Printer Friendly

Effects of hemodynamic forces on gene expression and signal transduction in endothelial cells.

Vascular endothelial cells respond to mechanical forces, such as shear stress, by expressing a number of immediate early genes. One of these genes encodes monocyte chemotactic protein-1 (MCP-1, Shyy et al., 1994), which plays a significant role in atherogenesis. This presentation summarizes the work done in our laboratory on the effects of shear stress on signal transduction and on the expression of the MCP-1 gene. Human umbilical vein endothelial cells and many other cell types respond to arterial level of shear stress (10-30 dynes/[cm.sup.2]) with a transient increase of MCP-1 gene expression that peaks at 1.5 h (Shyy et al., 1994). Sequential deletion of the 5[prime] promoter region of the MCP-1 gene and site-specific mutation of the cis-elements show that one of the two copies of the putative TPA-responsive elements (TRE), with the sequence TGACTCC, is critical for shear-stress induction of the MCP-1 gene (Shyy et al., 1994). Transactivation assays indicate that activating protein-1 (AP-1, composed of Jun-Fos heterodimer or Jun-Jun homodimer) is the nuclear binding protein responsible for shear activation of MCP-1.

The signal transduction pathways leading to the activation of AP-1/TRE by shear stress have been investigated with protein kinase assays and dominant negative mutants of signaling molecules in the pathways of the c-jun N[H.sub.2] terminal kinases (JNK) and the extracellular signal-regulated kinases (ERK). JNK(K-R) and MEKK(K-M), the catalytically inactive mutants of, respectively, JNK1 and MEKK in the JNK pathway, attenuate the shear-induced TRE responses. The dominant negative mutant of Ha-Ras blocks the shear-activation of JNK and the downstream TRE. These results indicate that shear stress activates primarily the Ras-MEKK-JNK pathway in inducing MCP-1 gene expression (Li et al., 1996, 1997).

Shear stress rapidly and transiently increases the association between growth factor receptor-2 (Grb2) and Son of sevenless (Sos) in bovine aortic endothelial cells. Shear stress also augments the tyrosine phosphorylation of FAK and its association with Grb2. FAK(F397Y) and FAK(F925Y), the negative mutants of FAK, attenuate the shear-stress induction of the kinase activity of HA-JNK (Li et al., 1997). Similarly, the shear-stress-induced activities of luciferase (Luc) reporter gene linked to MCP-1 or 4xTRE promoters are decreased by these FAK mutants. Thus, the Tyr-397 (autophosphorylation site) and the Tyr-925 (binding site for Grb2 src homology domain 2 [SH2]) of FAK are critical for its activation in response to shear stress. pGrb2-SH2, which encodes the SH2 domain of Grb2, and p[Delta]mSos1, in which the guanine nucleotide exchange domain has been deleted, also attenuate induction of HA-JNK, MCP1-Luc, and 4XTRE-Luc by shear stress. These results indicate that the FAK-Grb2-Sos-Ras-MEKK-JNK system is a major signaling pathway mediating the shear-induced gene expression [ILLUSTRATION FOR FIGURE 1 OMITTED]. Other signaling mechanisms may also be involved; e.g., we have found that shear stress causes an increase of protein kinase C, especially in the cortical region, in the endothelial cell (Hu and Chien, 1997).

Studies on the tissue factor gene show that the shear-stress-responsive element is Sp1 in the GC rich region of its promoter and that the two copies of TREs there are not critical (Lin et al., 1997). Coupled with the finding by Resnick et al. (1993) that the nucleotide sequence GAGACC is the shear-stress-responsive element for the platelet-derived growth factor B chain, our results indicate that shear stress activates different cis-elements in different genes (Shyy et al., 1995). The orchestration of various cis-elements and signaling pathways may play an important role in determining the complex gene regulation in response to mechanical forces in health and disease, including those induced by microgravity.


Supported by NIH Research Grant HL-44147.


Hu, Y.L., and S. Chien. 1997. Effects of shear stress on protein kinase C distribution in endothelial cells. J. Histochem. Cytochem. 45: 237-249.

Li, Y. S., J. Y. J. Shyy, S. Li, J.D. Lee, B. Su, M. Karin, and S. Chien. 1996. The Ras/JNK pathway is involved in shear-induced gene expression. Mol. Cell. Biol. 16: 5947-5954.

Li, S., M. Kim, D. D. Schlaepfer, T. Hunder, J. Y. J. Shyy, and S. Chien. 1997. The fluid shear stress induction of JNK pathway is mediated through FAK-Grb2-sos. J. Biol. Chem. 272: 30455-30462.

Lin, M. C., F. Almus-Jacobs, H.-H. Chen, G. C. N. Parry, N. Mackman, J. Y. J. Shyy, and S. Chien. 1997. Shear stress induction of tissue factor gene expression and phosphorylation of SPI. J. Clin. Invest. 99: 737-744.

Resnick, N., T. Collins, W. Atkinson, D. J. Bonthron, F. Dewey, Jr., and M. A. Gimbrone, Jr. 1993. Platelet-derived growth factor B chain promoter contains a cis-acting fluid shear-stress-responsive element. Proc. Natl. Acad. Sci. USA. 90: 4591-4595.

Shyy, Y.-J., and S. Chien. 1997. Role of integrins in cellular response to mechanical stress and adhesion. Curr. Opin. Cell Biol. 9: 707-713.

Shyy, Y.-J., H.-J. Hsieh, S. Usami, and S. Chien. 1994. Fluid shear stress induces a biphasic response of human monocyte chemotactic protein-1 gene expression in vascular endothelium. Proc. Nat. Acad. Sci. USA 91: 4678-4682.

Shyy, Y. J., Y. S. Li, M. C. Lin, H. Gregersen, W. Chen, S. Yuan, F. Almus, S. Usami, and S. Chien. 1995. The shear stress-mediated gene expression and the cis-elements involved. J. Biomechanics 28: 1451-1457.

Shyy, Y.-J., M. C. Lin, J. Han, Y. Lu, M. Petrime, and S. Chien. 1995. TPA responsive element as a common cis-acting element for mechanical and chemical stimuli. Proc. Nat. Acad. Sci. USA. 92: 8069-8073.


KOWALCZYK: Are the cells exposed to fresh medium or serum at the beginning of the experiment? Do you rinse them before flow is initiated? A lot of the changes you describe resemble transient effects associated with growth factor addition. Perhaps the initiation of flow over the cells is changing local concentrations of growth factors or metabolites near the cell surface.

CHIEN: Cells were washed with serum-free culture medium before initiation of flow. The responses I have described are not serum-dependent. Chemical stimuli such as TPA and mechanical stimulation can mutually affect each other. If you pretreat the cell with TPA to down-regulate the pathway to JNK, and then stimulate with mechanical shear, you get much less response, suggesting that both effects share the same pathway.

COULOMBE: Would you repeat the changes that are taking place in the three filament systems of the cells, and also speculate on the kinetics as these relate to the signaling events that you described?

CHIEN: The mRNA response starts within an hour or so, but cytoskeletal reorganization takes a long time. Thus, the shear-induced gene expression occurs much earlier than detectable reorganization of the cytoskeletal fibers. All the responses in the pathway occur in the correct sequence. Ras is activated upstream in less than 1 min. Activation of JNK occurs within a few minutes, and the MCP-1 gene is activated later. Thus, there is a sequential activation. The morphological changes of cytoskeletal reorganization are the long-term change; they do not precede the changes in signal transduction and gene expression. Now there may be some subtle changes in the cytoskeletal proteins that are responsible for the signaling and gene expression, but these cannot be seen morphologically at an early stage. Actin seems to be playing a role here. The actin filaments, which were present as peripheral bands in the static condition, gradually disappear after shear and reorganize themselves into long stress fibers. These fibers tend to move from the basal side toward the top of the cell. Using confocal microscopy we observed that the microtubules and the intermediate filaments, which were originally around the nuclei, began to move toward the base after shearing. The nature of the association of these cytoskeletal elements with each other has been mentioned several times during this workshop.

BAKER: You showed a peak of activity for many factors right after you start your shear stress. Is there a physiological static state, and does this apply to starting exercise, where the body sees a sudden, greater stress?

CHIEN: What we have shown here is not a physiological response, because we start from a static situation. This only applies to a pathological state, like re-perfusion after stoppage of flow. In the physiological state, long-term shear occurs in the straight part of the vasculature. As shown in my first slide, the vasculature in the straight part of the thoracic aorta is sheared all the time, and the MCP-1 gene is therefore down-regulated, and monocytes tend not to be attracted there. At branch points, the flow is unsteady and there are flow reattachment areas near the bifurcation. These reattachment areas don't stay at the same spot; they move back and forth. The spatial effects were discussed by Dr. Fujiwara. It is my belief that the spatial and temporal variations are very important. At one moment there is no shear; then there is shear in one direction; then there is shear in another direction. The endothelial cells in these areas can sense what is going on. Our preliminary results, using an in vitro analog of those branch points, do confirm the hypothesis that the reattachment areas in the branch regions are vulnerable. Concerning your question on exercise, I think during exercise shear increases, not only in the straight part of the aorta, but this increase also invades the bifurcation. The exposure of these regions to fairly steady high shear may down-regulate the genes. This may be why physicians always tell us to exercise three times a day, at least 30 min each time. I think that's what it takes for the down-regulation to occur.

CHEN: I'm wondering why the experiments do not start with a constant shear flow which is then changed. Do you use turbulent flows? I would imagine there is recirculation at those bifurcation points.

CHIEN: We are doing these experiments with recirculation and reattachment.

CHEN: Is there anything known about that?

CHIEN: Yes. The reattachment area seems to be very important. In the vicinity there are eddies, secondary flows, and stagnation. These are the sites of action. You ask why we don't start with a baseline flow. We are doing these experiments, and other groups are also doing them. Those studies have more physiological relevance, but it is much easier to see the effect if you start from zero. If you want to work out the pathway from the point of view of molecular biology, again it is simpler to start from zero. You can then go to a more physiologically relevant system.

BRUINSMA: You showed a physical stimulus that produces a gene response; then you show two parallel pathways, each containing eight steps. Why is this so complex? I can imagine that a number of amplification steps cross-link to other stimuli. Can you comment on this?

CHIEN: They are cross-linked to other pathways. I have presented a simplification which you say is complex. In fact, it is like a neural network, where there are layers and layers of interactions. The cells in our body have complex interacting pathways, and we are not dealing with simple situations. To deal with our daily environment and to adjust to microgravity, we must have a very finely tuned system, with redundancy and delicacy of controls.

BRUINSMA: Is there intertalk between the pathways? Do you see saturation?

CHIEN: If you vary the shear stress, you do see a saturation effect. We don't see any differences beyond about 10 dynes/[cm.sup.2]. We are very interested in the kinetics of this process. We are interested in modeling all of this, but it's too complex, and we don't have sufficient data to do the modeling. There is crosstalk among the pathways. Although we have some kinetic data describing how quickly each step happens in one pathway, we have no good means of manipulating every step so as to examine how the next step behaves. We need to establish the transfer function for each station. We're not getting that yet. With data from experiments being done in a number of labs, we hope that we can start to look into that.

STEWART: Continuing on with the complexity of the signaling pathways, one of the difficulties, always, is knowing which part is being stimulated, because quite often one pathway will stimulate an adjacent pathway. I was wondering whether two classes of experiment that you have done might help to distinguish where the signal is actually coming from. The first one would be to introduce the constitutively active Ras mutants. Strictly, putting in the N17 is wiping out the GEF, rather than actually showing the positive involvement of Ras. Then, wondering backwards, is the signal actually deriving from a surface receptor, or not? I know you mentioned, for example, that the EGF receptor was phosphorylated. That would make me wonder whether, if you treated these sort of cells with EGF, you would actually get the induction of the MCP-1 protein that you are measuring. Or is it that there is some other part of the cell that is sensing the pressure and that there might be a difference, then, in the response in stimulating a surface receptor to say, FAK?

CHIEN: These are excellent questions. To answer the question about whether the protein is induced, it is indeed. Not only is the protein induced, it is also functional. We get enhancement of monocyte adhesion after applications of shear stress. We are doing experiments with over-expression of the wild type, and the results do fit in.

GOLDMAN: I just want to make one comment. We have done experiments with Peter Davies in which we shear cells and then detect very significant changes in the cytoskeleton within 60 min. There are profound changes in the organization. It does not take 12 to 24 hours for us to see this effect.

CHIEN: What I meant was that you need that long to see the kind of morphological changes I showed.

GOLDMAN: The shape changes take a long time, but the cytoskeletal changes begin very quickly.

CHIEN: That's right. Actually they do not go to this final state; they go through a whole series of contortions. They may first align perpendicular to flow or swirling patterns, as mentioned by Dr. Fujiwara. That's what I meant. I agree with you completely, but to get to the picture I showed, it did take 1224 hours of shearing. In fact, I said that the actin plays a role, and there may be changes occurring in less than 60 min that we cannot even see morphologically.

GOLDMAN: I would say within minutes.

CHIEN: I agree with you completely. One other thing I wanted to mention is that the mechanical force probably activates many types of molecules, including receptors and channels. These effects must be summed to give the final response. That's why, whenever we try to block these responses, we can never block them completely. It is a summation effect of many responses. Each may be nonspecific and weak, but the sum total gives us a significant response. This is one of the differences from chemical stimulation, where the ligand-receptor interaction is specific.
COPYRIGHT 1998 University of Chicago Press
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 1998 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Title Annotation:The Cytoskeleton: Mechanical, Physical, and Biological Interactions; includes panel discussion
Author:Chien, Shu; Shyy, John Y.J.
Publication:The Biological Bulletin
Date:Jun 1, 1998
Previous Article:A role for pp125FAK in suppression of apoptosis in fibroblasts.
Next Article:Effect of flow on gene regulation in smooth muscle cells and macromolecular transport across endothelial cell monolayers.

Terms of use | Privacy policy | Copyright © 2019 Farlex, Inc. | Feedback | For webmasters