Human Endothelial And Smooth Muscle Cells Biology Essay

Published: Last Edited:

This essay has been submitted by a student. This is not an example of the work written by our professional essay writers.

Angiogenesis is a complex process requiring the participation of endothelial and smooth muscle cells migrating and assembling in a well-coordinated manner to form capillaries and larger vessels. Migration of human aortic smooth muscle cells (SMC) is induced optimally by low (10-20 nM) levels of sphingosine 1-phosphate (S1P) SPP is now widely used for sphingosine phosphate phosphatase, S1P is now used for the substrate and to name the receptors, but is strongly inhibited at the levels of S1P most stimulatory for endothelial cell chemotaxis (0.5-1.0 mM). We proposed that this differential response to S1P is the reflection of the ability of cells in vivo to recognize and respond to a concentration gradient of S1P that is formed when the lipid is released by platelets at the site of trauma 1. Here we demonstrate that the inhibitory effect of SPP could be mimicked by forskolin, while treatment of cells with a protein kinase A (PKA) inhibitor, H89, resulted in complete restoration of their migratory ability. Similarly, the propensity of SMCs to form tube-like structures on the Matrigel substrate was dramatically blocked by S1P, but was totally restored in the presence of H89. The inhibitory effect of elevated S1P on SMC migrationwas paralleled by inhibition of Rac and Cdc42 activation. Strikingly, these small GTPases were not affected by either SPP or cAMP in endothelial cells. Rather, the two proteins appear to be constitutively activated in these cells. These results demonstrate a differential regulationb of G-protein activation by cAMP that co-ordinated to assembly of endothelial and smoth muscle cells in forming new blood vessels. Or something like this to finish off the abstract and make it appealing.

Introduction? What journal is this formatted for?

We showed previously that S1P is a powerful stimulant of endothelial cell migration and capillary tube formation 2,3, both occurring maximally at 0.5-1.0 mM SPP. In contrast, smooth muscle cells migrate optimally at a much lower (20 nM) concentration of S1P, and the process is completely blocked by 500 nM SPP 1. In general, SMCs are strongly attracted to protein growth factors such as PDGF, or EGF 1,4, whereas ECs migrate much less robustly toward protein factors like VEGF (2,5, Fig. 1c). These striking differences in responses between ECs and SMCs led us to examine the biochemical underpinnings of the behavior of these two cell types. Previous work showed that 1 mM SPP causes accumulation of cAMP in SMCs 6. We also have cAMP values for Ecs and SMCs after S1P treatement that we sent to you. In agreement with this observation, we found that the inhibition of SMC migration by SPP is completely overcome by H89, and that forskolin (FSK) has a strongly inhibitory effect on SMC migration, but not on the EC response (Fig. 1a,b). Perplexingly, wortmannin, an inhibitor of phosphatidylinositol 3-kinase (PI 3-K), was completely without effect on EC migration, while totally blocking chemotactic responses of SMCs (Fig. 1 c). These results further emphasize the differences in signaling responses of ECs and smSMCs.

Cell migration is a necessary, but not sufficient, condition of angiogenesis. Once the cells reach their destination, they must assemble in structures that are precursors to mature vessels. We examined the ability of ECs and SMCs to assemble, as a function of S1P concentrations, into capillary-like forms on Matrigel substrate. We found that the propensity to form networks of capillary-like structures correlated with the migratory responses of cells: ECs produce the tubes efficiently with higher S1P concentrations, while SMCs are very sensitive to S1P and show complete breakdown of defined structures at 500 nM SPP (Fig. 2). The ability to form a cellular network is dramatically restored to SMCs upon the addition of H89, suggesting that in these cells both migration and capillary morphogenesis are dependent on the same pathway that is negatively regulated by cAMP and protein kinase A (PKA).

A well-orchestrated recruitment of SMCs to the developing vascular network of ECs is an essential element of angiogenic response 7,8. Thus, one expects that co-incubation of the two cell types on Matrigel will result in an ordered alignment of ECs and SMCs relative to each other. Figure 3 shows differentially stained Ecs and SMCs that have been incubated on the surface of Matrigel in the presence of 20 nM SPP. It is clear from the superimposed images of the cells that the SMCs arrange themselves in a manner that is far from random, enveloping the endothelial cells within a capillary-like structure. Individual nuclei are easily discerned in both cell types. This orderly arrangement is in keeping with the developmental pattern of vasculature formation 9.

Recent studies have linked the activity of S1P receptors, S1P1 and S1P3, and EC chemotaxis, to the PI3-K/Akt pathway. In particular, Hla and coworkers 10-12 established that S1P activation of the migratory pathway in Ecs involves phosphorylation of Akt and that Akt associates with S1P1 receptors, bringing about phosphorylation of S1P1 and the orderly chemotactic response. We compared the pattern of Akt phosphorylation in ECs and SMCs challenged with S1P and VEGF, or S1P and PDGF, respectively, and found that in both cell types Akt phosphorylation was completely eliminated by wortmannin, a PI3-K inhibitor (Fig. 4). The wortmannin inhibition of Akt phosphorylation is consistent with the established signaling pathway, but is at odds with the observed insensitivity of EC chemotaxis to this inhibitor. It is notable that H89 consistently enhanced Akt phosphorylation, indicating that PKA may have a negative regulatory effect on Akt/PI3-K interactions. As expected, PDGF potently stimulated Akt phosphorylation in SMCs while S1P stimulated phosphorylation in ECs, again consistent with the respective migratory responses in chemotactic assays. I think this needs more discussion, e.g., forskolin effects not dealt with and integrated with the observations.

An integral element of chemotactic response is the activation of small GTPases, Rac and Cdc42. These two proteins are involved in cell cycle progression, cytoskeletal reorganization, formation of lamellipodia and filopodia, and cell movement 13. Rac is a known downstream target of Akt 14, and can be regulated by phosphorylation. We, therefore examined the potential effect of S1P and of cAMP-enhancing treatments on the activation of Rac and Cdc42 proteins, since they are essential for cell migration. Figure 5 shows that in SMCs activation of Rac and Cdc42 (formation of GTP-Rac and GTP-Cdc42) was enhanced by 20 nM S1P and was blocked by 500 nM S1P (panels a and b), consistent with the chemotactic behavior of these cells. Ryu and coworkers 15 reported similar inhibition of Rac activation by elevated S1P concentrations, but did not observe Cdc42 activation by either PDGF or SPP in rat aortic smooth muscle cells. They did not examine Cdc42 activation in human SMCs. We do not know the reason for the discrepancy between our results. Dibutyryl-cAMP prevented Cdc42 and Rac? activation by 20 nM S1P. Furthermore, activation by PDGF was inhibited by 500 nM SPP and by forskolin, and restored by H89. These results indicate that cAMP-dependent protein kinase A pathway must play a role in controlling the activation of the Rac and Cdc42 in smooth muscle cells. As expected, wortmannin was strongly inhibitory for the activation of Rac and Cdc42.

In contrast, in endothelial cells, activation of Rac and Cdc42 (Fig. 5, panels c and d) appeared largely unaffected by S1P, VEGF, dibutyryl-cAMP or wortmannin (a slight increase was, in fact, noted sometimes in cells treated with the latter compound). Other investigators have shown that in non-adherent ECs, Rac is fully activated and does not respond to S1P stimulation 11. The same researchers showed that even the unstimulated adherent ECs possess relatively high basal level of GTP-Rac 11. The same is true for GTP-Cdc42, although not GTP-Rac, in S1P1-, S1P3 , or S1P2-transfected CHO cells 16. Our experiments, were performed with cells in suspension and demonstrated that the two GTPases are activated in unstimulated human aortic ECs. This observation helps to explain the apparently contradictory result: while Akt phosphorylation is abolished in the presence of PI3-K inhibitor, the migration of endothelial cells is not. We propose that, under our experimental conditions, the substantial unstimulated level activation for Rac and Cdc42 in ECs serves to bypass the need for Akt activity and allows migration even when PI3-K is blocked. In SMCs, the basal level of activated Rac and Cdc42 is low, the activation of these GTPases requires PI3-K and Akt, and therefore migration is sensitive to wortmannin. Alternatively, ECs utilize a PI3-K-independent pathway of chemotaxis, a pathway that is absent or not utilized in SMCs. Is this true? I do not hink this is established. All we can say is that basal activation is enough.

Previous studies showed that co-cultured ECs and SMCs organize spontaneously into 3-dimensional spheroids in which the SMCs occupy the internal core and the ECs line the surface of the spheroid 17,18. These spheroids have an ability to produce capillary-like sprouts upon co-stimulation with VEGF and Ang2 18. Other investigators demonstrated that TGF-b induces differentiation of mesenchymal cells toward the pericyte/SMC lineage and that TGF-b is also instrumental in stabilizing capillary-like structures in 3-dimensional model of in vitro angiogenesis 19,20. However, ours is the first demonstration that S1P, a bioactive phospholipid released by platelets, differentially modulates migration and capillary network formation by separately cultured ECs and SMCs (Fig. 1 and 2). High concentrations of S1P are profoundly inhibitory to SMC migration and morphological differentiation (because of cAMP accumulation and stimulation of PKA, and because of the resulting inactivation of Rac and Cdc42, Fig. 5), but the inhibition is completely abolished when the activity of protein kinase A is blocked with H89. The accumulation of cAMP is not sensitive to pertussis toxin (not shown), paralleling the inability of the toxin to overcome the migration-inhibitory effect of elevated S1P 1. On the other hand, the Ca2+-chelating agent, BAPTA, abolished cAMP production, consistent with the role for the Ca2+-dependent activation of adenylyl cyclase AC3 21 (data not shown).

Thus, we have correlated chemotactic responses of ECs and SMCs with the ability of these cells to form capillary-like networks on Matrigel support. We have also demonstrated that when ECs and SMCs are co-cultured on Matrigel, the cells organize themselves in a manner expected of a blood vessel, the ECs facing the "lumen", and SMCs lining the outside of the "capillary" (Fig. 3). This is consistent with the manner in which vascular smooth muscle cells organize in vivo around EC tubes 9,22. In contrast to previous reports 19,20, supplementation with TGF-b is not required to promote the association of ECs and SMCs, or for the formation of capillary-like structures.

The observation that the basal level of Rac and Cdc42 (activation or concentration, we need the relative levels of G-proteins in the two cell types) is high in ECs raises a question as to why these cells are not constantly on the move. We note, firstly, that in cells that have attached to the substratum, S1P does induce Rac activation, albeit starting from an apparently high level 11. Secondly, while Rac and Cdc42 are the necessary elements of migratory response, they are, obviously, not the only ones. We propose that the initial stages of eliciting a chemotactic reaction are similar in the two cell types, and require the participation of relevant receptors. These initial stages include ligand binding, receptor phosphorylation and activation, G protein activation, induction of a Ca2+ flux, and the like?. Without receptor-ligand interaction, these steps could not be carried out, and, therefore, migration could not occur. Thus, our data do not abrogate the essentiality of Akt in controlling the various signal transduction pathways, including the specific interaction with the S1P receptors 10-12,16. Even though ECs appear to dispense with the direct participation of PI3-K/Akt in migration, PI3-K is still likely to be an essential component of signaling, in particular that involving Ca2+ flux, adenylyl cyclase activation and interactions with PKA see above.

Our results support the notion of cAMP-dependent protein kinase A acting through PI3-K/Akt to influence migratory behavior of SMCs. First, forskolin (and dibutyryl-cAMP, not shown) inhibits chemotaxis toward S1P. Second, H89, a potent inhibitor of PKA, restores migration of SMCs exposed to high concentration of S1P and the ability of SMCs to form tubular networks (Fig. 1 and 2). These pharmaco-physiological observations have their counterparts in biochemical assays (Figs. 4 and 5). Akt phosphorylation appears to be enhanced in cells treated with H89 and at least slightly diminished in FSK-treated cells; this is also true of Akt in ECs. The sensitivity of Akt phosphorylation to cAMP, or cAMP-enhancing agents is in keeping with the reported interference by cAMP with localization and interaction between Akt and phosphoinositide-dependent kinase PDK1 23. Most importantly, our results have helped us to discern at least in part the molecular basis for the difference in migration between the two cell types: apparently constitutive activation of Rac and Cdc42 GTPases in non-adherent ECs, and regulated activation of these proteins in identically treated SMCs. These differences most likely reflect the physiological milieu in which the cells reside and the temporal sequence of events leading to formation of blood vessels that follows trauma or injury same thing? 1.


Cell cultures. Human aortic smooth muscle cells (HASMC) were obtained from Cell Systems (Seattle, WA, USA) and grown to near confluence in T150 flasks 1 in a medium containing DMEM-10 % fetal bovine serum (FBS) and penicillin-streptomycin-amphotericin B supplement (Life Technologies, Gaithersburg, MD). The cells were harvested by trypsinization, washed with DMEM and suspended in DMEM. Human aortic endothelial cells (HAEC) were obtained from Clonetics (San Diego, CA, USA) and were grown according to the Supplier's instructions. To measure migration, 105 cells in 0.1 ml of DMEM were placed in the top well of a Transwell chemotaxis chamber and allowed to settle for 30 min. The top wells were then inserted into the bottom chambers containing 0.3 ml of DMEM and the chemoattractant. Migration was allowed to proceed for 2 h (EC) or 6 hrs (SMC) and was quantitated by DAPI staining as described 24.

Formation of capillary-like structures on Matrigel. The methods described previously were used to examine the ability of ECs and SMCs to form a network of capillary-like structures 25,26. Growth factor-reduced Matrigel was obtained from Collaborative Biomedical Products, Bedford, MA. The liquefied Matrigel was used to coat wells of a 24-well plate (200 ml per well) and allowed to solidify for 30 min at 37oC. Trypsinized and washed cells (3x104 EC and 7.5x104 SMC) were suspended in serum-free medium (200 ml total volume), placed in wells and incubated at 37oC and 5% CO2 for 16-20 h in the presence of test compound. The media were removed and the Matrigel layer was analyzed under 40X magnification for the presence of tube-like structures.

Differential staining of endothelial and smooth muscle cells. To observe formation of capillary networks created by co-culturing of ECs and SMCs, cells (105 SMC and 4.25x104 EC) were suspended in the total volume of 400 ml and layered on 200 ml of solidified Matrigel in a well of a 2-chamber slide. The compounds under study were added and the slide was incubated at 37o C in 5% CO2 atmosphere. Before the assay, SMCs were harvested by trypsinization, washed in serum-free medium and labeled using the PKH2-GL green fluorescent cell linker kit (Sigma Chemical Co. St. Louis, MO, USA). The cells (5x106) were suspended in 250 ml of Diluent A and mixed with an equal volume of PKH2-GL dye diluted 250-fold just before use. After mixing continuously for 5 min, 50 ml of fetal bovine serum was added for 1 min, the labeled cells were washed rapidly with complete medium, centrifuged, washed 3 more times and used in the slide chamber assay. Following incubation and formation of capillary network, endothelial cells in the structures were visualized with a red fluorescent dye, DiI-AC-LDL (Biomedical Technologies, Stoughton, MA, USA). This dye is specific for endothelial cells and does not recognize smooth muscle cells. The dye (4 ml of a 200 mg ml-1 solution) was added directly to the well of the microscope slide and staining was observed after 5 hrs of incubation. The red and green images were superimposed using Adobe Photoshop 5.5 computer program (Adobe Systems, San Jose, CA, USA) to show the alignment of the two cell types relative to each other.

Detection of Akt phosphorylation and Rac and Cdc42 activation assay. Antibodies directed against P-Ser473 Akt and total Akt were purchased from Cell Signaling Technologies (Beverly, MA) or Upstate Biotechnology (Lake Placid, NY). A Rac/Cdc42 activation assay kit was purchased from Upstate Biotechnology and was used according to manufacturer's instructions. Smooth muscle cells or endothelial cells were grown as described above, harvested and suspended in DMEM. Approximately 1-3 x 106 cells in 0.2 ml aliquots were distributed into Eppendorf microcentrifuge tubes and incubated at 37oC in the presence or absence of agents specified in the Figures. After incubation, the cells were lysed with an equal volume of 2x lysis buffer (1x buffer is: 25 mM HEPES, pH 7.5, 150 mM NaCl, 1% NP-40, 10 mM MgCl2, 1 mM EDTA, and 10% glycerol). The buffer was supplemented with leupeptin (20 mg ml-1), aprotinin (20 mg ml-1), sodium pervanadate (200 mM), NaF (2 mM), and one tablet of Complete (a protease inhibitor cocktail, Roche Biochemicals, Indianapolis, IN) for every 10 ml of buffer. The lysates (100 ml) were mixed with 30 ml of 4x sample buffer for SDS-PAGE (Invitrogen Corporation, Carlsbad, CA), boiled for 5 min and stored frozen at 80oC till needed. The remaining portions of the lysates (0.3 ml) were treated with 10 ml aliquots of the PAK-1 PBD (GST fused to p21-binding domain of human PAK-1) immobilized on glutathione-agarose beads. The beads with bound Rac and Cdc42 were washed with 1x lysis buffer containing 10 % glycerol, suspended in 40 ml of 2x sample buffer, and boiled for 5 min. Aliquots of all samples were applied to 4-12% gradient polyacrylamide gels (Invitrogen) and subjected to electrophoresis. The proteins were transferred to nitrocellulose membrane, probed with appropriate antibodies, and the signal was developed using the Amersham Biosciences (Piscataway, NJ) ECL detection kit.