Published online before print February 5, 2008
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||



,1
* Departments of Clinical Analysis and Toxicology, Faculty of Pharmaceutical Sciences, and
Immunology, Biomedical Science Institute, University of São Paulo, Brazil; and
Department of Biochemistry/Molecular Biology/UNIFESP, São Paulo, Brazil
1Correspondence: Department of Immunology, Biomedical Science Institute, University of São Paulo, Av: Prof. Lineu Prestes, 1730, CEP 05508-000, São Paulo, SP, Brazil. E-mail: momrusso{at}icb.usp.br
|
|
|---|
Key Words: Toll like receptors macrophages nitric oxide (NO)
|
|
|---|
. These cytokines are produced mainly by activated macrophages and monocytes but also by endothelial cells, lymphocytes, and other cells [2
]. SAA appears to have a broad range of biological activities that includes immunomodulation, tissue remodeling [3
], and immunosurveillance [2
]. We have characterized the immunomodulatory properties of SAA in neutrophils by showing that SAA induces the expression and release of inflammatory cytokines and enhanced response to opsonized particles [4
5
6
7
]. Others have shown that SAA mobilizes intracellular calcium and stimulates cyclooxygenase-2 expression and PGE2 synthesis [8
9
10
]. A plethora of reports in recent years has shown that a number of endogenous molecules might also be potent activators of TLRs (reviewed in ref. [11 ]). For instance, Ohashi et al. [12 ], using macrophages from C3H/HeJ mice with point mutation in the Tlr4 gene, demonstrated that macrophage stimulation by recombinant human heat shock protein 60 (rhHsp60) required functional TLR4.
In the present study, we determined whether SAA is capable of activating TLR4. For this, we monitored the production of NO by peritoneal macrophages stimulated with SAA in TLR4-sufficent and TLR4-deficient mice strains. As TLR4 signaling includes the adaptor molecule MyD88 and activation of MAPKs, we also investigated whether MyD88 deficiency or inhibition of specific MAPKs influences NO production induced by SAA.
Here, we show that macrophage activation by SAA is TLR4-dependent. The downstream pathways involved in SAA-induced NO production include ERK1/2 and p38 activation but not the MyD88 pathway. We suggest that SAA might be an endogenous TLR4 agonist.
|
|
|---|
antibody.
Animals
Female C57BL/6, C3H/HePas, C3H/HeJ, C57BL/10ScCr, and MyD88–/–, 4- to 8-week-old mice under standard pathogen-free conditions, were bred in the breeding unit of the Institute of Biomedical Sciences (São Paulo, Brazil). The Faculty of Pharmaceutical Sciences, University of São Paulo Ethical Committee (Brazil), approved all experiments.
Administration of Con A
When indicated, the animals were injected i.p. with 10 µg Con A (Con A-primed macrophages) or sterile saline (resident macrophages) and were killed 48 h after treatment.
Cells and cell culture
Macrophages were obtained from mice by washing their peritoneal cavities with 5 ml ice-cold PBS. The cells from individual mice were centrifuged (160 g, 10 min, 4°C), ressuspended in complete RPMI-1640 medium, and adjusted to 2.0 x 106 cells/ml. Total and differential cell counts were performed on fixed and stained cell suspensions with 0.05% crystal violet dissolved in 3% acetic acid (final solution). Cell suspensions used in all experiments consisted of at least 85% macrophages, as judged by morphological criteria. For culturing, 100 µl cell suspension was plated into each well of 96-well, flat-bottom, tissue-culture plates (Corning, Corning, NY, USA), and cultures were incubated at 37°C and 5% CO2.
NO production
NO production was quantified by measuring the accumulation of the stable, oxidative metabolite, nitrite (NO2–) in culture supernatants by the standard Griess reaction as described previously [13
]. All determinations were performed in duplicate and expressed as micromoles concentration of NO.
Inhibitors of MAPK
The experiments with MAPK include the use of 10 µM SB203580 (a p38 MAPK inhibitor) or 50 µM PD98059 (a MAPK ERK, MEK-1). These inhibitors were preincubated with macrophages for 15 min before the administration of SAA and then cultured for 24 h.
Inducible NO synthase (iNOS) expression
iNOS expression was measured by Western blot analysis of total cell lysates. Briefly, cells were lysed in buffer containing 20 mM Hepes, 150 mM NaCl, 1.5 mM MgCl2, 1 mM EGTA, 10% glycerol, 1% Triton X-100, aprotinin (1 µg/mL), leupeptin (1 µg/mL), and 1 mM PMSF, pH 7.5, containing phosphatase inhibitors. Total cell lysates (20 µg/lane) were resolved in 8% SDS-polyacrylamide gel and blotted onto nitrocellulose sheets. Blots were probed using polyclonal antibody against iNOS (BD Transduction Laboratories, BD Biosciences, San Jose, CA, USA), diluted at 1:1000. After incubation with an appropriate HRP-conjugated secondary antibody, blots were developed using the Pierce system.
Measurement of endotoxin activity
The endotoxin activities of rhSAA and LPS from Escherichia coli 0111:B4 (Sigma-Aldrich, St. Louis, MO, USA) were determined using the Limulus amebocyte lysate (LAL) assay kit (BioWhittaker, Walkersville, MD, USA), which was performed according to the manufacturers recommendations. We found that 10 µg/mL rhSAA preparation contained 3.6 endotoxin units (EU)/mL, whereas 1 ng/mL LPS contained 5 EU/mL.
Removal of endotoxin from rhSAA
To remove residual endotoxin of rhSAA, we submitted our preparation to the Triton X-114 method as described by Aida and Pabst [14
]. The rhSAA was diluted in free endotoxin water (1 mg/mL), and the solution was submitted to two to three cycles of Triton X-114 extractions. After each cycle, the rhSAA content was determined by SDS-PAGE (18%). We found that after three cycles, the protein yield was very poor (we lost more than 60% of the protein) and that the endotoxin level of rhSAA was similar to the original preparation. We emphasized that we succeeded to remove endotoxin from other LPS-contaminated protein, such as commercial OVA, rhHsp60, or rhIDO (data not shown). Therefore, we used original rhSAA preparation throughout the experiments.
Proteinase K digestion and polymyxin B treatment
To exclude the effect of possible endotoxin contaminants, proteinase K (50 µg/mL) or polymyxin B (25 µg/mL) was added to SAA (10 µg/mL) for 45 min at 37°C in PBS (containing Ca++ and Mg++). The samples were then added to macrophage cultures, and 48 h later, the nitrite concentrations were determined.
Statistical analysis
The statistical analysis consisted of the one-way ANOVA with the Student-Newman-Keuls multiple comparisons test.
|
|
|---|
![]() View larger version (8K): [in a new window] |
Figure 1. (A) Dose response of SAA-induced NO production by peritoneal macrophages (r=0.99). (B) NO production found in the supernatants of 24 h- and 48 h-cultured macrophages in the absence and in the presence of SAA (10 µg/mL). *, P < 0.05; **, P < 0.01, versus control. There was no significant difference between 24- and 48-h groups. (C) The NO production in the supernatants in resident and Con A-primed in vivo macrophage cultures incubated with SAA (10 µg/mL) for 48 h (**, P<0.01). The data represent the average ± SD of three to five experiments performed in duplicate.
|
Involvement of MAPKs but not TNF-
in NO production
We have previously shown that SAA induces the expression and production of TNF-
in human neutrophils [6
]. Given the fact that TNF-
is involved in NO production (reviewed in ref. [18
]), we evaluated whether the addition of anti-TNF-
antibodies could affect NO production. As shown in Figure 2A
, treatment of macrophage cultures with anti-TNF-
antibody did not inhibit NO production. Next, we examined the involvement of MAPKs in NO production, as MAPKs are involved in TNF-
production [6
] and other inflammatory mediators [19
, 20
]. We used two highly selective p38 and ERK1/2 inhibitors (SB203580 or PD98059, respectively) to treat macrophages.After 15 min of incubation, macrophage cultures were exposed to SAA for 24 h, and the NO production and iNOS expression were determined. Before performing these experiments, we determined cell viability after incubation with 10 µM SB203580 or 50 µM PD98059. We found that at these concentrations, the imidazole inhibitors were not cytotoxic, as assessed by the trypan blue exclusion test (data not shown). Figure 2B
and 2C
, shows that both inhibitors caused a significant decrease in SAA-induced NO production and iNOS expression, with SB being more effective.
|
View larger version (11K): [in a new window] |
Figure 2. (A) Effects of anti-TNF- (5 µg/mL) in the NO production induced by SAA (10 µg/mL) in peritoneal macrophage cultures maintained for 48 h. There was not a significant difference between groups. The data represent average ± SD of at least three experiments. (B and C) Effects of SB203580 (SB; 10 µM) and PD98059 (PD; 50 µM) in the NO production and iNOS induced by SAA (10 µg/mL) on peritoneal macrophages. The cells were preincubated with the inhibitors 15 min before SAA stimulation and cultured for 48 h. The data are representative of at least three experiments.
|
![]() View larger version (6K): [in a new window] |
Figure 3. (A) The effect of SAA (10 µg/mL) in the production of NO in macrophages from C3H/HePas (TLR4-sufficient) and C3H/HeJ (TLR4-deficient). (B) Effect of SAA in the production of NO in macrophages from B6 [TLR4-sufficient (TLR4+/+)] and C57BL/10ScCr [TLR4-deficient (TLR4–/–)]. The production of NO was abolished in macrophages from TLR4-deficient (C3H/HeJ and C57BL/10ScCr) when compared with TLR4-sufficient mice (**, P<0.01). (C) Effects of SAA in the production of NO in macrophages from B6 [wild-type (WT)] and MyD88–/–. The data represent the average ± SD of at least five experiments.
|
![]() View larger version (11K): [in a new window] |
Figure 4. Effect of different concentrations of SAA, LPS, and SAA plus LPS on the production of NO. Macrophages were stimulated with SAA (1 and 10 µg/mL), LPS (0.1, 1.0, and 10 ng/mL), or SAA (1 µg/mL) + LPS (0.1, 1.0, and 10 ng/mL) and maintained in culture for 48 h. The data represent the average ± SD of at least three experiments.
|
![]() View larger version (8K): [in a new window] |
Figure 5. Effects of polymyxin B (Poly B) and proteinase K (PK) in the production of NO induced by SAA or LPS. The production of NO by macrophages determined 48 h after incubation with SAA (10 µg/mL) or LPS (0.1 µg/mL). Before addition of SAA or LPS, macrophages were preincubated with either Poly B (25 µg/mL) or PK (50 µg/mL) for 45 min. (A) Poly B does not inhibit the SAA-induced NO production. (B) Poly B inhibits LPS-induced NO production (*, P<0.05). (C) PK treatment abrogated SAA-induced NO production (**, P<0.01). (D) PK treatment does not affect LPS-induced NO production. The data represent the average ±SD of at least three experiments.
|
|
|
|---|
The production of NO in response to SAA by macrophages from TLR4-suficient mice was dose-dependent. As purified, endogenous TLR4 agonists are usually obtained from recombinant bacteria, a major concern in relation to our results is whether the recombinant SAA was contaminated with endotoxin. Thus, we first determined the endotoxin activity present in SAA and found that 10 µg SAA contained 3.6 EU/mL, which is near 1 ng LPS. Then, we submitted our rhSAA preparation to the Triton X-114-phase separation method, but we could not remove LPS from our SAA preparation. This was not the case when we used other LPS-contaminated proteins. The low recovery obtained with rhSAA and the difficulty in purifying SAA by Triton X-114 treatment are probably a result of the fact that rhSAA has hydrophobic sites and low molecular weight. Thus, to rule out the possibility that NO production is a result of LPS contamination, we used two experimental approaches that have been used by other authors when the SAA preparation is not LPS-free [27 28 29 ]. First, we added LPS at a different concentration starting with a similar concentration that we detected in our SAA preparation and increasing the concentration to 10- to 100-fold. We found 5 EU/mL LPS, which is equivalent to 1 ng, induced negligible production of NO by resident macrophages. Significant production of NO was only observed with a tenfold-higher dose, and the amount of NO produced was roughly threefold lower than that induced by SAA. Even with concentrations of LPS 100-fold higher, the NO production was still lower than that obtained with SAA (data not shown). We also ruled out a possible synergism between SAA and LPS, as addition of LPS at different doses to 1 µg SAA did not enhance NO production. Second, pretreatment of SAA solution with polymyxin B did not interfere in NO production. In contrast, pretreatment of SAA with proteinase K decreased SAA-induced NO production significantly.
We also found that SAA stimulated the production of TNF-
[6
], which subsequently could induce NO. However, we excluded the participation of TNF-
in mediating NO production, as neutralizing antibodies against TNF did not inhibit SAA-induced NO release. The induction of iNOS expression and NO production by SAA was dependent on p38 and ERK1/2 pathways, as specific MAPK inhibitors SB203580 and PD98059 inhibited iNOS expression and NO production. This finding is in agreement with the recognized signaling pathway for iNOS expression promoted in macrophages by other stimuli [19
, 20
] and in other cell types [30
]. Moreover, priming of macrophages by in vivo injection of a plant lectin Con A resulted in an enhanced production of NO after stimulation in vitro with SAA. Con A is a polyclonal T cell activator and mimics an IFN-
-dependent macrophage activation as previously reported [15
16
17
] (reviewed in ref. [31
]). Thus, it is likely that enhanced NO production resulted from a synergy between IFN-
signaling and SAA signaling through the TLR4 molecule.
Results obtained in MyD88-deficient mice indicate that this pathway is dispensable for SAA-induced NO production. Thus, it is likely that the Toll/IL-1R domain-containing, adaptor-inducing IFN-β pathway might be responsible for NO production [32 ]. Although a functional TLR4 is required for NO production, we do not know whether additional receptors are involved in optimal TLR4 signaling. It is known that SAA interacts with at least four distinct receptors: formyl peptide-receptor like-1, found in neutrophils [33 ]; Tanis, expressed on the hepatic tissue [34 ]; CD36 and lysosomal integral membrane protein-II analogous-1, a scavenger of high-density lipoprotein, found in various types of cells [35 ]; and receptor for advanced glycation end-products, a multiligand receptor [28 ]. Whether these receptors are essential or dispensable for optimal TLR4 signaling remains to be determined, but they could act like the CD14 molecule in the activation of TLR4 [36 ]. It is also important to determine whether SAA itself binds to the TLR4 receptor or is a carrier of other substances that activate the receptor. Another possibility is that SAA binds to LPS, and the complex SAA/LPS decreases the threshold of pathogen-associated molecular pattern detection, which in turn, enhances TLR4 signaling. This was demonstrated recently to occur with Hsp60, a putative, endogenous TLR4 ligand [37 ].
SAA, like other acute-phase proteins, is thought to play a role in innate defense by binding Gram-negative bacteria and acting as an opsonin [38 ]. However, it is becoming clear that SAA fulfills other immunological functions more related with cellular immunology that include its ability to associate with leukocytes, to induce the release of cytokines [4 5 6 , 27 , 39 ], to prime leukocytes for the release of reactive oxygen species [7 ], and to inhibit apoptosis [29 ]. Our finding that SAA triggers NO release adds a new, immunological function to this protein. Most of the biological effects mediated by SAA, such as augmentation of microbicidal activity [9 ] and angiogenesis [40 ], could be partly mediated by NO.
In conclusion, the activities of SAA are far from being restricted to opsonin-like activity, as our work indicates SAA could be a key element in the progression and/or resolution of the inflammatory process triggered by infection or injury.
Received April 3, 2007; revised November 8, 2007; accepted November 20, 2007.
|
|
|---|
, interleukin-1β, and interleukin-8 by human blood neutrophil Biochem. Biophys. Res. Commun. 268,405-408[CrossRef][Medline]
(TNF-
) in human neutrophils Immunol. Lett. 91,33-37[CrossRef][Medline]
independent oxidation of melatonin by macrophages J. Pineal Res. 34,69-74[CrossRef][Medline]
and TLR signaling pathways in macrophages Immunobiology 211,511-524[CrossRef][Medline]This article has been cited by other articles:
![]() |
R. D. Ye, F. Boulay, J. M. Wang, C. Dahlgren, C. Gerard, M. Parmentier, C. N. Serhan, and P. M. Murphy International Union of Basic and Clinical Pharmacology. LXXIII. Nomenclature for the Formyl Peptide Receptor (FPR) Family Pharmacol. Rev., June 1, 2009; 61(2): 119 - 161. [Abstract] [Full Text] [PDF] |
||||
![]() |
R. L. He, J. Zhou, C. Z. Hanson, J. Chen, N. Cheng, and R. D. Ye Serum amyloid A induces G-CSF expression and neutrophilia via Toll-like receptor 2 Blood, January 8, 2009; 113(2): 429 - 437. [Abstract] [Full Text] [PDF] |
||||
![]() |
N. Cheng, R. He, J. Tian, P. P. Ye, and R. D. Ye Cutting Edge: TLR2 Is a Functional Receptor for Acute-Phase Serum Amyloid A J. Immunol., July 1, 2008; 181(1): 22 - 26. [Abstract] [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||