Bone Morphogenic Protein Antagonists Are Coexpressed With Bone Morphogenic Protein 4 in Endothelial Cells Exposed to Unstable Flow In Vitro in Mouse Aortas and in Human Coronary Arteries
Role of Bone Morphogenic Protein Antagonists in Inflammation and Atherosclerosis
Background— Exposure to disturbed flow, including oscillatory shear stress, stimulates endothelial cells (ECs) to produce bone morphogenic protein (BMP) 4, which in turn activates inflammation, a critical atherogenic step. BMP activity is regulated by the level of BMP antagonists. Until now it was not known whether shear also regulates the expression of BMP antagonists and whether they play a role in EC pathophysiology.
Methods and Results— BMP antagonists follistatin, noggin, and matrix Gla protein were expressed in cultured bovine and human arterial ECs. Surprisingly, oscillatory shear stress increased expression of the BMP antagonists in ECs, whereas unidirectional laminar shear decreased such expression. Immunohistochemical studies with mouse aortas showed data consistent with in vitro findings: Only ECs in the lesser curvature exposed to disturbed flow, but not those in the greater curvature and straight arterial regions exposed to undisturbed flow, showed coexpression of BMP4 and the BMP antagonists. Similarly, in human coronary arteries, expression of BMP4 and BMP antagonists in ECs positively correlated with the severity of atherosclerosis. Monocyte adhesion induced by oscillatory shear stress was inhibited by knockdown of BMP4 or treatment with recombinant follistatin or noggin, whereas it was increased by knockdown of follistatin and/or noggin.
Conclusions— The present results suggest that ECs coexpress BMP antagonists along with BMP4 in an attempt to minimize the inflammatory response by oscillatory shear stress as part of a negative feedback mechanism. The balance between the agonist, BMP4, and its antagonists may play an important role in the overall control of inflammation and atherosclerosis.
Received December 11, 2006; accepted June 29, 2007.
Atherosclerosis is an inflammatory disease that occurs preferentially in branched or curved arterial regions exposed to disturbed flow conditions such as low mean and oscillatory shear stress (OS).1–6 In contrast, straight portions exposed to undisturbed laminar shear stress (LS) are relatively lesion free. The mechanisms by which disturbed and undisturbed flow conditions initiate or prevent inflammation and atherosclerosis are likely to involve both acute and chronic responses of the cells.5,7 Recently, numerous groups, including ours, have carried out microarray studies to identify the genes that change in response to OS versus the genes that change in response to LS.1,8,9
Editorial p 1221
Clinical Perspective p 1266
Through a transcript profiling study and subsequent functional studies, we discovered that OS induces inflammation by production of bone morphogenic protein (BMP) 4 in endothelial cells (ECs).10 Further studies have shown that BMP4 stimulates the expression and activity of NADPH oxidase requiring p47phox and Nox-1 in an autocrine-like manner.10,11 The NADPH oxidase activation by BMP4 then leads to reactive oxygen species production, nuclear factor κB (NFκB) activation, intercellular adhesion molecule-1 expression, and subsequent increased monocyte adhesivity of ECs,3,10,11 a critical early atherogenic step.
BMPs are members of the transforming growth factor (TGF)-β superfamily and are potent bone-inducing morphogens.12 BMPs also regulate cell proliferation, differentiation, lineage determination, motility, and death.13–17 Unlike shear-sensitive BMP4, BMP2 is upregulated by TNFα, H2O2, and high pressure.18,19 Both BMP2 and BMP4 have been found in calcified atherosclerotic plaques and aortic valve diseases,20,21 which suggests their importance in cardiovascular diseases. Moreover, we have shown that ECs overlying foam cell lesions expressed BMP4, but nondiseased ECs did not.10 More recently, we have shown that chronic BMP4 infusion causes endothelial dysfunction in a vascular NADPH oxidase–dependent manner, which induces hypertension in mice.22 In our in vitro and in vivo studies, the critical methods that we used to demonstrate the specific effect of BMP4 on endothelial function were targeted knockdown of BMP4 with small interference RNA (siRNA) and BMP inhibition by the antagonist noggin.10,11,22 It was not known, however, whether ECs expressed endogenous levels of noggin or any of the other BMP antagonists.
BMP antagonists are secreted glycoproteins and include noggin, chordin, crossveinless-2 (CV2), follistatin, matrix Gla protein (MGP), the DAN/Cerberus, gremlin, and twisted gastrulation (Tsg).23 Each antagonist has a different affinity for the various BMPs, and, once bound, they inhibit the interaction of BMPs to their cognate receptors.24–29
Thus, it is the goal of the present study to define whether the BMP antagonists are expressed in ECs, and, if they are, whether their expressions are regulated by shear stress and whether they play a role in inflammation and atherosclerosis. In the present study we used cultured ECs from both bovine aortas and human coronary arteries to demonstrate the expression of BMP antagonists in flow-, BMP4-, and atherosclerosis-dependent manners, as well as their implication in inflammation and atherosclerosis.
Recombinant human BMP4 (r-BMP4), recombinant noggin (r-noggin), and recombinant follistatin (r-follistatin) were purchased from R&D Systems (Minneapolis, Minn), and sources of antibodies are listed in the online-only Data Supplement. See the online-only Data Supplement for expanded Methods.
Endothelial Cell Culture
Bovine aortic endothelial cells (BAECs) were grown to confluence at 37°C and 5% CO2 in Dulbecco’s minimum Eagle’s medium (DMEM; Gibco, Grand Island, NY) supplemented with 10% fetal bovine serum (FBS, Atlanta Biologicals, Norcross, Ga), heparin sodium (American Pharmaceutical Partners, Los Angeles, Calif), endothelial cell growth supplement, and minimum nonessential amino acids (Gibco). Cells from passages 7 to 11 were used. Human coronary artery ECs obtained from Clonetics (San Diego, Calif) were cultured in the EGM-2 bullet kit.11
Reverse-Transcriptase Polymerase Chain Reaction Analysis
For reverse-transcriptase polymerase chain reaction analysis, total RNAs from BAECs were isolated and purified with use of the RNeasy kit (Qiagen, Valencia, Calif). Two micrograms of total RNA were subjected to reverse transcription with Superscript II (Invitrogen, Carlsbad, Calif). Aliquots of the cDNA pool obtained were subjected to polymerase chain reaction with use of specific primers (Data Supplement Table I), and reaction products were resolved by electrophoresis on agarose gels.
After experimental treatments, cell lysates and conditioned media were prepared and analyzed by Western blot analysis as described previously.30,31
Confluent monolayers of BAECs were exposed to physiologically relevant arterial levels of shear stress: steady unidirectional LS (15 dyne/cm2), pulsatile LS (P-LS, 15±5 dyne/cm2), or bidirectional OS without forward movement (OS, 0±5 dyne/cm2 at 1-Hz cycle) or with forward movement (F-OS, 2±5 dyne/cm2 at 1-Hz cycle) in serum-free media. This was done by rotation of a Teflon cone (0.5° cone angle), as described previously by our group.3,32
Immunohistochemical Staining of Frozen Sections and En Face Preparations
For staining of frozen sections, fresh mouse aortas were excised from C57BL/6 mice (8-week-old males), placed in Tissue-Tek OCT compound, snap-frozen in liquid nitrogen, and stored at −80°C until used. Frozen sections of human coronary arteries were obtained from patients who underwent heart transplantations, as described previously.10 Frozen sections (10 μm thick) of mouse aortic arches and thoracic aortas as well as human coronary arteries were fixed in ice-cold acetone for 5 minutes, blocked for 1 hour with 10% donkey serum, and incubated with primary antibodies overnight at 4°C followed by rhodamine-conjugated secondary antibodies for 2 hours at room temperature. Nuclei were counter-stained with DAPI (Vector Laboratories, Burlingame, Calif). All photographs were taken with a Zeiss (Jena, Germany) epifluorescent microscope. For en face staining, mice were euthanized by CO2 inhalation, and the aortas were pressure-fixed with saline that contained 10% formalin. The aortas were carefully cleaned in situ, and the arches and thoracic aortas were dissected and stained with BMP4 and noggin antibodies (a gift from LabFrontier, Seoul, Korea), followed by rhodamine-conjugated secondary antibodies for 2 hours at room temperature. The aortas were then mounted on glass slides with Vectashield that contained DAPI (Vector Laboratories). The aortas were opened, and the lesser curvature (LC) and greater curvature (GC) of each arch were separated. En face images were collected with a Zeiss LSM 510 META confocal microscope.
Small Interference RNA Experiments
Nonsilencing siRNA was purchased from Qiagen. The sequences of specific siRNAs for bovine BMP4, noggin, and follistatin (MWG Biotech, Ebersberg, Germany) are shown in Data Supplement Table II. Transfection was carried out with Oligofectamine (Invitrogen), as we have described.11
Monocyte Adhesion Assay
After treatments, BAECs were washed with fresh media and incubated with fluorescently labeled THP-1 monocytes for 45 minutes at 37°C under no-flow conditions. Bound monocytes were determined by fluorescent microscopy as previously described.10
Statistical analyses were performed with the SPSS software package (SPSS, Chicago, Ill). Data are presented as mean±SEM (n=3 to 4). When significant differences were indicated by ANOVA, Bonferroni post hoc tests were performed for multiple comparisons. In this study, P<0.05 was considered statistically significant.
The authors had full access to and take full responsibility for the integrity of the data. All authors have read and agree to the manuscript as written.
Expression of BMP Antagonists in ECs
To determine which BMP antagonists were expressed in ECs at the mRNA and protein levels, reverse-transcriptase polymerase chain reaction and Western blot analysis were carried out with cultured BAECs and human coronary artery ECs. We detected mRNA for chordin, CV-2, DAN, follistatin, gremlin, MGP, noggin, and Tsg but not cerberus in BAECs (Figure 1A). At the protein level we detected expression of follistatin, noggin (30 kDa), and MGP (17 kDa). Follistatin proteins were detected as 3 distinct bands between 40 and 45 kDa, as expected from the 3 known alternatively spliced forms.33 Follistatin, noggin, MGP, and BMP4 were also easily detected in human coronary artery ECs (Figure 1C). As in BAECs, more follistatin, noggin, and MGP were found in the human coronary artery EC conditioned media than the cell lysate. These results suggest that follistatin, noggin, and MGP are expressed in ECs, and some are secreted into the media.
Differential Regulation of BMP Antagonists and BMP4 Expression by Shear Stress in ECs
To determine whether the expression of the BMP antagonists follistatin, noggin, or MGP was regulated by shear stress, we exposed BAECs to static, LS (15 dyne/cm2), OS (0±5 dyne/cm2), or alternative shear stress profiles: pulsatile LS (P-LS, 15±5 dyne/cm2) or OS with net forward flow (F-OS, 2±5 dyne/cm2) for 1 day (Figure 2, A and D). BMP4 was nearly undetectable in the conditioned media of BAECs exposed to LS, whereas it was abundantly expressed by the cells exposed to OS or static conditions (Figure 2, B and C). BAECs exposed to OS expressed ≈20% more BMP4 than that found in static cells. To our surprise, however, we found that the expression patterns of the 3 BMP antagonists were nearly identical to that of BMP4. The BMP antagonists were barely detectable in LS-exposed cells, whereas they were easily detected in static- and OS-exposed cells. Approximately 24% and 38% more noggin and MGP, respectively, were found in cells exposed to OS than in static cells, whereas follistatin protein levels were similar in OS and static cells (Figure 2, B and C). To further determine which component of shear stress (direction change, amplitude, or pulsatility) plays a critical role in regulation of the expression of BMP4 and the BMP antagonists, the conditioned media of BAECs exposed to P-LS or F-OS was compared with that of static. As shown in Figure 2, nonpulsatile LS was more inhibitory than P-LS at the same mean shear level, whereas bidirectional OS, regardless of forward-flow movement, had similar effects on the expression of BMP4 and the antagonists. The effect of F-OS on their expression was comparable to that of OS (Figure 2, B, C, E, and F). In contrast, P-LS showed less potent inhibitory effects on the expression of BMP4 and the antagonists: P-LS inhibited expression of BMP4, noggin, follistatin, and MGP by 32% to 46% of the static controls (Figure 2, B, C, E, and F). These results suggest that the unidirectionality and pulsatility of LS and the bidirectionality of OS, regardless of net-forward fluid movement, are critical shear components that play a key role in regulation of the levels of BMP4 and its antagonists. BMP4 and the BMP antagonists are coexpressed in mouse aortic endothelium located in disturbed flow regions but not in undisturbed flow regions.
Figure 3A is a schematic diagram of the mouse aortic arch that illustrates the mean wall shear stress distribution. The vector diagrams indicate the time-varying changes in the direction and magnitude of wall shear stress throughout the cardiac cycle, as recently reported by our groups.34 To show the time-varying shear directions and magnitudes in the locations where we collected samples used for the present staining studies, we carried out an additional computational study and modified our previously reported figure (see Figure 3 in the work by Suo et al34), as shown in Figure 3A. The LC of the arch, which corresponds to proinflammatory and atheroprone regions,34,35 showed a remarkably OS pattern with a relatively low mean wall shear level, as expected (Figure 3A).34 In contrast, the GC of the arch and the thoracic aorta (not shown, see Suo et al34) displayed a mostly undisturbed flow profile with predominantly unidirectional and relatively high mean wall shear stress, which corresponds to the atheroresistant region.34,35
To determine whether BMP antagonists were expressed in a flow-dependent manner in arterial ECs, we carried out immunohistochemical analyses with the specific BMP4 and antagonist antibodies in 3 different locations of mouse aortas: the LC, the GC, and the straight part of the thoracic aorta. Detectable levels of BMP4, follistatin, noggin, and MGP were found in ECs that line the LC (arrows in the upper panels of Figure 3B) but not in those in the GC (lower panels in Figure 3B) and straight thoracic aortas (Figure 3C). Unlike in the endothelium, BMP4 and several BMP antagonists (follistatin, noggin, MGP, chordin, CV-2, and DAN) were all easily detected in the medial layers of the aortas (indicated by arrowheads in Figure 3, B and C). Staining with a platelet endothelial adhesion molecule-1 antibody indicated the presence of ECs (Figure 3C). In addition, we performed en face staining for BMP4 and noggin in the same 3 different locations (GC, LC, and the thoracic aorta) of mouse aortas (Figure 3D). BMP4 and noggin were abundantly expressed in the LC compared with the GC, which showed faint but still detectable staining, and the thoracic aorta, which was almost undetectable. The BMP4 and noggin antibodies stained only the endothelium of the mouse aortas and not smooth muscle cells as shown by the orthogonal image (Data Supplement Figure I). Because the follistatin and MGP antibodies were not effective on formalin-fixed tissues, we were not able to determine their expression by en face staining. These results demonstrate that BMP4 and 3 BMP antagonists (follistatin, noggin, and MGP) were coexpressed in mouse aortic endothelium exposed to disturbed flow conditions but not in endothelium exposed to undisturbed flow. These data from mice are consistent with the in vitro findings shown in Figure 2.
Coexpression of BMP4 and the BMP Antagonists Follistatin and Noggin in Human Atherosclerotic Lesions
We next determined whether the BMP antagonists follistatin and noggin were expressed along with BMP4 in human coronary arteries and whether their expression levels correlated with the degree of atherosclerotic lesion development. Serial sections of human coronary arteries that contained various stages of atherosclerotic lesions were stained with anti-BMP4, anti-follistatin, anti-noggin, and anti–platelet endothelial adhesion molecule-1. As shown in the low-magnification images (Figure 4A), BMP4, follistatin, and noggin were easily detected in the medial layer regardless of the disease stage, and their expression levels tended to increase as the disease progressed. Examination of high-magnification images (Figure 4B) revealed the expression of BMP4, follistatin, and noggin in the intimal endothelium (arrows) as well. Although type I lesions (American Heart Association classification) showed faint BMP4 staining in the intimal endothelium, the staining intensity showed an apparent increase as the lesion progressed to types II through VI. We have shown previously that BMP4 expression is difficult to detect in nondiseased human coronary arterial ECs.10 Follistatin and noggin were not detected in early lesions (type I and II), but they were detected in advanced lesions (types IV and VI). Consistent with the mouse immunohistochemistry data (Figure 3), human samples also showed an apparent correlation between BMP4 and BMP antagonist expression in the endothelium overlying atherosclerotic lesions, especially in advanced lesions.
OS Increases Noggin and Follistatin Expression in BAECs by the BMP4-Dependent and -Independent Mechanisms, Respectively
The in vitro and mouse and human arterial immunohistochemical data indicated that BMP4 and BMP antagonists are coexpressed in ECs exposed to OS in vitro and in the endothelium exposed to disturbed flow conditions in the atheroprone regions. These were surprising findings, as we initially hypothesized the opposite—that OS would stimulate the expression of BMP4 while it downregulated BMP antagonists. One possibility is that BMP4, induced by OS in ECs, would stimulate the expression of BMP antagonists as part of a negative feedback mechanism. To examine this hypothesis, we measured the mRNA (Data Supplement Figure II) and protein levels of BMP4, follistatin, or noggin in BAECs treated with either siRNA or recombinant proteins.
First, treatment of BAECs with siRNA specific for BMP4, noggin, or follistatin, but not by nonsilencing control siRNA, knocked down the expression of cognate protein levels as determined by Western blot analyses with the cell lysates or conditioned media (Figure 5A). We then determined whether knockdown of each protein had any effect on the expression of the other 2 antagonists (Figure 5A). BMP4 siRNA reduced noggin expression but not follistatin. Noggin siRNA had no effect on BMP4 or follistatin expression. Similarly, follistatin siRNA had no effect on BMP4 or noggin expression.
Conversely, treatment of ECs with r-BMP4 increased noggin expression but not follistatin expression in a dose-dependent manner (Figure 5B). However, treatment with r-noggin had no effect on BMP4 or follistatin expression, and treatment with r-follistatin had no effect on BMP4 or noggin expression. These results suggested that BMP4 regulates expression of noggin but not follistatin.
Next, we determined whether OS-dependent follistatin and noggin expressions were regulated in a BMP4-dependent manner. For the present study, BAECs treated with BMP4 siRNA or nonsilencing siRNA were exposed to OS for 1 day. BMP4 siRNA knocked down BMP4 levels as expected, and it also inhibited noggin expression but not follistatin (Figure 5C). These results suggest that OS induces the expression of BMP4, which in turn increases noggin expression. In contrast, OS increases follistatin expression in a BMP4-independent manner.
Role of BMP Antagonists as a Negative Feedback Mechanism Against Excessive Inflammation by OS
As described above (Figure 5), we examined the first half of our hypothesis that BMP4, produced in response to OS, would stimulate the expression of noggin and follistatin as part of a negative feedback mechanism to prevent ECs from overactivation. We tested the second half of the hypothesis with regard to the negative feedback mechanism. To that end, we treated BAECs with noggin or follistatin siRNA, as shown in Figure 5, and examined whether knockdown of the BMP antagonists enhanced monocyte adhesion induced by OS or BMP4. As expected,10,11 exposure of ECs to LS inhibited monocyte adhesion, whereas exposure to OS or BMP4 for 1 day increased it above the static control (Figure 6, A and B). OS-induced monocyte adhesion was blocked by BMP4 siRNA treatment (Figure 5C), as shown previously.10,11 In contrast to the BMP4 siRNA, knockdown of either follistatin or noggin significantly increased monocyte adhesion above those of nonsilencing control groups in both static and OS-exposed cells (Figure 6C). Concurrent knockdown of follistatin and noggin further increased monocyte adhesion in both static and OS-exposed cells above those of individual knockdowns, which suggested an additive effect of the 2 antagonists.
Conversely, r-noggin added to the culture media inhibited monocyte adhesion induced by OS, which confirmed our previous findings.10,11 Treatment with r-follistatin and/or r-noggin also decreased monocyte adhesion to a similar degree in both OS and static cells (Figure 6B). Together, these results suggest that noggin and follistatin coexpressed with BMP4 in ECs in response to OS acted as inhibitors of the BMP4-dependent monocyte adhesion, which supports the hypothesis of a negative feedback mechanism.
We present several novel and unexpected findings with regard to the expression of BMP antagonists in ECs. First, 3 BMP antagonists (noggin, follistatin, and MGP) were coexpressed with BMP4 in cultured ECs exposed to OS, whereas LS decreased their expression. Second, and similarly, BMP4 and the 3 BMP antagonists (follistatin, noggin, and MGP) were expressed in the endothelium located in disturbed flow regions (atheroprone regions), but not in stable flow regions (atheroresistant regions) in mouse aortas. Third, in human coronary arterial endothelium, BMP4 and 2 BMP antagonists (follistatin and noggin) were also coexpressed, showing an apparent increase as the atheromatous lesions became more advanced. Fourth, OS upregulated noggin and follistatin expression in BMP4-dependent and -independent mechanisms, respectively. Fifth, the BMP antagonists coexpressed with BMP4 acted as inhibitors of a BMP4-dependent inflammatory response, which suggested negative feedback regulation of BMP4 action by the antagonists.
Although noggin is known as a specific regulator of BMP4 activity, follistatin has not been investigated as a BMP4 antagonist. It is interesting to find that BMP4 regulates only noggin but not follistatin (Figure 5). Although we do not understand the underlying mechanisms responsible for the differential regulation, it suggests a potential specific relationship of BMP4 with noggin but not with follistatin. Noggin is well established as a specific BMP antagonist with a dissociation constant (Kd) of 20 pM,26 whereas follistatin is known as an activin antagonist.28 Nevertheless, our present study suggests that follistatin can also act as a BMP4 antagonist. Our finding that follistatin expression is upregulated along with BMP4 in ECs exposed to unstable flow conditions in vitro and in vivo suggests that it may have a pathophysiological role as a specific BMP4 antagonist.
In the present in vitro study, as in most other studies in this field, we used the static culture condition as a “control”, “normal”, or “100%” and compared the effects of LS and OS to it. However, it is important to note that the static condition is not “normal”; static cells are partially inflamed or activated (Figure 6A) and may represent pathological conditions such as flow cessation because ECs are constantly exposed to flow conditions in vivo. In vivo studies, including the present study, refer to the data obtained from stable flow regions as “normal”, whereas those from disturbed flow regions are referred to as “activated” or “inflamed”. Therefore we feel that it is most appropriate to use LS as the control against which to compare the OS results. If we do, then the in vitro effects of OS versus LS reported in the present study become much more pronounced. For example, Figure 6 would show that OS stimulates monocyte adhesion by >4-fold above that of LS-exposed cells, rather than an ≈90% increase above the static control as we discussed in the Results. Similarly, Figure 2 would show that OS stimulates expression of BMP4 and BMP antagonists by >20-fold above that of LS control.
The importance of LS as a control becomes more apparent when we compare the in vitro results to in vivo findings with mouse aortas (Figure 3). The mouse data clearly demonstrate distinct differences in the expression levels of BMP4, follistatin, noggin, and MGP in ECs according to the flow conditions (ie, disturbed versus stable flow regions). These proteins were detected only in ECs located in the flow-disturbed regions (LC), not in the stable flow areas (GC and the straight thoracic aorta). The flow-disturbed areas also corresponded to atheroprone regions,34,35 which suggests a link between disturbed flow, BMP4, its antagonists, and atherosclerosis. The human atheromatous coronary artery data (Figure 4) provide further supporting evidence for this link. As we have shown previously,10 BMP4 expression was identified in ECs of diseased human coronary arteries but not in nondiseased arteries. The present study further shows an apparent correlation between endothelial BMP4 expression and lesion progression. Moreover, it is interesting to note that noggin and follistatin expressions were apparent only in advanced lesions (types IV to VI) but not in earlier lesions. These findings raise an interesting question as to whether there is any pathophysiological relationship among disturbed flow, the balance between BMP4 and BMP antagonists, and atherosclerosis.
What is the pathophysiological significance of the coexpression of BMP4 and BMP antagonists in ECs exposed to disturbed flow conditions both in vitro and in vivo? The BMP antagonists appear to serve a negative feedback role, by partial inhibition of the inflammatory responses mediated by BMP4 in ECs exposed to disturbed flow. The most direct line of evidence to support this hypothesis is the finding that knockdown of noggin or follistatin increased monocyte adhesion induced by OS (Figure 6C).
On the basis of our findings, we propose the following working hypothesis: ECs exposed to chronic disturbed flow conditions produce BMP4, which increases reactive oxygen species production from the NADPH oxidase and subsequently induces intercellular adhesion molecule-1 expression and monocyte adhesion. Perhaps coproduction of BMP antagonists in response to disturbed flow partially limits the inflammatory response and atherosclerotic lesion development to a minimum. Under continuous proatherogenic insults, which include disturbed flow and other factors such as hypercholesterolemia, hypertension, diabetes, and smoking, the balance between BMP4 and BMP antagonists may change to favor more proinflammatory BMP4 actions. Foam cell lesions and more advanced atherosclerotic lesions, such as BMP4-mediated calcification of medial cells, would then ensue.
In conclusion, we have shown for the first time that ECs coexpress BMP antagonists along with BMP4 in response to disturbed flow conditions and that these antagonists seem to play a negative feedback role against the inflammatory response of BMP4. In addition, coexpression of BMP4 and the antagonists in ECs located in the flow-disturbed regions and atherosclerotic lesions suggests that they may be used as novel biomarkers of inflammation and atherosclerosis.
We thank Michelle Sykes for reading the manuscript and Dr Hyuk Sang Kwon for help with statistical analysis.
Sources of Funding
This work was supported by funding from National Institute of Health grants HL75209 (Dr Jo), HL70531 (Drs Jo, Giddens, and Taylor), and UO1HL80711 (Drs Jo, Giddens, and Taylor).
Dai G, Kaazempur-Mofrad MR, Natarajan S, Zhang Y, Vaughn S, Blackman BR, Kamm RD, Garcia-Cardena G, Gimbrone MA Jr. Distinct endothelial phenotypes evoked by arterial waveforms derived from atherosclerosis-susceptible and -resistant regions of human vasculature. Proc Natl Acad Sci U S A. 2004; 101: 14871–14876.
Davies PF, Shi C, Depaola N, Helmke BP, Polacek DC. Hemodynamics and the focal origin of atherosclerosis: a spatial approach to endothelial structure, gene expression, and function. Ann N Y Acad Sci. 947: 7–16, 2001; discussion 16–17.
Ku DN, Giddens DP, Zarins CK, Glagov S. Pulsatile flow and atherosclerosis in the human carotid bifurcation: positive correlation between plaque location and low oscillating shear stress. Arteriosclerosis. 1985; 5: 293–302.
Traub O, Berk BC. Laminar shear stress: mechanisms by which endothelial cells transduce an atheroprotective force. Arterioscler Thromb Vasc Biol. 1998; 18: 677–685.
Zarins CK, Giddens DP, Bharadvaj BK, Sottiurai VS, Mabon RF, Glagov S. Carotid bifurcation atherosclerosis: quantitative correlation of plaque localization with flow velocity profiles and wall shear stress. Circ Res. 1983; 53: 502–514.
Davies PF. Flow-mediated endothelial mechanotransduction. Physiol Rev. 1995; 75: 519–560.
Brooks AR, Lelkes PI, Rubanyi GM. Gene expression profiling of human aortic endothelial cells exposed to disturbed flow and steady laminar flow. Physiol Genomics. 2002; 9: 27–41.
Butcher JT, Tressel S, Johnson T, Turner D, Sorescu G, Jo H, Nerem RM. Transcriptional profiles of valvular and vascular endothelial cells reveal phenotypic differences: influence of shear stress. Arterioscler Thromb Vasc Biol. 2006; 26: 69–77.
Sorescu GP, Sykes M, Weiss D, Platt MO, Saha A, Hwang J, Boyd N, Boo YC, Vega JD, Taylor WR, Jo H. Bone morphogenic protein 4 produced in endothelial cells by oscillatory shear stress stimulates an inflammatory response. J Biol Chem. 2003; 278: 31128–31135.
Sorescu GP, Song H, Tressel SL, Hwang J, Dikalov S, Smith DA, Boyd NL, Platt MO, Lassegue B, Griendling KK, Jo H. Bone morphogenic protein 4 produced in endothelial cells by oscillatory shear stress induces monocyte adhesion by stimulating reactive oxygen species production from a nox1-based NADPH oxidase. Circ Res. 2004; 95: 773–779.
Massague J, Chen YG. Controlling TGF-beta signaling. Genes Dev. 2000; 14: 627–644.
Hogan BL. Bone morphogenetic proteins: multifunctional regulators of vertebrate development. Genes Dev. 1996; 10: 1580–1594.
Kiyono M, Shibuya M. Bone morphogenetic protein 4 mediates apoptosis of capillary endothelial cells during rat pupillary membrane regression. Mol Cell Biol. 2003; 23: 4627–4636.
Wozney JM, Rosen V Bone morphogenetic protein and bone morphogenetic protein gene family in bone formation and repair. Clin Orthop Relat Res. 1998: 26–37.
Csiszar A, Smith KE, Koller A, Kaley G, Edwards JG, Ungvari Z. Regulation of bone morphogenetic protein-2 expression in endothelial cells: role of nuclear factor-kappaB activation by tumor necrosis factor-alpha, H2O2, and high intravascular pressure. Circulation. 2005; 111: 2364–2372.
Mohler ER 3rd, Gannon F, Reynolds C, Zimmerman R, Keane MG, Kaplan FS. Bone formation and inflammation in cardiac valves. Circulation. 2001; 103: 1522–1528.
Miriyala S, Gongora Nieto MC, Mingone C, Smith D, Dikalov S, Harrison DG, Jo H. Bone morphogenic protein-4 induces hypertension in mice: role of noggin, vascular NADPH oxidases, and impaired vasorelaxation. Circulation. 2006; 113: 2818–2825.
Nakamura T, Takio K, Eto Y, Shibai H, Titani K, Sugino H. Activin-binding protein from rat ovary is follistatin. Science. 1990; 247: 836–838.
Sudo S, Avsian-Kretchmer O, Wang LS, Hsueh AJ. Protein related to DAN and cerberus is a bone morphogenetic protein antagonist that participates in ovarian paracrine regulation. J Biol Chem. 2004; 279: 23134–23141.
Boo YC, Sorescu G, Boyd N, Shiojima I, Walsh K, Du J, Jo H. Shear stress stimulates phosphorylation of endothelial nitric-oxide synthase at Ser1179 by Akt-independent mechanisms: role of protein kinase A. J Biol Chem. 2002; 277: 3388–3396.
Jo H, Sipos K, Go YM, Law R, Rong J, McDonald JM. Differential effect of shear stress on extracellular signal-regulated kinase and N-terminal Jun kinase in endothelial cells: Gi2- and Gbeta/gamma-dependent signaling pathways. J Biol Chem. 1997; 272: 1395–1401.
Go YM, Boo YC, Park H, Maland MC, Patel R, Pritchard KA, Jr., Fujio Y, Walsh K, Darley-Usmar V, Jo H. Protein kinase B/Akt activates c-Jun NH(2)-terminal kinase by increasing NO production in response to shear stress. J Appl Physiol. 2001; 91: 1574–1581.
Suo J, Ferrara DE, Sorescu D, Guldberg RE, Taylor WR, Giddens DP. Hemodynamic shear stresses in mouse aortas: implications for atherogenesis. Arterioscler Thromb Vasc Biol. 2007; 27: 346–351.
VanderLaan PA, Reardon CA, Getz GS. Site specificity of atherosclerosis: site-selective responses to atherosclerotic modulators. Arterioscler Thromb Vasc Biol. 2004; 24: 12–22.
Atherosclerosis is an inflammatory disease that occurs preferentially in arterial regions exposed to disturbed flow such as oscillatory shear stress (OS). Exposure to OS stimulates endothelial cells (ECs) to produce bone morphogenic protein (BMP) 4, which in turn activates inflammation, a critical atherogenic step. Moreover, chronic BMP4 infusion in mice induces hypertension, which can be inhibited by the BMP inhibitor noggin. In the present study, we examined whether BMP antagonists are expressed in ECs in a flow-dependent manner and whether they play a role in EC pathophysiology. Several BMP antagonists, such as noggin, follistatin, matrix Gla protein, DAN, gremlin, crossveinless-2, and twisted gastrulin, were expressed in cultured ECs. Surprisingly, OS increased expression of noggin, follistatin, and matrix Gla protein, along with BMP4, in ECs, whereas atheroprotective laminar shear decreased such expression. Immunohistochemical studies with mouse aortas showed that ECs exposed to disturbed flow but not aortas exposed to undisturbed flow showed coexpression of BMP4 and the BMP antagonists. Similarly, in human coronary arteries, expression of BMP4 and BMP antagonists in ECs positively correlated with the severity of atherosclerosis. Monocyte adhesion, a key inflammatory step, induced by OS was inhibited by BMP4 knockdown or treatment with recombinant follistatin or noggin, whereas it was increased by follistatin or noggin knockdown. These results suggest that ECs coexpress BMP antagonists along with BMP4 in an attempt to minimize the inflammatory response by OS as part of a negative feedback mechanism. The balance between the agonist BMP4 and its antagonists may play an important role in the overall control of inflammation and atherosclerosis.
The online-only Data Supplement, which consists of methods, figures, and tables, is available online with this article at http://circ.ahajournals.org/cgi/content/full/CIRCULATIONAHA.106.683227/DC1.