The wall volume in the rabbit abdominal aorta, an indicator of atherosclerotic plaque burden, was analyzed by serial in vivo MRI scanning. Comparison with age and sex-matched New Zealand White rabbits confirmed a significantly higher wall volume of abdominal aorta in WHHL rabbits after 6 weeks of feeding with high cholesterol chow (Supplementary Figure 3). After 6 weeks of dietary supplementation with LA, wall volume in the abdominal aorta was significantly lower compared to the control arm (Figure 1A and B). These effects of LA were more pronounced at the end-point (12 weeks). At the end of 12 weeks, there was a significant slowing of the rate of progression of plaque in the LA group, while abdominal aorta wall volume continued to increase in the control group (Figure 1A and B). Morphometric assessment of plaque burden in the thoracic aorta at sacrifice, corroborated the anti-atherosclerotic effects of LA on the abdominal aorta (Figure 1C and D). Immunohistochemical analysis of thoracic aortic sections revealed a reduction in CD68⁺ cells in atherosclerotic plaque and increased smooth muscle content (Figure 2A-D). LA also decreased lipid accumulation and fibrillar collagen deposition (Figure 2E-H).

Figure 3 and Supplementary Table 2 illustrates the responses to various agonists in the thoracic aorta of WHHL rabbits. Figure 3A and B depict LA effects on acetylcholine and insulin mediated responses. In contrast to significant effects on acetylcholine and insulin mediated dilation, LA had no effects in response to SNP, an endothelium-independent vasodilator (Supplementary Table 2). LA reduced vasoconstrictor responses to angiotensin II (Figure 3C) but did not alter responses to phenylephrine or endothelin-1 (Supplementary Table 2). In view of decreased contractility to angiotensin II in the aortic segments, we examined the expression of the angiotensin Type 1 (AT1) receptor. Real time PCR analysis of revealed that AT1 receptor was reduced by LA treatment (Supplementary Figure 4).

LA reduced oxidative stress as evidenced by lower 8-isoprostane levels in plasma and aortic homogenates (Figure 4A and B). These results were confirmed by in situ aortic O₂^•- measurement using dihydroethidium (DHE) staining (Figure 4C and D). Because various studies have indicated that NADPH oxidase is the major source of vascular O₂^•-, its activity in aortic homogenates was analyzed.(Mehta and Griendling, 2006) Consistent with the results of 8-isoprostane measurement and DHE staining, basal O₂^•- production in aortic homogenates was significantly lower in the LA group compared to the control groups (Figure 4E). In response to exogenously added NADPH, there was a 1.75 fold lower O₂^•- generation in LA treated aortas compared to controls (Figure 4E).

As vascular inflammation plays a pivotal role in atherosclerosis we assessed pro-inflammatory gene expression and cellular correlates of inflammation. LA decreased mRNA expression of VCAM-1, ICAM-1 and L-selectin (CD62L) in thoracic aorta (Figure 5A). The mRNA expression of P-selectin and MCP-1 (CCL2) showed non-significant decreases in the LA group (Figure 5A). We corroborated the effect of LA on VCAM-1 and ICAM-1 expression by measuring expression of these adhesion molecules at the protein level (Figure 5B and C). Since NF-κB regulates expression of pro-inflammatory genes, we examined the content of the p65 sub-unit of NF-κB in the nuclear and cytosol as a surrogate for its activation. LA decreased the translocation of the p65 sub-unit of NF-κB from cytosol to nucleus (Figure 5D and 5E). The inhibition of NF-κB signaling by LA was confirmed by assessing the DNA binding activity of NF-κB with electrophoretic mobility shift assay (EMSA. Figure 5F and G). Consistent with these effects on NF-κB, acute treatment with LA inhibited adhesion of leukocytes in the cremasteric circulation when acutely administered to mice 24 hours prior to an inflammatory challenge with TNFα (Supplemental Figure 5).

Since T lymphocytes are critical components of atherosclerotic inflammation, we examined whether supplementation with LA chronically decreased T cell infiltration in aortic wall. Figures 6A and B revealed that LA significantly decreased T cell content (CD3+ cells) at the level of the thoracic aorta. Adhesion to endothelium is a very important early step for T cell infiltration. To investigate the effect of LA on the interaction between T cells and endothelial cells, T cells labeled with the fluorescent dye CFSE were injected into high fat fed ApoE^-/- mice with or without LA pre-treatment over the preceding 8 weeks. LA pre-treatment significantly decreased the rolling and adhesion of labeled T cells to the cremasteric endothelium (Supplementary Figure 7A and B).

MRI scanning demonstrated that 6 weeks of dietary supplementation with LA was sufficient to inhibit atherosclerotic lesion development in WHHL rabbits. We induced rapid progression of plaque by high cholesterol feeding for 6 weeks prior to randomization. LA treatment inhibited progression of plaque in the absence of notable changes in plasma lipoprotein levels. Our data are consistent with the effects noted by Zhang et al but differ in dosing and timing of LA treatment (Zhang et al fed mice with 0.2% (wt/wt) LA for a period of 10 weeks.). In contrast, our studies were in established atherosclerosis with much lower doses (Zhang, et al., 2008). Zhang et al demonstrated an effect of LA in reducing TG concentrations in the ApoE^-/- mouse model while total cholesterol increased in the ApoE^-/-/LDLR^-/-strain. These disparate effects on lipoprotein metabolism raise the question of context dependent effects of LA that may relate to degree and genesis of lipoprotein abnormalities in the animal model studied. In our studies, high fat diet administration in the setting of additional genetic predisposition lead to extremely high values of most lipid parameters and precluded detection of any changes that may have been induced by LA.

An additional important observation in this study pertained to the weight loss noted with LA. LA has been described to have appetite suppressant effect through an AMP kinase mechanism in the hypothalamus (Kim, et al., 2004). The only definitive way to discount an effect on food intake is to perform pair-feeding experiments. Our experiments were very similar to a pair fed experimental situation, in that the LA and control groups consumed equal amount of pre-apportioned diet. Thus the weight loss effects are unlikely to relate to reduced food intake. While it may be difficult to dissociate the effects on weight loss from its direct effects in reducing plaque progression, we believe that latter effects are distinct and dominant. In the study by Zhang et al that did employ pair feeding, the attenuation of lesion area was not accounted by weight loss alone (Zhang, et al., 2008). Studies by Kim et al in rats, demonstrated weight loss at doses that ranged from 0.25, 0.5 and 1%, wt/wt (Kim, et al., 2004). Our studies were designed cognizant of these earlier studies and therefore employed far lower doses (20 mg/d) than that employed by either study to limit the weight loss effects. Nevertheless it appears that weight loss is seen in the rabbit model and may potentially explain the effects on insulin sensitivity. The effects of LA in potentiating insulin mediated dilation are important as LA has been suggested to be an insulin sensitizing agent through PI3 kinase/Akt dependent pathways (Jacob, et al., 1999; Yaworsky, et al., 2000). LA has been shown to enhance IRS-1 expression and insulin-stimulated IRS-1 association with PI3K (Saengsirisuwan, et al., 2004). Our results indeed do confirm a marked insulin sensitizing effect of LA in the vasculature although. Steady state, levels (non-insulin stimulated) of phosphorylated Akt and eNOS in the aorta were no different. The lack of difference in phosphorylation states by westerns does not by itself argue against the in-vivo functional data as the phosphorylation data was obtained in the absence of agonist stimulation. Thus there could still be important differences in the presence of insulin or acetylcholine. Additionally eNOS and Akt is regulated by phosphorylation at multiple sites and the lack of difference at one site does not preclude differences at other site.

Our results demonstrating a favorable effect on endothelial function is consistent with our prior observations suggesting an effect of LA even in the context of multiple risk factors in diseases such as metabolic syndrome and Type II diabetes (Sola, et al., 2005). Notably, although LA treatment did not affect the vasoconstriction to PE and endothelin-1, it reduced the constriction to angiotensin II through down-regulation of AT1 mRNA, similar to its effects in the renal vasculature (Mervaala, et al., 2003). These effects are consistent with therapeutic synergism seen in our human studies evaluating the effects of AT-1 receptor blockade in conjunction with LA supplementation (Sola, et al., 2005). The renin-angiotensin-aldosterone system, a crucial regulator of vascular homeostasis, has consistently been shown to play a prominent role in atherogenesis (Daugherty, et al., 2000; Munzel, et al., 2008; Weiss, et al., 2001). Conversely, blockade of AT1 receptors in atherosclerosis normalizes NADPH oxidase activity, improves endothelial function, and reduces plaque area and macrophage infiltration (Hornig, et al., 2001; Sorescu, et al., 2001; Warnholtz, et al., 1999). Thus LA may potentially exert additional synergistic effects with drugs that target the RAAS system in atherosclerosis, a finding that may be worthy of exploration in the future.

A novel finding was the pronounced effects of LA on T cell transmigration. Our results show that LA markedly reduced T cell infiltration in the atherosclerotic lesion. This may be attributed to reduced migration or “homing” of T cells presumably through its effects on reducing adhesion molecules such as CD62L (L-Selectin) and ICAM and/or chemotactic influences such as RANTES and SDF-1, factors critically important in the migration of T cell populations to the plaque. It is interesting to note that RANTES is regulated by NFkB and the effects of LA on RANTES expression may relate to these effects (Lee, et al., 2000; Munzel, et al., 2000). These observations are similar to the effect of LA in modulating neural inflammation and reducing the migration of T cells to the central nervous system in experimental encephalomyelitis, through effects on cyclic AMP pathways (Chaudhary, et al., 2006; Schillace, et al., 2007). Thus treatment with LA has broad anti-inflammatory and metabolic effects that may be beneficial in inflammatory diseases.

01

02