Fluid shear stress regulates placental growth factor expression via heme oxygenase 1 and iron

Increased fluid shear stress (FSS) is a key initiating stimulus for arteriogenesis, the outward remodeling of collateral arterioles in response to upstream occlusion. Placental growth factor (PLGF) is an important arteriogenic mediator. We previously showed that elevated FSS increases PLGF in a reactive oxygen species (ROS)-dependent fashion both in vitro and ex vivo. Heme oxygenase 1 (HO-1) is a cytoprotective enzyme that is upregulated by stress and has arteriogenic effects. In the current study, we used isolated murine mesentery arterioles and co-cultures of human coronary artery endothelial cells (EC) and smooth muscle cells (SMC) to test the hypothesis that HO-1 mediates the effects of FSS on PLGF. HO-1 mRNA was increased by conditions of increased flow and shear stress in both co-cultures and vessels. Both inhibition of HO-1 with zinc protoporphyrin and HO-1 knockdown abolished the effect of FSS on PLGF. Conversely, induction of HO-1 activity increased PLGF. To determine which HO-1 product upregulates PLGF, co-cultures were treated with a CO donor (CORM-A1), biliverdin, ferric ammonium citrate (FAC), or iron-nitrilotriacetic acid (iron-NTA). Of these FAC and iron-NTA induced an increase PLGF expression. This study demonstrates that FSS acts through iron to induce pro-arteriogenic PLGF, suggesting iron supplementation as a novel potential treatment for revascularization.


Increased fluid shear stress (FSS) is a key initiating stimulus for arteriogenesis, the outward remodeling of collateral arterioles in response to upstream occlusion. Placental growth factor (PLGF) is an important arteriogenic mediator. We previously showed that elevated FSS increases PLGF in a reactive oxygen species (ROS)-dependent fashion both in vitro and ex vivo. Heme oxygenase 1 (HO-1) is a cytoprotective enzyme that is upregulated by stress and has arteriogenic effects. In the current study, we used isolated murine mesentery arterioles and co-cultures of human coronary artery endothelial cells (EC) and smooth muscle cells (SMC) to test the hypothesis that HO-1 mediates the effects of FSS on PLGF. HO-1 mRNA was increased by conditions of increased flow and shear stress in both co-cultures and vessels. Both inhibition of HO-1 with zinc protoporphyrin and HO-1 knockdown abolished the effect of FSS on PLGF. Conversely, induction of HO-1 activity increased PLGF. To determine which HO-1 product upregulates PLGF, co-cultures were treated with a CO donor (CORM-A1), biliverdin, ferric ammonium citrate (FAC), or iron-nitrilotriacetic acid (iron-NTA). Of these FAC and iron-NTA induced an increase PLGF expression. This study demonstrates that FSS acts through iron to induce pro-arteriogenic PLGF, suggesting iron supplementation as a novel potential treatment for revascularization.
Coronary artery disease (CAD) is a major cause of death worldwide 1,2 . A primary predictor of survival for CAD patients is the number of preexisting collateral vessels (arterial-arterial anastomoses) and the degree to which they have remodeled outward to increase their flow capacity 3,4 . This outward remodeling occurs in response to increased flow through the vessels, which is generated by an increase in the pressure gradient across the vessels due to decreased downstream pressure in the occluded branch. This remodeling process is termed arteriogenesis. Pharmacological stimulation of arteriogenesis has been a long sought after goal, because of the potential of arteriogenesis to reduce mortality and morbidity in CAD. However, early attempts to induce arteriogenesis via administration of single exogenous growth factors were marred with failure 5 , and it is clear that a deeper understanding of the myriad of signaling events contributing to arteriogenesis is necessary to develop safe and effective pro-arteriogenic treatments.
Placental growth factor (PLGF) is a member of the vascular endothelial growth factor (VEGF) family. PLGF is a potent arteriogenic agent, even more so than VEGF-A 6,7 . PLGF exclusively binds fms-like tyrosine kinase-1 (VEGFR-1) and elicits distinct downstream signaling events than those induced by VEGF-A binding of VEGFR-1 8 . PLGF exacts its arteriogenic effect by recruitment of monocytes (which only express VEGFR-1) to the vascular wall 6,9,10 . PLGF knockout mice exhibit a blunted arteriogenic response to hindlimb ischemia 6 , whereas PLGF protein levels in coronary artery plasma are positively correlated with improved patient outcome following myocardial infarction 11 .
Heme oxygenase (HO) catabolizes heme into equimolar quantities of CO, divalent iron, and biliverdin. There are two isoforms of HO; an inducible isoform (HO-1) and a non-inducible, constitutively expressed isoform (HO-2). HO-2 is constitutively expressed in the testes 12 and the brain 13 . HO-1, on the other hand, is strongly induced in response to cellular stresses such as increased reactive oxygen species 14,15 (ROS) and radiation 16,17 . There are only two reported cases of HO-1 deficiency in humans. These patients exhibited severe growth retardation and endothelial dysfunction, along with abnormal hemostasis and an increased susceptibility to oxidative stress [18][19][20] . HO-1 knockout mice present with similar growth abnormalities and appear to be in a chronic inflammatory state 21 . Arteriogenesis is diminished with advanced age 22  www.nature.com/scientificreports/ expression and/or signaling, since induction of HO-1 in aged rats with blunted arteriogenic potential restores outward vascular remodeling to levels comparable with young rats 23 . We previously reported that exposure to arteriogenic FSS increases PLGF protein and mRNA both in vitro in a co-culture model of the vessel wall and ex vivo in isolated mouse mesenteric arterioles. We also reported that this increase is dependent on FSS-induced production of hydrogen peroxide by endothelial NADPH oxidase 4 (Nox4) 24 . It is established that FSS also increases activation and expression of HO-1 in endothelial cells [25][26][27] , and that these effects are dependent on ROS produced by Nox isoforms 25,27 . Similarly, HO-1 is upregulated in hindlimb skeletal muscle following femoral artery ligation 28,29 in a Nox-dependent manner 29 . A possible link between HO-1 and PLGF is suggested by the observation that PLGF and HO-1 expression are both significantly increased in hindlimb skeletal muscle early after femoral artery ligation 30 . Furthermore, HO-1 haploinsufficiency in mice causes a decrease in PLGF expression and reduces the extent of revascularization following induction of hindlimb ischemia 31 . Lastly, HO-1 knockout increases the expression of antiarteriogenic soluble VEGFR-1 32 . Therefore, we hypothesized that the effects of FSS on PLGF are mediated by HO-1. We tested this hypothesis in an endothelial cell/smooth muscle cell co-culture model and in isolated mouse mesenteric arterioles in order to characterize the role of HO-1 in FSS-mediated regulation of PLGF.

Results
First, we determined the relative mRNA levels of HO-1 in our models following experimental treatment. In endothelial cells (Fig. 1A), shear stress significantly increased HO-1 mRNA immediately after exposure (1.40 ± 0.08-fold of static control), and this increase remained evident 4 h after exposure (1.54 ± 0.18-fold of static). By 10 h after exposure, HO-1 mRNA was not significantly different from control. In smooth muscle cells We previously reported that FSS increases PLGF expression both in our co-culture model and in perfused vessels 24 . To determine HO-1's role in the FSS induced expression of PLGF, we inhibited HO-1 with zinc protoporphyrin IX (ZPP). ZPP (30 µM) blocked the effect of FSS on PLGF, and even decreased PLGF protein below static control levels ( Fig. 2A). Similarly, in perfused vessels ZPP (30 µM) prevented the effect of increased flow on PLGF mRNA (Fig. 2B).
To determine the relative importance of endothelial cell vs smooth muscle cell HO-1 activity in mediating the effects of FSS on PLGF expression, HO-1 was separately knocked down in each cell type in the co-culture model using siRNA (Fig. 4). HO-1 knockdown in endothelial cells prevented the FSS-induced expression of PLGF (Fig. 4A). In contrast, knockdown of HO-1 in smooth muscle cells did not affect the FSS-mediated increase in PLGF (Fig. 4B). Knockdown in each cell type was confirmed by real time PCR (Fig. 4C,D). HO-1 knockdown in endothelial cells did not affect HO-1 mRNA in co-cultured smooth muscle cells, or vice versa (not shown).

Discussion
In this study we demonstrated that FSS increases HO-1 expression both in an in vitro co-culture model of the cell wall and ex vivo in intact vessels. We further showed that the FSS mediated increase in PLGF expression which we previously reported is dependent on HO-1 activity, we also identified endothelial cell HO-1 activity as necessary for this response. FSS-independent activation of HO-1 (by hemin) was sufficient to induce PLGF expression. Furthermore, treatment of vascular cell co-cultures with the three products of HO-1 activity identified iron as the mediator of the effects of HO-1 on PLGF expression.
Blood flow recovery in the ischemic hindlimb of HO-1 knockout mice is significantly impaired 33 and inhibition of HO-1 activity following hindlimb ischemia in mice results in poor flow recovery and diminished recruitment of circulating cells to the ischemic hindlimb 28 . Consistent with these observations, we found that inhibition of HO-1 activity attenuates the effects of shear stress on PLGF. PLGF stimulates collateral remodeling by recruitment of monocytes 6 and other circulating cells to the vessel wall 34 . Therefore, our findings suggest a novel mechanism for the above-described effects of HO-1 inhibition on arteriogenesis in rodent models. Stimulation of HO-1 with cobalt protoporphyrin IX following myocardial infarction has been shown to result in improved outcome and greater neovascularization in rats 35 . Likewise, overexpression of HO-1 in both mice and rats results in improved re-establishment of blood flow following hindlimb ischemia 36,37 . Similarly, overexpression of PLGF has been shown to improve cardiac performance and vascularization following myocardial infarction in mice 38 .
We report that arteriogenic FSS increases HO-1 expression in endothelial cells and smooth muscle cells, consistent with previous findings that laminar FSS induces HO-1 in endothelial cells 39  www.nature.com/scientificreports/ HO-1 more robustly in intact vessels than did increased shear in our in vitro co-culture model. The co-culture model much more closely approximates the vessel wall than do monocultures of vascular cells 40 . However, there are limitations to the model, including the lack of exposure of the cells to cyclic tangential or circumferential stretch. Furthermore, the model lacks the complexity of the in vivo extracellular matrix, including the extensive glycocalyx of intact vessels (which is important in the mechanosensing machinery of the vessel wall) [41][42][43][44] . Despite these shortcomings, the co-culture model offers practical advantages with regard to cell type specificity for both genetic manipulation and assays, which are not easily achieved in whole tissue. Comparison of the results from the intact vessel model with those from the co-culture model suggests that the role of FSS in regulation of PLGF may be even more pronounced in vivo, highlighting the potential physiological significance of this pathway. HO-1 catabolizes heme into three products: biliverdin, CO, and divalent iron. Biliverdin is subsequently metabolized into bilirubin by biliverdin reductase. Biliverdin exhibits strong antioxidant effects [45][46][47] and has been   48 . Furthermore, biliverdin inhibits neointimal thickening following vascular injury in rats 49 . These beneficial effects have been attributed to biliverdin's antioxidant properties. Consistent with its antioxidant action and its ability to inhibit HO-1, we found that biliverdin inhibited rather than enhanced PLGF expression in static co-cultures. We previously demonstrated that the effects of shear stress on PLGF expression are mediated by hydrogen peroxide produced by endothelial NADPH oxidase 4 24 ; thus, antioxidants would not be expected to stimulate PLGF expression. Several studies have demonstrated a cytoprotective role of HO-1 50,51 . The cytoprotective effects of HO-1 have been attributed to CO. Rats pretreated with the CO donor methylene chloride before myocardial infarction developed more intermediate and large collateral arteries in the infarct area 35 . Furthermore, CO has been reported to induce VEGF expression in endothelial cells 52 and smooth muscle cells 53 . Despite these arteriogenic effects, the CO donor CORM-A1 did not induce PLGF expression in our co-culture model; indeed, it slightly but significantly reduced PLGF. CO has also been shown to have antioxidant properties [54][55][56] , which may act to reduce PLGF expression.
HO-1 activity results in the release of labile iron, which is cytotoxic in excess amounts. Increased labile iron leads to oxidative stress within the cell through Fenton reactions. We were able to demonstrate that the effects of iron on PLGF expression were not due to oxidative stress, as catalase treatment had no effect on the FAC-induced increase in PLGF. We conclude that iron is the HO-1 product that mediates upregulation of PLGF by FSS.
Increased iron levels in different pathological conditions including sickle cell disease (SCD) 57,58 and hereditary hemochromatosis (HH) 59 have been linked to significantly increased PLGF levels. More interestingly, PLGF levels in HH patients decreased after iron overload relief via phlebotomy 59 . Lower iron levels were associated with the risk of preeclampsia 60,61 , where low serum PLGF level is used as an important indicator for disease diagnosis. These studies provide an important aspect of the role iron plays in PLGF upregulation.
Iron metabolism is complex, and further alterations in iron homeostasis in our model system following FSS exposure remain to be determined. For example, as a protective response against pro-oxidant Fenton reactions, increased labile iron induces an increase in ferritin translation 62 and iron sequestration. Labile iron also induces expression of an iron efflux pump colocalized with HO-1 63 . Ferroportin 1, an iron exporter, is also upregulated by labile iron generated by HO-1 64 . Indeed, HO-1 knockout in mice and HO-1 deficiency in humans is linked to anemia, due to accumulation of iron in tissues and decreased iron recycling 18,20,65 , whereas overexpression of HO-1 is associated with increased cellular iron efflux and decreased influx 66 . FAC has also been reported to increase ferritin synthesis 67 and ferroportin expression 68,69 . Taken together, these data suggest that an increase in HO-1 expression/activity would be expected to result in a secondary decrease in labile or "chelatable" iron, which has implications for downstream signaling.
In conclusion, we demonstrate that the key arteriogenic factor PLGF is regulated by HO-1 in response to the physiological stimuli of FSS, which is considered to be an important signal for collateral development in the coronary circulation. This study builds on the increased interest in HO-1 and its metabolites in vascular remodeling by shedding light into a novel possible mechanism by which HO-1 exerts its arteriogenic effects.

Methods
Reagents. All reagents were purchased from Sigma-Aldrich unless otherwise specified.
Perfused arterioles. All animal experimental procedures were approved by the Institutional Animal Care and Use Committee at Oklahoma State University (assurance number A3722-01, protocol number ACUP VM-  and the reporting in this manuscript follows the recommendations in the ARRIVE guidelines. 6-8-week-old C57BL/6J male mice were purchased from Jackson Laboratories. Mice were deeply anesthetized  After cannulation of the proximal (upstream) end of the vessel, the intraluminal pressure was gradually raised (less than 20 mmHg) to clear the lumen of clotted blood. Once cleared, the distal (downstream) end of the vessel was also cannulated. Time from euthanasia to complete cannulation was under 60 min. The temperature of the bath was then raised to 37 °C and pressure was gradually raised to 60 mmHg (~ 10 mmHg/10 min). The pressure increase was achieved by gradually raising two reservoirs connected by silicone tubing to each cannula. Perfusion buffer consisted of 1% bovine serum albumin in PSS. Once equilibrated at 60 mmHg, the longitudinal pressure gradient was increased from zero to 20 mmHg ("control") or 50 mmHg ("proarteriogenic") 70 . This was achieved by lowering the distal reservoir and raising the proximal reservoir, allowing for the average intraluminal pressure to be maintained at 60 mmHg. The control flow rate was ~ 75 µL/min, and the "proarteriogenic" flow rate was ~ 170 µL/min. Vessels were then perfused for 2 h. Function of the vessel wall was determined at the end of perfusion by assessing the vasoconstrictive response to epinephrine and the vasodilator response to acetylcholine, as assessed by changes in vessel diameter measured by video micrometer.    24 . Inserts were inverted and the bottom surface was coated with 0.1% gelatin in DMEM, then were placed in a humidified incubator in 5% CO 2 at 37 °C for 1 h. HCASMC (10 4 cells/cm 2 ) were then seeded onto the inverted insert, and inserts were returned to the incubator overnight (Fig. 5A). The following day, the inserts were placed into 6 well plates containing SmGM-2 media and incubated for an additional 24 h (Fig. 5B). The top surface of the insert was then coated with 0.1% gelatin in DMEM and incubated for 1 h. HCAEC (25,000/cm 2 ) were then seeded on the top surface of the insert. EGM-2MV was added to the insert and the system was again incubated overnight (Fig. 5C). When possible, HCASMC and HCAEC donors were matched. Confluence of the co-cultures was confirmed by Hoffman modulation contrast microscopy (Olympus IX71). Lastly, the confluent co-cultures were incubated in serum reduced media for 24 h prior to experiments. For HCAEC mono-culture experiments, no HCASMC were seeded on the bottom of the insert, but cultures were otherwise processed as described above.
Shear stress exposure. Only the HCAEC layer of the co-culture was directly exposed to FSS. FSS was applied using a cone and plate viscometer shearing system as we previously described 24 . Co-cultures were then exposed to a pulsatile FSS waveform that had a time-averaged FSS of 1.24 Pa (Fig. 5D). This waveform is based on previously published models of coronary collateral flow 71 . Shear experiments were performed on a laboratory benchtop in HEPES-buffered media with temperature maintained at 37 °C. Co-cultures were exposed to FSS for 2 h. Culture media and/or cell lysates were collected for analysis at various time points, from pre-shear up to 24 h post-shear.
siRNA knockdown experiments. HCAEC were seeded into 6-well plates at a density of 210,000 cells/ well. After 24 h, cells were transfected with either predesigned HO-1 siRNA (Silencer Select; s194530) or negative control siRNA (Silencer no.1 siRNA; scRNA), all purchased from Invitrogen. Prior to addition to cells, 5 nM of siRNA was precomplexed with lipofectamine RNAiMAX transfection reagent (Invitrogen) in Opti-MEM media (Gibco) for 20 min. Cells were exposed to transfection media (DMEM + 10% FBS containing precomplexed siRNA) for 6 h, after which cells were trypsinized and seeded onto the upper surface of inserts precoated with 0.1% gelatin. The lower surface of the inserts had been previously seeded with wild type HCASMC, as described above. Co-cultures were incubated overnight in reduced serum media as described above before exposure to shear stress. In a separate group of co-cultures, HCASMC were transfected similarly to HCAEC, after seeding onto the lower surface of inserts. At the end of the transfection period, HCASMC were washed and untreated HCAEC were then seeded onto the upper surface as above. Cell specificity and efficacy of target mRNA knockdown was determined by real time PCR.
PLGF ELISA. Media samples were collected from treated cells and their corresponding controls, treated with protease inhibitor cocktail (1 mM PMSF, 1 mM Na 3 VO 4 , 1 µg/mL leupeptin, 1 mM benzamindine-HCl, 1 µg/mL aprotinin, 1 µg/mL pepstatin A) and stored at − 80 °C until further processing. Media samples were collected prior to treatment and 24 h after treatment (The time at which FSS had the greatest effect on PLGF protein) PLGF was measured using the DuoSet ELISA development kit (R&D Systems) according to manufacturer's pro- After a further overnight incubation in serum reduced media, cells were exposed to FSS using a microstepper driven cone.