Sepsis continues to be an ever present problem in the medical world with costs rising and a still unacceptable mortality despite new antibiotics or interventions (1, 2). Studies have shown that the number of severe sepsis cases has tripled from a prevalence of 0.5 to 1.5 cases per 1,000 persons (3). The rate of septic shock has increased more than sixfold in the United States in the past 11 years (4), although sepsis mortality has recently been stable (5).
One of the problems complicating the delivery of health care is the abundance of over-the-counter substances that people consume on a daily basis, many of which have unknown effects on the immune system. Caffeine is one such substance, consumed daily for its stimulatory effects in the form of coffee, tea, and energy drinks. In the United States, the average person will consume approximately 3 mg/kg per day of caffeine, mostly coming from coffee (6). Caffeine consumption by adults has been estimated to be 106 to 170 mg/d, when considering all sources (7). An updated survey published in 2014 of 37,602 people in the United States demonstrated that the daily caffeine intake was 165 mg/d (8). Caffeine alters the immune system by acting as an antagonist on all four adenosine receptors: A1, A2A, A2B, and A3 (9). Adenosine acting through its receptors is able to modulate cytokine and chemokine release and affect cell function (10).
Adenosine is released by tissues such as damaged blood vessels when a hypoxic environment is created. The purpose of this increase in extracellular adenosine is to inhibit the overactive immune cells and reduce bystander damage during the response to damage-associated molecular patterns (11). This effect was shown in a study where a hypoxic environment in the lung resulted in a lesser degree of lung injury when compared with mice who breathed air with higher partial pressures of oxygen (12). This group also showed the protective effect of adenosine communicating through the A2A receptor because mice genetically modified to knock out the receptor had worse lung injury after lipopolysaccharide exposure. Others have examined the role of adenosine receptors in polymicrobial septic challenges and have shown improved survival when specifically antagonizing the A2A or A2B receptors (11, 13).
Caffeine has been shown by 1 group to promote inflammation and liver injury in a model utilizing concanavalin A. In this study, mice were given a dose of caffeine with liver injury or allowed to imbibe caffeine on their own. Both methods resulted in worse liver injury and higher levels of proinflammatory cytokines (14). This same group also showed that caffeine at 100 mg/kg prevented the liver injury produced by concanavalin A, but this action was not performed through adenosine receptors. This protective effect in other models of liver injury has been documented through decreased transaminase levels (15). A recent epidemiology study showed that consumption of beverages containing caffeine increased lower urinary tract symptoms typically found in patients with urinary tract infections (16). Coffee consumption has also been linked to other deleterious health outcomes, such as increasing the risk of myocardial infarction (17), although a study in rats demonstrated that caffeine provided protection from ischemia-reperfusion injury (18).
If an important role of adenosine is to serve as a stop signal for the immune response and thus prevent bystander damage, then consumption of caffeine would place the immune system in a proinflammatory state by blocking the biological action(s) of adenosine. The current study tested the hypothesis that ingesting caffeine may have deleterious consequences on multiple aspects of the immune system in a polymicrobial sepsis model, which may result in decreased survival.
MATERIALS AND METHODS
Adult female ICR mice (from Harlan-Sprague Dawley, Inc, Indianapolis, Ind) were used. Mice were acclimatized to our housing room for at least 5 days before surgery. Animals were housed in a temperature- and humidity-controlled room with 12-h light–12-h dark diurnal cycle. They were provided food and water ad libitum for the entire duration of the experiment. The experiments were approved by Boston University Animal Care and Use Committee.
Caffeine (anhydrous) was purchased from MP Pharmaceuticals (Solon, Ohio).
Sepsis and blood sampling
Bacterial peritonitis was induced in mice using our standard model of cecal ligation and puncture (CLP) using our previous methods (19) based on modifications of the original description of the procedure (20). Anesthesia was induced with 5% isoflurane and maintained with 3% isoflurane. The mouse abdomen was washed with chlorhexidine, and a midline incision was made. The cecum was identified, extruded from the abdomen, and ligated at one-third the distance from the ileocecal valve to the end of the cecum. The cecum was punctured twice with an 18-gauge needle that was inserted from the tip of the cecum toward the ligated portion; a small amount of stool was extruded to ensure the patency of the perforations. The cecum was then returned into the abdomen, and the abdominal wall was closed with 4-0 silk sutures. The skin was the closed using Vetbond (3M, St Paul, Minn). The mouse was allowed to recover in a clean cage. One milliliter of normal saline containing 0.05 mg/kg of buprenorphine and 25 mg/kg of imipenem was given via the subcutaneous route immediately postoperatively. Buprenorphine was given for a total of four doses spaced 12 h apart. Imipenem was given for a total of 5 days in lactated Ringer’s solution with 5% dextrose. Blood samples of 20 μL in volume were collected at 6 and 24 h after CLP via facial vein sampling using a 23-gauge needle. Blood was placed into 180 μL of 1× phosphate-buffered saline with EDTA and centrifuged for 5 min, and the supernatants were stored at −20°C until they were ready to be analyzed.
Mice were divided into two experimental groups. For the first experimental group, mice received a single dose of 20 mg/kg of caffeine included in the first dose of pain medication and antibiotics, which were given subcutaneously. The controls received normal saline in place of caffeine in a similar manner. In the second experimental group, mice had an osmotic pump placed subcutaneously the day before CLP, which gave the mice a constant infusion of caffeine (10 mg/kg per hour) for a total of 24 h. Normal saline was infused using the osmotic pumps in a separate group of control mice. Because the experiments were designed to evaluate the impact of caffeine on the acute phase of sepsis, mice were followed up for survival and vital signs measurements for 5 days. A group of mice from the single-dose caffeine group was killed at 24 h after CLP in order to perform peritoneal lavages. The absorption and pharmacokinetics of caffeine have been reported to be similar in animals and humans (21).
Osmotic infusion pump placement
Osmotic infusion pumps were purchased from Alzet (model 2001D; Cupertino, Calif). Mice were induced with 5% isoflurane, and anesthesia was maintained with 3% isoflurane. The mouse was laid on its abdomen, and the dorsal surface washed with chlorhexidine. A small incision was made, and a pocket created in the subcutaneous tissue. The osmotic pump was inserted into the pocket, and the skin was closed with Vetbond (3M).
Mice were given a mixture of ketamine/xylaxine intraperitoneally, and blood was collected from the retro-orbital venous plexus. The abdominal area was washed with chlorhexidine, and an incision was made parallel to the initial incision for CLP. The abdominal cavity was irrigated with a total of 10 mL of Hanks’ balanced salt solution containing EDTA. The lavage fluid was then processed for cell and bacterial counts and for cytokine analysis. The peritoneal cells collected were then centrifuged, red blood cells were lyzed with 3 mL of ACK lysis buffer (Invitrogen, Grand Island, NY), and cells were resuspended in Hanks’ balanced salt solution and counted using Beckman-Coulter particle counter model ZF (Coulter Electronics, Hialeah, Fla). Supernatants were stored at −20°C until they were analyzed.
Measurement of physiologic parameters
Heart rate, respiratory rate, pulse distension, and pulse oximetry were recorded using a mouse collar clip device from Starr Life Science (Allison Park, Pa). Readings were taken at baseline, 16 h after placement of osmotic infusion pump but prior to CLP, and then at 2, 6, and 24 h after CLP.
Survival was analyzed by log-rank analysis, and differences between groups were determined by Student t test using GraphPad (San Diego, Calif). To determine the impact of both caffeine and survival, a two-sample analysis of variance was used and analyzed with the GraphPad software. GraphPad was also used to determine the area under the curve for the receiver operator characteristic. Sample size calculation was performed with the G*Power3 software (22).
Acute caffeine administration does not alter sepsis survival
The first experiment sought to determine if acute exposure to caffeine would alter the mortality of polymicrobial sepsis induced by CLP. Caffeine was given subcutaneously to ensure uniform delivery of the full dose. Caffeine has been shown to have similar pharmacokinetics in humans and mice (21). It should be noted that these experiments were not designed to evaluate adenosine receptors, but rather to determine if a commonly consumed substance alters the inflammatory response. Because the initial immune response to a polymicrobial challenge during CLP develops within the first 6 h, caffeine was given immediately after surgery to block the adenosine anti-inflammatory signal (23). Caffeine administered at this time did not affect overall survival, nor did it change the time at which the deaths occurred as shown in Figure 1. Because there was no impact on the mortality, we sought to determine if caffeine would modulate aspects of the inflammatory response, in order to verify that an immunomodulatory dose had been given.
Caffeine increases chemokine production
Plasma levels of several inflammatory mediators were measured in blood collected 6 h after the onset of sepsis, a time when several proinflammatory and anti-inflammatory mediators are increased (19, 24). Interleukin 6 (IL-6) has been widely used as a marker of the severity of the inflammatory response in CLP (23, 25). The cytokine response is dramatically different in mice that live compared with those who die, and blood was collected without killing the mice. This allowed us to separate the mice into those who lived to day 5 compared with those who died by day 5. We measured cytokines in the peripheral blood, and in the first analysis, mice were grouped into survival status at day 5 (alive versus dead) and not treatment status (normal saline versus caffeine). As expected, mice that will die during the acute phase of sepsis have higher levels of IL-6, the CXC chemokine KC, and the anti-inflammatory cytokine IL-10 (Fig. 2). Even when the mice had been given caffeine, a receiver operator characteristic curve analysis showed that IL-6 levels predicted mortality.
These data were further analyzed to determine if caffeine administration given at the onset of sepsis would alter cytokine production, even though there was no impact on survival. Mice were divided into their status at day 5 (alive or dead) and whether they were given normal saline or caffeine. Caffeine administration did not alter the plasma levels of IL-6 or IL-10 (Fig. 3). A sample size analysis showed that 71 mice would need to be examined to show a significant difference in IL-6 in the normal saline versus caffeine mice in the dead group, and 156 mice would be needed to show a difference in IL-10 in the dead group. However, caffeine did cause a significant increase in KC among those mice who survived. Previous reports using a cancer cell line demonstrated that pretreatment with caffeine would decrease chemokine production (26), but our in vivo results did not confirm this finding. Although the in vivo studies did not replicate the in vitro studies, our results did show that the dose and route of administration of the caffeine did modulate the inflammatory response.
Local (peritoneal) inflammatory response
Cytokines were also measured in the peritoneal fluid 24 h after CLP to determine if caffeine would alter cytokine levels. Because the mice were killed, we could not separate the mice into alive and dead, as was done in Figure 3. Additional cytokines could be measured because of the larger sample volume obtained from the peritoneal lavage. The proinflammatory cytokines IL-6 and tumor necrosis factor, the chemokines KC and macrophage inflammatory protein 2, and the cytokine inhibitors IL-10 and IL-1 receptor antagonist were measured, and none of the peritoneal cytokine levels were different in those mice who received caffeine (Fig. 4).
We also evaluated whether caffeine had any effect on cell recruitment or bacterial growth at the nidus of infection. The total number of peritoneal cells and bacterial counts were quantified to see if cell recruitment or bacterial killing had been affected by administration of caffeine. While mice that received caffeine had more cells within the peritoneal cavity, this was not significantly higher. Sample size analysis showed that 57 mice would be needed to demonstrate a difference between the groups. Bacterial counts were found to be comparable between the groups (Fig. 5).
Caffeine delivered via infusion—second experimental protocol
Caffeine has a short half-life, and it was possible that a single dose of caffeine given at the initiation of sepsis would not replicate that in a frequent coffee drinker who develops an inflammatory insult. In order to achieve a sufficient and constant dose, we implanted osmotic infusion pumps subcutaneously that would deliver a dose of 10 mg/kg per hour for a total of 24 h. This dose is half of what the mice in the previous experimental set of mice received, but it is still considered a proinflammatory dose (14). The change in dose was necessary because of the constraint in the size of infusion pump that was able to be implanted in the mice. The pump was placed in the mice 16 h prior to performing CLP and delivered caffeine up to 6 h after CLP. Caffeine administered via an infusion pump at 10 mg/kg per hour for 24 h did not change the survival of CLP when compared with mice that received normal saline via the osmotic pump (Fig. 6). While there was a slight decrease in survival with caffeine infusion, a sample size analysis showed that more than 150 mice would need to be studied to show a statistically significant decline. The survival in the saline group appears to be better than that in the saline group in Figure 1, but the actual survival was not statistically significant.
Physiological effects of caffeine
Because caffeine did not alter mortality, we examined if caffeine, a stimulatory drug, would affect the hemodynamic parameters of the mice prior to CLP. We also examined if caffeine would alter the postseptic physiological response. The purpose of these studies was to demonstrate that administration of caffeine was sufficient to impact the host, so we did not attempt to reproduce the immunological parameters in the first study. Physiological parameters were measured prior to pump placement and 16 h later immediately before CLP was performed. Mice that received the infusion pump containing caffeine experienced a significant increase in their heart rate of 94 beats/min (Fig. 7, A and B). In fact, every mouse that received caffeine increased their heart rate, whereas the mice that received infusion pumps filled with normal saline did not have a change in their heart rate. After CLP, mice that received caffeine through the pump had smaller changes in hemodynamic parameters than did the normal saline mice. In addition to the initial increase in heart rate, caffeine prevented the early drop in blood pressure after CLP. Up to 24 h after CLP, caffeine-treated mice still had higher heart rates compared with normal saline mice (Fig. 7C).
With the consumption of caffeine on a daily basis by people, we sought to understand its effects on the septic response, because coffee drinkers may develop sepsis. In the age of antibiotic-resistant bacteria, it is important to elucidate the effect of caffeine on the immune system, especially with its interactions with the adenosine receptors (27, 28). These receptors have been shown to play important roles in antitumor research because blocking the A2AR improved tumor rejection in a mouse model (29).
Utilizing a CLP model of polymicrobial sepsis, we demonstrated that caffeine at doses typically ingested by normal individuals does not affect the survival during the acute phase of CLP. These results differ from others that have shown that antagonizing individual adenosine receptors, specifically A2A, has been shown to improve survival in CLP (30). This difference could be due the nature of selective versus global antagonism of adenosine receptors. In the peripheral blood, it appears that caffeine promotes a slight proinflammatory state as higher levels of the chemokine (KC) were present with equivalent levels of anti-inflammatory cytokines. Caffeine levels above what a typical patient would ingest were not utilized because these have been shown to act in pathways other than the adenosine receptors. Our data did show increases in heart rate after caffeine with minimal changes in inflammatory parameters. The ability to influence cardiovascular parameters but not immune parameters may relate to differences in adenosine receptors on different cell types. A limitation of the study is that only female mice were examined.
Our results do not agree with previous published works in which caffeine is mostly labeled as an anti-inflammatory substance at doses that are normally consumed by humans. These effects have been defined by decreased antibody production in murine splenocytes, decreased lymphocyte proliferation, and decreased proinflammatory cytokine production (31). Most of the models reviewed use a single type of challenge compared with our model of polymicrobial sepsis. To add to the discrepancies, one group administered caffeine to humans after lipopolysaccharide exposure and found no difference in cytokine profiles or measured end organ damage when comparing to placebo (32), whereas others have shown the deleterious effects of caffeine during ischemia reperfusion, again in humans (33).
Adenosine and caffeine have been both shown to impact the vascular system. Adenosine functions as a vasodilator, whereas coadministration of caffeine can block this effect (34). Our experiments utilizing caffeine resulted in improved hemodynamic parameters when compared with the mice that received only normal saline. These results agree with previous work using rats given caffeine after CLP, which showed improved cardiac function when compared with controls. However, this work also showed improved survival following CLP (35). The improved survival with caffeine in this report is probably due to the high mortality in the vehicle group.
Caffeine has been shown to be a nonselective adenosine receptor antagonist capable of exerting its effect at concentrations that typical coffee consumers will achieve. Adenosine has been shown to be protective in tissues during inflammatory responses with caffeine blocking this action and maintaining the inflammatory response to a particular stimulus. Data reported in the current article indicate that the typical caffeine intake in a human will have a small effect on inflammatory and physiologic parameters, but not survival in sepsis.
1. Angus DC, Linde-Zwirble WT, Lidicker J, Clermont G, Carcillo J, Pinsky MR: Epidemiology of severe sepsis in the United States: analysis of incidence, outcome, and associated costs of care. Crit Care Med
29 (7): 1303–1310, 2001.
2. Martin GS, Mannino DM, Eaton S, Moss M: The epidemiology of sepsis in the United States from 1979 through 2000. N Engl J Med
348 (16): 1546–1554, 2003.
3. Linde-Zwirble WT, Angus DC: Severe sepsis epidemiology: sampling, selection, and society. Crit Care
8 (4): 222–226, 2004.
4. Walkey AJ, Wiener RS, Lindenauer K: Utilization patterns and outcomes associated with central venous catheter in septic shock: a population-based study. Crit Care Med
41 (6): 1450–1457, 2013.
5. Stevenson EK, Rubenstein AR, Radin GT, Wiener RS, Walkey AJ: Two decades of mortality trends among patients with severe sepsis: a comparative meta-analysis. Crit Care Med
42 (3): 625–631, 2014.
6. Barone JJ, Roberts HR: Caffeine consumption. Food Chem Toxicol
34 (1): 119–129, 1996.
7. Knight CA, Knight I, Mitchell DC, Zepp JE: Beverage caffeine intake in US consumers and subpopulations of interest: estimates from the Share of Intake Panel survey. Food Chem Toxicol
42 (12): 1923–1930, 2004.
8. Mitchell DC, Knight CA, Hockenberry J, Teplansky R, Hartman TJ: Beverage caffeine intakes in the U.S. Food Chem Toxicol
63: 136–142, 2014.
9. Fredholm BB: Astra Award Lecture. Adenosine
receptors and the actions of caffeine. Pharmacol Toxicol
76 (2): 93–101, 1995.
10. Jacobson KA, Gao ZG: Adenosine
receptors as therapeutic targets. Nat Rev Drug Discov
5 (3): 247–264, 2006.
11. Belikoff B, Hatfield S, Sitkovsky M, Remick DG: Adenosine
negative feedback on A2A adenosine
receptors mediates hyporesponsiveness in chronically septic mice. Shock
35 (4): 382–387, 2011.
12. Hatfield S, Belikoff B, Lukashev D, Sitkovsky M, Ohta A: The antihypoxia-adenosinergic pathogenesis as a result of collateral damage by overactive immune cells. J Leukoc Biol
86 (3): 545–548, 2009.
13. Belikoff BG, Hatfield S, Georgiev P, Ohta A, Lukashev D, Buras JA, Remick DG, Sitkovsky M: A2B adenosine
receptor blockade enhances macrophage-mediated bacterial phagocytosis and improves polymicrobial sepsis survival in mice. J Immunol
186 (4): 2444–2453, 2011.
14. Ohta A, Lukashev D, Jackson EK, Fredholm BB, Sitkovsky M: 1,3,7-Trimethylxanthine (caffeine) may exacerbate acute inflammatory liver injury by weakening the physiological immunosuppressive mechanism. J Immunol
179 (11): 7431–7438, 2007.
15. He P, Noda Y, Sugiyama K: Suppressive effect of coffee on lipopolysaccharide-induced hepatitis in d-galactosamine-sensitized rats. Biosci Biotechnol Biochem
65 (8): 1924–1927, 2001.
16. Maserejian NN, Wager CG, Giovannucci EL, Curto TM, McVary KT, McKinlay JB: Intake of caffeinated, carbonated, or citrus beverage types and development of lower urinary tract symptoms in men and women. Am J Epidemiol
177 (12): 1399–1410, 2013.
17. Nawrot TS, Perez L, Kunzli N, Munters E, Nemery B: Public health importance of triggers of myocardial infarction: a comparative risk assessment. Lancet
377 (9767): 732–740, 2011.
18. Li XY, Xu L, Lin GS, Jiang XJ, Wang T, Lu JJ, Zeng B: Protective effect of caffeine administration on myocardial ischemia/reperfusion injury in rats. Shock
36 (3): 289–294, 2011.
19. Osuchowski MF, Craciun F, Weixelbaumer KM, Duffy ER, Remick DG: Sepsis chronically in MARS: systemic cytokine responses are always mixed regardless of the outcome, magnitude, or phase of sepsis. J Immunol
189 (9): 4648–4656, 2012.
20. Baker CC, Chaudry IH, Gaines HO, Baue AE: Evaluation of factors affecting mortality rate after sepsis in a murine cecal ligation and puncture model. Surgery
94: 331–335, 1983.
21. Fredholm BB, Battig K, Holmen J, Nehlig A, Zvartau EE: Actions of caffeine in the brain with special reference to factors that contribute to its widespread use. Pharmacol Rev
51 (1): 83–133, 1999.
22. Faul F, Erdfelder E, Lang AG, Buchner A: G*Power 3: a flexible statistical power analysis program for the social, behavioral, and biomedical sciences. Behav Res Methods
39 (2): 175–191, 2007.
23. Remick DG, Bolgos GR, Siddiqui J, Shin J, Nemzek JA: Six at six: interleukin-6 measured 6 h after the initiation of sepsis predicts mortality over 3 days. Shock
17 (6): 463–467, 2002.
24. Moitra R, Beal DR, Belikoff BG, Remick DG: Presence of preexisting antibodies mediates survival in sepsis. Shock
37 (1): 56–62, 2012.
25. Turnbull IR, Javadi P, Buchman TG, Hotchkiss RS, Karl IE, Coopersmith CM: Antibiotics improve survival in sepsis independent of injury severity but do not change mortality in mice with markedly elevated interleukin 6 levels. Shock
21 (2): 121–125, 2004.
26. Merighi S, Benini A, Mirandola P, Gessi S, Varani K, Simioni C, Leung E, Maclennan S, Baraldi PG, Borea PA: Caffeine inhibits adenosine
-induced accumulation of hypoxia-inducible factor-1alpha, vascular endothelial growth factor, and interleukin-8 expression in hypoxic human colon cancer cells. Mol Pharmacol
72 (2): 395–406, 2007.
27. Olivares J, Bernardini A, Garcia-Leon G, Corona F, B Sanchez M, Martinez JL: The intrinsic resistome of bacterial pathogens. Front Microbiol
4: 103, 2013.
28. Aminov RI: A brief history of the antibiotic era: lessons learned and challenges for the future. Front Microbiol
1: 134, 2010.
29. Sitkovsky M, Lukashev D, Deaglio S, Dwyer K, Robson SC, Ohta A: Adenosine
receptor antagonists: blockade of adenosinergic effects and T regulatory cells. Br J Pharmacol
153 (Suppl 1): S457–S464, 2008.
30. Nemeth ZH, Csoka B, Wilmanski J, Xu D, Lu Q, Ledent C, Deitch EA, Pacher P, Spolarics Z, Hasko G: Adenosine
receptor inactivation increases survival in polymicrobial sepsis. J Immunol
176 (9): 5616–5626, 2006.
31. Horrigan LA, Kelly JP, Connor TJ: Immunomodulatory effects of caffeine: friend or foe? Pharmacol Ther
111 (3): 877–892, 2006.
32. Ramakers BP, Riksen NP, van den Broek P, Franke B, Peters WH, van der Hoeven JG, Smits P, Pickkers P: Circulating adenosine
increases during human experimental endotoxemia but blockade of its receptor does not influence the immune response and subsequent organ injury. Crit Care
15 (1): R3, 2011.
33. Riksen NP, Zhou Z, Oyen WJ, Jaspers R, Ramakers BP, Brouwer RM, Boerman OC, Steinmetz N, Smits P, Rongen GA: Caffeine prevents protection in two human models of ischemic preconditioning. J Am Coll Cardiol
48 (4): 700–707, 2006.
34. Smits P, Lenders JW, Thien T: Caffeine and theophylline attenuate adenosine
-induced vasodilation in humans. Clin Pharmacol Ther
48 (4): 410–418, 1990.
35. Verma R, Huang Z, Deutschman CS, Levy RJ: Caffeine restores myocardial cytochrome oxidase activity and improves cardiac function during sepsis. Crit Care Med
37 (4): 1397–1402, 2009.