Background
Approximately 900,000 cases of sepsis occur annually in the United States, causing roughly 210,000 deaths and costing almost 17 billion dollars [
1]. The overwhelming inflammation that occurs along with infection during sepsis has been the target of several therapeutic interventions [
2]. Unfortunately, despite successful treatment in animal models, antibody neutralization of individual components of this inflammation has not proved beneficial for the majority of patients in clinical sepsis trials [
3].
Tissue hypoxia, as occurs in sepsis, enhances breakdown of adenosine triphosphate (ATP) to adenosine monophosphate (AMP), which is then dephosphorylated by the cytosolic 5'nucloeotidase to adenosine [
4]. Adenosine can bind to four G protein coupled receptors, A
1, A
2A, A
2B, and A
3. The A
2A adenosine receptor (AR) is present on inflammatory cells including neutrophils, mast cells, monocytes, macrophages, eosinophils, platelets, and T-cells, and is involved in anti-inflammatory activities [
5]. Activation of A
2A ARs results in an increase in cyclic AMP concentration in inflammatory cells which is increased further by concurrent type IV phosphodiesterase inhibitors. A
2AAR agonists decrease superoxide production in neutrophils, degranulation of neutrophils, TNF-α production by monocytes and macrophages, and neutrophil-endothelial cell adherence [
6]. Death occurs in mice deficient for
Adora2a, the A
2A AR gene, after exposure to Concanavalin A liver injury within 8 hours compared to complete survival in wild-type mice. Pro-inflammatory cytokines are present in higher concentrations in
Adora2a
-/- mice when compared to wild-type mice. Similar findings are observed in experimental animals exposed to LPS [
7].
To assess the broad applicability of A
2A AR agonists, specifically ATL313, in the treatment of experimental sepsis due to different infections we contrasted survival results in mice challenged with a gram-negative (
Escherichia coli) vs. gram-positive (
Staphylococcus aureus) pathogen or purified LPS. T cells are increasingly recognized as important in the pathogenesis of experimental and clinical sepsis [
8]. Cytokine expression, inflammation, and therefore outcomes may differ during experimental sepsis depending upon the mouse strain and its T cell repertoire. This may have implications for sepsis therapy including ATL313. Therefore, we used a mouse strain, C57BL/6, with a predominantly Th1 phenotype and a mouse strain, Balb/C, with a predominantly Th2 phenotype to see if there were differences in outcomes based on underlying T cell phenotypes [
9].
Sepsis starts as a process of intravascular inflammation mediated by pro-inflammatory cytokines/chemokines including TNF-α, MIP-1α, MCP-1, IFN-γ, and IL-17 as well as anti-inflammatory cytokines, e.g. IL-10 [
10]. Therefore, to better understand the underlying protective effect of A
2A AR agonists, we evaluated cytokines in animals undergoing experimental sepsis with and without the addition of ATL313. ATL313 is a hundred fold more selective for the A
2A AR than for the A
1 AR and twenty fold more selective than for the A
3 AR. Furthermore, ATL313 is more selective and has a longer half-life (approximately 30 minutes in rodents) than its A
2A AR agonist predecessors. The A
2A AR agonists are potentially useful therapeutic agents because, unlike nonspecific AR agonists, A
2A AR agonists do not induce hypotension [
11]. We also studied the peripheral blood of the animals to assess the impact of an A
2A AR agonist on circulating white blood cell concentrations.
Methods
Mice
Female C57BL/6 and BALB/c mice (≅20 g; Jackson Laboratories, Bar Harbor, ME) were housed at 68–72°F with a 12 h light/dark cycle, fed standard laboratory food and water ad libitum, and were kept under specific pathogen-free conditions. The protocol used in this study was approved by the Animal Care and Use Committee of the University of Virginia.
Reagents and drugs
LPS (E. coli O111:B4) was purchased from Sigma (St. Louis, MO). The A2A AR agonist, ATL313, (4-{3-(6-amino-9-(5-cyclopropylcarbamoyl-3,4-dihydroxytetrahydrofuran-2-yl)-9H-purin-2-yl)prop-2-ynyl}piperidine-1-carboxylic acid methyl ester) was supplied by Adenosine Therapeutics, LLC (Charlottesville, VA). Ketamine and xylazine were purchased from Vedco, Inc., (St. Joseph, MO). In all cases vehicle used was phosphate buffered saline (PBS).
Dose effect
The appropriate dose of ATL313 for survival studies was determined by a dose response curve where increasing doses of ATL313 (5–200 μg/kg) were administered every 6 hours to female BALB/c and C57BL/6 mice after the intraperitoneal (IP) injection of a lethal dose of O111:B4 LPS (20 mg/kg). A similar dose response curve was used to determine the optimal dose of LPS for BALB/c and C57BL/6 mice.
Survival studies
Female C57BL/6 mice were inoculated with 1 × 108 colony forming units (CFU) live K12, O26:B6 E. coli or 8 × 108 CFU S. aureus at t = 0 and inoculated with the A2A AR agonist, ATL313 (5 μg/kg), or PBS, at t = 8, and then every 6 hours spanning 48 hours. To determine an optimal dosing interval, ATL313 was administered every 12 hours spanning 48 hours for S. aureus inoculated mice. A dose of ceftriaxone (25 mg/kg; t = 8) was given to a subset of treated and control mice to create a total of four experimental groups (bacterial infection + PBS; bacterial infection + ATL313; bacterial infection + Ceftriaxone; bacterial infection + ATL313 + Ceftriaxone). Female BALB/c mice were also tested in the E. coli model (data not shown). Female BALB/c mice undergoing endotoxemia were injected IP with LPS one half hour prior to injection with ATL313 (5 μg/kg) or PBS which occurred at t = 0 and every 6 hrs thereafter, for a total of eight doses spanning 48 hours (N = 20–29 per group). To provide equal fluid resuscitation in the setting of sepsis, all injections were I mL in volume. Moribund mice were anesthetized using ketamine and xylazine and euthanized via cervical dislocation.
Time course studies
Female C57BL/6 and BALB/c mice were injected intraperitoneally (IP) with O111:B4 LPS and ATL313 or PBS as described above. Experimental and control animals were sequentially sacrificed at times 1, 2, 4, 8, and 24 hours after LPS injection. The mice were anesthetized with ketamine and xylazine and killed by cervical dislocation. Immediately prior to sacrifice the animal underwent cardiac puncture and ventricular blood was aspirated for white blood cell analysis via the Hemavet 850 veterinary multi-species hematology system (Drew Scientific, Inc., Oxford, CT) and future cytokine analysis.
Cytokine quantification
Cytokine investigations were carried out using a protein bead-based multiplex immunoassay system (Bio-Rad Laboratories, Hercules, CA). With this system we are able to measure IL-1α and β, TNF-α, IFN-γ, IL-2, IL-3, IL-4, IL-5, IL-6, IL-10, IL-12 p40 and p70, IL-17, KC, MIP-1α, and RANTES through use of pre-packaged bead arrays. This panel allows for evaluation of early and late appearing as well as pro- and anti-inflammatory cytokines.
Statistical analysis
Statistical comparisons of cytokine values were done by a 2 tailed Student's t-Test (Microsoft Excel software, Microsoft Corporation, Redmond, WA). Survival data were plotted and the survival curves compared with a log-rank test (GraphPad PRISM software, San Diego, CA). Data are displayed as means ± SEM unless otherwise stated. Differences were considered significant at p < 0.05.
Discussion
This study documents: 1) a survival benefit with administration of ATL313 during live E. coli and S. aureus sepsis (in conjunction with an antibiotic) and in LPS-induced sepsis; 2) a decrease of TNF-α, MIP-1α, MCP-1, IFN-γ, and IL-17 blood concentrations in animals receiving A2A AR agonists with a concordant increase in IL-10 concentrations; and 3) a relative increase in circulating peripheral blood leukocyte concentrations in mice receiving ATL313 when compared to controls. The first observation bears on the generalizability of our observations to different etiologies of sepsis, including both gram positive and gram negative infections, while the latter two emphasize the mechanism(s) of protection. While these results are encouraging, further work is needed to prove the efficacy of these compounds in clinical sepsis and other models of sepsis including cecal ligation and puncture.
In our murine model of LPS-induced sepsis, TNF-α blood concentrations rise rapidly and are profoundly decreased by the co-administration of an A
2A AR agonist, ATL313. We have previously shown that the administration of an A
2A AR agonist up to 24 hours after LPS challenge is still protective (i.e. improved survival) when TNF-α concentrations in blood have returned to baseline; therefore the benefit cannot be attributed to effects on TNF-α alone [
12]. Given the same argument, it is unlikely that the decrease in MIP-1α, MCP-1, IFN-γ or IL-17 completely explains the protective efficacy of A
2A AR agonists in experimental murine sepsis. However, the effect of A
2A AR agonists on multiple cytokines is encouraging, since the clinical presentation of sepsis in humans is highly variable with regard to the cytokine cascade, but other as yet undetermined targets may play a role as well.
Inoculation of experimental animals with TNF-α alone reproduces many of the fundamental pathophysiologic alterations typical of sepsis [
13]. In vivo, TNF-α alters endothelium and is a potent chemoattractant for neutrophils which contributes to the pathology of venous thromboses, arteriosclerosis, vasculitis, and disseminated intravascular coagulation [
14,
15]. By decreasing TNF-α expression, A
2A AR agonists may decrease neutrophil recruitment and therefore inflammatory damage to the endothelium. This conjecture is bolstered by the increased number of neutrophils in the circulation of mice treated with ATL313 after LPS challenge. Our findings are mirrored by in vitro and in vivo work which revealed that A
2A AR agonists decreased neutrophil extravasation and their release of oxidative and non-oxidative products in experimental gram negative bacterial meningitis and
S. aureus septic arthritis [
16,
17].
Cytokines can cause inflammation via upregulation of other pro-inflammatory cytokines in the case of TNF-α, MIP-1α and IL-17, modulation of T-cells by IFN-γ, and activation of monocytes by MCP-1 [
18‐
21]. These pro-inflammatory mechanisms result in recruitment and activation of neutrophils, NK cells, and macrophages which produce deleterious reactive oxygen species and lysosomal enzymes [
22‐
25]. TNF-α, MIP-1α, MCP-1, and IFN-γ have been measured in patients with sepsis and acute respiratory distress syndrome and are correlated with poor outcomes [
14,
26,
27]. Antibodies to MIP-1α, IL-17, and IFN-γ have ameliorative in vivo effects in animal sepsis studies [
28‐
32]. As in the case of TNF-α, A
2A AR agonists down-regulate the expression of MIP-1α, MCP-1, IFN-γ, and IL-17 with a concurrent survival benefit.
While pro-inflammatory cytokines such as TNF-α, MIP-1α, MCP-1, IFN-γ, and IL-17 increase the inflammatory response, anti-inflammatory cytokines such as IL-10 temper this response. IL-10 is produced primarily by Th-2 cells and decreases concentrations of TNF-α by degrading cytokine mRNA [
33]. IFN-γ production is thought to be inhibited by IL-10 in a more indirect manner through the suppression of IL-12 production [
34].
IL-10 reduces the release of TNF-α into the circulation which hinders the development of a systemic inflammatory response syndrome, and this correlates with an increase in survival [
35‐
38]. We have found that elevated IL-10 and decreased TNF in the first few hours of experimental sepsis is correlated with increased survival. It is possible that the decrease in pro-inflammatory cytokines occurs as a direct result of the increase in IL-10 when an animal is treated with ATL313. Given the changes in IL-10 expression in endotoxemic mice after treatment with an A
2A AR agonist in C57BL/6 mice, we studied survival in a predominantly Th2 T-lymphocyte populated mouse, e.g. BALB/c. Again, we showed a significant survival benefit in these animals treated with an A
2A AR agonist after live
E. coli or LPS O111:B4 injection.
Therefore, the benefit of ATL313 in these sepsis models is equally beneficial in mice with a Th1 or Th2 phenotype, perhaps implying that ATL313 and similar compounds exert their anti-inflammatory effect through modification of the innate immune system. We are actively pursuing this line of investigation. Additionally, targeting individual cytokines alone has not been clinically successful, but a more pluripotent approach such as is offered by A2A AR agonists may be more successful. The influence of A2A AR agonists on late mediators including triggering receptor expressed on myeloid cells (TREM-1) and High mobility group box 1 (HMGB1) should be investigated and is the subject of our ongoing research efforts.
Competing interests
WMS and JL have equity interests in Adenosine Therapeutics, LLC who provided the A2A Adenosine Receptor agonist compound, ATL313
Authors' contributions
CCM designed and performed all experiments and drafted the manuscript. ENM assisted in design and performance of some experiments. GL performed some experiments. TO assisted in design of some experiments. JL assisted in design of some experiments. WMS oversaw all elements of experimental design and execution and manuscript preparation.