Serum amyloid A (SAA) levels are elevated highly in acute phase response and elevated slightly and persistently in chronic diseases such as rheumatoid arthritis and diabetes. Given that fibroblasts exert profound effects on progression of inflammatory chronic diseases, the aim of this study was to investigate the response of fibroblasts to SAA. A dose-dependent increase in O2- levels was observed by treatment of fibroblasts with SAA (r = 0·99 and P ≤ 0·001). In addition, the expression of p47-phox was up-regulated by SAA (P < 0·001) and diphenyliodonium (DPI), a nicotinamide adenine dinucleotide phosphate (NADPH) oxidase inhibitor, reduced the release of O2- by 50%. Also, SAA raised fibroblast proliferation (P < 0·001) and this effect was completely abolished by the addition of anti-oxidants (P < 0·001). These findings support the notion that, in chronic inflammatory sites, SAA activated fibroblast proliferation and ROS production.
Serum amyloid A (SAA), a classical acute phase protein, plays a role in the inflammatory process by increasing production of cytokines [1–4] nitric oxide  and reactive oxygen species (ROS) by leucocytes . These SAA functions are particularly relevant in conditions of persistent elevated plasma SAA levels, as observed in chronic diseases such as rheumatic arthritis , cancer  and diabetes .
Several SAA receptors have been described, including CD36 and LIMPII analogous-1 (CLA-1),  lipoxin A1 receptor/formyl peptide receptor 1 (FPRL1) , tanis, a hepatic receptor activated by glucose  and Toll-like receptor-4 (TLR-4)  and TLR-2 . It has been reported recently that SAA activates rheumatoid synovial fibroblasts by binding to receptors of advanced glycation end products (RAGE) . Also, a high-density lipoprotein receptor, the scavenger receptor class B type 1 (SR-B1), is expressed in RA synovial tissue and is apparently involved in SAA-induced inflammation in arthritis .
Persistent inflammation, fibroblast deficient/excessive proliferation and endothelial cell hypertrophy have been implicated in various aspects of chronic diseases . Under these conditions the role of ROS in cellular dysfunction, in particular in signal transduction, is also increasingly recognized .
Identification of SAA cellular targets is very important for the full understanding of the SAA effects. Persistently elevated SAA levels induce a reactive form of amyloidosis in peripheral tissues, leading to progressive organ failure associated with amyloidal accumulation . Amyloidosis is associated commonly with ROS and the formation/development of excess fibrous connective tissue in an organ or tissue by fibroblast. As a result, fibroblasts exert profound effects on progression of chronic degenerative diseases. Thus, the purpose of this study was to investigate the effect of SAA on Swiss 3T3 fibroblasts ROS production and proliferation.
Materials and methods
Phenol red, hydroethidine and streptomycin were purchased from Invitrogen (Carlsbad, CA, USA). Lucigenin was supplied by Sigma Chemical Co. (St Louis, MO, USA). Recombinant human Apo-SAA 1 was obtained from Peprotech (Rocky Hill, NJ, USA).
Swiss 3T3 fibroblasts, obtained from the American Type Culture Collection, were maintained at 37°C in Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% (v : v) fetal bovine serum (FBS), ampicilin (25 µg/ml) and streptomycin (100 µg/ml). Adherent fibroblasts were treated with trypsin, washed gently, and resuspended in phosphate-buffered saline (PBS). Cell viability was always greater than 98%, as indicated by flow cytometry.
Superoxide anion determination
Lucigenin (1 mM) was added to fibroblast (2·5 × 106 cells/ml) incubation medium. Immediately afterwards, cells were treated with SAA (0, 5, 10 and 15 µg/ml). ROS release was monitored for 30 min in a Microplate Luminometer (EG&G Berthold LB96V, New Haven, CT, USA). The assays were run in PBS buffer supplemented with CaCl2 (1 mM), MgCl2 (1·5 mM) and glucose (10 mM) at 37°C .
Hydrogen peroxide determination
Fibroblasts (2·5 × 106 cells/ml) were treated with 15 µg/ml of SAA in the presence of phenol red (0·28 mM) and 1 U/ml horseradish peroxidase (HRP) type II. H2O2 release was monitored for 30 min by measurement at 610 nm. The reaction was stopped by addition of 10 µl 1 N NaOH .
Flow cytometric measurement of reactive oxygen metabolites using hydroethidine
Fibroblasts (2·5 × 106 cells/ml) were treated with 15 µg/ml of SAA in the presence of hydroethidine (1 µM). Fluorescence was measured using the FL3 (480 nm) channel in a fluorescence activated cell sorter (FACSCalibur) flow cytometer (Becton Dickinson, San Juan, CA, USA). Ten thousand events were analysed per experiment .
RNA extraction and quantitative reverse transcriptase–polymerase chain reaction (qRT–PCR)
After incubation at 37°C for 1 h in the absence or presence of SAA, RNA of fibroblast was extracted by using Trizol reagent (Invitrogen Life Technologies). Total RNA was then reverse-transcribed into cDNA by using Superscript II RT (Invitrogen) with oligo random hexamers. Prepared cDNA was subjected to qPCR analysis by using ROTOR GENE 3000 equipment (Corbett Research, Mortlake, Australia) with SYBR Green Master Mix (Applied Biosystems, Warrington, UK). Quantification of gene expression was based on cycle threshold (Ct) values, using β2 myosin genes as inner control .
Fibroblasts were plated in DMEM with 10% fetal calf serum (FCS). Six hours later, the medium was removed and replaced by DMEM with 0·5% FCS. Thirty-six hours afterwards, the medium was removed and cells were stimulated with SAA to trigger DNA synthesis. Twelve hours later, [3H]-methyl-thymidine (1 mCi/ml) was added and incorporation measurement was detected 12 h later. Cells then were lysed, DNA was precipitated and collected on filter paper and counted in a scintillation counter.
Fibroblasts (2 × 104 cells/cm2) were plated in DMEM with 0·5% FCS and SAA (1 µg/ml), changing the medium every third day, and the total number of cells was counted in a Fuchs–Rosenthal camera.
Comparisons were performed by one-way analysis of variance (anova) and Dunnett test. Results were obtained from three to five separate experiments and are expressed as mean ± standard error of the mean (s.e.m.).
Results and discussion
To test if SAA was able to trigger ROS production in 3T3 cells we used three assays: hydroethidine reduction, phenol red reduction and lucigenin-amplified chemiluminescence. The ROS measurements were taken under conditions in which the interference of the other effects was minimized; appropriate controls were carried out using SAA in the assays without cells. The SAA did not directly affect the luminol, lucigenin and phenol red assays. The possible interference of SAA in the reaction of ROS with the reagents was tested in a previous study conducted by our group by adding xanthine and xanthine oxidase to lucigenin and luminol assays, without cells. These tests did not indicate a direct effect of SAA on the lucigenin, luminol and phenol red ROS detecting systems. Furthermore, the lucigenin-amplified chemiluminescence probe is the most sensitive test to measure the rise in O2- levels and causes less ROS cycling interference than other probes for measuring ROS. Due probably to the higher sensitivity of the lucigenin-amplified chemiluminescence probe, we were able to measure the rise in O2- levels only by the lucigenin-amplified chemiluminescence assay (Fig. 1a). No detectable ROS production was observed with assays using hydroethidine reduction and phenol red reduction.
There was a positive correlation between SAA concentration and O2- production (Fig. 1b), and Pearson's correlation found r = 0·99 and P = 0·001 (Fig. 1a).
ROS production seems to occur, at least in part, due to activation of the nicotinamide adenine dinucleotide phosphate (NADPH) oxidase enzymes, as diphenyleneiodonium (DPI), a specific inhibitor of NADPH oxidase, reduced O2- production induced by SAA by 50% (Fig. 1c). Also, the expression of p47-phox was up-regulated by SAA, as evaluated by qPCR (Table 1). The mechanism of O2- production by a membrane-bound NADPH oxidase involves translocation of the cytosolic subunits to the membrane. Non-phagocytic cells express a specific pattern of NADPH oxidase subunits. In fibroblasts, p47phox is one of the cytosolic subunits that have a regulatory function. In these cells the active enzyme migrates from the cytosol to the plasma membrane during the activation process .
Table 1. Effect of serum amyloid A (SAA) (15 µg/ml) on mRNA expression of the nicotinamide adenine dinucleotide phosphate (NADPH) oxidase component p47-phox in Swiss 3T3 fibroblasts.
p47-phox mRNA (relative units) (mean ± standard deviation)
P < 0·001 for comparison between control and SAA treatment.
In neutrophils, we have already shown that SAA primes cells, increasing the amount of ROS produced in response to opsonized stimulus . Recently it has been described that SAA directly induces activation of the neutrophil NADPH oxidase . Activation of the NADPH oxidase complex in neutrophils and other phagocytes has been implicated in killing activity . However, in non-phagocytic cells, the production of ROS triggered by a variety of ligands may be important as second messengers. Ligands such as several growth factors and cytokines induce production of ROS by fibroblasts, endothelial and other cell types . This process is relevant depending upon the amount and duration of ROS modulation.
SAA stimulated [3H]-thymidine uptake into DNA (Fig. 2a), directly indicating an increase in the relative rate of DNA synthesis, implying that SAA promotes Swiss 3T3 cell proliferation. These observations were confirmed in the cell growth curve experiment (Fig. 2b) after 6–8 days of incubation. Fibroblast proliferation and ROS production induced by SAA were correlated events, as indicated by the inhibitory effect of the antioxidant agents, N-acetyl-l-cysteine (NAC) and α-tocoferol, on cell proliferation (Fig. 2c). ROS activates cell proliferation by several pathways, as follows: (i) through the action of protein tyrosine phosphatases (PTPs). PTPs control the phosphorylation state of numerous signal-transducing proteins and are therefore involved in the regulation of cell proliferation, differentiation, survival, metabolism and motility. The catalytic region of PTPs includes cysteines, which are susceptible to oxidative inactivation. Thus, ROS decrease phosphatase activity that enhances protein tyrosine phosphorylation, thereby influencing signal transduction. (ii) Treatment of cells with hydrogen peroxide leads to phosphorylation and activation of p38 mitogen-activated protein (MAP) kinase. This probably occurs due to the activation of upstream signalling pathways of extracellular-regulated kinase (ERK1/2), or it may be an indirect effect of the inhibition of phosphatase activity by ROS. (iii) ROS can regulate intracellular and plasma membrane ion channels. Such regulation of ion channels may occur either directly or through ROS-sensitive signalling systems. Cytosol-free Ca+2 concentrations act as important intracellular messenger systems . The effect of SAA on cell proliferation has been shown previously on human fibroblast-like synoviocytes. Moreover, SAA stimulated the proliferation, migration and tube formation of endothelial cells . Further studies are needed to identify the specific receptor(s) involved in SAA-induced ROS production and fibroblast proliferation.
Regulation of the intracellular redox state by growth factor-induced changes of NADPH oxidase activity is thought to have an important impact on redox-sensitive signalling cascades. Activation of growth factor-stimulated signalling cascades by low levels of ROS results in increased cell cycle progression. For example, the proliferative state of fibroblasts is associated closely with intracellular ROS levels. Low ROS levels lead to cell growth arrest, as induced by contact inhibition. Conversely, overproduction or insufficient scavenging of ROS can result in enhanced oxidative stress and membrane and DNA damage, which have been implicated in cancer initiation and promotion, apoptosis and necrosis .
High concentrations of SAA associated with increased microvascular permeability observed in some diseases  may define specific loci, such as the interstitial space and body cavities, for SAA action. In these sites, SAA may activate neutrophils and macrophages and promote fibroblast proliferation. Furthermore, it has to be considered that local production of SAA, especially by endothelial cells, adipocytes, macrophages and own fibroblasts, may supply SAA even without a remarkable inflammatory process. SAA serum levels are normally very low (1–3 µg/ml). In circulation, SAA associates predominantly with the third fraction of high-density lipoproteins (HDL3). SAA associated with HDL3 is not considered a proinflammatory agent . In inflammatory foci, SAA may dissociate from HDL3 and be active. The concentration of free SAA in an inflammatory site is not known. Nevertheless, we showed that SAA is active on fibroblasts in concentrations where it is possible to find SAA. After the dissolution of HDL, SAA is absorbed by macrophages and in lysosomes it is cleaved by cathepsins. An impaired SAA degradation process leads to accumulation of the amino acid intermediate. The acid environment of lysosomes facilitates the formation and polymerization of these intermediates. After deposition of accumulated intermediates in the extracellular space, glycosaminoglycans, serum amyloid P (SAP) and lipid components bind to the fibrils and confer resistance to proteolysis. In normal circumstances, SAA is degraded completely. In fact, any alteration in the genes responsible for maintaining the normal process or that persistently elevates the SAA serum level may influence the formation of AA amyloidosis. A large part of human AA proteins isolated from amyloid deposits derives from SAA1. AA amyloidosis is, in part, influenced by genetic factors. For example, Japanese patients with rheumatoid arthritis show a higher association of SAA1 polymorphisms and AA amyloidosis, specifically C13T and C2995T polymorphisms .
At the same concentration of SAA, fibroblasts synthesize more ROS than neutrophils. SAA alone does not trigger ROS production in neutrophils, but augments the maximum ROS production rate when a second stimulus is added. This effect defines SAA as a priming agent of neutrophils . Neutrophils play an essential role in host defences against microbial pathogens and their anti-microbial arsenal is composed of ROS. In phagocytes, superoxide is generated mainly by the reaction of oxygen and NADPH through the NADPH oxidase complex. Superoxide anions and hydrogen peroxide (H2O2) generated by NADPH oxidase give rise to other ROS that are strong cytolytic agents, such as hypochlorous acid [formed by the action of myeloperoxidase (MPO) released from neutrophil granules] and hydroxyl radicals. The finding that SAA primes neutrophils suggests that SAA has a concerted mode of action, driving a more powerful response of innate host defence. In fact, neutrophils are classic and more potent cell producers of ROS than fibroblasts. Fibroblast functions involve synthesis of the extracellular matrix and collagen. These cells play a critical role in wound healing. Fibroblasts are not classic cells ‘programmed’ to produce ROS. In fibroblasts, ROS are signalling molecules for other cell functions, such as differentiation, proliferation, death, senescence and production of cytokines.
Fibroblast proliferation and fibrogenesis are important factors that lead to complications of various diseases, such as atherosclerosis, rheumatic arthritis, diabetic nephropathy and retinopathy . Data from this study highlight SAA as an important inducer of ROS production and proliferation of fibroblasts. This effect may be critical during physiological processes such as would healing, for example. However, it can be deleterious in promoting uncontrolled inflammation in proliferative diseases that involve fibroblast dysfunction, such as fibrosis . This is more relevant in pathological conditions with a persistent increase of SAA serum levels, such as diabetes , rheumatic diseases  and cancer .
All authors thank Fundação de Amparo à Pesquisa do Estado de São Paulo, Brasil (FAPESP) and Conselho Nacional do Desenvolvimento Científico e Tecnológico, Brasil (CNPq) for grants and fellowships.
The authors declare that there is no conflict of interest.