First published online October 22, 2003
doi: 10.1242/10.1242/jcs.00755
Potential role of MCP-1 in endothelial cell tight junction `opening': signaling via Rho and Rho kinase
Svetlana M. Stamatovic1,
Richard F. Keep1,3,
Steven L. Kunkel2 and
Anuska V. Andjelkovic1,2,*
1 Department of Neurosurgery, University of Michigan Medical School, Ann Arbor, MI 48109, USA
2 Department of Pathology, University of Michigan Medical School, Ann Arbor, MI 48109, USA
3 Department of Physiology, University of Michigan Medical School, Ann Arbor, MI 48109, USA
View larger version (44K):
[in a new window]
|
Fig. 1. mBMEC express the CCR2 receptor. (A) Expression of CCR2 mRNA in quiescent and IL-1ß-treated cells examined by RT-PCR. GADPH, glyceraldehyde phosphate dehydrogenase; PC, positive control: a mixture of cDNA provided by the supplier of the chemokine receptors and GADPH primers. Volumes of CCR2 bands were expressed as percentage of control gene (GADPH) bands amplified in same PCR reactions. Each bar represents the mean±s.e.m. of three independent experiments. Asterisks indicated significant differences (P<0.001) from corresponding control levels. (B) Western blot analysis of CCR2 receptors. Cells were either quiescent mBMEC in first passage or were treated with 10 ng/ml IL-1ß for 6, 12 and 24 hours. PC, positive control: lysate of murine peritoneal macrophages. Results are presented as means±s.e.m. of three independent experiments; *P<0.01. (C) CCR2+/+ or CCR2–/– mBMEC were incubated with rabbit anti-CCR2 antibody followed by fluorescein-conjugated secondary antibody. Scale bar 80 µm. (D) mBMEC chemotaxis. Different concentrations of MCP-1 were placed in the lower wells of chemotaxis chambers and mBMEC were placed in the upper wells. After incubation at 37°C the cells on the Neuroprobes filter were counted. Chemotaxis index represents the ratio of migrating endothelial cells in the presence of MCP-1 and in the absence of MCP-1 (control medium). In a separate set of experiments, MCP-1 was also added to a suspension of mBMEC or anti-MCP-1 antibody (1 µg/ml) was added to the lower chamber.
|
|
View larger version (21K):
[in a new window]
|
Fig. 2. Effect of MCP-1 on brain endothelial barrier function. (A) Confluent CCR2+/+ or CCR2–/– mBMEC, were grown on the filters in a Transwells chamber system and TEER measurements made. There were no differences in TEER between the two strains in the absence of MCP-1. However, after exposure to 100 nM MCP-1, the TEER in wild-type mouse cells decreased dramatically while that in CCR2–/– cells was unaffected. Addition of a neutralizing antibody to MCP-1 partially restored the TEER in CCR2+/+ mice. (B) Dose response for the effects of MCP-1 on TEER in wild-type cells. (C) The effect of MCP-1 (100 nM) on [14C]inulin permeability in CCR2+/+ or CCR2–/– mice. Results are presented as means ± s.e.m., *P<0.001.
|
|
View larger version (91K):
[in a new window]
|
Fig. 3. Effect of MCP-1 on actin cytoskeleton and intercellular TJs. Confluent CCR2+/+ and CCR2–/– mBMEC, were treated with recombinant mouse MCP-1 (100 nM) for the indicated time period (15 minutes and 2 hours) or served as normal controls (control). The cells were then fixed and stained with anti-occludin, ZO-1, ZO-2, claudin-5 antibodies or phalloidin Alexa 488 for F-actin. Untreated quiescent mBMEC (control, 7 days after initial plating) showed a typical polygonal shape, with actin filament distributed primarily in the cortical ring with a few stress fiber spanning the cells. They also had very specific continuous staining for occludin, ZO-1, ZO-2 and claudin-5 localized along the cell margins, possibly at the sites of cell-cell contact. In CCR2+/+ cells, treatment with MCP-1 induced marked structural alterations in the distribution of actin filaments and TJ proteins in a time-dependent manner. In the absence of CCR2 receptors, the effect of MCP-1 on the actin cytoskeleton and TJ proteins were abrogated. Scale bar: 200 µm.
|
|
View larger version (29K):
[in a new window]
|
Fig. 4. Shift of TJ proteins from soluble to insoluble phase. Confluent CCR2+/+ and CCR2–/– mBMEC were subjected to 100 nM MCP-1 for 15 minutes or 2 hours. Triton X-100-soluble and Triton X-100-insoluble fractions were collected. Immunoblots of those fractions were then probed with anti-occludin, anti-claudin-5, anti-ZO-1 and anti-ZO-2 antibodies. (B) Immunoblots were analyzed and quantified with NIH Image software. Data represent means±s.e.m. of three independent experiments.
|
|
View larger version (71K):
[in a new window]
|
Fig. 5. Analysis of signal pathways in response to MCP-1. (A) cDNA microarray analysis of signal pathways induced by MCP-1. Data represent relative expression levels of specific genes normalized using a housekeeping gene control and compared with the group of untreated cells. Arrowheads indicate several representative genes whose levels increased in response to MCP-1. (B) Confluent mBMEC were pretreated with the following inhibitors: PD98059, SB 203580, LY294002, Ro-37840, Y27632, W7 and U7322 for 1 hour at 37°C. Recombinant mouse MCP-1 (100 nM) was then added for 2 hours. The cells were then fixed and processed for immunocytochemistry using anti-ZO-1 and Alexa 488 Phalloidin. The samples were viewed on a laser scanning Zeiss confocal microscope. Scale bar: 200 µm. (C) Changes in TEER during treatment with specific inhibitors of different signal pathways. Confluent mBMEC were grown on the filters in a Transwells chamber system and pretreated with the following inhibitors: PD98059, SB 203580, LY294002, Ro-37840, Y27632, W7 and U7322 for 1 hour at 37°C. After that the cells were exposed to MCP-1 (100 nM) in the presence of inhibitors. TEER was measured every 15 minutes over a time period of 2 hours. Results are presented as means±s.e.m., *P<0.01; **P<0.001.
|
|
View larger version (22K):
[in a new window]
|
Fig. 6. MCP-1 transiently activates Rho in mouse BMEC. (A) Confluent mBMEC were treated with 100 nM murine MCP-1 for 10, 20, 30 minutes, 1 or 2 hours. Cell lysates were subject to affinity precipitation using Rhotek in recombinant protein agarose conjugated which specifically precipitates active RhoA (Rho-GTP). Total Rho indicates total amount of active and inactive Rho in the mBMEC. The immunoblot represents one of three independent experiments. (B) Densitometric analysis of MCP-1-induced activation of RhoA. Data are means±s.e.m., n=3 independent experiments *P<0.001; (C) Confluent mBMEC were subject to treatment with a Rho kinase inhibitor Y27632 10 µM for 30 minutes) or with specific inhibitor of RhoA, C3 exoenzyme (5 µg/ml for 18 hours), or transiently transfected with dominant negative mutant T19NRho. Western blot was performed using an antibody specific for RhoA.
|
|
View larger version (27K):
[in a new window]
|
Fig. 7. Effect of inhibition of RhoA and Rho kinase on MCP-1-induced changes in brain endothelial permeability. (A) Western blot analysis of TJ proteins (occludin, claudin-5, ZO-1 and ZO-2) after treatment of cell with Rho kinase and RhoA inhibitors, Y27632 and C3 or transiently transfected with dominant negative mutant T19NRho. Effect of Rho and Rho kinase inhibitors on changes (B) TEER and (C) [14C]inulin permeability (P) induced by MCP-1. Results are presented as means±s.e.m., n=3 independent experiments.
|
|
View larger version (84K):
[in a new window]
|
Fig. 8. Effect of inhibition of RhoA and Rho kinase on MCP-1-induced alterations in actin and TJ proteins. Confluent mBMEC were pretreated for 30 minutes with 10 µM Y27632, or 18 hours with 5 µg/ml C3 exoenzyme, or transiently transfected with T19NRho and then were exposed to 100 nM MCP-1 for 2 hours. The cells were then fixed and processed for immunocytochemistry using anti-ZO-1, -ZO-2, -occudin, and claudin-5 antibodies and Alexa 488 Phalloidin for F-actin. Arrows indicate organization of actin and TJ proteins in presented experimental groups. Scale bar: 200 µm.
|
|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati 
Twitter What's this?
© The Company of Biologists Ltd 2003
