Serum Amyloid A Stimulates Vascular and Renal Dysfunction in Apolipoprotein E-Deficient Mice Fed a Normal Chow Diet

Abstract

Elevated serum amyloid A (SAA) levels may promote endothelial dysfunction, which is linked to cardiovascular and renal pathologies. We investigated the effect of SAA on vascular and renal function in apolipoprotein E-deficient (ApoE^-/-) mice. Male ApoE^-/- mice received vehicle (control), low-level lipopolysaccharide (LPS), or recombinant human SAA by i.p. injection every third day for 2 weeks. Heart, aorta and kidney were harvested between 3 days and 18 weeks after treatment. SAA administration increased vascular cell adhesion molecule (VCAM)-1 expression and circulating monocyte chemotactic protein (MCP)-1 and decreased aortic cyclic guanosine monophosphate (cGMP), consistent with SAA inhibiting nitric oxide bioactivity. In addition, binding of labeled leukocytes to excised aorta increased as monitored using an ex vivo leukocyte adhesion assay. Renal injury was evident 4 weeks after commencement of SAA treatment, manifesting as increased plasma urea, urinary protein, oxidized lipids, urinary kidney injury molecule (KIM)-1 and multiple cytokines and chemokines in kidney tissue, relative to controls. Phosphorylation of nuclear-factor-kappa-beta (NFκB-p-P65), tissue factor (TF), and macrophage recruitment increased in kidneys from ApoE^-/- mice 4 weeks after SAA treatment, confirming that SAA elicited a pro-inflammatory and pro-thrombotic phenotype. These data indicate that SAA impairs endothelial and renal function in ApoE^-/- mice in the absence of a high-fat diet.

Keywords: acute-phase protein; endothelial function; inflammation; nitric oxide; renal dysfunction.

Figure 1

Plasma MCP-1 and renal antioxidant Gpx-1 protein and activity are elevated in mice treated with SAA. ApoE^−/− mice were administered SAA, LPS or sterile PBS (vehicle control) by i.p. injection every 3 days over 2 weeks (as described in Study 1). Plasma was isolated 4 weeks after commencement of SAA treatment and levels of (A) MCP-1 by commercial ELISA. Data represent mean ± SD, n = 8 (control and LPS groups), or 10 (SAA group) animals. ^*Different to vehicle-control and LPS-treated mice in the absence of SAA; P < 0.05. Renal sections were then stained with anti-GPx1 antibody in batches so that all reagents and handling were identical, as described in Methods. Immuno-reactivity of representative kidney sections from control, LPS, and SAA-treated mice are shown. Figures represent at least two independent samples from each treatment group; magnification, x 200; scale bar, 200 μm. Renal homogenates were tested for total GPx activity (B). Data expressed as means ± SD for control (n = 5), LPS (n = 8), or SAA (n = 4) treatments. *Different to the vehicle-control group; P < 0.05; ^#Different to the LPS group; P < 0.05.

Figure 2

SAA enhances vascular cell adhesion molecule (VCAM) expression in ApoE^−/− mice leading to increased adhesion of circulating leukocytes. Upper panels: ApoE^−/− mice were injected i.p. with SAA, LPS or sterile PBS vehicle and sacrificed 2 weeks after treatment was finalized (as described under Study 1; experimental design). This sections of thoracic aorta from vehicle (control)-, LPS-, and SAA-treated mice were stained with anti-VCAM-1, counterstained with hematoxylin then imaged by light microscopy (arrow indicates immune-reactive VCAM-1). Data is representative of aortae from n = 6 mice in control and SAA groups and n = 4 mice in the LPS-treatment group. Scale bar = 50 μm. Lower panels (A,B): Aortae were isolated from ApoE^−/− mice treated with sterile PBS vehicle or SAA (as described under Study 2; experimental design). Aortae were carefully cleaned of fat and placed in an ex vivo dynamic flow adhesion system to assess the adherence of fluorescently labeled leukocytes using a live stage fluorescent microscope as described in Methods. (C) The total number of cells adherent to the vessel wall quantified time-dependently. Data represent mean ± SD, n = 4 mice per group. ^**Different to the vehicle control; P < 0.05; ^***Different to the vehicle control; P < 0.05.

Figure 3

SAA increases circulating urea levels in ApoE^−/− mice. Male ApoE^−/− mice were administered SAA, LPS or sterile PBS (vehicle control) by i.p. injection over 2 weeks and sacrificed a further 2 weeks after cessation of treatment (as described in Study 1). Aortae were excised and homogenized and then the homogenate samples tested for the levels of (A) cGMP as a surrogate marker for vaso-dilating NO. Plasma levels of (B) urea were quantified and their corresponding isolated aortae (C) assessed for total arginase activity as described in Methods. Data represent mean ± SD, n = 8 (control and LPS groups), or 10 (SAA group) animals. *Different to vehicle-control and LPS-treated mice in the absence of SAA; P < 0.05.

Figure 4

Urinary protein and kidney injury molecule-1 increase in mice after SAA administration. ApoE^−/− mice were treated with SAA, LPS or sterile PBS by i.p. injection over 2 weeks (as described in Study 1). Total protein (A) was quantified in urine samples from mice 4 or 18 weeks after commencement of SAA-treatment; open, black and gray bars represent control, LPS and SAA-treated groups. To assess acute renal damage, (B) urinary KIM-1 concentrations were determined in samples obtained 4 weeks after commencement of SAA-treatment using a commercial ELISA kit. Data represent mean ± SD; n = 8 (control and LPS) or 10 (SAA) mice. *Different to control and LPS-treated mice in the absence of SAA; P < 0.05.

Figure 5

SAA promotes kidney inflammation in ApoE^−/− mice. ApoE^−/− mice were administered SAA, LPS or sterile PBS for 2 weeks (as described in Study 1). After a further 2 weeks mice were sacrificed and kidney sections from control, LPS- and SAA-treated mice were stained with PAS. Data represent at least 3 different fields within each kidney section obtained from n = 8 (control or LPS) or 10 (SAA) animals. Arrows highlight condensed glomeruli in renal samples from SAA-treated mice. Magnification x200, scale bar = 200 μm. The change in Bowman's space (A) expressed as a percentage of the corresponding total glomerular area was calculated as defined in the Methods. Each point represents mean data from glomeruli present in a field of view (at least 3 fields of view; FOV) obtained at 200x magnification as assessed with Image-pro Plus (V6). Data represent mean ± SD; n = 8 (Control) or 8 (LPS-) and 10 SAA-treated mice. ^*Different to vehicle- or LPS-treated mice in the absence of SAA; P < 0.05; ns, not significant. Renal homogenates were assayed simultaneously for 3-chlorotyrosine and total tyrosine content (B). Quantitative mass data represent mean ± SD; n = 5 (Control) or 4 (LPS- and SAA-treated) mice. ^*Different to vehicle- or LPS-treated mice in the absence of SAA; P < 0.05.

Figure 6

NFκB phospho-p65 is increased in SAA-treated ApoE^−/− mice kidney. ApoE^−/− mice were administered SAA, LPS or sterile PBS and kidneys harvested 3 days later. Tissue sections were incubated with anti-mouse NFκB P-P65 Ser267 (FITC) and anti-mouse F4/80 (Alexafluor 594). Representative views (x200 magnification) are shown for renal medulla (Upper and cortex (Lower) from control, LPS- and SAA-treated mice. Where NFκB P-P65⁺ immunoreactivity (Ser267) labeling (green) co-registered with F4/80⁺ immunoreactivity labeling (red) cells appear yellow. Figure inset shows magnified region highlighting NF-κB P-P65⁺ and F4/80⁺ immunoreactivity and co-registration of these antigens in the renal medulla. Scale bar = 100 μm. NK-κB P-P65 levels in SAA-treated ApoE^−/− mouse kidney tissue homogenates were examined by Western blot (A) and normalized to total (in-gel) protein as shown in (B). Data are means ± SD; n = 2 experiments performed in duplicate.

Figure 7

Cytokines and chemokines in kidneys from control, LPS and SAA-treated mice. ApoE^−/− mice were administered SAA, LPS or sterile PBS for 2 weeks (as described in Study 1). After a further 2 weeks mice were sacrificed and levels of (A) IL-1α, (B) IL-1β, (C) IL-2, (D) MCP-1, (E) IFN-γ, and (F) GM-CSF were determined in kidney homogenates with a commercial Multiplex kit. Data expressed as mean ± SD: control (n = 7), LPS (n = 8) and SAA (n = 6). *Different to the control; P < 0.05. ^#Different to the LPS group; P < 0.05. Note, renal levels of IL-3, IL-4, IL-5, IL-6, IL-10, IL-12p70, IL-17, TNFα, MIP-1α, and RANTES that were determined on the same ELISA plate were all markedly elevated above those detected in vehicle- and LPS-treated (control) mice albeit this was above the maximum detection limit and therefore, not quantified.

Figure 8

SAA promotes TF accumulation in renal tissues. ApoE^−/− mice were administered SAA, LPS or sterile PBS for 2 weeks (as described in Study 1). Tissue sections obtained from vehicle- (control), LPS- and SAA–treated mice were co-stained with an FITC-conjugated anti-TF antibody (Upper) or with DAPI alone (corresponding Lower) to highlight the nuclear envelope then imaged with an Olympus fluorescence microscope. No TF⁺ immune-reactivity was detected in kidney tissue from control or LPS-treated mice. Arrows indicate clusters of DAPI-labeled nuclei localized to glomeruli that are detected in lower panels and reflected as a corresponding arrow in the corresponding panel above to indicate glomeruli in the same renal section. Data are representative of at least three different fields of view (×200 magnification) from each kidney section taken from n = 8 (control or LPS) or 10 (SAA) mice. Digital inset shows a representative overlay of DAPI and FITC fluorescence within a glomerulus (400× magnification); scale bar, 200 μm.