Shear Stress–Dependent Regulation of the Human β-Tubulin Folding Cofactor D Gene
Abstract—The flowing blood generates shear stress at the endothelial cell surface. The endothelial cells modify their phenotype by alterations in gene expression in response to different levels of fluid shear stress. To identify genes involved in this process, human umbilical vein endothelial cells were exposed to laminar shear stress (venous or arterial levels) in a cone-and-plate apparatus for 24 hours. Using the method of RNA arbitrarily primed polymerase chain reaction, we cloned a polymerase chain reaction fragment representing an mRNA species downregulated by arterial compared with venous shear stress (shear stress downregulated gene-1, SSD-1). According to Northern blot analysis, corresponding SSD-1 cDNA clones revealed a similar, time-dependent downregulation after 24 hours of arterial shear stress compared with venous shear stress or static controls. Three SSD-1 mRNA species of 2.8, 4.1, and 4.6 kb were expressed in a tissue-specific manner. The encoded amino acid sequence of the human endothelial SSD-1 isoform (4.1-kb mRNA species) revealed 80.4% identity and 90.9% homology to the bovine β-tubulin folding cofactor D (tfcD) gene. Downregulation of tfcD mRNA expression by shear stress was defined at the level of transcription by nuclear run-on assays. The tfcD protein was downregulated by arterial shear stress. The shear stress–dependent downregulation of tfcD mRNA and protein was attenuated by the NO synthase inhibitor Nω-nitro-l-arginine methyl ester. Furthermore, the NO donor DETA-NO downregulated tfcD mRNA. Because tfcD was shown to be a microtubule-destabilizing protein, our data suggest a shear stress–dependent regulation of the microtubular dynamics in human endothelial cells.
- endothelial cells
- shear stress
- RNA arbitrarily primed polymerase chain reaction
- β-tubulin folding cofactor D
- nitric oxide
The flowing blood generates shear stress at the endothelial cell surface. Physiological shear stress is not only involved in the regulation of vascular tone but is also considered a protective mechanism against the localization of arteriosclerotic plaques.1 2
The mean physiological shear stress acting on endothelial cells is higher in arterial vessels (≈15 dyne/cm2) compared with venous vessels (≈1 dyne/cm2). This difference in shear stress affects the shape and differentiation of endothelial cells in arteries and veins. Endothelial cells align their shape and reorganize their cytoskeleton in response to the direction and degree of shear stress.3 4 Long-term application of arterial levels of laminar shear stress results in alignment of actin stress fibers in the direction of flow.5 However, the molecular mechanism underlying this cytoskeletal reorganization is not well understood. Furthermore, the regulation of gene expression by arterial levels of shear stress could be mediated by a mechanism involving vasoactive substances released from endothelial cells in a flow-dependent manner. High shear stress stimulates flow-dependent dilation of large vessels6 7 by release of endothelium-derived NO in animal studies and induces endothelial cell NO synthase expression.8 9 10
In view of these differences, we compared the gene expression of human endothelial cells exposed to high arterial or low venous levels of shear stress by RNA arbitrarily primed (RAP) polymerase chain reaction (PCR). One cDNA species showing very strong regulation by high shear stress was cloned and studied in more detail. We also tested the hypothesis that shear stress–dependent regulation of this gene may be mediated by NO.
Materials and Methods
Cell Culture and Application of Shear Stress
All cell culture reagents and chemicals were purchased from Sigma Chemical Co if not indicated otherwise. Human umbilical vein endothelial cells (HUVECs) were isolated as described previously.11 12 HUVECs were subjected to laminar shear stress in a cone-and-plate viscometer13 with minor modifications as described.14 Laminar shear stress of 1 dyne/cm2 (0.1 N/m2 [venous or low shear stress]) or 15 or 30 dyne/cm2 (1.5 or 3 N/m2, respectively; arterial or high shear stress) was applied in a humidified environment with 5% CO2 at 37°C. Cell culture medium containing dextran did not affect mRNA expression in this study (Figure 1⇓ online; available in the online data supplement at http://www.circresaha.org). Shear stress did not increase medium temperature or lactate dehydrogenase release (Figures 2⇓ and 3⇓ online).
RNA Isolation and RAP-PCR
Total RNA from endothelial cells was isolated by guanidinium thiocyanate/cesium chloride centrifugation.15 The mRNA was then obtained using an mRNA purification kit (Pharmacia).
RAP-PCR was used to identify differentially expressed transcripts (RAP-PCR kit, Stratagene).
Northern Blot Analysis
Northern blot analysis was done as previously described.16 The human multiple tissue Northern blot was purchased from Clontech. The expression of control gene GAPDH was not affected by laminar shear stress (Figure 4⇓ online).
cDNA Cloning and DNA Sequence Analysis
Selected RAP-PCR fragments were cloned with the pCR-Script Amp SK(+) cloning kit (Stratagene). Single clones showing an identical shear stress–dependent regulation by Northern blot analysis were used to screen 1×106 plaque-forming units of an oligo(dT) and randomly primed human heart–Lambda ZAP II–cDNA library (Stratagene).
The cDNA sequence was determined by cycle sequencing on an automated ABI PRISM 373A DNA sequencer (ABI/Perkin Elmer). The DNA and deduced protein sequence was analyzed by database searches of GenBank.17
Nuclear Run-On Assays
Nuclear run-on assays were performed as previously described.18
Inhibitor and NO Donor Studies
HUVECs were cultured under static conditions or exposed to laminar shear stress for 24 hours with or without the NO synthase inhibitor Nω-nitro-l-arginine methyl ester (L-NAME, 1 or 400 μmol/L).
In further studies, static cultures of HUVECs were incubated with the NO donor DETA-NO (0.1 to 2 mmol/L) for 24 hours and analyzed by Northern blotting.
Production of Antibodies Directed Against Shear Stress Downregulated Gene-1 (SSD-1) Peptide
One strongly immunogenic peptide of SSD-1 protein was selected using Gene Runner software (Hastings Software, Inc). A 14-mer immunogenic peptide (VKKEIKNSKDIQKL, SSD-1 residues 1075 to 1088) was synthesized and conjugated with an additional N-terminal cysteine residue to KLH and used for immunization of rabbits (Eurogentec, Berlin, Germany). Specificity of preimmune and SSD-1 antiserum was tested in slot blot and Western blot analysis.
Protein Isolation and Western Blot Analysis
Protein isolation and Western blot analysis using SSD-1 antiserum was performed as described.16
Data are given as mean±SEM (n≥3 in all cases). Statistical analysis was performed with ANOVA procedure followed by the Bonferroni t test (multiple comparison) or the Student t test (SigmaStat software, Jandel Corp). Differences were taken as statistically significant at P<0.05.
An expanded Materials and Methods section can be found in an online data supplement available at http://www.circresaha.org.
Molecular Cloning of SSD-1, a Gene Downregulated by Arterial but Not Venous Shear Stress
After exposure of HUVECs to low venous levels of shear stress (1 dyne/cm2) or high arterial levels of shear stress (15 dyne/cm2) for 24 hours, 30 cDNA fragments differentially expressed at these 2 levels of shear stress were isolated by the method of RAP-PCR. Using this unbiased approach, we selected one mRNA species showing the greatest downregulation by long-term (24 hours) shear stress of 15 dyne/cm2 (Figure 1⇑). We termed this mRNA species “shear stress downregulated gene-1” (SSD-1). SSD-1 RAP-PCR fragments were cloned into the pCR-Script Amp SK(+) plasmid.
Single SSD-1 clones were tested by Northern blot analysis of HUVECs exposed to shear stress of 1 or 15 dyne/cm2 (24 hours). SSD-1 clones showing a similar downregulation by shear stress of 15 dyne/cm2 were used to isolate 3 independent, overlapping cDNA clones from a human heart–Lambda ZAP II–cDNA library. A SSD-1 fragment present in all 3 cDNA clones was hybridized with RNA of HUVECs exposed to shear stress of 15 dyne/cm2 for varying periods of time (Figure 2⇑). This SSD-1 fragment detects an RNA transcript of 4.1 kb in HUVECs. The endothelial SSD-1 transcript was downregulated by shear stress of 15 dyne/cm2 for 24 hours by 70% as compared with the static control. In contrast, long-term application of low venous shear stress (1 dyne/cm2, 24 hours) had no effect on SSD-1 mRNA expression. Independently, in separate experiments, we isolated fragments with sequence identity to SSD-1 showing a similar downregulation by long-term arterial shear stress compared with venous shear stress or static control (not shown).
SSD-1 Encodes Human β-Tubulin Folding Cofactor D (tfcD)
The DNA sequence of 3 SSD-1 cDNA clones suggested that these clones originated from the same gene because of large overlapping sequences. The sequence of the SSD-1 cDNA fragments revealed 4049 bp with a 3744-bp open reading frame encoding a protein of 1248 amino acids (Figure 3⇑). The SSD-1 amino acid sequence revealed 2 putative cell adhesion sequences (RGD sites) and 9 potential N-myristoylation sites. We found in the SSD-1 protein putative phosphorylation sites for cAMP- and cGMP-dependent protein kinase (1 site), tyrosine kinase (1 site), protein kinase C (8 sites), and casein kinase II (22 sites).
The SSD-1 amino acid sequence revealed 80.4% identity and 90.9% homology to the recently cloned bovine tfcD.19 Alignment of the amino acid sequences of human SSD-1 encoding tfcD with bovine tfcD is available in the online data supplement at http://www.circresaha.org (Figure 5⇓ online).
Tissue-Specific Expression of Human tfcD Gene
The human tfcD was hybridized to RNA from different human tissues (Figure 4⇑). The tfcD probe detected mRNA species of 4.6, 4.1, and 2.8 kb. The 4.1-kb transcript, found in endothelial cells, was the most abundant tfcD mRNA in all of the human tissues we tested (heart, brain, placenta, lung, liver, skeletal muscle, kidney, and pancreas). In addition, 2 tfcD mRNA species of 4.6 and 2.8 kb, expressed at lower levels, were detected in human heart, brain (4.6 kb only), placenta, lung (2.8 kb only), skeletal muscle (2.8 kb only), and pancreas, suggesting the presence of tissue-specific tfcD isoforms.
Transcriptional Regulation of tfcD mRNA by Shear Stress
Downregulation of tfcD mRNA by shear stress could be due to downregulation at the transcriptional level or to decreased mRNA stability. Therefore, we performed nuclear run-on assays with nuclei isolated from HUVECs with or without application of arterial levels of laminar shear stress. Using a new reverse transcriptase (RT)–PCR–based run-on assay, downregulation of tfcD mRNA expression (to 11±7% of the level in static controls; n=3, P<0.001) by arterial levels of shear stress was found at the level of transcription (Figure 5⇑). As a control, endothelial NO synthase (eNOS), a gene well known to be upregulated by laminar shear stress, was induced at the transcriptional level in the same nuclei (to 273±61% of the level in static controls; n=3, P<0.05), whereas mRNA expression of the housekeeping gene GAPDH was not affected by shear stress. RT-PCR fragments amplified with this procedure reflect de novo mRNA synthesis in isolated nuclei during the incubation period because RT-PCR fragments were usually not detectable or were less abundant (maximum 20% of control) in control nuclei lysed immediately after isolation.
Downregulation of tfcD by Arterial Levels of Shear Stress Is Mediated by a NO-Dependent Pathway
Because arterial shear stress induces endothelial NO synthesis, the effect of eNOS inhibition (L-NAME, 400 μmol/L) on shear stress–dependent downregulation of tfcD mRNA was tested (Figure 6⇓). The inhibitor had no significant effect on basal tfcD mRNA expression but prevented the downregulation of tfcD mRNA by arterial shear stress.
A similar downregulation of tfcD protein was observed after exposure to different levels of arterial shear stress (15 dyne/cm2, 47.6±6.7%; 30 dyne/cm2, 40.6±5.1%; n≥5 each, P<0.05 versus static control in each) (Figure 7⇓), whereas application of long-term venous shear stress (1 dyne/cm2, 24 hours) did not affect tfcD protein expression. The downregulation of tfcD protein expression by arterial levels of shear stress was attenuated by eNOS inhibition with L-NAME (400 μmol/L) (Figure 7⇓).
To get direct evidence for NO-dependent regulation of tfcD expression, HUVECs were incubated under static conditions with different concentrations of the NO donor DETA-NO for 24 hours (Figure 8⇓). At DETA-NO concentrations higher than 1 mmol/L, tfcD mRNA was downregulated by 30% as compared with control.
Laminar shear stress induces a variety of molecular changes in endothelial cells.2 4 Cellular changes in response to shear stress include alignment of cells in the direction of flow, reorganization of the cell surface, downregulation of fibronectin expression, and increased mechanical stiffness.4 Our interest was to identify changes in endothelial gene expression by long-term exposure to venous or arterial levels of laminar shear stress. We initially identified 30 cDNA fragments differentially expressed after application of venous or arterial shear stress in endothelial cells. However, for an overall estimation of the number of genes differentially expressed by long-term venous or arterial levels of laminar shear stress, initially identified fragments have to be confirmed by isolation of individual cDNA clones and alternative techniques (eg, subtractive hybridization, cDNA expression array). We focused in this study on one gene with apparently the most prominent downregulation by arterial shear stress. This gene was identified as the human tfcD gene.
The recently cloned bovine tfcD is involved in the generation of exchange-competent β-tubulin.19 The folding of newly synthesized β-tubulin into its native conformation is the first posttranslational step in the pathway leading to the tubulin heterodimer (consisting of one α- and one β-tubulin polypeptide). The correct folding of α- and β-tubulin requires the assistance of a cytosolic chaperonin and several additional tubulin folding cofactors (A through E).20 21 This tubulin-specific chaperone/tfc supercomplex is considered as a dimer-making machine.20 Because tubulin generates the microtubules as essential elements of the cytoskeleton, this process might affect the reorganization of the cytoskeleton in response to arterial laminar shear stress. Apart from folding β-tubulin, tfcD can capture β-tubulin in tfcD/β-tubulin complexes by disrupting native αβ-tubulin heterodimers.22 Recently, tfcD was shown to modulate microtubule dynamics by sequestering β-tubulin from GTP-bound αβ-tubulin heterodimers, suggesting a role as a microtubule destabilizing protein.23 This is supported by overexpression of tfcD in transfected HeLa cells resulting in a progressive loss of microtubules.23 24 Therefore, downregulation of tfcD expression in response to arterial laminar shear stress would probably increase the number of microtubules, resulting in stabilization of the microtubular network.
We identified in the human tfcD protein several interesting structural features. The existence of 2 RGD sites usually found in extracellular matrix proteins suggests a role of tfcD in cell adhesion.25 From the 9 potential myristoylation sites detected by tfcD sequence analysis, most probably only 1 N-terminal site exists in vivo. Furthermore, the putative phosphorylation sites for cAMP- and cGMP-dependent protein kinases, tyrosine kinase, protein kinase C, and casein kinase II represent potential targets for regulation of tfcD activity. The functional importance of each of these structural features must be confirmed in further studies.
A critical role of microtubules in shear stress–dependent reorganization of the cytoskeleton was previously described.3 Disruption of microtubular network and inhibition of tyrosine kinase activity blocked the shear-induced alignment of cell shape and actin stress fibers. These data suggest a dynamic interaction between the microtubular and actin fiber network in response to mechanical forces.26 The role of microtubules in this tensegrity model is a stabilization of cytoplasm and nucleus against lateral compression. Shear stress of low degree (eg, venous shear stress) might induce only minor changes in the microtubular network, resulting in a more flexible cytoskeleton, compared with higher degrees of arterial shear stress with development of actin stress fibers and reorganization of microtubular network, resulting in increased mechanical stiffness. The key role of tyrosine kinases in shear stress–induced reorganization of cytoskeleton3 could involve modulation of tfcD activity by phosphorylation of the evolutionarily conserved tyrosine kinase site.
Further studies were aimed at understanding the molecular mechanism underlying this downregulation of tfcD by arterial laminar shear stress. Because application of arterial laminar shear stress was found to induce the endothelial isoform of NO synthase and NO formation as an atheroprotective mechanism,8 9 10 we analyzed the effect of NO synthase inhibition on shear stress–dependent downregulation of tfcD mRNA. We also studied the effect of the NO donor DETA-NO on tfcD expression. Our data provide evidence for a NO-dependent downregulation of tfcD. The downregulation of tfcD mRNA by shear stress was much more marked than the effect shown in response to the exogenous NO donor DETA-NO. One possible explanation for this difference could be the need of a long-lasting NO release to get a sustained downregulation of tfcD mRNA. Even while DETA-NO is considered as a compound mediating a long-lasting NO release, the 24-hour incubation period in the cell culture medium containing 5% dextran could decrease the NO level. In contrast, high laminar shear stress is considered as a strong physiological stimulus causing a sustained upregulation of eNOS expression and NO release.10 This could explain the lower level of downregulation of tfcD mRNA by NO donor DETA-NO, compared with high laminar shear stress. Furthermore, shear stress elicits the activation of other signaling pathways and the generation of additional endothelial autacoids that affect gene regulation without the involvement of NO. NO can affect expression of other genes via cGMP by activation or deactivation of transcription factors.27 Disruption of microtubular network attenuates flow-dependent NO release in perfused vessels.28 Our data suggest an additional signal-transduction pathway leading to the opposite direction with NO-mediated regulation of microtubular turnover through β-tubulin folding.
In summary, our data show a NO-dependent downregulation of the tfcD gene by arterial levels of laminar shear stress in human endothelial cells. The tfcD gene might be involved in modulation of endothelial microtubule dynamics in response to shear stress. The higher degree of shear stress in arterial vessels reduces the expression of the microtubule-destabilizing protein tfcD in endothelial cells and makes the cytoskeleton more rigid than the more flexible cytoskeleton in venous endothelial cells. This change in microtubule dynamics seems to be mediated by NO. Therefore, our data suggest a new role of NO as a signaling molecule that transduces mechanical forces into functional changes of the microtubular network in human endothelial cells.
This work was supported by the Deutsche Forschungsgemeinschaft and the Oskar Lapp Award of the German Cardiac Society. We thank M. Schultz, D. Barowsky, H. Lehnich, and H.-D. Pauer for their help in the development of the cone-and-plate apparatus; G. Kaltenborn for his support in DNA sequencing; and E. Heinke, R. Gall, and R. Busath for excellent technical assistance. We are grateful to H.E. Ives (Cardiovascular Research Institute and Division of Nephrology, University of California, San Francisco, Calif) for critically reading the manuscript.
- Received May 31, 2000.
- Revision received October 11, 2000.
- Accepted October 11, 2000.
- © 2000 American Heart Association, Inc.
Asakura T, Karino T. Flow patterns and spatial distribution of atherosclerotic lesions in human coronary arteries. Circ Res. 1990;66:1045–1066.
Traub O, Berk BC. Laminar shear stress: mechanisms by which endothelial cells transduce an atheroprotective force. Arterioscler Thromb Vasc Biol. 1998;18:677–685.
Malek AM, Izumo S. Mechanism of endothelial cell shape change and cytoskeletal remodeling in response to fluid shear stress. J Cell Sci. 1996;109:713–726.
Wong AJ, Pollard TD, Herman IM. Actin filament stress fibers in vascular endothelial cells in vivo. Science. 1983;219:867–869.
Hintze TH, Vatner SF. Reactive dilation of large coronary arteries in conscious dogs. Circ Res. 1984;54:50–57.
Nishida K, Harrison DG, Navas JP, Fisher AA, Dockery SP, Uematsu M, Nerem RM, Alexander RW, Murphy TJ. Molecular cloning and characterization of the constitutive bovine aortic endothelial cell nitric oxide synthase. J Clin Invest. 1992;90:2092–2096.
Topper JN, Cai J, Falb D, Gimbrone MA Jr. Identification of vascular endothelial genes differentially responsive to fluid mechanical stimuli: cyclooxygenase-2, manganese superoxide dismutase, and endothelial cell nitric oxide synthase are selectively up-regulated by steady laminar shear stress. Proc Natl Acad Sci U S A. 1996;93:10417–10422.
Jaffe EA, Nachman RL, Becker CG, Minick CR. Culture of human endothelial cells derived from umbilical veins: identification by morphologic and immunologic criteria. J Clin Invest. 1973;52:2745–2756.
Morawietz H, Rueckschloss U, Niemann B, Duerrschmidt N, Galle J, Hakim K, Zerkowski HR, Sawamura T, Holtz J. Angiotensin II induces LOX-1, the human endothelial receptor for oxidized low-density lipoprotein. Circulation. 1999;100:899–902.
Morawietz H, Ma YH, Vives F, Wilson E, Sukhatme VP, Holtz J, Ives HE. Rapid induction and translocation of Egr-1 in response to mechanical strain in vascular smooth muscle cells. Circ Res. 1999;84:678–687.
Tian G, Bhamidipati A, Cowan NJ, Lewis SA. Tubulin folding cofactors as GTPase-activating proteins: GTP hydrolysis and the assembly of the α/β-tubulin heterodimer. J Biol Chem. 1999;274:24054–24058.
Lewis SA, Tian G, Vainberg IE, Cowan NJ. Chaperonin-mediated folding of actin and tubulin. J Cell Biol. 1996;132:1–4.
Bhamidipati A, Lewis SA, Cowan NJ. ADP ribosylation factor-like protein 2 (Arl2) regulates the interaction of tubulin-folding cofactor D with native tubulin. J Cell Biol. 2000;149:1087–1096.
Forstermann U, Boissel JP, Kleinert H. Expressional control of the “constitutive” isoforms of nitric oxide synthase (NOS I and NOS III). FASEB J. 1998;12:773–790.