br Adenosine as a mediator of
Adenosine as a mediator of procedures used to manage pain
The issue of caffeine Caffeine, from dietary sources, is perhaps the most widely consumed behaviorally active agent in the world (Fredholm et al., 1999). Initial characterization indicated caffeine had a higher affinity at A1−, A2A− and A2BRs (KD 2–20μM) than at A3Rs (KD 80–190μM), and pharmacological effects at dietary levels were primarily attributed to block of A1Rs and A2ARs (Fredholm et al., 1999, Karcz-Kubicha et al., 2003, Yang et al., 2009b). Additional effects (phosphodiesterase inhibition, block of GABAA receptors, Ca2+ release) occur at higher concentrations (Fredholm et al., 1999), but their role in caffeine actions at dietary levels is unclear. With respect to pain, caffeine (⩾100mg) is an adjuvant analgesic in humans when given in combination with common analgesics (acetylsalicylic acid, acetaminophen/paracetamol, ibuprofen) (Derry et al., 2014). In preclinical studies, caffeine produces intrinsic antinociception in several models at higher doses (⩾35mg/kg, but sometimes lower), and this results from block of A2ARs and A2BRs, as selective antagonists for these receptors (SCH58261 and PSB115, respectively) mimic the effect of caffeine (Sawynok, 2011). At lower doses, which lack intrinsic effects (3–10mg/kg), caffeine inhibits the antinociceptive effect of acetaminophen, amitriptyline, carbamazepine and oxcarbazepine, allopurinol, tramadol, levetiracetam and sumatriptan (Section Adenosine as a mediator of pharmacological antinociception). This effect results from inhibition of A1Rs, as it is generally mimicked by systemic administration of a selective A1R antagonist (DPCPX). Compartmental analysis reveals that A1Rs at both peripheral and/or spinal sites contribute to antinociception by several of these agents (Section Adenosine as a mediator of pharmacological antinociception). In more recent overviews of adenosine receptors, caffeine has been noted to have a similar affinity at all human adenosine receptors (Ki: A1R 10.7μM, A2AR 23.4μM, A2BR 33.8μM, A3R 13.3μM) (Fredholm et al., 2011). The affinity of adenosine for human receptors is similar at A1−, A2A− and A3Rs (Ki ∼100–300nM) but lower at A2BRs (Ki 15,000nM) (Fredholm et al., 2011). This suggests that the actions of caffeine might also reflect block of A3Rs, in addition to other subtypes that were the earlier focus of attention. Given that the EPZ005687 of A3Rs for pain is now recognized to be antinociception in several paradigms (Section Adenosine A3Rs and pain), it will be important to determine whether effects of caffeine on nociception also involve A3Rs by examining whether selective A3R antagonists can mimic the action of caffeine, and in which compartments such effects occur. In humans, caffeine is consumed chronically, and experimental protocols to deliver caffeine in the drinking water over 7–10days have been developed to better model this intake pattern (Yang et al., 2009a). Such protocols result in inhibition of antinociception by systemic amitriptyline (Esser et al., 2001, Liu et al., 2013a) and acetaminophen (Sawynok and Reid, 2012), and this mimics the effect of acute administration of caffeine. Furthermore, the caffeine drinking water protocol results in inhibition of electroacupuncture delivered to the SP6 acupoint in mice (Moré et al., 2013). These observations raise the possibility that dietary caffeine might interfere with the actions of some currently used analgesics, as well as acupuncture. Variations in caffeine consumption might account for differences in acupuncture responses between individuals, as well as between populations at different study sites, as there can be prominent differences in intake between countries (Fredholm et al., 1999, Moré et al., 2013). The effects of caffeine on multiple adenosine receptors indicate that caffeine intake may need to be monitored in clinical trials. Thus, there is exploration of A1R, A2ARs and A3R agonists and modulators as potential novel analgesics (Sections Adenosine A1Rs and pain and Adenosine A3Rs and pain), and caffeine intake could potentially impact on results. Given that >85% of adults in the United States consume caffeine regularly (Frary et al., 2005), and that prior caffeine consumption may alter receptor function (Yang et al., 2009a, Sousa et al., 2011), a pragmatic approach might be the most relevant. Such designs have been used in studies of methotrexate effects in rheumatoid arthritis, whereby participants were stratified into low-, medium- and high-caffeine intake levels. In one study, caffeine was reported to influence the clinical efficacy of methotrexate in newly recruited subjects (Nesher et al., 2003). However, subsequent studies examining those on existing methotrexate regimens found no differences in outcomes between different caffeine intake groups in rheumatoid arthritis (Benito-Garcia et al., 2006) and in psoriatic arthritis (Swanson et al., 2007). There are some data indicating that caffeine intake modifies baseline pain levels as detected by quantitative sensory testing and thermal threshold tests, and contributes to subsequent analgesic efficacy (Karunathilake et al., 2012). There are also gender differences in pain being recognized in human studies (Greenspan et al., 2007, Fillingim et al., 2009, Racine et al., 2012), and recent data indicate mechanistic differences, including involvement of glia and immune cells, in chronic pain between genders (Vacca et al., 2014, Sorge et al., 2015). Given that multiple ARs on glia are now implicated in mechanisms of antinociception (Sections Adenosine A1Rs and pain, Adenosine A2ARs and pain, Adenosine A2BRs and pain and Adenosine A3Rs and pain), and that there are some preclinical data indicating differences in caffeine effects in males and females (Yang et al., 2009b), gender is an additional factor that will require attention in future exploratory trials.