this injury. Moreover, effects of the compounds 1 - 7, 29, and 52 - 54 on the endothelial GJIC function were examined.

Materials and Methods

Materials

5,6-Carboxyfluorescence diacetate (CFDA), a molecular probe for GJIC, was purchased from Sigma Chem. Co. (St. Louis, MO, USA). Mouse anti-connexin 43 antibody was obtained from Transduction Laboratory (Lexington, KS, USA). SOD was from Sigma. Others were same as those in Part 1 [14].

Test compounds

See Part 1 (previous paper [14]).

Preparations of test samples

See Part 1 (previous paper [14]).

Cell culture

See Part 1 (previous paper [14]).

In vitro H/R model

See Part 1 (previous paper [14]).

GJIC analysis

The GJIC assay was performed according to a modified procedure [4] using a technique known as gap-FRAP (fluorescence recovery after photobleaching). This technique is able to yield quantitative data.

Cells were rinsed twice with PBS containing 1.25 mM CaCl₂ and 0.5 mM MgCl₂ [PBS (+)] and incubated with 5 to 10 μM CFDA in PBS (+) at 37 °C with 5 % CO₂ for 10 to 15 min. No staining differences were found among all experimental cells. The cells were then rinsed five times with PBS (+) to remove extracellular dye and covered with PBS (+) for FRAP analysis. The fluorescence of randomly selected cells that had been stained with CFDA was bleached with 488-nm 600 mW visible laser beams, and then the recovery of fluorescence was monitored over the subsequent 15 min. Unbleached cells served as the 100 % fluorescence control, and were used to correct for the loss of fluorescence due to background leakage and photobleaching.

Western blotting of connexin 43

Immunoblotting analyses were performed as described previously [4]. Cultured cells were promptly washed twice with PBS and scraped in the same solution on ice. The pellets were separated by 800 rpm centrifugation at 4 °C for 5 min, then suspended in the sample loading buffer (25 mM Tris-HCl pH 6.8, 1 % SDS, 5 % glycerol, 0.04 % bromophenol blue, 1 % β-mercaptoethanol), sonicated, and boiled for 3 to 5 min. Proteins were quantified by using a Bio-Rad Protein assay kit (Bio-Rad Corp., Richmond, CA, USA). Cell lysates containing 10 μg protein were loaded onto each lane of a 10 % SDS-polyacrylamide gel, electrophoresed, and transferred to polyvinyl difluoride nitrocellulose membranes. The membranes were blocked in a solution (NIP 551 dissolved in 20 mM Tris-HCl pH 8.0, 150 mM NaCl, 0.1 % Tween 20, T-TBS) for 15 to 20 min, extensively washed with T-TBS, and incubated with the monoclonal mouse anti-connexin 43 antibody (1 : 1000) at 4 °C overnight. The membranes were then washed several times with T-TBS, and incubated with peroxidase-labeled anti-mouse antibody (1 : 1000) for 1 h at room temperature, washed extensively, then processed with an enhanced chemilunescence kit, and exposed to X-ray films.

Results

GJIC assay by the FRAP technique

Intercellular communication through the gap junctions was monitored using the FRAP technique which involves quantitating the increase of fluorescence over time in a photobleached cell adjacent to unbleached dye-labeled cells with the ACAS 570 laser cytometer.

The cells were stained with CFDA, a chemical molecule that can freely pass through the gap junctions. In each image, several cells were randomly selected, and the fluorescence of prebleaching was regarded as 100 %. The fluorescence was then photobleached by laser beams, and recoveries of fluorescence were monitored over the following 15 min. As shown in Figure [1], at normoxia, the fluorescence nearly recovered to the prebleaching level after 15 min of photobleaching, indicating that perfect GJIC was present in normal HUVEC and regulated endothelial integrity.

H/R-induced GJIC injury

Next the influence of H/R on GJIC was investigated. As shown in Figure [1], hypoxia treatment alone [R (0-h)] did not interfere with GJIC, the fluorescence was recovered to about 96 % after 15 min of photobleaching. However, after 2 h of reoxygenation [R (2-h)], GJIC was significantly inhibited, the fluorescence was only recovered to 76 % after 15 min of photobleaching. But after 4 h of reoxygenation [R (4-h)], GJIC was improved, the fluorescence was recovered to about 86 % after 15 min of photobleaching; and after 6 h of reoxygenation [R (6-h)], it nearly recovered to the normal level. These results suggested that H/R induced a temporary GJIC injury in cultured HUVEC, and which reached the maximal level after 2 h of reoxygenation, too.

Relationship between this GJIC injury and PTK activation

Then we investigated what caused this GJIC injury. The free radical scavenger DMSO (0.5 % v/v), antioxidant SOD (300 U/ml), protein kinase C (PKC) inhibitor calphostin C (1 × 10^-7 M), PTK inhibitor genistein (10 μM), and tyrosine phosphatase inhibitor vanadate (1.5 mM), were examined for their effects on H/R-injured GJIC. As demonstrated in Figure [2] a, except vanadate, the other agents did not interfere with normal GJIC (fluorescence recoveries after 15 min of photobleaching range from 92 % to 95 %). At 2-h reoxygenation (Fig. [2] b), GJIC in control cells was reduced to about 74 %, DMSO and SOD showed no protective effects (75 % and 72 %, respectively). Calphostin C partly inhibited this reduction in GJIC (86 %). Vanadate enhanced the reduction in GJIC further (56 %), whereas genistein completely blocked this H/R-induced GJIC injury. These data suggested that PTK activation was the major cause that led to the H/R-induced GJIC injury.

Effects of natural compounds on GJIC

Subsequently we assessed the effects of compounds 1 - 7, 29, and 52 - 54 on H/R-injured GJIC. Their influence on GJIC under normoxia was first examined; as shown in Figure [3] a, compounds 1 - 7 did not change GJIC (91 % - 96 %); cells treated with compounds 52 - 54 demonstrated relatively weak GJIC levels (87 % - 90 %) compared with those of control and compounds 1 - 7, but the differences were not so clear. Compound 29 exhibited a small but obviously reduced level of GJIC (86 %). At 2-h reoxygenation (Fig. [3] b), GJIC in control cells declined to 77 %. Although compounds 2 - 4 protected this GJIC injury in part (about 85 %), an analogue of them, i.e., compound 1, however, reduced the GJIC level more severely (68 %) than that of control (77 %). Compounds 5 and 6 exerted complete protections on H/R-injured GJIC (95 % and 93 %, respectively). Compounds 7 and 52 - 54 showed similar protections against the reduction in GJIC (86 % - 90 %). In contrast, compound 29 significantly enhanced this reduction in GJIC further (65 %).

Interference with the gap junctional protein connexin 43

In order to find out the most directly related reasons, compounds 2, 5, 29, and 53 were selected according to both their structures and inhibitory effects (on PTK activation and GJIC reduction), and explored for their influence on connexin 43 [Cx43, a connexin protein with 43,000 Da molecular weight, is an important connexin for modulating GJIC in HUVEC [8]].

Cx43 usually exists as three forms, i.e., the non-phosphorylated form (NP, about 42 Kda), the phosphorylated form (P1, about 43 Kda), and the hyperphosphorylated form (P2, about 46 Kda). Among them, NP is beneficial for GJIC, but P1 and P2 (particularly P2) block GJIC [9]. As shown in Figure [4], the normoxic Cx43 mainly manifested as NP and P1, together with a small amount of P2. Treatment with compounds did not change the normoxic Cx43 (data not shown). After 2 h of reoxygenation, the amount of P1 and P2 obviously increased; similarly, compound 29 stimulated the increases of P1 and particularly P2. However, these H/R-induced increases in P1 and P2 were reduced by treatment with compound 5, the Cx43 protein mainly manifested as the NP form. Treatment with compound 2 inhibited H/R-induced increases in P1 and P2, although still showing slightly larger amounts of P1 and P2 than under normoxia. Treatment with compound 53 made the Cx43 protein degrade, all of them (NP, P1 and P2) decreased.

Discussion

In the present study, we showed that H/R induced a GJIC injury that demonstrated the same time course as that of H/R-induced PTK activation. When PTK was maximally activated after 2 h of reoxygenation, GJIC was most severely injured; and after 6 h of reoxygenation, both of them nearly recovered to the normal level. Furthermore, the results in Figure [2] indicated that the major cause of this GJIC injury was just the H/R-activated PTK, in that PTK inhibitor genistein completely blocked the reduction in GJIC. Moreover, Figure [2] also showed that this GJIC injury was not related to free radicals and ROS, but in a certain sense it resulted from activation of PKC.

Our ultimate goal was to seek anti-I/R or anti-H/R bioactive agents from constituents of traditional Chinese medicines. In Part 1, we have demonstrated that compounds 2 - 7 and 52 - 54 inhibited H/R-induced PTK activation. Since H/R-induced injury was mainly ascribed to PTK activation (Fig. [2]), it was, therefore, necessary and valuable to explore whether these compounds can protect H/R-induced injury in GJIC. The results in Figure [3] b indicate that they really inhibited H/R-injured GJIC, though complete inhibitory effects like that of compound 5 were not observed for all of them. In contrast, compound 29 showed a similar effect to vanadate [a classical tyrosine phosphatase inhibitor, and a PTK stimulator as well [10]]; it reduced the GJIC levels both under normoxia and under H/R, probably because it enhanced PTK activation (Part 1). As to compound 1, although it did not interfere with H/R-induced PTK activation, it caused a more serious injury in GJIC (68 %, Fig. [3] b). This might be due to its cytotoxicity under H/R, because cell death phenomena were visible at 2-h reoxygenation even at the C concentration (1.9 μg/ml).

The basis of GJIC is the so-called connexin protein. Six connexins construct the hemichannel of a gap junction, and two hemichannels from two opposing cells form an integrated gap junction. Changes of connexin proteins directly regulate the GJIC level. In HUVEC, Cx43 is an important connexin for modulating GJIC [8]; it modulates the GJIC level largely through phosphorylation and de-phosphorylation. Phosphorylation on tyr or ser/thr residues of the C-terminal of Cx43 blocks GJIC. The results in Figure [4] showed that H/R increased the amount of P1 and P2 of Cx43, suggesting that Cx43 was phosphorylated by H/R, so inhibition of GJIC was observed (Fig. [1]). Moreover, Figure [2] b indicated that activated PTK was the major cause for inducing H/R-injured GJIC; therefore, the H/R-induced increased phosphorylation (P1 and P2) of Cx43 might be mainly mediated by PTK. This concept was supported by treatments with compounds 2, 5, 29, and 53. Compound 29 enhanced PTK activation (Part 1), so it promoted the H/R-induced increase in P2 of Cx43 (Fig. [4]); as a result, GJIC was significantly inhibited (Fig. [3] b). In contrast, compound 5 inhibited PTK activation (Part 1), so the H/R-increased P1 and P2 of Cx43 were blocked by it, and Cx43 mainly manifested as the NP form (Fig. [4]); thus, cells treated with compound 5 exhibited perfect GJIC (Fig. [3] b). On the other hand, treatment with compound 2 did not completely inhibit H/R-induced PTK activation (Part 1), therefore, even the H/R-induced increases in P1 and P2 were obviously inhibited by it, the Cx43 protein still exhibited a slightly increased amount of P1 compared with that under normoxia (Fig. [4]), so GJIC was inhibited (Fig. [3] b). As to compound 53, although treatment with it did not strongly inhibit PTK activation (Part 1), the H/R-induced phosphorylation of Cx43 was obviously inhibited by it; nevertheless, NP of Cx43 was also significantly reduced (Fig. [4]), suggesting that compound 53 induced protein degradation of Cx43, though the mechanism was not clear. Unlike compound 5, Cx43 from cells treated with compound 53 mainly manifested as P1 and P2 but in lower amounts than under normoxia. Since NP of Cx43 is beneficial for GJIC, it is not unexpected that GJIC of cells treated with compound 53 failed to recover to the normal level (Fig. [3] b). In other words, the influence of these compounds on H/R-injured GJIC may be principally through their effects on PTK activation, thereafter, on tyrosine phosphorylation of Cx43. However, the involvement of other kinases should not be neglected. The results in Figure [2] b implied that PKC played a small part in H/R-induced GJIC injury, so the H/R-phosphorylated Cx43 may also contain those from ser/thr phosphorylation by PKC, and which might also be modulated by compound treatment. But they were far less important than PTK (Fig. [2] b).

The data in Part 1 and Part 2 suggested that if some compound could inhibit H/R-induced PTK activation, then it might inhibit the H/R-induced reduction in GJIC, and thereby protect HUVEC from H/R injury. Based on this concept, compound 5 (i.e., β-sitosterol) was thought to show the strongest activity against H/R. In fact, many reports about the action of β-sitosterol against cardiovascular diseases [11] [12] [13] have been published. Therefore, we concluded that it was feasible to screen for and detect anti-I/R or anti-H/R bioactive agents by means of this in vitro H/R model through examining PTK activation, and it was also suitable as an activity-guided fractionating assay system. However, in order to clarify the potential of this assay system, further studies on animal models are required.

Acknowledgements

The authors wish to thank Yong Jiang, Naili Wang, Junpong Peng, Zhenrong Fan, Gexia Qu, Feng Qiu, Yueping Pei, and Fakui Chen for their kind cooperation and for providing us with the test compounds.

References

• 1 Pinsky  D,, Oz  M,, Liao  H,, Morris  S,, Brett  J,, Sciarra  R,, Karakurum  M,, Van Lookerren-Campagne  M,, Platt  J,, Vowygrod  R.. .  J. Clin. Invets.. 1993;;  92 2994-3002
  PubMedSearch in Google Scholar
• 2 Hagar  H,, Ueda  N,, Shah  S V.. .  Kidney Int.. 1997;;  51 1747-1753
  PubMedSearch in Google Scholar
• 3 Polacek  D,, Bech  F,, McKinsey  L F,, Davies  P F.. .  J. Vasc. Res.. 1997;;  34 19-30
  PubMedSearch in Google Scholar
• 4 Xie  H Q,, Hu  V W.. .  Exp. Cell Res.. 1994;;  214 172-176
  CrossrefPubMedSearch in Google Scholar
• 5 Nicholson  S M,, Bruzzone  R.. .  Cur. Biology. 1997;;  7 R340-R344
  PubMedSearch in Google Scholar
• 6 Spray  D C.. .  Circ. Res.. 1998;;  83 679-681
  PubMedSearch in Google Scholar
• 7 Inoguchi  T,, Ueda  F,, Umeda  F,, Yamashita  T,, Nawata  H.. .  Biochem. Biophys. Res. Commun.. 1995;;  208 492-497
  CrossrefPubMedSearch in Google Scholar
• 8 Harold  V R,, Marjan  V K,, Cuc  A,, Martin  R,, Antoni  V G,, David  G,, Habo  J.. .  Am. J. Physiol.. 1997;;  272 C117-C130
  PubMedSearch in Google Scholar
• 9 Mohammad  Z H,, Peng  A,, Alton  L B.. .  J. Cell. Physiol.. 1998;;  174 66-77
  CrossrefPubMedSearch in Google Scholar
• 10 Zipser  Y,, Piade  A,, Kosowers  N S.. .  FEBS Lett.. 1997;;  406 126-130
  CrossrefPubMedSearch in Google Scholar
• 11 Faizi  S,, Siddiqui  B S,, Sallem  R,, Aflab  K,, Shaheen  F,, Gilani  A H.. .  Planta Med.. 1998;;  64 225-228
  PubMedSearch in Google Scholar
• 12 Field  F J,, Born  E,, Mathur  S N.. .  J. Lipid Res.. 1997;;  38 348-360
  PubMedSearch in Google Scholar
• 13 Weizel  A,, Richter  W O.. .  Eur. J. Med. Res.. 1997;;  2 265-269
  PubMedSearch in Google Scholar
• 14 Zhang  YW,, Morita  I,, Shao  G,, Yao  XS,, Murota  Si.. .  Planta Medica. 2000;;  66 114-118
  Thieme ConnectPubMedSearch in Google Scholar

Prof. Dr. Ikuo Morita

Section of Cellular Physiological Chemistry

Graduate School

Tokyo Medical & Dental University

1-5-45 Yushima

Bunkyo-Ku

Tokyo 113-8549

Japan

Email: morita.cell@dent.tmd.ac.jp

Phone: +81-3-5803-0212

Fax: +81-3-5803-5575

References

• 1 Pinsky  D,, Oz  M,, Liao  H,, Morris  S,, Brett  J,, Sciarra  R,, Karakurum  M,, Van Lookerren-Campagne  M,, Platt  J,, Vowygrod  R.. .  J. Clin. Invets.. 1993;;  92 2994-3002
  PubMedSearch in Google Scholar
• 2 Hagar  H,, Ueda  N,, Shah  S V.. .  Kidney Int.. 1997;;  51 1747-1753
  PubMedSearch in Google Scholar
• 3 Polacek  D,, Bech  F,, McKinsey  L F,, Davies  P F.. .  J. Vasc. Res.. 1997;;  34 19-30
  PubMedSearch in Google Scholar
• 4 Xie  H Q,, Hu  V W.. .  Exp. Cell Res.. 1994;;  214 172-176
  CrossrefPubMedSearch in Google Scholar
• 5 Nicholson  S M,, Bruzzone  R.. .  Cur. Biology. 1997;;  7 R340-R344
  PubMedSearch in Google Scholar
• 6 Spray  D C.. .  Circ. Res.. 1998;;  83 679-681
  PubMedSearch in Google Scholar
• 7 Inoguchi  T,, Ueda  F,, Umeda  F,, Yamashita  T,, Nawata  H.. .  Biochem. Biophys. Res. Commun.. 1995;;  208 492-497
  CrossrefPubMedSearch in Google Scholar
• 8 Harold  V R,, Marjan  V K,, Cuc  A,, Martin  R,, Antoni  V G,, David  G,, Habo  J.. .  Am. J. Physiol.. 1997;;  272 C117-C130
  PubMedSearch in Google Scholar
• 9 Mohammad  Z H,, Peng  A,, Alton  L B.. .  J. Cell. Physiol.. 1998;;  174 66-77
  CrossrefPubMedSearch in Google Scholar
• 10 Zipser  Y,, Piade  A,, Kosowers  N S.. .  FEBS Lett.. 1997;;  406 126-130
  CrossrefPubMedSearch in Google Scholar
• 11 Faizi  S,, Siddiqui  B S,, Sallem  R,, Aflab  K,, Shaheen  F,, Gilani  A H.. .  Planta Med.. 1998;;  64 225-228
  PubMedSearch in Google Scholar
• 12 Field  F J,, Born  E,, Mathur  S N.. .  J. Lipid Res.. 1997;;  38 348-360
  PubMedSearch in Google Scholar
• 13 Weizel  A,, Richter  W O.. .  Eur. J. Med. Res.. 1997;;  2 265-269
  PubMedSearch in Google Scholar
• 14 Zhang  YW,, Morita  I,, Shao  G,, Yao  XS,, Murota  Si.. .  Planta Medica. 2000;;  66 114-118
  Thieme ConnectPubMedSearch in Google Scholar

Prof. Dr. Ikuo Morita

Section of Cellular Physiological Chemistry

Graduate School

Tokyo Medical & Dental University

1-5-45 Yushima

Bunkyo-Ku

Tokyo 113-8549

Japan

Email: morita.cell@dent.tmd.ac.jp

Phone: +81-3-5803-0212

Fax: +81-3-5803-5575