Febuxostat, a Xanthine Oxidase Inhibitor, Decreased Macrophage Matrix Metalloproteinase Expression in Hypoxia

Macrophages in the atheroma region produce matrix metalloproteinases (MMPs) and decrease plaque stability. Tissue oxygen tension decreases in the arterial wall of the atherosclerotic region. Hypoxia inducible factor (HIF)-1α plays a critical role in the transcriptional activation of hypoxia inducible genes. However, the precise roles of HIF-1α independent pathways in hypoxic responses are largely unknown. Xanthine oxidase (XO) is an enzyme that utilizes molecular oxygen and produces reactive oxygen species (ROS). Here, we show that ROS derived from XO increases MMP-3, -10, and -13 expression in murine macrophages. We found that the transcript levels of macrophage MMP-3, -10, and -13 were increased in hypoxic conditions. Hypoxia induced MMP expression in HIF-1α deficient macrophages. N-acetylcysteine (NAC) or febuxostat, an XO inhibitor, suppressed MMP expression in murine macrophages. Febuxostat decreased the incidence of plaque rupture in apolipoprotein-E-deficient mice. Our results indicate that febuxostat stabilized atherosclerotic plaque via suppressing the activities of macrophage MMP-9 and -13. Febuxostat administration is a potential therapeutic option in the management of atherosclerotic patients.


Introduction
The incidence rate of cardiovascular events, including ischemic heart diseases and stroke, is increasing annually worldwide. Thrombosis caused by atherosclerotic plaque rupture leads to acute myocardial infarction, unstable angina, and sudden death from ischemic heart disease, known as acute coronary syndrome [1][2][3]. Therefore, it is important to understand the mechanisms of atherosclerotic plaque vulnerability and rupture to prevent acute coronary syndrome. The features of plaque vulnerability include a large necrotic core, high infiltration of inflammatory macrophages, and a thin fibrous cap composed of a small number of collagen fibers. Macrophages are immune cells Macrophages in atherosclerotic plaque are exposed to hypoxic conditions [20]. Therefore, to examine the expression level of various MMP mRNAs under hypoxic conditions (1% oxygen concentration), primary TEPMs were isolated from mice. It was found that mRNAs of MMP-3, -10, and -13 were significantly increased under hypoxia (Figure 1), whereas the expression of other MMPs was unaffected ( Figure 1 and data not shown). HIF-1α is the central regulator of the cellular response to hypoxia. Thus, to examine whether HIF-1α was involved in hypoxia-induced elevation of these MMP gene expressions, primary TEPMs were isolated from hematopoietic/endothelial-specific HIF-1α knockout mice (HIF-1α flox/flox ; Tie2-cre +/− mice; HIF-1α KO) or cre-negative littermate controls. Pgk1 mRNA, well-known as a HIF-1α target gene, was significantly reduced in HIF-1α KO TEPMs ( Figure 2a). However, MMP-3, -10, and -13 mRNAs were induced under hypoxic conditions in HIF-1α KO TEPMs, as in wild-type TEPMs (Figure 2a,b). These results showed that hypoxia increased MMP-3, -10, and -13 mRNAs in a HIF-1α-independent manner. Biomedicines 2019, 7, x FOR PEER REVIEW 3 of 15 Figure 1. MMP-3, -10, and -13 mRNAs in TEPMs are induced by hypoxia. TEPMs were exposed to hypoxia (1% O2), and the relative expression level of MMPs mRNAs was analyzed. Data are the means and SE of at least three independent experiments. The one-way ANOVA and Tukey's multiple comparisons test were performed for the statistical analysis. * p < 0.05. Hypoxic induction of MMP-3, -10, and -13 mRNAs is independent of HIF-1α activity. (a) TEPMs were isolated from hematopoietic/endothelial-specific HIF-1α knockout mice (Cre (+)) or crenegative littermates (Cre (-)) as a control. TEPMs were exposed to hypoxia (1% O2), and the relative expression levels of MMP-3, -10, and -13 mRNAs were analyzed. Quantitative PCR analysis was repeated in at least three independent experiments. The two-way ANOVA and Tukey's multiple comparisons test were performed for the statistical analysis. * p < 0.05. (b) The deletion efficiency of HIF-1α mRNA in TEPMs was examined. Data are the means and SE of at least three independent experiments. The unpaired t-test was performed for the statistical analysis. * p < 0.05.

The Induction of MMP-3 and -10 mRNAs by Hypoxia Was Dependent on ROS Generated through XO Activity
Cells cultured under hypoxic conditions have elevated intracellular ROS [12,21]. We hypothesized that such elevated ROS in TEPMs might be responsible for the increased expression of MMP-3, -10, and -13 mRNAs. To test this, we asked whether the ROS scavenger N-acetyl-L-cysteine (NAC) could attenuate the expression of these MMPs under hypoxic conditions. The treatment with NAC significantly suppressed hypoxic elevation of MMP-3, -10, and -13 mRNAs (Figure 3a). It is well-known that intracellular ROS is mainly derived from mitochondria, NADPH oxidase, and (a) TEPMs were isolated from hematopoietic/endothelial-specific HIF-1α knockout mice (Cre (+)) or cre-negative littermates (Cre (-)) as a control. TEPMs were exposed to hypoxia (1% O 2 ), and the relative expression levels of MMP-3, -10, and -13 mRNAs were analyzed. Quantitative PCR analysis was repeated in at least three independent experiments. The two-way ANOVA and Tukey's multiple comparisons test were performed for the statistical analysis. * p < 0.05. (b) The deletion efficiency of HIF-1α mRNA in TEPMs was examined. Data are the means and SE of at least three independent experiments. The unpaired t-test was performed for the statistical analysis. * p < 0.05.

The Induction of MMP-3 and -10 mRNAs by Hypoxia Was Dependent on ROS Generated through XO Activity
Cells cultured under hypoxic conditions have elevated intracellular ROS [12,21]. We hypothesized that such elevated ROS in TEPMs might be responsible for the increased expression of MMP-3, -10, and -13 mRNAs. To test this, we asked whether the ROS scavenger N-acetyl-l-cysteine (NAC) could attenuate the expression of these MMPs under hypoxic conditions. The treatment with NAC significantly suppressed hypoxic elevation of MMP-3, -10, and -13 mRNAs (Figure 3a). It is well-known that intracellular ROS is mainly derived from mitochondria, NADPH oxidase, and xanthine oxidase (XO) [22,23]. To identify which ROS sources were involved in the up-regulation of MMP-3, -10, and -13 mRNAs under hypoxic conditions, selective inhibitions of mitochondrial ROS with Mito-TEMPO, NADPH oxidase with VAS2870, and XO with febuxostat were performed. While MMP-13 was not suppressed by any inhibitors, hypoxic induction of MMP-3 and -10 mRNAs was significantly suppressed by febuxostat, but not Mito-TEMPO and VAS2870 (Figure 3b-d). Moreover, ROS signaling by XO under hypoxia did not affect the accumulation and transcriptional activity of HIF-1α protein (Supplementary Figure S1a-c). These findings suggest that ROS derived from XO contributes to the hypoxia-induced elevation of MMP-3 and -10 mRNAs in TEPMs independent of HIF-1α signaling.
Biomedicines 2019, 7, x FOR PEER REVIEW 5 of 15 xanthine oxidase (XO) [22,23]. To identify which ROS sources were involved in the up-regulation of MMP-3, -10, and -13 mRNAs under hypoxic conditions, selective inhibitions of mitochondrial ROS with Mito-TEMPO, NADPH oxidase with VAS2870, and XO with febuxostat were performed. While MMP-13 was not suppressed by any inhibitors, hypoxic induction of MMP-3 and -10 mRNAs was significantly suppressed by febuxostat, but not Mito-TEMPO and VAS2870 (Figure 3b-d). Moreover, ROS signaling by XO under hypoxia did not affect the accumulation and transcriptional activity of HIF-1α protein (Supplementary Figure 1a-c). These findings suggest that ROS derived from XO contributes to the hypoxia-induced elevation of MMP-3 and -10 mRNAs in TEPMs independent of HIF-1α signaling.

Hypoxia Activates MMPs via ROS Derived from XO
It was previously reported that MMP-3 and -10, also known as stromelysin-1 and 2, digest several extracellular matrix proteins but not interstitial collagen. Furthermore, these MMPs participate in proMMP activation for tissue remodeling. Stromelysin-1 (MMP-3) activates gelatinase B (MMP-9) in vascular smooth muscle cells (VSMCs), and contributes to neointima formation [24]. On the other hand, Stromelysin-2 (MMP-10) is expressed in macrophages. MMP-10-deficient macrophages reduce collagenolytic activity, particularly MMP-13 [25]. Therefore, to examine the hypothesis that febuxostat attenuates collagenolytic and gelatinolytic activity in macrophages, we measured MMP activity in culture media of macrophages by in-gel zymography [9,26]. The zymographic analysis showed that febuxostat significantly suppressed collagenolytic and gelatinolytic activity in macrophages (Figure 4a,b). These results suggest that inhibition of XO signaling attenuated the activity of collagenase and gelatinase in TEPMs.

Hypoxia Activates MMPs via ROS Derived from XO
It was previously reported that MMP-3 and -10, also known as stromelysin-1 and 2, digest several extracellular matrix proteins but not interstitial collagen. Furthermore, these MMPs participate in proMMP activation for tissue remodeling. Stromelysin-1 (MMP-3) activates gelatinase B (MMP-9) in vascular smooth muscle cells (VSMCs), and contributes to neointima formation [24]. On the other hand, Stromelysin-2 (MMP-10) is expressed in macrophages. MMP-10-deficient macrophages reduce collagenolytic activity, particularly MMP-13 [25]. Therefore, to examine the hypothesis that febuxostat attenuates collagenolytic and gelatinolytic activity in macrophages, we measured MMP activity in culture media of macrophages by in-gel zymography [9,26]. The zymographic analysis showed that febuxostat significantly suppressed collagenolytic and gelatinolytic activity in macrophages (Figure 4a,b). These results suggest that inhibition of XO signaling attenuated the activity of collagenase and gelatinase in TEPMs.   Figure S2). Data are expressed as the means ± SD of three independent experiments. The two-way ANOVA and Tukey's multiple comparisons test were performed for the statistical analysis. * p < 0.05.

XO Inhibitor Suppresses the Rupture of Atherosclerotic Plaques in Mice
MMP production from macrophages contributes to atherosclerotic plaque instability and rupture [4]. It was examined whether XO signaling contributed to the vulnerability of atherosclerotic plaques using an atherosclerotic plaque rupture mouse model generated by the ligation and cuff placement method (Figure 5a) [27]. Atherosclerotic lesions were formed (Figure 5b) as follows: 15 mg/L febuxostat was continuously administered to these mice via their drinking water for 32 days. The mice were then euthanized to perform histological analyses. As shown in Figure 5c, the average number of plaque rupture lesions was significantly decreased in mice that drank the water containing febuxostat. These findings suggest that XO activity contributed to the instability of atherosclerotic plaques through MMP.
Biomedicines 2019, 7, x FOR PEER REVIEW 7 of 15 of MMP-9 and 13 were assayed by collagen zymography. The quantification of MMP activity was performed using a densitometry analytic tool (Supplementary Figure 2). Data are expressed as the means ± SD of three independent experiments. The two-way ANOVA and Tukey's multiple comparisons test were performed for the statistical analysis. * p < 0.05.

XO Inhibitor Suppresses the Rupture of Atherosclerotic Plaques in Mice
MMP production from macrophages contributes to atherosclerotic plaque instability and rupture [4]. It was examined whether XO signaling contributed to the vulnerability of atherosclerotic plaques using an atherosclerotic plaque rupture mouse model generated by the ligation and cuff placement method (Figure 5a) [27]. Atherosclerotic lesions were formed (Figure 5b) as follows: 15 mg/L febuxostat was continuously administered to these mice via their drinking water for 32 days. The mice were then euthanized to perform histological analyses. As shown in Figure 5c, the average number of plaque rupture lesions was significantly decreased in mice that drank the water containing febuxostat. These findings suggest that XO activity contributed to the instability of atherosclerotic plaques through MMP.

Discussion
It is well understood that inflammatory macrophages contribute to atherosclerotic plaques' formation and vulnerability [28]. The microenvironment in atherosclerotic plaques contains various oxidized lipids and cytokines. Macrophages are exposed to these stimuli, resulting in the inflammatory phenotype. These inflammatory macrophages express and secrete pro-inflammatory cytokines and MMPs. Macrophages in atherosclerotic lesions have strong proteolytic activity and cause the vulnerability and rupture of plaques. At the same time, it has also been reported that the macrophage-rich region in human carotid atherosclerotic lesions is a hypoxic environment [20]. Therefore, macrophages within atherosclerotic plaques are exposed to not only inflammatory but also hypoxic environments. However, little is known about whether hypoxic stress on macrophages affects atherosclerotic plaque progression.
The present study reveals that hypoxic stress to macrophages induces the expression of MMP-3, -10, and -13 mRNAs, and activates gelatinolytic and collagenolytic activity. It was previously reported that MMP-3, -10, and -13 were involved in the development of atherosclerotic lesions and plaque progression [29][30][31][32]. MMP-3 was one of the first MMPs identified in atherosclerotic plaques, and it is expressed in macrophages, as well as lymphocytes and activated VSMCs [5]. The MMP-3 knockout mice showed a reduction in intimal growth caused by the ligation of carotid arteries. MMP-10 expression in the atherosclerotic lesion is mainly localized in macrophage-like cells, and it is necessary for collagenolytic activity. MMP-10/Apoe double knockout (DKO)mice show decreased atherosclerotic lesion size and plaque calcification. MMP-13 mRNA was abundantly expressed in macrophages. MMP-13/Apoe DKO mice showed the accumulation of collagen in the plaque and resulted in plaque stabilization. These previous reports and the present study suggest that hypoxic signaling in macrophages causes instability and rupture of atherosclerotic plaques by expressing MMP-3, -10, and -13.
Intriguingly, the hypoxic induction of MMP-3 and -10 mRNAs was dependent on ROS generated by XO, but not HIF-1α, which is known as the central regulator of the hypoxic response. Previous studies have reported that hypoxia activates p38/SAPK2, JNK/SAPK, and Akt/PKB signaling [33][34][35]. In addition, it has been reported that ROS also activates p38, JNK, and Akt kinases [36,37]. Actually, we confirmed that inhibitors of p38 and JNK suppress the hypoxic induction of MMP-3, -10, and -13 mRNAs (data not shown). Therefore, we consider that ROS production under the hypoxic condition induces MMP-3, -10, and -13 mRNAs via the activation of these kinases.
At the same time, it is known that ROS interacts with the Cys thiol within the catalytic domain of MMPs, which disrupts the interaction with the catalytic Zn 2+ ion and leads to autocatalytic activation. Previous studies report that ROS activates various proMMPs, including MMP-1, MMP-2, MMP-7, MMP-8, and MMP-9 [38][39][40][41][42][43]. Therefore, there is a possibility that our in-gel zymographic analysis shows the activation of gelatinase and collagenase by not only MMP-3 and -10 but also ROS as the direct effect. Further investigations are needed to elucidate the detailed mechanisms of the hypoxic induction and activation of these MMPs.
Moreover, febuxostat significantly inhibited the production of MMP-9 and the activation of MMP-13 even under normoxic conditions. Previous studies reported that ROS activates MMP secretion [44,45]. In general, the mitochondria respiratory chain (MRC) consumes most oxygen in cells, and 2~4% of total oxygen consumption goes toward the generation of ROS instead of ATP synthesis [46]. Therefore, mitochondria constitutively produce ROS. A recent study has reported that inhibition of xanthine oxide reductase (XDH/XO) by febuxostat causes the increment of intracellular ATP by purine salvage pathway and reduces the production of ROS from MRC under normoxic conditions [47]. Thus, febuxostat may affect the production and activation of MMPs by reducing ROS derived from MRC under normoxic conditions. However, future studies are required to uncover the precise mechanism of inhibitory effects of febuxostat on MMP expression without transcriptional regulations.
Using the atherosclerotic plaque rupture mouse model, this study demonstrated that febuxostat administration significantly reduced the number of plaque ruptures. Febuxostat also attenuated Biomedicines 2020, 8, 470 9 of 13 collagenolytic (MMP-13) and gelatinolytic activity (MMP-9) in the culture media of macrophages. These results may indicate that the hypoxic environment accelerates atherosclerotic plaque vulnerability and rupture by activating MMPs via XO. Previous clinical trials reported that the administration of allopurinol, a purine inhibitor of XO, was associated with a lower rate of stroke and cardiac events [48,49]. Febuxostat, a nonpurine inhibitor of XO, as well as allopurinol, is used for the management of hyperuricemia in patients with gout [50]. In addition to these beneficial effects, this study suggests that the administration of an XO inhibitor could be useful for suppressing atherosclerotic plaque instability and rupture in patients.

Hypoxia Treatment
Cells were placed in a personal CO 2 multi-gas incubator (Model APM-30D, ASTEC, Fukuoka, Japan) calibrated to maintain a hypoxic atmosphere (1% O 2 , 5% CO 2 , and 94% N 2 ) under a continuous flow of nitrogen. A standard tissue culture incubator was used for a normoxic environment (21% O 2 , 5% CO 2 ).

Mouse Models of Atherosclerotic Plaque Rupture
All animal experiments were approved by the Institutional Animal Care Committee of Nagasaki University (approval number: 1701251356). B6.129P2-Apoe tm1Unc /J (ApoE−/−) mice were kindly provided by Dr. J. Aruga (Nagasaki University, Nagasaki, Japan). To generate mouse models of atherosclerotic plaque rupture, 9-week-old male ApoE −/− mice were fed a standard diet. The left common carotid artery of these mice was dissected and ligated proximal to the bifurcation under anesthesia. After the ligation, the mice were randomly divided into two groups, including the experimental group (febuxostat) and the normal control group (Control), and fed a high-cholesterol diet (Oriental Yeast, Tokyo, Japan). The experimental group was given drinking water containing 0.015 mg/mL of febuxostat after the ligation. At 4 weeks after the ligation of the left common carotid artery, a polyethylene cuff (length 2 mm) was placed around the left common carotid artery tightly adjacent to the site of ligation. At 4 days after cuff placement, mice were anesthetized and injected intraperitoneally with heparin. Lesions were then obtained after perfusion from the left ventricle using a normal saline solution. The vessels after cuff removal were isolated and fixed in 10% formalin overnight.

Histological and Immunohistochemical Staining
For staining, vessels were fixed in 4% formaldehyde and embedded in paraffin. Specimens were serially sectioned at 5 µm intervals and mounted on three consecutive slides with 60-µm intervals on glass slides. Two slides were individually stained with hematoxylin and eosin (H&E) and Masson's trichrome stains. Five serial, identically stained slides were selected from the middle part of the cuff placement site of each vessel to count the number of internal elastic lamina breaks. The number of ruptures was taken as the average of 5 sections.

Quantitative Real-Time RT-PCR Analysis
Total RNA was purified from culture cells using the RNeasy Mini Kit (Qiagen, Thermo Fisher Scientific, Inc., Waltham, MA, USA). Complementary DNA (cDNA) was synthesized using the ReverTra Ace ® qPCR RT Master Mix (TOYOBO, Osaka, Japan). Quantitative real-time PCR (qPCR) was performed using the THUNDERBIRD ® SYBR ® qPCR MIX kit (TOYOBO, Osaka, Japan). A melting curve analysis was performed to confirm the detection of a single PCR product. We performed the standard curve method for the quantitative analysis of qPCR. Then, 5, 25, 125, and 625-fold dilutions of a sample were prepared for each gene. The standard curve was drawn from the Ct values of the samples. The expression levels of each sample were calculated using the standard curve. The relative abundance of all genes of interest was normalized to the housekeeping control 18S ribosomal RNA (18S rRNA). The primer sequences of the analyzed genes are shown as follows: MMP-3 (NCBI accession number NM_010809.

Zymography and Quantification
TEPMs were exposed to hypoxia for 24 h in the presence or absence of febuxostat, and supernatants were collected. Cell supernatants were clarified by brief centrifugation and concentrated using the Amicon Ultra-2, 3-kDa nominal molecular mass cutoff filters (Millipore, Billerica, MA, USA). The protein concentration of supernatants was determined by the BCA Protein Assay kit (Pierce, Thermo Fisher Scientific, Inc.). Equal amounts (5 µg/well) were loaded onto 10% zymogram plus (gelatin) protein gels (Novex; Thermo Fisher Scientific, Inc., Waltham, MA, USA) for the gelatin zymogram or 10% SDS-PAGE containing 0.6 mg/mL rat type I collagen (Gibco, Thermo Fisher Scientific, Inc., Waltham, MA, USA) for the collagen zymogram, and run for 90 min under a constant voltage of 125 V. Gels were incubated in a renaturing buffer (Novex; Thermo Fisher Scientific, Inc., Waltham, MA, USA), followed by incubation at 37 • C for 24 h in developing buffer (Novex; Thermo Fisher Scientific, Inc., Waltham, MA, USA). Gels were stained with the SimplyBlue SafeStain (Invitrogen, Thermo Fisher Scientific, Inc., Waltham, MA, USA). Zymograms were analyzed by densitometry analysis using the ChemiDoc Touch Imaging System (BIO RAD Laboratories Inc., Hercules, CA, USA) (Supplementary Figure S2).

Statistical Analysis
All data are shown as means with SD. Comparisons between two groups were performed using the unpaired t-test. p values of less than 0.05 were considered significant. Differences among more than two groups were analyzed using one-way or two-way ANOVA followed by Dunn's multiple comparisons test or Tukey's multiple comparisons test.