Minocycline reduces plaque size in diet induced atherosclerosis via p27Kip1

Atherosclerosis 219 (2011) 74–83

Contents lists available at ScienceDirect

Atherosclerosis journal homepage: www.elsevier.com/locate/atherosclerosis

Minocycline reduces plaque size in diet induced atherosclerosis via p27Kip1 Khurrum Shahzad a , Madhusudhan Thati a , Hongjie Wang a,b , Muhammed Kashif a , Juliane Wolter a , Satish Ranjan a , Tao He a,b , Qianxing Zhou b,c , Erwin Blessing c , Angelika Bierhaus a , Peter P. Nawroth a , Berend Isermann a,d,∗ a

Internal Medicine I and Clinical Chemistry, University of Heidelberg, INF 410, 69120 Heidelberg, Germany Department of Cardiology, Tongji Hospital, Tongji Medical College, Huazhong University of Science and Technology, 430030 Wuhan, China c Internal Medicine III (Cardiology, Angiology, Pneumology), University of Heidelberg, INF 410, 69120 Heidelberg, Germany d Otto-von-Guericke-University Magdeburg, Department of Clinical Chemistry and Pathobiochemistry, Leipziger Str. 44, 39120 Magdeburg, Germany b

a r t i c l e

i n f o

Article history: Received 27 February 2011 Received in revised form 23 May 2011 Accepted 26 May 2011 Available online 13 June 2011 Keywords: Minocycline PARP-1 Vascular smooth muscle cells p27 Atherosclerosis

a b s t r a c t Objective: Minocycline, a tetracycline derivate, mediates vasculoprotective effects independent of its antimicrobial properties. Thus, minocycline protects against diabetic nephropathy and reduces neointima formation following vascular injury through inhibition of apoptosis or migration, respectively. Whether minocycline has an effect on primary atherogenesis remains unknown. Methods: Using morphological and immunohistochemical analyses we determined de novo atherogenesis in ApoE−/− mice receiving a high fat diet (HFD) with or without minocycline treatment. The effect of minocycline on proliferation, expression of p27Kip1 or PARP-1 (Poly [ADP-ribose] polymerase 1), or on PAR (poly ADP-ribosylation) modification in vascular smooth muscle cells (VSMC) was analyzed in ex vivo and in vitro (primary human and mouse VSMC). Results and conclusion: Minocycline reduced plaque size and stenosis in ApoE−/− HFD mice. This was associated with a lower number and less proliferation of VSMC, reduced PAR (poly ADP-ribosylation) modification and increased p27Kip1 expression within the plaques. In agreement with the ex vivo data minocycline reduced proliferation, PARP-1 expression, PAR modification while inducing p27 expression in human and mouse VSMC in vitro. These effects were observed at a low minocycline concentration (10 ␮M), which had no effect on VSMC migration or apoptosis. Minocycline inhibited PARP-1 and induced p27Kip1 expression in VSMC as efficiently as the specific PARP-1 inhibitor PJ 34. Knock down of p27Kip1 abolished the antiproliferative effect of minocycline. These data establish a novel antiatherosclerotic mechanism of minocycline during de novo atherogenesis, which depends on p27Kip1 mediated inhibition of VSMC proliferation. © 2011 Elsevier Ireland Ltd. All rights reserved.

1. Introduction Minocycline, a tetracycline derivate, mediates cytoprotective effects unrelated to its antimicrobial properties. In particular, neuroprotective properties of minocycline have been demonstrated in a number of animal models of neurodegeneration or brain injury [1,2]. These studies have led to clinical evaluations of minocycline in patients with neuronal disease, which showed that minocycline

Abbreviations: GOT, glutamate-oxalacetate transaminase (AST); GPT, glutamate pyruvate transaminase (ALT); HASMC, human aortic smooth muscle cell; HFD, high fat diet; MMP, matrix metalloproteinase; PAR, poly ADP-ribosylation; PARP-1, poly [ADP-ribose] polymerase 1; VSMC, vascular smooth muscle cell. ∗ Corresponding author at: Otto-von-Guericke-University Magdeburg, Department of Clinical Chemistry and Pathobiochemistry, Leipziger Str. 44, D-39120 Magdeburg, Germany. Tel.: +49 (0)391 67 13900; fax: +49 (0)391 67 13902. E-mail addresses: [email protected], [email protected] (B. Isermann). 0021-9150/$ – see front matter © 2011 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.atherosclerosis.2011.05.041

has promising neuroprotective properties in humans [3]. Considering the potential long-term use of minocycline in patients with neurodegenerative disorders minocycline’s effect on primary atherogenesis is relevant, but remains to be explored. Recent studies demonstrated that minocycline conveys cytoprotective effects in vascular cells. For example, minocycline protects against diabetic microvascular complication [4,5]. In regard to macrovascular disease tetracycline derivates like minocycline or doxycycline reduce neointima formation following an acute vascular injury of the rat carotid artery [6]. In these studies the reduced number of vascular smooth muscle cells (VSMC) has been attributed to an inhibition of MMP activity and cytokine induced VSMC migration [6–8]. However, in vitro data demonstrated that minocycline directly reduces VSMC number, suggesting that minocycline may modulate cellular proliferation, although the underlying mechanism remains unknown [9–11]. Proliferation is increased in primary human atherosclerotic lesions, in particular in early lesions [12]. In undiseased arteries

K. Shahzad et al. / Atherosclerosis 219 (2011) 74–83

VSMCs proliferate at a low frequency and are arrested in the G0 /G1 phase of the cell cycle. Endothelial dysfunction or arterial injury induce proliferation of VSMCs [13]. Cell proliferation is negatively regulated by cyclin-dependent kinase (CDK) inhibitors (CKI), which associate with and inhibit the activity of CDK/cyclin holoenzymes, leading to a G1 arrest [14]. The CKI p27Kip1 , which is expressed in healthy and atherosclerotic arteries, has evolved as an important regulator of VSMC proliferation [15]. Mice lacking p27Kip1 display enhanced arterial cell proliferation and larger atherosclerotic plaques in a murine model of diet induced atherogenesis [12], establishing not only a role of p27Kip1 , but of proliferation in general during primary atherogenesis. Cell cycle progression is also regulated by PARP-1 (Poly [ADP-ribose] polymerase 1) [16], an enzyme potently inhibited by minocycline [17] and hence providing a potential mechanistic link between minocycline and regulation of proliferation. Considering that minocycline has vasculoprotective effects we addressed the question whether minocycline modulates PARP-1 activity and proliferation during atherogenesis. To this end we induced atherogenesis using a high fat diet (HFD) in atherosclerosis prone ApoE-deficient (ApoE−/−) mice, treating a subgroup of mice for the entire study period with minocycline. We show that minocycline reduces plaque size and inhibits proliferation of smooth muscle cells through a PARP-1 and p27Kip1 dependent mechanism. 2. Materials and methods See the materials and methods section in Supplementary data for details. 2.1. Mice Animal experiments were carried out in accordance with the local Animal Care and Use Committee (Regierungspraesidium Karlsruhe, Germany). ApoE−/− mice (C57Bl/6 background) were obtained from the Jackson Laboratory (Bar Harbor, ME, USA). Mice (age 6–8 weeks) received either a Western type diet or a normal chow diet (controls). A subset of the Western type diet-fed mice received minocycline intraperitoneally (5 mg/kg body weight) once per day from the first day of the diet until 1 day before analysis. Control mice were injected with an equal volume of PBS intraperitoneally once per day. Mice were sacrificed and analyzed after 20 weeks of treatment (see the materials and methods section in Supplementary data). 2.2. Analysis of blood lipids, cytokines, and transaminases Plasma levels of lipids were measured as previously described using the Advia 2400 Chemistry System Siemens (Eschborn, Germany) [18]. Plasma levels of TNF␣ and IL-6 were determined by ELISA following manufacturer instructions (eBioscience, Germany). Plasma levels of GOT (glutamate-oxalacetate transaminase, AST, U/l) and GPT (glutamate pyruvate transaminase, ALT, U/l) were measured using a two step enzymatic reaction on the Advia 2400 Chemistry System. 2.3. Histology and immunohistochemistry Thoracic aortas were stained with Oil-red-O. MOVATs stain was performed on frozen sections of brachiocephalic arteries. Samples were analyzed by a blinded investigator using the Image Pro Plus software [18] (see the materials and methods section in Supplementary data).

75

2.4. Cell culture All experiments performed with human umbilical vein endothelial cells (HUVECs) or human aortic smooth muscle cells (HASMCs) were done with passages 4–6. Cells were maintained at 37 ◦ C in a humidified 5% CO2 incubator using endothelial or smooth muscle cell growth medium in the presence of growth factors and supplements. Cells were subcultured at confluence by trypsinization with 0.05% trypsin and 0.02% EDTA. Medium was changed every other day. For additional information regarding THP-1 cells and primary mouse aorta VSMC see the materials and methods section in Supplementary data. 2.5. Apoptosis and proliferation assays Apoptosis was determined by TUNEL assay [4]. Proliferation was determined in vitro by BrdU labelling (see the materials and methods section in Supplementary data). 2.6. Immunoblotting Immunoblotting was essentially done as previously described [4] (see the materials and methods section in Supplementary data). 2.7. Statistical analyses All in vitro experiments were performed at least in triplicates. The data are summarized as the mean ± s.e.m. (standard error of the mean). Statistical analyses were performed using Student’s ttest, ANOVA, or 2 test (as indicated in the figure legend or the text). StatistiXL software (http://www.statistixl.com) was used for statistical analyses. Statistical significance was accepted at the P < 0.05 level. 3. Results 3.1. Minocycline reduces atherosclerotic plaque size and vascular stenosis En face analysis of Oil-red-O stained thoracic aortas showed that minocycline significantly reduced the plaque burden of the thoracic aorta in ApoE−/− HFD mice, in particular in the portion of the aortic arch (7.1% vs. 11.1%, P = 0.039, Fig. 1A and B). Likewise, we observed a significant reduction of plaque burden within the aortic root of minocycline treated ApoE−/− HFD mice (plaque size 43.3 × 103 ␮m2 vs. 80.1 × 103 ␮m2 , P < 0.001, Fig. 1C). Histological analysis of cross-sections of the truncus brachiocephalicus, which exhibits a highly consistent rate of lesion progression and develops characteristics of advanced lesions [19], likewise showed smaller plaques in minocycline treated ApoE−/− HFD mice in comparison to ApoE−/− HFD (50.1 × 103 ␮m2 vs. 159.2 × 103 ␮m2 P = 0.02, Fig. 1D and E). The total vessel cross section area was increased in ApoE−/− HFD with and without minocycline treatment, reflecting positive vascular remodelling. Vascular remodelling was somewhat less pronounced in minocycline treated ApoE−/− HFD mice, but this difference was not significant (181.1 × 103 ␮m2 vs. 253.1 × 103 ␮m2 , P = 0.192, Fig. 1F). Vascular stenosis was less severe in ApoE−/− HFD diet receiving minocycline (34% vs. 59% in ApoE−/− HFD control mice, P = 0.038, Fig. 1G). Treatment with minocycline was well tolerated and did not impair the apparent health or survival of the mice. Analysis of serum samples showed no significant differences of total cholesterol (439.0 mg/dl vs. 538.1 mg/dl, P = 0.16, Supplementary Fig. S1B), triglyceride (43.0 mg/dl vs. 44.1 mg/dl, P = 0.16, Supplementary Fig. S1A) or HDL levels (17.5 mg/dl vs. 14 mg/dl,

76

K. Shahzad et al. / Atherosclerosis 219 (2011) 74–83

Fig. 1. Minocycline reduces plaque size and vascular stenosis. (A) and (B) En face analysis of Oil-red-O stained thoracic aortas from ApoE−/− mice receiving a control diet (C), a high fat diet (HFD), or a high fat diet and minocycline (HFD + mino). Treatment with minocycline significantly reduces the total plaque area. (C) Analyses of Oil-red-O stained aortic roots, showing a marked reduction of plaque size in aortic roots obtained from ApoE−/− mice receiving a HFD and being treated with minocycline (HFD + mino) as compared to ApoE−/−HFD mice. Exemplary images (left) and bar graph summarizing results (right). (D)–(G) Minocycline (ApoE−/− HFD + mino) reduces the plaque size in the truncus brachiocephalicus of ApoE−/− HFD mice (MOVAT stain, D, E). Positive vascular remodelling, reflected by a larger total vessel cross section area (F), is increased in ApoE−/− HFD mice with and without minocycline treatment. The degree of vascular stenosis is significantly enhanced in ApoE−/− HFD mice as compared to ApoE−/− HFD + mino mice (G). C: normal chow control diet (N = 10); HFD: high fat diet (N = 10); mino: minocycline HFD + mino, (N = 12); mean value ± s.e.m.; scale bar = 150 ␮m; *P < 0.05, **P < 0.01, ANOVA.

K. Shahzad et al. / Atherosclerosis 219 (2011) 74–83

77

Fig. 2. Minocycline reduces the frequency of VSMC within atherosclerotic plaques. (A)–(D) Plaque stability, as determined by the necrotic core area (A), fibrous cap thickness (B), extracellular matrix content (C), and frequency of ruptured shoulders (D) does not significantly differ between ApoE−/− HFD and ApoE−/− HFD + mino mice. (E) and (F) The frequency of VSMC is significantly reduced in ApoE−/− HFD + mino mice as compared to ApoE−/− HFD mice (E), while no significant difference is observed in the frequency of MOMA-2 positive macrophages (F). Smooth muscle cell immunohistochemistry, brown: calponin-positive cells detected by HRP-DAB reaction, blue: hematoxylin counterstain; MOMA-2 immunohistochemistry, green: MOMA-2 positive cells detected by FITC, blue: DAPI counterstain. C: normal chow control diet (N = 10); HFD: high fat diet (N = 10); mino: minocycline (HFD + mino, N = 12); mean value ± s.e.m.; scale bar = 250 ␮m; *P < 0.05, **P < 0.01, by t-test (A)–(C) or Chi-square test (D), ANOVA (E) and (F).

78

K. Shahzad et al. / Atherosclerosis 219 (2011) 74–83

Fig. 3. Minocycline reduces proliferation of VSMC. (A) and (B) Treatment with minocycline reduces the frequency of proliferating (red) vascular smooth muscle cells (green) within arteriosclerotic plaques of ApoE−/− HFD mice; co-immunofluorescence staining of brachiocephalic arteries for VSMC (anti-calponin antibody, green) and the proliferation marker Ki-67 (red). Insets show area marked at left at higher magnification (A) and bar graph (B) summarizing results. (C) and (D) Minocycline reduces proliferation of HASMC (C) and of murine primary aortic VSMC (D) in vitro both in the presence or absence of LDL (BrdU incorporation). Bar graph summarizing results of at least three independent experiments performed in duplicates. Scale bar = 200 ␮m (A); C: normal chow control diet (N = 10); HFD: high fat diet (N = 10); mino: minocycline (HFD + mino, N = 12); mean value ± s.e.m.; *P < 0.05, **P < 0.01, by ANOVA ((C) and (D)) or t-test (B).

P = 0.27, Supplementary Fig. S1C). Consistent with the known anti-inflammatory effects of minocycline we observed a reduction of cytokines (TNF-␣ and IL-6) in the plasma of minocycline treated ApoE−/− HFD mice (4c21 pg/ml vs. 1089 pg/ml for TNF-␣ and 328 pg/ml vs. 806 pg/ml for IL-6, P<0.001 for both, Supplementary Fig. S1D and E). Since short term treatment with higher dosages of minocycline (35 mg/kg twice daily) resulted in liver toxicity in rats [6] we determined serum values of GOT and GPT, which did not differ between ApoE−/− HFD mice with or without minocycline (105 U/l vs. 90 U/l, P = 0.62, and 34 U/l vs. 36 U/l, P = 0.78 for GOT and GPT, respectively, Supplementary Fig.S1F and G). Likewise, histological analysis of liver tissues did not reveal any obvious pathology (H and E stain, data not shown). Thus, long term

exposure to minocycline is well tolerated and reduces plaque size during de novo atherogenesis. 3.2. Minocycline reduces the frequency of intraplaque smooth muscle cells We next compared indices of plaque stability in plaques from ApoE−/− HFD mice with or without minocycline treatment. No difference between minocycline treated and control mice was observed in regard to cross-sectional area of necrotic cores (17.8% vs. 13.5%, P = 0.50), fibrous cup thickness (9.7 ␮m vs. 12.5 ␮m, P = 0.19), extracellular matrix accumulation within the plaques (34.5 × 103 ␮m2 vs. 38.8 × 103 ␮m2 , P = 0.81),

K. Shahzad et al. / Atherosclerosis 219 (2011) 74–83

79

Fig. 4. Minocycline induces p27Kip1 expression in HASMC and in atherosclerotic plaques. (A) and (B) Minocycline induces the expression of p27Kip1 in HASMC in vitro, both in the presence and in the absence of LDL. Representative images of semiquantitative RT-PCR (A) and immunoblots (B) for p27Kip1 and corresponding bar graphs summarizing results of four independent experiments. (C) and (D) Minocycline increases the frequency of p27Kip1 positive VSMC in atherosclerotic plaques of ApoE−/− HFD mice. Exemplary images with insets showing area marked at left at higher magnification (C), immunofluorescence staining for p27Kip1 (red) and VMSC (calponin, green), and bar graph (D) summarizing results. (E) and (F) Transient transfection of murine primary aortic VSMC with p27Kip1 shRNA efficiently reduces p27Kip1 mRNA expression (E). In p27Kip1 knock down VSMC minocycline fails to reduce proliferation (F). Representative image of semiquantitative RT-PCR (E) and bar graph summarizing results of three independent experiments (F). Scale bar = 200 ␮m (C); ApoE−/− HFD: N = 10, ApoE−/− HFD + mino: N = 12; mean value ± s.e.m.; scale bar = 150 ␮m, *P < 0.05, **P < 0.01, by ANOVA ((A), (B), and (F) or t-test ((D) and (E)).

80

K. Shahzad et al. / Atherosclerosis 219 (2011) 74–83

Fig. 5. Minocycline inhibits PARP expression and PAR formation in vitro and in vivo. (A) and (B) Treatment of HASMC with the specific PARP-1 inhibitor PJ 34 is sufficient to induce p27Kip1 expression. Representative images of semiquantitative RT-PCR (left) or immunoblot (right) and bar graph summarizing results of four independent experiments. (C) and (D) LDL induces PARP-1 expression (C) and PAR formation (D), reflecting PARP-1 activity, in HASMC. Minocycline significantly decreases LDL induced PARP-1 expression and reduces PAR formation to the same extent as observed with the PARP-1 inhibitor PJ34. Representative immunoblot using antibodies against PARP-1 (C) or PAR (D) and bar graph summarizing results from four independent repeat experiments. (E) and (F) Reduced immunohistochemical staining of PAR in minocycline treated

K. Shahzad et al. / Atherosclerosis 219 (2011) 74–83

the frequency of ruptured shoulders (51.7% vs. 40.2%, P = 0.16 Fig. 2A–D), or intraplaque hemorrhage (frequency of intraplaque erythrocytes, determined by TER-119 immunohistochemistry: 5.0% vs. 6.0%, P = 0.34, data not shown). Plaque size is in part determined by the cellular components. Hence, we determined the frequency of vascular smooth muscle cells (VSMC) and macrophages within the plaques. Immunohistochemical analysis revealed a significant lower staining intensity for VSMC calponin within plaques of minocycline treated mice compared to control ApoE−/− HFD mice (score 0.66 vs. 1.69; P = 0.005, Fig. 2E). The same effect was observed when using an antibody against ␣-SMC actin (data not shown). Conversely, based on immunohistochemical analysis the expression of MOMA-2, a macrophage marker, did not differ between groups (6.6% vs. 7.8%; P = 0.17, Fig. 2F). These results indicate that minocycline primarily reduces the VSMC frequency. 3.3. Minocycline reduces proliferation of VSMC To evaluate whether minocycline may regulate proliferation of VSMC in atherosclerotic lesions we determined the frequency of Ki-67 and calponin double positive cells. Based on immunohistochemical co-staining for these markers minocycline reduces VSMC proliferation by 37% in atherosclerotic plaques compared to ApoE−/− HFD mice without minocycline (1.9% vs. 3.0% of total VSMC, P = 0.004, Fig. 3A and B and Supplementary Fig. S2). We next determined the effect of minocycline on proliferation of various cell types in vitro. Treatment of HASMC with minocycline reduced proliferation both in the absence and presence of LDL (166% in LDL only vs. 82% LDL plus minocycline, P = 0.001, Fig. 3C). We likewise observed an antiproliferative effect of minocycline in murine primary aortic VSMC, in particular in the presence of LDL (153% in LDL only vs. 111% in LDL plus minocycline, P = 0.02, Fig. 3D). Using FACS analyses of propidium iodide stained HASMC we determined that minocycline promoted an arrest in the G1 phase of the cell cycle (data not shown). Conversely, minocycline had no effect on proliferation of macrophage like PMA-treated THP-1 cells (Supplementary Fig. S3A) or of HUVECs (Supplementary Fig. S3B). Furthermore, while minocycline markedly reduced VSMC proliferation at 10 ␮M, it had no effect on apoptosis or migration of VSMC at this concentration (Supplementary Fig. S3C and D). Consistent with previous reports [20] minocycline inhibited VSMC migration at higher concentrations (≥20 ␮M, data not shown). Furthermore, at the chosen dosage minocycline failed to reduce expression of MMP-9 or MMP-2 within the thoracic aorta (Supplementary Fig. S3E and F). Together these data establish that minocycline inhibits proliferation in VSMC both in vitro and in vivo, providing a rationale for the lower frequency of VSMC within atherosclerotic plaques of ApoE−/− HFD mice. 3.4. Minocycline induces p27Kip1 expression in VSMC To explore the mechanism through which minocycline reduces VSMC proliferation we determined the expression of the cyclindependent kinase inhibitors (CKIs) p27Kip1 and p21Cip1 , which have been previously reported to regulate VSMC proliferation [15]. While minocycline had no effect on the expression of p21Cip1 (data not shown), expression of p27Kip1 was markedly induced by minocycline in HASMC. Minocycline increased p27Kip1 expression in HASMC both at the mRNA and protein level in the absence of

81

LDL (253% vs. 100%, P = 0.01, and 270% vs. 100%, P = 0.02, for mRNA and protein, respectively; Fig. 4A and B). The induction of p27Kip1 expression by minocycline was somewhat blunted in the presence of LDL, but remained significantly higher as compared to cells treated with LDL only (204% vs. 81%, P = 0.041, and 168% vs. 67%, P = 0.039, for mRNA and protein, respectively; Fig. 4A and B). Ex vivo analysis of atherosclerotic plaques of the truncus brachiocephalicus revealed a significantly increased frequency of p27Kip1 positive VSMC cells in minocycline treated ApoE−/− HFD mice as compared to control ApoE−/− HFD mice (8.8% vs. 4.8%, P = 0.01, Fig. 4C and D and Supplementary Fig. S4). The in vitro and in vivo data suggests that minocycline inhibits VSMC proliferation via p27Kip1 . To this end we reduced expression of p27Kip1 in primary mouse aortic VSMC (Fig. 4E). Knock down of p27Kip1 abolished the antiproliferative effect of minocycline in murine VSMC (162% in LDL plus minocycline vs.168% in LDL, Fig. 4F) and HASMC (data not shown), establishing that the antiproliferative effect of minocycline in VSMC depends on p27Kip1 . 3.5. Minocycline inhibits PARP-1 expression and PAR formation in atherosclerotic plaques Minocycline is a potent inhibitor of PARP-1 and PAR formation [17]. Hence we explored whether the minocycline dependent induction of p27Kip1 expression may dependent on PARP-1 inhibition. Treatment of LDL-challenged HASMC with the specific PARP-1 inhibitor PJ 34 had the same effect as minocycline in regard to p27Kip1 expression, inducing p27Kip1 levels (increase to 229% at mRNA level, P<0.001 and to 177% at protein level, P < 0.001, Fig. 5A and B). This establishes that inhibition of PARP-1 is sufficient to induce p27Kip1 in HASMC. We next determined whether minocycline at the concentration used to inhibit proliferation (10 ␮M) inhibits PARP-1 in HASMC. LDL treatment almost doubled PARP-1 expression in HASMC in vitro (189% vs. 100%, P = 0.0008, Fig. 5C). Minocycline markedly reduced PARP-1 expression, in particular in LDL treated HASMC (122% vs. 189%, P = 0.003, Fig. 5C). The altered expression of PARP-1 was associated with concomitant changes of poly ADP-ribosylaton (PAR), a marker of PARP-1 activity. Thus, exposure to LDL increased PAR formation (170.4% vs. 100%, P = 0.06, Fig. 5D) and treatment with minocycline reduced PAR formation both in control and in LDL treated HASMC (44% vs. 100%, P = 0.002, in control cells, and 46% vs. 170%, P = 0.002, in LDL-treated cells; Fig. 5D). The effect of minocycline was comparable to that of PJ34. Hence, minocycline is a potent inhibitor of PARP-1 expression and activity in HASMC in vitro. To determine whether minocycline reduces PAR-formation in atherosclerotic plaques in vivo immunohistochemical analyses were next performed. Consistent with the in vitro results minocycline reduced PAR-formation in plaques of the truncus brachiocephalicus of ApoE−/− HFD mice treated with minocycline as compared to ApoE−/− HFD control mice (3.3% vs. 5.8%, P = 0.03, Fig. 5E and F and Supplementary Fig. S5). Collectively, this data establish that minocycline inhibits PARP-1 and induces p27Kip1 expression in atherosclerotic plaques, and that PARP-1 inhibition is sufficient to induce p27Kip1 expression in VSMC. 4. Discussion Minocycline, a tetracycline derived antibiotic, mediates direct cellular functions independent of its antibiotic properties. Here we show that minocycline inhibits proliferation of VSMC by

ApoE−/− HFD (ApoE−/− HFD + mino) mice as compared to control mice (ApoE−/− HFD). Exemplary images with insets showing area marked at left at higher magnification (E), immunofluorescence staining for PAR (red) and VMSC (calponin, green), and bar graph (F) summarizing results. Scale bar = 200 ␮m (D); ApoE−/− HFD: N = 10, ApoE−/− HFD + mino: N = 12; mean value ± s.e.m.; scale bar = 200 ␮m, *P < 0.05, **P < 0.01, by ANOVA ((C) and (D)) or t-test ((A), (B), and (F)).

82

K. Shahzad et al. / Atherosclerosis 219 (2011) 74–83

inducing p27Kip1 expression. In a murine model of diet induced de novo atherogenesis minocycline induced p27Kip1 expression in VSMC, reducing proliferation of intraplaque VSMC, atherosclerotic plaque size, and vascular stenosis. This identifies a novel mechanism through which minocycline modulates vascular diseases. Beneficial effects of tetracycline derivates, such as minocycline and doxycycline, have been demonstrated in animal models of acute vascular injury. Thus, minocycline or doxycycline reduce neointima formation, MMP activity and VSMC migration after balloon injury of the carotic artery in rats and inhibit VEGF induced MMP-9 mRNA transcription and protein activation in human aortic VSMC in vitro [6,8,21]. The anti-inflammatory properties of minocycline, such as reduction of IL-1␤ and TNF␣ or inhibition of NF-kB activation [22,23], likely contribute to reduced neointima formation. In addition, minocycline and doxycycline have been shown to reduce the cell number of VSMC and of carcinoma cells, although the underlying mechanism remained undefined [6,23–25]. The current study shows that minocycline directly inhibits VSMC proliferation by regulating p27Kip1 expression. Minocycline induced p27Kip1 expression and reduced proliferation at a relatively low dose (10 ␮M). Of note, minocycline failed to inhibit VSMC migration and expression or activity of MMPs at such a low dose [20] (and current study). Hence, we propose that minocycline modulates atherosclerotic disease not only by inhibiting VSMC migration, as previously proposed, but in addition through directly regulating VSMC proliferation via p27Kip1 . Enhanced proliferation contributes to atherosclerotic plaque development [26]. In human atherosclerotic tissue expression of p27Kip1 is inversely associated with proliferation of VSMCs [27], implying a role of p27Kip1 in human atherosclerosis. A causal role of p27Kip1 during primary atherogenesis is established by the development of larger plaques in p27Kip1 deficient mice [12]. We observed an approximately 2-fold increase of p27Kip1 levels in primary human or mouse VSMC treated with minocycline and of p27Kip1 staining intensity in atherosclerotic plaques of ApoE−/− mice treated with minocycline. Of note, haploinsufficiency of p27Kip1 is sufficient to promote diet induced atherosclerosis in mice, consistent with a dose dependent effect of p27Kip1 in atherosclerosis [12]. Given this dose-dependent effect the observed doubling of p27Kip1 expression is likely mechanistically relevant. A mechanistic relevance of the increased p27Kip1 expression for the regulation of VSMC proliferation is finally confirmed by the loss of minocycline’s antiproliferative effect in p27Kip1 knock down cells. How does minocycline regulate p27Kip1 expression? Expression of p27Kip1 can be regulated by PARP-1 [28] and minocycline inhibits PARP-1 at very low concentrations [17]. Indeed, the potent PARP-1 inhibitory effects of minocycline and the regulation of p27Kip1 through PARP-1 provide a rationale, why we observed a specific inhibition of VSMC proliferation at a low dose of minocycline (10 ␮M), while the same concentration of minocycline failed to inhibit VSMC migration. Interestingly, PARP-1 functions as a corepressor of Foxo1 [28], and Foxo proteins inhibit growth factor dependent VSMC proliferation in vitro. Constitutively active Foxo3 (FKHRL1) increases p27Kip1 expression in VSMC in vivo, resulting in reduced VSMC proliferation and neointimal hyperplasia following balloon injury in rats [29]. Future studies are required to determine whether minocycline interacts with Foxo proteins and whether this contributes to the smaller plaque size observed in primary atherogenesis. The current study shows that minocycline reduces plaque size during de novo atherogenesis. Given our previous data showing protection against microvascular disease in an in vivo model of diabetic nephropathy, minocycline, a well tolerated tetracycline derivate, conveys vasculoprotective effects both in regard to microand macrovascular complications. However, in regard to primary atherogenesis, the smaller plaque size secondary to a reduced pro-

liferation of VSMC may come at a price, as we observed a tendency to impaired plaque stability. Plaque stability is in part determined by the frequency of VSMCs and ECM components [30]. Hence, the benefit of a smaller plaque size secondary to a reduced VSMC frequency, as observed in minocycline treated ApoE−/− mice may be offset by impaired plaque stability. This “side-effect” is likely not specific for minocycline, but for any therapeutic intervention reducing VSMC proliferation during primary atherogenesis. Interestingly, reduction of PARP-1 activity using a PARP-1 inhibitor or in heterozygous PARP-1 (PARP-1 +/−) mice results not only in smaller, but also in more stable plaques [31]. This indicates that minocycline and PARP-1 inhibition have partially distinct effects during atherosclerotic plaque development. Delineating the differential mechanism may allow to increase the safety of minocycline or other drugs inhibiting VSMC proliferation during atherosclerosis. In this respect direct PARP-1 inhibition appear to be superior to the use of minocycline or minocycline may not be suitable for a monotherapeutic approach, but rather be a valuable adjunct to existing vasculoprotective medical interventions, such as ACE-inhibitors and HMG-CoA reductase inhibitors. The feasibility and therapeutic benefit of such a combined treatment approach has been demonstrated in experimental autoimmune encephalomyelitis (EAE), a murine model of multiple sclerosis. Combined treatment of EAE mice with the HMG-CoA reductase inhibitor atorvastatin and minocycline was superior to using each medication alone [32]. Studies have been initiated to evaluate whether such a combined therapeutic approach provides likewise a superior therapeutic benefit in an atherosclerosis model in mice. Disclosures The authors have nothing to disclose. Acknowledgments This work was supported by grants of the Deutsche Forschungsgemeinschaft (IS 67/5-1) to BI, a grant from the Dietmar Hopp Stiftung to BI, AB, and PPN, and a grant from the Stiftung Pathobiochemie to BI. KS has a scholarship from the Pakistan Higher Education Commission, TH and QZ have scholarships from the China Scholarship Council, and TM has a postdoctoral fellowship from the Medical Faculty at the University of Heidelberg. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.atherosclerosis.2011.05.041. References [1] Brundula V, Rewcastle NB, Metz LM, Bernard CC, Yong VW. Targeting leukocyte MMPs and transmigration: minocycline as a potential therapy for multiple sclerosis. Brain 2002;125:1297–308. [2] Zhu S, Stavrovskaya IG, Drozda M, et al. Minocycline inhibits cytochrome c release and delays progression of amyotrophic lateral sclerosis in mice. Nature 2002;417:74–8. [3] Lampl Y, Boaz M, Gilad R, et al. Minocycline treatment in acute stroke: an openlabel, evaluator-blinded study. Neurology 2007;69:1404–10. [4] Isermann B, Vinnikov IA, Madhusudhan T, et al. Activated protein C protects against diabetic nephropathy by inhibiting endothelial and podocyte apoptosis. Nat Med 2007;13:1349–58. [5] Krady JK, Basu A, Allen CM, et al. Minocycline reduces proinflammatory cytokine expression, microglial activation, and caspase-3 activation in a rodent model of diabetic retinopathy. Diabetes 2005;54:1559–65. [6] Pinney SP, Chen HJ, Liang D, et al. Minocycline inhibits smooth muscle cell proliferation, migration and neointima formation after arterial injury. J Cardiovasc Pharmacol 2003;42:469–76. [7] Ohshima S, Fujimoto S, Petrov A, et al. Effect of an antimicrobial agent on atherosclerotic plaques: assessment of metalloproteinase activity by molecular imaging. J Am Coll Cardiol 2010;55:1240–9.

K. Shahzad et al. / Atherosclerosis 219 (2011) 74–83 [8] Yao JS, Chen Y, Zhai W, et al. Minocycline exerts multiple inhibitory effects on vascular endothelial growth factor-induced smooth muscle cell migration: the role of ERK1/2, PI3K, and matrix metalloproteinases. Circ Res 2004;95: 364–71. [9] Bicknell KA, Surry EL, Brooks G. Targeting the cell cycle machinery for the treatment of cardiovascular disease. J Pharm Pharmacol 2003;55:571–91. [10] Lutgens E, de Muinck ED, Kitslaar PJ, et al. Biphasic pattern of cell turnover characterizes the progression from fatty streaks to ruptured human atherosclerotic plaques. Cardiovasc Res 1999;41:473–9. [11] McAllister 2nd JP, Miller JM. Minocycline inhibits glial proliferation in the H-Tx rat model of congenital hydrocephalus. Cerebrospinal Fluid Res 2010;7:7. [12] Diez-Juan A, Andres V. The growth suppressor p27(Kip1) protects against dietinduced atherosclerosis. FASEB J 2001;15:1989–95. [13] Libby P. Inflammation in atherosclerosis. Nature 2002;420:868–74. [14] Morgan DO. Principles of CDK regulation. Nature 1995;374:131–4. [15] Tanner FC, Boehm M, Akyurek LM, et al. Differential effects of the cyclindependent kinase inhibitors p27(Kip1), p21(Cip1), and p16(Ink4) on vascular smooth muscle cell proliferation. Circulation 2000;101:2022–5. [16] Carbone M, Rossi MN, Cavaldesi M, et al. Poly(ADP-ribosyl)ation is implicated in the G0-G1 transition of resting cells. Oncogene 2008;27:6083–92. [17] Alano CC, Kauppinen TM, Valls AV, Swanson RA. Minocycline inhibits poly(ADPribose) polymerase-1 at nanomolar concentrations. Proc Natl Acad Sci U S A 2006;103:9685–90. [18] Seehaus S, Shahzad K, Kashif M, et al. Frommhold, D., Nawroth, P.P. and Isermann, B., Hypercoagulability inhibits monocyte transendothelial migration through protease-activated receptor-1-, phospholipase-Cbeta-, phosphoinositide 3-kinase-, and nitric oxide-dependent signaling in monocytes and promotes plaque stability. Circulation 2009;120:774–84. [19] Rosenfeld ME, Polinsky P, Virmani R, et al. Advanced atherosclerotic lesions in the innominate artery of the ApoE knockout mouse. Arterioscler Thromb Vasc Biol 2000;20:2587–92. [20] Yao JS, Shen F, Young WL, Yang GY. Comparison of doxycycline and minocycline in the inhibition of VEGF-induced smooth muscle cell migration. Neurochem Int 2007;50:524–30.

83

[21] Bendeck MP, Conte M, Zhang M, et al. Doxycycline modulates smooth muscle cell growth, migration, and matrix remodeling after arterial injury. Am J Pathol 2002;160:1089–95. [22] Huang TY, Chu HC, Lin YL, et al. Minocycline attenuates 5-fluorouracil-induced small intestinal mucositis in mouse model. Biochem Biophys Res Commun 2009;389:634–9. [23] Tikka T, Fiebich BL, Goldsteins G, Keinanen R, Koistinaho J. Minocycline, a tetracycline derivative, is neuroprotective against excitotoxicity by inhibiting activation and proliferation of microglia. J Neurosci 2001;21:2580–8. [24] Fife RS, Rougraff BT, Proctor C, Sledge Jr GW. Inhibition of proliferation and induction of apoptosis by doxycycline in cultured human osteosarcoma cells. J Lab Clin Med 1997;130:530–4. [25] Fife RS, Sledge Jr GW. Effects of doxycycline on in vitro growth, migration, and gelatinase activity of breast carcinoma cells. J Lab Clin Med 1995;125:407–11. [26] Orr AW, Hastings NE, Blackman BR, Wamhoff BR. Complex regulation and function of the inflammatory smooth muscle cell phenotype in atherosclerosis. J Vasc Res 2010;47:168–80. [27] Tanner FC, Yang ZY, Duckers E, et al. Expression of cyclin-dependent kinase inhibitors in vascular disease. Circ Res 1998;82:396–403. [28] Sakamaki J, Daitoku H, Yoshimochi K, Miwa M, Fukamizu A. Regulation of FOXO1-mediated transcription and cell proliferation by PARP-1. Biochem Biophys Res Commun 2009;382:497–502. [29] Abid MR, Yano K, Guo S, et al. Forkhead transcription factors inhibit vascular smooth muscle cell proliferation and neointimal hyperplasia. J Biol Chem 2005;280:29864–73. [30] Libby P, Tanaka H. The molecular bases of restenosis. Prog Cardiovasc Dis 1997;40:97–106. [31] Oumouna-Benachour K, Hans CP, Suzuki Y, et al. Poly(ADP-ribose) polymerase inhibition reduces atherosclerotic plaque size and promotes factors of plaque stability in apolipoprotein E-deficient mice: effects on macrophage recruitment, nuclear factor-kappaB nuclear translocation, and foam cell death. Circulation 2007;115:2442–50. [32] Luccarini I, Ballerini C, Biagioli T, et al. Combined treatment with atorvastatin and minocycline suppresses severity of EAE. Exp Neurol 2008;211:214–26.