Arachidonate 15-Lipoxygenase Type B Knockdown Leads to Reduced Lipid Accumulation and Inflammation in Atherosclerosis

Inflammation in the vascular wall is important for development of atherosclerosis. We have shown previously that arachidonate 15-lipoxygenase type B (ALOX15B) is more highly expressed in human atherosclerotic lesions than in healthy arteries. This enzyme oxidizes fatty acids to substances that promote local inflammation and is expressed in lipid-loaded macrophages (foam cells) present in the atherosclerotic lesions. Here, we investigated the role of ALOX15B in foam cell formation in human primary macrophages and found that silencing of human ALOX15B decreased cellular lipid accumulation as well as proinflammatory cytokine secretion from macrophages. To investigate the role of ALOX15B in promoting the development of atherosclerosis in vivo, we used lentiviral shRNA silencing and bone marrow transplantation to knockdown mouse Alox15b gene expression in LDL-receptor-deficient (Ldlr −/−) mice. Knockdown of mouse Alox15b in vivo decreased plaque lipid content and markers of inflammation. In summary, we have shown that ALOX15B influences progression of atherosclerosis, indicating that this enzyme has an active proatherogenic role.


Introduction
Atherosclerosis is a complex disease that involves chronic inflammation at every stage from initiation to progression and eventually to plaque rupture [1]. Macrophage-derived foam cells play integral roles in atherosclerosis and are key regulators of the lipid-driven proinflammatory responses that promote atherosclerosis [2]. An inflammatory subset of macrophages accumulates in atherosclerotic plaques and produces proinflammatory cytokines [2][3] as well as lipoxygenases, which oxygenate polyunsaturated fatty acids to proinflammatory mediators [4]. Arachidonate 15lipoxygenase (ALOX15) catalyzes the production of eicosanoids that activate monocyte integrins and thereby enhance the adhesion of monocytes to the endothelium [5][6]. In vitro, ALOX15 appears to be proatherogenic, as it can be directly involved in LDL oxidation [7][8][9]. In vivo, however, both pro-and anti-atherogenic effects have been reported (reviewed in [10]). For example, atherosclerotic lesion formation is increased in Ldlr 2/2 mice that overexpress human ALOX15 in the endothelium [11]. In contrast, transgenic rabbits overexpressing human ALOX15 in macrophages display reduced atherosclerotic lesion area in the thoracic aorta [12].
Previous studies of lipoxygenases in the development of atherosclerotic plaques have largely focused on ALOX15 type A (ALOX15A) [13]. However, ALOX15 type B (ALOX15B) is expressed at considerably higher levels than ALOX15A in human carotid plaque macrophages [14][15] and levels of ALOX15B have been shown to be higher in symptomatic compared with asymptomatic lesions [14] suggesting a role for this enzyme in the atherosclerotic process. Furthermore, we have previously shown that overexpression of ALOX15B in macrophages increases secretion of the chemokines CXCL10 and CCL2, which are chemoattractants for T cells and monocytes [16]. In addition, products released in response to increased ALOX15B activation lead to increased expression of the T-cell activation marker CD69 as well as increased T-cell migration [16].
Here, we investigated the role of ALOX15B in foam cell formation in human monocyte-derived macrophages and found that silencing of the gene encoding ALOX15B (ALOX15B) decreases cellular lipid accumulation as well as proinflammatory signaling from macrophages. We also investigated the role of the mouse ortholog of ALOX15B (encoded by the gene Alox15b, also known as Alox8 [17]) in Ldlr 2/2 mice. In contrast to human ALOX15B, which catalyzes the formation of 15(S)-hydroperoxy eicosatetraenoic acid (HPETE), mouse ALOX15B produces 8(S)-HPETE and 8(S)-, 15(S)-diHPETE from arachidonic acid. However, the two enzymes display the highest sequence identity among all the mammalian lipoxygenases and substitution of only two amino acids of mouse ALOX15B is required to change the eicosanoids generated to 90% 15S-HETE [18]. We observed that mice transplanted with mouse Alox15b knockdown bone marrow displayed decreased atherosclerosis and impaired immune signal-ing. On the basis of our results, we suggest that ALOX15B has a proatherogenic and proinflammatory role during atherogenesis.

Primary macrophages
Buffy coats were obtained from healthy adult volunteer blood donors at Kungä lv Hospital, Sweden, and samples were deidentified before handling. Human mononuclear cells were isolated by centrifugation in a discontinuous gradient of Ficoll-Paque (GE Healthcare). Cells were seeded in Macrophage-SFM medium (Gibco) containing granulocyte macrophage colony stimulating factor (GM-CSF). After 3 days, the medium was changed to RPMI medium without GM-CSF and cells were cultured for 7 days before transfection. Macrophages were transfected with 20 nmol/L ALOX15B siRNA (Qiagen, SI03076206) or nonsilencing control siRNA (Qiagen, 1027280) in HiPerFect transfection reagent (Qiagen) according to the manufacturer's recommendations. Cells were washed after 24 h, siRNA was added again and cells were incubated with or without dimethyloxalylglycine (DMOG) for 24 h before extraction of RNA. DMOG is an inhibitor of HIF-prolyl hydroxylase, and acts  to stabilize HIF-1a expression causing induction of ALOX15B [15].
Mice C57BL/6 Ldlr 2/2 mice (7-week-old males) were purchased from The Jackson Laboratory (Bar Harbor, ME, USA) and housed in a pathogen-free barrier facility. Bone marrow donor mice of the same genotype (13-week-old males) were also purchased. All animal studies were approved by the regional animal ethics committee (animal ethics committee of Gothenburg).

Analysis of gene and protein expression
Total RNA was isolated with the RNeasy kit (Qiagen). Expression of human ALOX15B and mouse Alox15b mRNA was determined and normalized to b-actin mRNA expression using quantitative real-time PCR. The reverse transcription reaction was set up using a cDNA reverse transcription kit (#4368814) and performed with a Gene Amp PCR system 9700 (Applied Biosystems). Real time PCR amplification was set up using Taq man gene expression assays for human ALOX15A (Hs00609608_m1), human ALOX15B (Hs00153988_m1), mouse Alox15b (Mm01325281_m1), human ACTB (Hs99999903_m1) mouse Emr1 (Mm00802529_m1) and mouse ActB (Mm00607939_s1) respectively in combination with Universal PCR master mix (#4324018) and performed for 50 cycles on an ABI PRISM 7700 sequence detection system (Applied Biosystems).
Immunoblots were prepared as described [20]. Total cellular lysates were prepared from human macrophages, mouse bone marrow macrophages and aortic tissue. Antibodies are listed in Table 1. The polyclonal ALOX15B antibody (LX25) has been well characterized and found to detect human as well as mouse ALOX15B but not ALOX15A [21].

Quantification of Oil Red O-stained lipid droplets
Human macrophages were grown on chamber slides (Lab-Tek Systems), stained with Oil Red O and hematoxylin, and viewed with a Zeiss Axioplan 2 microscope. Twenty images per slide were obtained and analyzed for red pixels (lipid droplets), cell size and cell number using BioPix iQ 2.2.1 (Gothenburg, Sweden, see www.biopix.se for further information) [22].

Quantification of atherosclerosis
The aortic roots were embedded in OCT Tissue-Tec medium and immediately frozen in liquid nitrogen. Cryostat sections (10 mm thick) were cut from the level of the aortic valves and distally, and stained in 0.5% Oil Red O or used for immunohistochemistry (see next section). Distal aortae were dissected free from adipose and connective tissue, cut open longitudinally, and pinned flat on silicone-coated dishes. The aorta was then fixed in 70% ethanol for 5 minutes, stained with 0.5% Sudan IV for 6 minutes, and differentiated for 3 minutes in 80% ethanol. Digital images of the entire vessel were captured and the outline of the aortic surface was defined manually, whereas the stained lesion area was defined using computerized color selection in BioPix iQ 2.2.1 (Gothenburg, Sweden). The extent of atherosclerosis was calculated as the percentage of the aortic surface covered by lesions.

Immunohistochemistry
Cryostat sections of the aortic root were fixed for 10 minutes in ice-cold acetone, dried, and blocked in 5% goat serum in Trisbuffered saline. Sections were incubated (60 minutes, room temperature) with antibodies according to Table 1. Endogenous peroxidase was silenced with 0.3% H 2 O 2 in H 2 O for 10 minutes. 3-amino-9-ethylcarbazole (AEC) was used for detection. a-actin was stained using a Cy3 conjugated primary antibody. Collagen was stained using Masson's trichrome stains according to the manufacturer's instructions (Sigma; HT-15). Pictures were taking using a Zeiss Axioplan 2 epifluorescence microscope with an Axiocam camera and Axiovision software. Quantification of stained lesion area was defined using computerized color selection in BioPix iQ 2.2.1 (Gothenburg, Sweden, see www.biopix.se for further information) [22].

Knockdown of ALOX15B in human macrophages decreases cellular lipid levels, foam cell size and cytokine secretion
Transfection of human primary macrophages with ALOX15B siRNA resulted in knockdown of ALOX15B mRNA to approximately 50% of nonsilenced levels (Fig. 1A), but did not affect the expression of ALOX15A mRNA (Fig. 1B) or ALOX15A protein (too low levels to detect, data not shown). ALOX15B knockdown resulted in decreased lipid droplet accumulation as shown by reduced Oil Red O staining (Fig. 1C and D) and decreased foam cell size (Fig. 1E). These data indicate that ALOX15B can promote lipid droplet formation. In vitro experiments with human primary macrophages also showed lower levels of secreted IFN-c, IL-6, IL-8 and IL-12 (total) while there was no change in the levels of IL-1b (Fig. 1F) in the medium when ALOX15B was knocked down using siRNA, indicating that ALOX15B affects secretion of some proinflammatory cytokines.

Knockdown of mouse Alox15b decreases atherosclerosis in mice aorta
Immunohistochemical analysis of serial sections of aorta from cholesterol-fed Ldlr 2/2 mice showed that mouse ALOX15B was expressed in macrophage-rich areas of atherosclerotic plaques in mice ( Fig. 2A).
To investigate the importance of mouse ALOX15B in atherogenesis, lethally irradiated Ldlr 2/2 mice were transplanted with bone marrow cells infected with lentivirus containing Alox15b shRNA or nonsilencing control shRNA and fed a western diet for 20 weeks to induce atherosclerosis. Knockdown of Alox15b was confirmed by Q-PCR of mRNA and western blotting of protein from aortic tissue ( Fig. 2B and C) and bone marrow macrophages ( Fig. 2D and E) from recipient mice and by immunohistochemical evaluation of the proximal aorta from recipient mice (Fig. 2F).
Body weight, plasma cholesterol and triglycerides did not differ between the two groups (Fig. 3A, B, C). Atherosclerotic lesions both in the whole aorta ( Fig. 3D and E) and in the aortic root ( Fig. 3F and G) were significantly decreased in mice that received Alox15b knockdown bone marrow compared to nonsilencing control.
These data show that atherosclerotic lesions are decreased in Alox15b knockdown mice.

Knockdown of mouse Alox15b reduces T cell number and IL-2 levels in vivo
The number of T cells was significantly reduced in lesion areas in Alox15b knockdown mice (Fig. 4A) while no significant change was found in plaque macrophage content (Fig. 4B). Markers of plaque stability (collagen and a-actin) showed no change ( Fig. 4C  and D). To evaluate the possible proinflammatory role of mouse ALOX15B, we also measured cytokine levels in plasma from Alox15b knockdown mice compared to nonsilencing shRNA control-animals. The knockdown animals displayed significantly reduced plasma levels of IL-2 ( Fig. 4E) indicating that Alox15b knockdown may cause decreased systemic inflammation.

Discussion
In the present report, we studied the effects of decreased ALOX15B levels in human macrophages as well as in an atherosclerotic mouse model. Silencing of ALOX15B in human macrophages decreased cellular lipid accumulation and reduced proinflammatory cytokine secretion. In vivo experiments in an atherosclerotic mouse model showed that mouse Alox15b knockdown resulted in decreased atherosclerosis (measured as plaque area) as well as decreased inflammation, providing new evidence for ALOX15B as a proatherogenic protein.
ALOX15B knockdown in human macrophages caused markedly decreased cellular lipid accumulation and foam cell size, suggesting that ALOX15B is involved in lipid accumulation and foam cell formation. Silencing of ALOX15B decreased secretion of the proinflammatory cytokines IFN-c IL-6, IL-8 and IL-12 from the macrophages. IFN-c is known to promote foam cell formation [2] suggesting that the anti-foam cell formation effect of ALOX15B knockdown could be partly due to lower levels of IFN-c. Previous data show that ALOX15B expression in human macrophages induces chemokine secretion and that the preconditioned medium from these macrophages positively affects T-cell migration [23].
To understand the role of ALOX15B in atherosclerosis in vivo, we studied the consequence of Alox15b knockdown in an atherosclerotic mouse model. Although mouse ALOX15B differs from human ALOX15B, the two enzymes display 78% sequence identity and overproduction of human or mouse ALOX15B inhibited growth in mouse keratinocytes to the same extent, suggesting common signaling pathways [24]. Upon knockdown of Alox15b in the atherosclerotic mouse model, we found that subendothelial lipid accumulation was decreased in aortas from Ldlr 2/2 mice that received Alox15b knockdown bone marrow. The difference in lipid accumulation was not due to differences in plaque macrophage content, as no significant change in MAC-2 staining between the two groups was detected. Since silencing of human ALOX15B affected lipid accumulation in human macrophages, it is reasonable to propose that the effect on subendothelial lipid accumulation in mouse plaques is at least partly due to direct effects on foam cell formation.
Plaque T-cell content and plasma levels of IL-2 were significantly decreased in mice that received Alox15b-deficient bone marrow, indicating that mouse ALOX15B may also affect immunoregulation. Systemic administration of IL-2 to ApoE 2/2 mice has been shown to cause increased atherogenesis while anti-IL-2 antibodies decreased development of atherosclerosis [25]. IL-2 plays a significant role in T-cell activation and proliferation [26]. In agreement, our previous results show that human ALOX15B overexpression in human macrophages induces activation and migration of T cells [16]. Decreased T-cell content as well as decreased IL-2 levels in mice that received Alox15B deficient bone marrow are consistent with a role for mouse ALOX15B and its products in recruiting T cells to the lesion. Lesional T cells mainly have properties of the proinflammatory Th1 subtype and secrete the cytokines IL-2, IFN-c, and tumor necrosis factor (TNF)-a.
These cytokines cause activation of macrophages and vascular cells, promote inflammation and also participate in cellular immunity [27]. Atherosclerosis is a Th1-cell-driven disease and Th1 cytokines stimulate plaque formation [28]. Furthermore IL-2 promotes angiogenesis by activation of Akt (protein kinase B) and increase of reactive oxygen species [29]. Angiogenesis in human lesions may be maladaptive as it promotes lesion instability, with increased risk of plaque rupture and cardiovascular events [30]. It is likely that the reduced T-cell content in plaques of mice that received Alox15b-deficient bone marrow is caused by decreased immunological signaling from plaque macrophages, since silencing of human ALOX15B caused reduced cytokine expression in human macrophages.
Our experiments show that reduction of ALOX15B decreases inflammation and lipid accumulation, both in human primary macrophages and in mice, suggesting an active proinflammatory and proatherogenic role of ALOX15B.