Upregulatory Expression of Furin and Transforming Growth Factor-β by Fluid Shear Stress in Vascular Endothelial Cells
Abstract—Furin, a yeast Kex2-family endoprotease, converts many vasoregulatory propeptides, including pro-transforming growth factor (TGF)-β to their mature forms. We examined whether furin expression is regulated by shear stress in vivo and in vitro. When an arteriovenous shunt was placed between the carotid artery and external jugular vein in rabbits, furin and TGF-β were highly expressed in shear stress–loaded endothelial cells. Exposure of bovine aortic endothelial cells in culture to shear stress induced furin and TGF-β expression in a similar manner. Molecular analysis of furin expression in bovine aortic endothelial cells revealed that shear stress increases the furin gene expression at transcriptional levels. Furthermore, TGF-β itself increased the furin mRNA levels. Shear-mediated furin expression was partly mediated by TGF-β because shear-induced furin mRNA levels were considerably decreased by overexpression of the truncated form of the TGF-β type II receptor. Likewise, blockade of furin activity by a furin inhibitor significantly decreased the endothelial production of mature TGF-β. Taken together, the results indicate that furin expression is induced and maintained by a coordination of shear stress and TGF-β. Increased furin expression may facilitate the formation of mature TGF-β, resulting in the enhanced effects of TGF-β on endothelial cells and vascular smooth muscle cells in the vasculature.
- Received October 10, 2000.
- Accepted January 22, 2001.
Shear stress is associated with the atheroprotective activities controlling vascular tone and maintaining the nonadhesive and nonthrombogenic properties of the endothelium.1 2 Shear stress controls the expression of a number of genes involved in the endothelial cell functions, including transforming growth factor (TGF)-β,3 C-type natriuretic peptide,4 adrenomedullin,4 platelet-derived growth factor,5 endothelin,6 and tissue plasminogen activator.7 These vasoregulatory factors are synthesized in a precursor that undergoes cleavage at an Arg−4-X−3-X−2-Arg−1 (RXXR) site by yeast Kex2-type proprotein-processing endoproteases, such as furin and PC5 (also known as PC6).8 The motif RXXR is also found in other growth factor precursors (such as activin A), in some growth factor proreceptors (such as insulin receptor and hepatocyte growth factor receptor), in matrix metalloproteinases (MMPs, such as stromelysin-3 and membrane-type MMP1), in adhesion molecules (such as the cadherin family and integrins α3 and α6), and in many plasma proteins (including coagulation factors VII, IX, and X and von Willebrand factor).8
Previously, we have found that when furin substrates, such as brain natriuretic peptide and TGF-β, are expressed, furin is also expressed in the cardiac muscle and the liver, respectively.9 10 In cardiac myocytes, furin is strongly expressed when the cells undergo hypertrophic growth by stretching.11 In the rat liver, furin and TGF-β mRNA levels are coelevated in partially hepatectomized liver.10 Thus, it seems that furin expression is closely linked with the expression of its substrates. Human embryonal venous endothelial cells express furin and PC5. When the human embryonal venous endothelial cells reach a confluent state, PC5 expression increases, but furin expression remains unchanged.12 However, it is unclear just which enzymes are induced in physiological conditions, such as mild shear stress loading.
Endothelial cells express TGF-β1 in a shear stress–dependent manner.3 TGF-β is known to produce proteoglycans in bovine aortic endothelial cells (BAECs)13 and to inhibit vascular smooth muscle cell proliferation in response to vascular injury.14 15 Crawford et al16 generated mice lacking the thrombospondin (TSP)-1 gene, an activator protein of the TGF-β precursor. The histological findings in TSP-1–null mice resemble those in TGF-β1 gene–lacking mice. In both mice, vascular smooth muscle hyperplasia and alveolar hemorrhage were especially pronounced in the lungs, suggesting that TGF-β regulates the proliferation of vascular smooth muscle cells and the formation of extracellular matrix.
TGF-β is secreted in an inactive precursor composed of a 25-kDa dimer and a propeptide dimer (latency-associated peptide) connected by the RHKR sequence.17 The precursor is then bound to a latent TGF-β binding protein at the amino-terminal side of the propeptide.18 This TGF-β complex requires 2 processing reactions for maturation: first, cleavage of RHKR by furin,17 although the latency-associated peptide and mature TGF-β remain bound noncovalently; second, removal of the latent TGF-β binding protein either by plasmin19 20 or by a conformational change by TSP16 20 or integrin αvβ6.21 With the second reactions, mature TGF-β is released from the complex. Yet, regardless of the type of second reaction involved, furin cleavage is always required for the first reaction.
To determine the physiological function of furin, we investigated the alteration of the furin mRNA levels and the mature TGF-β formation mediated by furin in response to shear stress in cultured endothelial cells. Our results showed that shear stress is the potent stimulator of furin and TGF-β expression. Molecular analysis of furin gene expression revealed the positive-feedback loop between furin and TGF-β expression. The present study demonstrates that shear stress leads the upregulatory formation of mature TGF-β by coordinated induction of furin and TGF-β.
Fifteen male Japanese White rabbits weighing ≈3.0 kg underwent surgery to create a carotid arteriovenous shunt between the left carotid artery and the left external jugular vein, as described previously.22 As a control, sham operations were performed by use of an identical procedure without shunt formation in 3 rabbits. The shunt rabbits were killed with a lethal injection of sodium pentobarbital at either 6, 12, 24, or 48 hours after the operation to obtain shear stress–loaded arterial tissues. At each time point, 3 rabbits were used to measure blood flow proximal to the arteriovenous shunt site on the left and, at the same level, on the right control side, with the use of an electromagnetic flowmeter. The arterial tissues were subjected to immunostaining.
The arterial tissue sections or cells were first incubated with primary rabbit antibody to furin, PC5, or TGF-β1. The antibody to furin was generated in rabbits by use of a synthetic peptide.9 11 The rabbit antibody to PC5 was a kind gift of Dr Nabil G. Seidah, Clinical Research Institute of Montréal, Québec, Canada. The rabbit antibody to TGF-β1 was purchased from Santa Cruz Biotechnology. The secondary antibody used was indodicarbocyanide-conjugated affinity-purified donkey anti-rabbit IgG (Jackson ImmunoResearch).
Cell Culture and Shear Stress Procedure
BAECs isolated from the thoracic aortas were cultured on a 0.5-mm-thick quartz cover glass and were placed on a parallel-plate–type flow chamber (inner space size 16 mm wide×35 mm long×0.2 mm deep). The apparatus was placed in a CO2 incubator at 37°C. The shear stress forces were calculated on the basis of an equation described previously.23 The flow rate was adjusted to 15 dyne/cm2, unless stated otherwise.
Reverse Transcription–PCR and Northern Blot Analyses
Total RNA was extracted from BAECs and was reverse-transcribed at 42°C for 1 hour with an oligo(dT)17 primer. The oligonucleotides used for polymerase chain reaction (PCR) were 5′-GTACGGCTAC GGGCTGTTGGA-3′ and 5′-TCGCCAGAGGGATCCTCGTC-3′ for furin (369 bp), 5′-GGACTACTACGCCAAGGAGGTCAC-3′ and 5′-GGTCAGCCACTGCCGCACAACT-3′ for TGF-β1 (325 bp), and 5′-ATGACCACTGTCCACGCCAT-3′ and 5′-GCCTGCT- TCACCACCTTCTT-3′ for GAPDH (272 bp). Northern blot analysis was performed by using 10 μg of total RNA from BAECs by probing with the mouse furin cDNA (924 bp), as described before.9 10 11
We used the TATA box–containing furin promoter as described before.24 Three 5′-upstream DNA fragments, −3633/+55 (PstI-PstI), −612/+55 (XbaI-PstI), and −56/+55 (BamHI-PstI), were placed before a firefly luciferase gene supplied with the Luciferase Reporter Assay System (Promega). Each of the 3 luciferase gene constructs was transfected to BAECs by using a TransFast liposomal transfection reagent (Promega). Twenty-four hours after the transfection, the medium was changed, and the culture was continued another 6 hours with or without shear stress. Then, the cells were harvested to prepare cell lysates for luciferase assay. Luciferase activity is expressed as multiplicity (fold value) against the value obtained by the (−56/+55) luciferase gene under shear stress.
Assessment of Mature TGF-β Formation
Mature TGF-β formation was assessed by 3 methods: immunoblotting, ELISA, and cell proliferation–inhibiting activity. For immunoblotting, the BAECs were cultured for 12 hours with or without a shear stress of 15 dyne/cm2. The medium (40 mL) conditioned by the culture was immunoprecipitated with mouse anti–TGF-β antibody (R & D Systems) at 4°C for 3 hours and then with protein G–Sepharose 4FF (Amersham Pharmacia) for 1.5 hours. Proteins eluted from the Sepharose were separated by a 10% SDS-PAGE under a nonreducing condition and then blotted onto a nitrocellulose membrane for probing with rabbit anti-human TGF-β1 antiserum (Santa Cruz Biotech) at a dilution of 1:350.
To suppress proteolytic activity in each cell lysate, we used the furin inhibitor decanoyl-arginyl-valyl-lysyl-arginyl-chloromethylketone (decRVKR-CMK, a kind gift of Dr W. Garten, Philips University, Marburg, Germany). The inhibitor was used at a final concentration of 25 μmol/L, which did not affect the cell viability up to 12 hours.11 The potency of the inhibitor was assessed by pyr-Arg-Thr-Lys-Arg-methylcoumarylamide (Peptide Institute) as a substrate.11
ELISA was performed by using a Quantikine Human TGF-β1 Assay Kit (R & D Systems) in which a soluble TGF-β type II receptor–precoated multiwell plate was used to absorb the TGF-β1.
TGF-β bioactivity was measured by a growth inhibition assay. Mink lung Mv1Lu cells (CCL64, American Tissue Culture Collection) were plated in 24-well plates at a density of 2×104 cells per well in 0.5 mL DMEM with 10% FBS. Sixteen hours after the addition of the conditioned medium, cells were incubated with 1.0 μCi/mL [3H]thymidine for 2 hours, and the radioactive counts incorporated into DNA were then determined.
Construction of an Adenovirus Vector for ΔTGFβRII
The truncated human type II TGF-β receptor (ΔTGFβRII) was generated by using PCR with the sense primer 5′-TCGGTCTATGACGAGCAGCGG-3′ and the antisense primer 5′-AGCGACCTTTCCCCACCAGG-3′.25 The ΔTGFβRII cDNA was subcloned into a pCEP4 vector (InVitrogen), and the expression unit was transferred to the cassette cosmid pAdex vector.25 This cosmid pAd-ΔTGFβRII was subjected to a homologous recombination with the EcoT22I-digested DNA-terminal protein complex of Ad5-dlX in human embryonal kidney 293 cells for generating a recombinant virus Ad-ΔTGFβRII. A virus titer was determined by plaque-forming units with the use of 293 cells and expressed as a multiplicity of infection (MOI). Because the adenoviral expression of β-galactosidase was observed for 70% of BAECs by MOI 20 and for 95% of BAECs by MOI 50, we performed Ad-ΔTGFβRII infection by MOI 50.
Coexpression of Furin and TGF-β After Arteriovenous Shunt in Rabbit Endothelial Cells
TGF-β is reportedly processed by furin in furin-deficient LoVo cells by infecting furin-expressing and pro-TGF-β–expressing vaccinia recombinants.17 We initially examined the coexpression of furin and TGF-β in arteriovenous shunt–generated rabbits. Although furin did not virtually stain in the endothelial cells of the control carotid artery, shear stress, induced via an arteriovenous shunt, caused a marked increase in staining with time, up to the 24-hour point (Figure 1⇓, left). The expression of furin was more marked in endothelial cells compared with cells in the surrounding smooth muscle.
TGF-β also stained increasingly up to 24 hours after the operation (Figure 1⇑, right), and as with furin expression, the staining was more marked in endothelial cells compared with cells in the smooth muscle. Thus, the coordinated increase of furin and TGF-β may suggest that pro-TGF-β is processed by furin.
Coexpression of Furin and TGF-β by Shear Stress in BAECs
BAECs were subjected to a shear stress of 15 dyne/cm2 for up to 24 hours. With shear stress, furin and TGF-β messages similarly increased in PCR-amplified band intensity from the 6-hour to 24-hour points (Figure 2A⇓). To examine a quantitative increase by PCR amplification, Northern blot was performed for a 6-hour point. The increase in furin mRNA by PCR and by Northern blot was similar (both an ≈2.3-fold increase, Figure 2B⇓). Thus, coelevation of furin and TGF-β by shear stress was also confirmed in BAECs.
We next examined shear stress–dependent expression of furin and PC5 by immunostaining. Both furin and PC5 are Kex2-family endoproteases and have similar domain structures. But only furin has a transmembrane domain, which makes it a resident protein of trans-Golgi networks, whereas PC5 is a cytosol protein.8 Furin was expressed only a little in the static condition, but the expression increased with shear stress at 15 and 60 dyne/cm2 (Figure 2C⇑, a through c). Furin immunostaining was localized adjacent to the nuclei, which is a typical feature of Golgi-resident proteins. In contrast, PC5 was distributed over the entire cytoplasm, and its staining intensity was similar up to 60 dyne/cm2 (Figure 2C⇑, d through f). Thus, furin expression is shear stress dependent, whereas PC5 is stress independent.
Transcriptional Control of Furin Expression by Shear Stress
To determine the shear stress–mediated induction of furin expression at transcriptional levels, we examined the shear stress effect in the presence of actinomycin D, a potent inhibitor of gene transcription. Shear-induced furin mRNA increase is significantly blunted by actinomycin D, suggesting the involvement of the transcriptional control of furin expression (Figure 3A⇓). We next examined the de novo protein synthesis for shear stress–induced furin expression. To this end, we treated the BAECs with a protein synthesis inhibitor, cycloheximide, before their exposure to shear stress. The furin mRNA levels were increased by shear stress at the 6-hour time point even in the presence of cycloheximide, whereas no measurable induction was observed at the 12-hour time point, suggesting the involvement of de novo protein synthesis for the late induction but not for the early induction of the furin mRNA levels. In the early phase, preexisting transcription factors are presumably activated by a shear stress–mediated signaling pathway, resulting in the activation of the furin promoter. However, in the late phase, these preexisting factors are spent out, and de novo factors are not replenished by cycloheximide, resulting in the loss of furin promoter activation despite shear stress stimulation.
Activation of the Furin Promoter by Shear Stress
To further confirm the transcriptional control of the furin gene, we performed transient transfection assays by using the reporter gene containing the furin promoter spanning from −3633 to +55 in front of the luciferase gene, which is referred to as (−3633/+55)Luc.24 Exposure of the transfected cells to shear stress for 6 hours increased the luciferase activity of the construct (−3633/+55)Luc by 2-fold compared with the static condition, indicating that the furin promoter is responsive to shear stress (Figure 3B⇑). This increase in luciferase activity was not artificial because transfection of BAECs with the promoterless luciferase construct did not yield any luciferase activity by shear stress. To narrow down the shear stress–responsive region, the constructs containing the shorter promoter were transfected into BAECs. (−612/+55)Luc exhibited a decrease in the shear stress–mediated induction of the promoter activity as well as in the basal promoter activity. Deletion to the position at −56 further attenuated the response to shear stress. These results suggest that shear stress–induced activity of the furin promoter is mediated through the region between −3633 and −612 as well as between −612 and −56.
TGF-β Induces the Expression of Furin
TGF-β is known to induce the expression of its own gene and the expression of furin.26 27 In BAECs, both messages increased dose-dependently with 10−10 and 10−9 mol/L TGF-β during a 6-hour treatment (Figure 4A⇓). We hypothesized that TGF-β induced by shear stress mediates the upregulation of furin and TGF-β. To examine the interactions between shear stress and TGF-β effects on furin expression, we used ΔTGFβRII-expressing endothelial cells to block the TGF-β signal pathway. After ΔTGFβRII expression in BAECs, the shear stress effect on the expression of furin was decreased compared with the effect on the expression in untreated cells (Figure 4B⇓ versus Figure 2A⇑). However, the elevation of furin expression was constantly observed 6 hours after shear stress even by expressing ΔTGFβRII with a high MOI number up to 100. Thus, the expression of furin appears to occur partly through a shear stress–specific signal pathway distinct from a TGF-β signal pathway.
Processing of TGF-β Precursor by Furin
After exposure to a static or shear stress–loaded culture for 12 hours, we examined the formation of a 25-kDa mature TGF-β dimer by PAGE under a nondenaturing condition. The high molecular weight band of 175 kDa appears to be a precursor complex of TGF-β with latent TGF-β binding protein. The 25-kDa TGF-β increased after shear stress (Figure 5A⇓), as reported previously.28 However, the 25-kDa form decreased toward the control level in the presence of the furin inhibitor decRVKR-CMK, suggesting that furin is involved in processing of the TGF-β precursor.
In a 12-hour static or shear stress–loaded culture, the media were activated with 0.1 mol/L HCl, and the protein levels of TGF-β were obtained by ELISA. In the static culture, the TGF-β1 secreted in the medium measured only 100 pg per 106 cells per 12-hour culture (Figure 5B⇑). In contrast, shear stress increased the receptor-bound TGF-β1 level to nearly 10 ng per 106 cells per 12-hour culture. However, the TGF-β1 level decreased to 1.5 ng per 106 cells per 12-hour culture in the presence of the furin inhibitor decRVKR-CMK, representing approximately one seventh of the level observed in the absence of the inhibitor. Thus, shear stress significantly increased the production of a mature form of TGF-β.
Shear stress increased the inhibitory effect of TGF-β on the incorporation of [3H]thymidine in Mv1Lu cells by ≈38% (Figure 5C⇑). However, this effect was reversed to control levels in the presence of the furin inhibitor decRVKR-CMK. Thus, the production of bioactive TGF-β is enhanced by furin in BAECs.
The shear stress–loaded culture system has been used to identify a number of vasoregulatory factors involved in the physiological functions of vascular endothelial cells.1 TGF-β1 is a factor that is elevated by shear stress.3 6 28 Exposure of BAECs to increased shear stress induces TGF-β1 via a signal transduction pathway modulated by K+ channel currents, and this induction is regulated at the transcription level.3 To generate a bioactive TGF-β1, the first processing reaction takes place by furin in the trans-Golgi network.17 We demonstrated the coordinated expression of TGF-β1 and furin in shear stress–loaded BAECs and endothelial cells of rabbits with a carotid arteriovenous shunt. Interestingly, TGF-β1 itself and furin are also induced in static culture on stimulation with TGF-β1.26 27 Furthermore, the furin inhibitor decRVKR-CMK extensively reduced the production of the mature form of TGF-β1 induced by shear stress. These results suggest that furin is a genuine processing enzyme for pro-TGF-β. Interestingly, in mouse embryos at early stages, a striking degree of overlap is noted in the distribution of the TGF-β1 family proteins and furin.29 Both proteins are highly expressed in extraembryonic mesoderm, in which vascular endothelial growth factor receptor Flk-1 is present and endothelial precursor cells are derived. Thus, the coexpression of furin and TGF-β1 is tightly coupled.
In the present study, we provide evidence that furin expression is induced by shear stress at the transcriptional level. A number of endothelial genes are reportedly regulated by shear stress.1 2 Resnick and Gimbrone6 have identified the shear stress response element (SSRE), 5′-GAGACC-3′, within the platelet-derived growth factor-B chain promoter and have shown that SSRE is present in many shear stress–inducible genes. Our 5′-deletion analysis suggests that a proximal promoter region up to −56 is not sufficient to be responsive to shear stress and that the region between −612 and −56 is necessary for shear stress response. Furthermore, the region upstream from −612 contributes to the increase in furin promoter activity by shear stress. Searching for the SSRE within the promoter region between −3633 and +55 revealed that SSREs are located at −419, −817, −2337, −3199, and −3626 (accession X15723), suggesting that these multiple SSREs are responsible for the shear stress–mediated furin gene expression.
Induction of the furin expression by TGF-β1 merits further discussion. Overexpression of the ΔTGFβRII did not abolish the early increase in furin mRNA levels by shear stress, but it did inhibit its later increase. These results suggest that the late increase is induced by TGF-β1. Because furin produces mature TGF-β1, the positive-feedback loop, which links shear stress–induced furin and TGF-β expression to a further generation of mature TGF-β by TGF-β–induced furin, may enhance the effect of TGF-β on endothelial cells and vascular smooth muscle cells. When TGF-β1 is overproduced in rat arterial endothelium by an adenovirus vector, transduced endothelium developed a cellular and matrix-rich neointima with cartilaginous metaplasia of the vascular media.30
It is well appreciated that TGF-β plays pleiotropic effects on vascular cells. TGF-β strongly accelerates lesion formation by increasing cellularity and markedly inducing extracellular matrix fibrosis. An increase in active TGF-β levels is associated with the progression of vascular lesions. A soluble form of TGFβRII was effective in preventing negative remodeling with adventitial fibrosis and neointima formation in an arterial balloon injury in rats.31 To block the negative effect of TGF-β, modification of TGF-β synthesis by a furin inhibitor may be therapeutically useful in the treatment of atherosclerotic vascular lesions. In contrast, in the lesion-free endothelium, TGF-β provides an atheroprotective function by stimulating proteoglycan formation13 and inhibiting vascular smooth muscle cell proliferation.14 Physiological levels of TGF-β1 appear to suppress cell division partly by inhibiting the expression of vascular endothelial growth factor receptors, such as Flk-1, in endothelial cells.32 We suggest that the coordinated upregulatory loops between furin and TGF-β expression in the shear stress–loaded vasculature favor an atheroprotective role of TGF-β in lesion-free vascular endothelial functions.
This work is supported by grants-in-aid from the Ministry of Education, Science, Sports, and Culture. We thank Eiko Hamana for her secretarial assistance.
Davis PF. Flow-mediated endothelial mechanotransduction. Physiol Rev. 1995;75:519–560.
Ohno M, Cooke JP, Dzau VJ, Gibbons GH. Fluid shear stress induces endothelial transforming growth factor beta-1 transcription and production. J Clin Invest. 1995;95:1363–1369.
Chun T-H, Ito H, Ogawa Y, Tamura N, Takaya K, Igaki T, Yamashita J, Doi K, Inoue M, Masatsugu K, et al. Shear stress augments expression of C-type natriuretic peptide and adrenomedullin. Hypertension. 1997;29:1296–1302.
Resnick N, Collins T, Atkinson W, Bonthron DT, Dewey CF Jr, Gimbrone MA Jr. Platelet-derived growth factor-B chain promoter contains a cis-acting fluid shear-stress-responsive element. Proc Natl Acad Sci U S A. 1993;90:4591–4595.
Resnick N, Gimbrone MA Jr. Hemodynamic forces are complex regulators of endothelial gene expression. FASEB J. 1995;9:874–882.
Nakayama K. Furin: a mammalian subtilisin/Kex2p-like endoprotease involved in processing of a wide variety of precursor proteins. Biochem J. 1997;327:625–635.
Sawada Y, Suda M, Yokoyama H, Kanda T, Sakamaki T, Tanaka S, Nagai R, Abe S, Takeuchi T. Stretch-induced hypertrophic growth of cardiocytes and processing of brain-type natriuretic peptide are controlled by proprotein-processing endoprotease furin. J Biol Chem. 1997;272:20545–20554.
Kaji T, Yamada A, Miyajima S, Yamamoto C, Fujiwara Y, Wight TN, Kinsella MG. Cell density-dependent regulation of proteoglycan synthesis by transforming growth factor β1 in cultured bovine aortic endothelial cells. J Biol Chem. 2000;275:1463–1470.
Ueba H, Kawakami M, Yaginuma T. Shear stress as an inhibitor of vascular smooth muscle cell proliferation: role of transforming growth factor-β1 and tissue-type plasminogen activator. Arterioscler Thromb Vasc Biol. 1997;17:1512–1516.
Song RH, Kocharyan HK, Fortunato JE, Glagov S, Bassiouny HS. Increased flow and shear stress enhance in vivo transforming growth factor-β1 after experimental arterial injury. Arterioscler Thromb Vasc Biol. 2000;20:923–930.
Dubois CM, Laprise M-H, Blanchette F, Gentry LE, Leduc R. Processing of transforming growth factor β1 precursor by human furin. J Biol Chem. 1998;270:10618–10624.
Sato Y, Rifkin DB. Inhibition of endothelial cell movement by pericytes and smooth muscle cells: activation of a latent transforming growth factor-β1-like molecule by plasmin during co-culture. J Cell Biol. 1989;109:309–315.
Schultz-Cherry S, Murphy-Ullich JE. Thrombospondin causes activation of latent transforming growth factor-β secreted by endothelial cells by a novel mechanism. J Cell Biol. 1993;122:923–932.
Masuda H, Zhuang Y-J, Singh TM, Kawamura K, Murakami M, Zarins CK, Glagov S. Adaptive remodeling of internal elastic lamina and endothelial lining during flow-induced arterial enlargement. Arterioscler Thromb Vasc Biol. 1999;19:2298–2307.
Kosaki K, Ando J, Korenaga R, Kurokawa T, Kamiya A. Fluid shear stress increases the production of granulocyte-macrophage colony-stimulating factor by endothelial cells via mRNA stabilization. Circ Res. 1998;82:794–802.
Kamimura H, Konda Y, Yokota H, Tkenoshita S-I, Nagamachi Y, Kuwano H, Takeuchi T. Kex2 family endoprotease furin is expressed specifically in pit-region parietal cells of the rat gastric mucosa. Am J Physiol. 1999;277:G183–G190.
Yamamoto H, Ueno H, Ooshima A, Takeshita A. Adenovirus-mediated transfer of a truncated transforming growth factor-β (TGF-β) type II receptor completely and specifically abolishes diverse signaling by TGF-β in vascular wall cells in primary culture. J Biol Chem. 1996;271:16253–16259.
Van Obberghen-Schilling E, Roche NS, Flanders KC, Sporn MB, Roberts AB. Transforming growth factor β1 positively regulates its own expression in normal and transformed cells. J Biol Chem. 1988;263:7741–7746.
Roebroek AJM, Umans L, Pauli IGL, Robertson EJ, van Leuven F, Van de Ven WJM, Constam DB. Failure of ventral closure and axial rotation in embryos lacking the proprotein convertase furin. Development. 1998;125:4863–4876.
Schulick AH, Taylor AJ, Zuo W, Qiu C-B, Dong G, Woodward RN, Agah R, Roberts AB, Virmani R, Dichek DA. Overexpression of transforming growth factor β1 in arterial endothelium causes hyperplasia, apoptosis, and cartilaginous metaplasia. Proc Natl Acad Sci U S A. 1998;95:6983–6988.
Smith JD, Bryant SR, Couper LL, Vary CPH, Gotwals PJ, Koteliansky VE, Lindner V. Soluble transforming growth factor-β type II receptor inhibits negative remodeling, fibroblast transdifferentiation, an intimal lesion formation but not endothelial growth. Circ Res.. 1999;84:1212–1222.
Mandriota SJ, Menoud P-A, Pepper MS. Transforming growth factor β1 down-regulates vascular endothelial growth factor receptor 2/flk-1 expression in vascular endothelial cells. J Biol Chem. 1996;271:11500–11505.