Acupuncture is an alternative treatment for wide spectrum chronic pain. However, its validity remains controversial due to the disputed efficacy assessed in various clinical studies. Moreover, variability amongst individuals complicates the predictability of outcome, which impedes the integration of acupuncture into mainstream pain management programs. In light of our previous finding that the analgesic effect of acupuncture is mediated by adenosine A1 receptor activation at the acupuncture point, we here report that in acute and chronic animal pain models, oral intake of caffeine, a potent adenosine receptor antagonist, interferes with acupuncture analgesia, even at a low dose. Local administration of caffeine at the acupuncture point was sufficient to eliminate the analgesic effect, dismissing the systemic action of caffeine. Such interference was reversible, as caffeine withdrawal fully restored the efficacy of acupuncture by the next day, and long-term exposure to caffeine did not alter A1 receptor expression at the acupuncture point. Combined, these data indicate that a trace amount of caffeine can reversibly block the analgesic effects of acupuncture, and controlling caffeine consumption during acupuncture may improve pain management outcomes.
Analgesia is one of the most widely accepted effects of acupuncture. However, acupuncture’s efficacy beyond the placebo effect still remains controversial, primarily because of an incomplete understanding of its biological basis. The credibility is further diminished by the high variability of the efficacy, ranging from complete pain relief to no effect at all. Thus, acupuncture is often used as a last treatment option, even though pain reduction as a result of acupuncture has been demonstrated in many human and veterinary patients, without adverse side effects that are often present with medications.
We previously showed that, in an animal model, the analgesic effect of acupuncture is mediated by a steep increase in extracellular adenosine levels at the acupuncture point and subsequent local activation of the adenosine A1 receptor1. The A1 receptors responsible for mediating analgesia are localized in the vicinity of the acupuncture point rather than at distant central nervous system. This is evident since local administration of the A1 receptor agonist, 2-Chloro-N6-cyclopentyladenosine (CCPA), at the acupuncture point also resulted in analgesia, while acupuncture applied to the contralateral leg to the site of pain failed to reduce pain1. Such a transient adenosine increase was also present with acupuncture in humans2.
Adenosine receptor activation during acupuncture analgesia is potentially impacted by caffeine, which is a potent adenosine receptor antagonist. Caffeine is present in a wide variety of foods and drinks, and it is estimated that around 90% of the US population consumes caffeine daily3. Even though the average half-life of caffeine in plasma is 4–5 hours, a detectable amount of caffeine stays in the body for more than 12 hours after drinking a cup of coffee4,5,6. Therefore, it is likely that a residual amount of caffeine could be present at the time of acupuncture treatment in those who ingest caffeine daily.
In our animal model, genetic removal of the adenosine A1 receptor completely abolished the analgesic effect of acupuncture1. While the inhibitory constant (Ki) of caffeine at the A1 receptor is reported to be 10–44 µM7, a cup of coffee would result in a peak caffeine concentration of ~10 µM in the plasma8. Thus, a low level of residual caffeine could still potentially interfere with the mechanism of adenosine receptor-mediated analgesia during acupuncture. Therefore, we examined whether caffeine consumption reduces the efficacy of acupuncture analgesia in both acute and chronic pain models. We also tested whether such inhibition could be reversed if caffeine was withdrawn.
Caffeine interferes with the analgesic effect of acupuncture in acute inflammatory pain
Acute inflammatory pain in the left hind limb ankle joints was induced by intra-articular administration of complete Freund’s adjuvant (CFA). Mechanical sensitivity was measured using the von Frey filament touch test 3 days after CFA injection (Fig. 1A). Mice that received drinking water without caffeine exhibited an increase in mechanosensitivity, defined as the percent of positive responses to the filament touch in a set of trials, from 8.1 ± 1.1% to 80.6 ± 3.1% in the ipsilateral leg to the inflammation (Fig. 1B). Similarly, increased mechanosensitivity was detected in a group of mice received drinking water containing 0.3 mg/ml of caffeine, which is comparable to 4–5 cups of coffee in a man per day9, as well as in another group that received 0.6 mg/ml of caffeine. Although all groups developed hyper-mechanosensitivity in the ipsilateral leg, the degree of response was dose-dependently decreased in the caffeine groups compared to the no caffeine group (p < 0.0001) (Fig. 1B). Following the mechanosensitivity measurement, animals received acupuncture treatment, in which an acupuncture needle was applied to ST36 acupuncture point of the ipsilateral leg for 20 min with gentle manipulation, and further measurements were taken to evaluate the effects of the treatment. The no caffeine group showed a 35.0 ± 3.5% reduction in sensitivity following acupuncture treatment (p < 0.0001 compared to Before acupuncture) (Fig. 1C,D). In contrast, the analgesic effect of acupuncture was not observed in either of the caffeine-drinking groups. Rather, acupuncture treatment increased the mechanosensitivity by 15.6 ± 4.5% in the 0.3 mg/ml caffeine group and 18.0 ± 5.0% in the 0.6 mg/ml caffeine group (p < 0.003 compared to Before acupuncture) (Fig. 1C,D). The responses to the filament touch in the contralateral limb were unaffected by caffeine (p = 0.13 compared to Before acupuncture; p = 0.94 among the three groups), suggesting that caffeine does not alter motor functions (Fig. 1E). The increase or decrease of mechanosensitivity was not due to the needle insertion or the stress of prolonged restraints during treatment, as sham acupuncture, in which the same acupuncture needle was inserted for 20 min without manual stimulation, did not alter the mechanosensitivity in all groups (−2.1 ± 4.5%, −4.8 ± 8.5% and −4.8 ± 5.9% change in No caffeine, 0.3 mg/ml caffeine and 0.6 mg/ml caffeine groups, respectively; p > 0.4 compared to Before sham acupuncture) (Fig. 1F). The amount of caffeine present in the plasma at the time of the acupuncture session was on average 2.3 ± 0.4 µg/ml in mice that received 0.3 mg/ml caffiene, and 3.5 ± 0.6 µg/ml in those that received 0.6 mg/ml caffeine (Fig. 1G).
Caffeine interferes with the analgesic effect of acupuncture in chronic pain in a dose dependent manner
As shown in Fig. 1, inhibition of acupuncture analgesia by caffeine was robust in the acute pain model. We therefore next tested the effect of lower caffeine doses in the chronic pain model. Knee osteoarthritis was induced by administering monosodium iodoacetate (MIA) in the hind limb knee joints of mice, in which chronic pain developed in 2.5 weeks. Mice were then given drinking water with either no or low caffeine (0.1, 0.2, or 0.3 mg/ml) (Fig. 2A). At 3 weeks after pain induction, all four groups showed an increase in mechanical sensitivity (Fig. 2B). Similar to the CFA-induced ankle joint pain, MIA-induced knee arthritis pain was attenuated by acupuncture treatment (42.9 ± 5.2% reduction, p < 0.0001 compared to Before acupuncture) (Fig. 2B,C). A low dose of 0.1 mg/ml caffeine in the drinking water resulted in a small but detectable caffeine level in the plasma (0.6 ± 0.1 µg/ml) (Fig. 1F). Even at this low concentration, the analgesic effect of acupuncture was blocked (−8.3 ± 6.1% change, p = 0.19 compared to Before acupuncture, p < 0.001 compared to After acupuncture in No caffeine group) (Fig. 2B,C). In the 0.2 and 0.3 mg/ml caffeine groups, acupuncture appeared to worsen the pain, with mechanosensitivity enhanced by 16.7 ± 13.9% (p = 0.30 compared to Before acupuncture) and 20.0 ± 6.2% (p = 0.03 compared to Before acupuncture), respectively (Fig. 2B,C). As in the acute pain model, sham acupuncture treatment did not alter the mechanosensitivity (3.1 ± 3.8% change, p = 0.42 compared to Before sham acupuncture) (Fig. 2C). Changes in mechanosensitivity after acupuncture plotted against the plasma caffeine levels demonstrate a dose-dependent effect, with an inhibitory concentration (IC50) calculated as 0.51 µg/ml (R2 = 0.985) (Fig. 2D).
The presence of caffeine at the acupuncture site abolishes the analgesic effect of acupuncture
A previous report demonstrated that the adenosine A1 receptor responsible for acupuncture-induced anti-nociception is located at the acupuncture point1. However, orally consumed caffeine also acts on adenosine receptors present in other tissues, including central nervous and cardiovascular systems. Therefore, it is unclear whether caffeine blocks activation of A1 receptors in the vicinity of the acupuncture point, or if there is a certain systemic effect of caffeine that suppresses acupuncture analgesia. To address this question, we next administered caffeine at the acupuncture point instead of oral administration (Fig. 3A). On the first day of pain measurement, mice in the no caffeine condition were given acupuncture treatment to confirm the presence of the analgesic effect (−36.9 ± 4.4% change by acupuncture) (Fig. 3A, B). As previously reported, acupuncture analgesia was temporal1, and by the next day the mechanosensitivity returned to pre-acupuncture levels (80.1 ± 3.0% to 78.1 ± 2.1%) (Fig. 3B). On the second day of pain assessment, a small volume of high concentration caffeine (100 µM; 5 µl) was injected to the acupuncture point prior to the acupuncture session. The administered concentration was high to ensure complete local blockage of adenosine receptors, without causing neurotoxicity10. The presence of caffeine at the acupuncture point abolished the analgesic effect (−4.4 ± 4.1% change, p = 0.30 compared to Before acupuncture, p < 0.0001 compared to After acupuncture on the day before) (Fig. 3B,C). In contrast, prior administration of vehicle instead of caffeine did not disrupt the acupuncture analgesia (−49.4 ± 3.9% change, p < 0.001 compared to Before acupuncture) (Fig. 3B,C). In situ administration of caffeine alone without subsequent manual acupuncture did not alter the mechanosensitivity of the animals (−8.3 ± 6.3% change, p = 0.23 compared to Before caffeine administration) (Fig. 3B insert). The effect of local administration of caffeine was not systemic, as caffeine was not detected in the plasma of these animals (p = 0.14 compared to No caffeine group) (Fig. 3D). Therefore, the local presence of caffeine was sufficient to eliminate the efficacy of acupuncture, dismissing the involvement of caffeine’s action in the central nervous and cardiovascular systems.
Caffeine habituation does not alter the mechanism of acupuncture-induced analgesia
Persistent inhibition of A1 receptors by long-term consumption of caffeine might decrease A1 receptor expression at the acupuncture point, which may influence the efficacy of acupuncture. Therefore, we next measured the expression of the A1 receptor gene, Adora1, both in the muscle tissue around the acupuncture point (acupoint) and in the sciatic nerve dorsal root ganglia (DRG), which send axons underneath the acupoint. Induction of pain by CFA administration as well as acupuncture treatment did not alter Adora1 expression at either site (Fig. 4A). In mice drinking 0.3 mg/ml of caffeine, Adora1 expression was also unaffected (Fig. 4A). At the high dose of caffeine (0.6 mg/ml), a slight non-significant decrease in Adora1 expression was detected in the DRG (p = 0.21), but there was no change at the acupoint (p = 0.75) (Fig. 4A). Thus, inhibition of acupuncture analgesia by caffeine was unlikely due to decreased A1 receptor expression at the DRG or the acupoint.
Next we tested whether the analgesic effects of acupuncture could be restored following caffeine removal. Caffeine is known to induce mild physical dependency associated with sustained physiological and molecular changes after daily consumption11,12,13. Thus, certain long-term changes caused by caffeine habituation may influence acupuncture treatment. First, mechanosensitivity in mice in the caffeine conditions (0.3 and 0.6 mg/ml) were assessed before and after acupuncture. We then replaced the drinking water containing caffeine with normal water and re-evaluated the sensitivity in the same mice on the second day following caffeine withdrawal (Fig. 4B). After 24 hours, there were no detectable levels of caffeine present in the plasma (p = 0.37 for 0.3 mg/ml and p = 0.34 for 0.6 mg/ml caffeine groups compared to No caffeine group) (Fig. 3D). Caffeine withdrawal increased mechanosensitivity before acupuncture treatment (51.4 ± 7.4% to 84.3 ± 5.3% and 31.3 ± 7.7% to 72.5 ± 5.9% in 0.3 and 0.6 mg/ml Caffeine withdrawal groups, respectively, p < 0.001) (Fig. 4C), to the level similar to mice who have never been exposed to caffeine (80.6 ± 3.1% as shown in Fig. 1C Before acupuncture in No caffeine group, p > 0.1). Full restoration of the analgesic effect of acupuncture was observed in both groups after caffeine withdrawal (−57.1 ± 6.1% and −46.2 ± 6.0% change in 0.3 and 0.6 mg/ml Caffeine withdrawal groups, respectively, p < 0.001 compared to Before acupuncture), whereas these same mice exhibited slight increases in hypersensitivity by acupuncture in the caffeine conditions on the previous day (+14.3 ± 5.7% and + 27.5 ± 8.8% change in 0.3 and 0.6 mg/ml With caffeine groups, p = 0.05 and p = 0.02 compared to Before acupuncture, respectively) (Fig. 4C,D). As with the previous experiments, sham acupuncture treatment after caffeine withdrawal did not alter the mechanosensitivity (2.4 ± 7.7% and −2.1 ± 9.1% change in 0.3 and 0.6 mg/ml Caffeine withdrawal groups, p = 0.77 and p = 0.83 compared to Before sham acupuncture, respectively) (Fig. 4D). Taken together, the presence of caffeine itself was responsible for inhibiting acupuncture analgesia even at low concentrations, and any enduring effect after the clearance of caffeine from the body does not affect the anti-nociceptive mechanism of acupuncture.
In the present study, we adapted both acute ankle joint and chronic knee joint inflammatory pain models and successfully demonstrated that manual acupuncture treatment induced acute analgesia. Unlike various clinical studies where placebo effects play a major role in overall efficacy of acupuncture treatment, the animals in our experiements had no prior expectations toward the possible outcomes of the procedures. Moreover, our animal models are free from complications in interpreting the effect of caffeine on anlagesia, which can arise from various side effects associated with changes in caffeine consumption in humans. Our data clearly showed that caffeine consumption, ranging from low to high amounts (equivalent to 1–10 cups of coffee a day in humans), interfered with the analgesic action of acupuncture9, 14. Systemically administered caffeine acts on adenosine receptors located in various tissues, including in the central nervous and cardiovascular systems, and modulation of these systems changes pain perception15. Caffeine also has effects unrelated to adenosine receptors such as inhibition of ryanodine receptors16, inositol trisphosphate receptors17, phosphodiesterase18, and acetylcholinesterase19. Caffeine at high doses has been reported to have anti-nociceptive effects in hot-plate and formalin tests, which affect central processing20, and was also reported to inhibit acupuncture in postoperative pain21. However, our data revealed a new action of caffeine – inhibition of acupuncture-induced analgesia that is mediated by antagonizing peripheral adenosine receptors at the acupuncture point, even at low doses. First, systemic caffeine did not alter the response to mechanical sensation given to the contralateral leg without pain (Fig. 1E), suggesting that caffeine does not alter motor function. Consistently, the dose of caffeine required to affect locomotor activity was much higher (ten times or more) than those used in the present study14, 22, 23. Second, a small volume of caffeine administered at the acupuncture point, without raising systemic caffeine concentration, also completely blocked the effect of acupuncture (Fig. 3), indicating that caffeine at the acupuncture point without systemic presence is sufficient to abolish the analgesic effect of acupuncture. Our previous data demonstrated that acupuncture induces analgesia through local activation of the adenosine A1 receptor1, and thus caffeine at the site of acupuncture likely blocks receptor activation. Lastly, caffeine withdrawal restored acupuncture’s anti-nociceptive effect (Fig. 4B-D), supporting the notion that the A1 receptor was competitively antagonized by caffeine at the acupuncture site, but was not indirectly inhibited by irreversible systemic changes.
We observed that caffeine’s IC50 was 0.51 µg/ml (2.6 µM) (Fig. 2D), which in a human is less than the predicted caffeine concentration after drinking one cup of coffee8, 22. This value is lower than the reported Ki of caffeine (44 µM)7, and it is presumed to block less than 15% of A1 receptors22. Previously, we estimated that the extracellular adenosine concentration at the acupuncture point in mice peaked around 1.5 µM, much higher than the Ki of the A1 receptor (0.1 µM)2, which indicates that the amount of systemic caffeine is not enough to completely block A1 receptor activation. Therefore, it was surprising that such a small amount of caffeine resulted in the complete inhibition of acupuncture analgesia. Perhaps caffeine accumulates in the local tissue where targeted adenosine A1 receptors reside, or a partial blockage of A1 receptors is sufficient to affect the sensory neuronal activity and inhibit the effect of acupuncture. Indeed, a low dose caffeine was reported to induce significant biological consequences such as cerebral glucose utilization24. Shen et al. reported that Gq protein-coupled rhodopsins control thermal sensory neurons via TRPA1 channel, which enables fine-tuning thermotaxis behaviors to respond to small changes in temperature less than 1 °C25. Perhaps the adenosine A1 receptor, also a Gq protein-coupled receptor, might similarly control nociceptive neuronal signaling. Partial blockage of adenosine receptors in peripheral nociceptive neurons may translate to a significant change in nociceptive signals and central pain perception.
In caffeine-drinking groups, we observed a lower degree of hyper-mechanosensitivity compared to the no caffeine group after pain induction (Figs 1B,C and 2B). This observation is consistent with the previously reported anti-nociceptive action of high doses caffeine in rats20, 26. It is unlikely that this reduction in the pre-acupuncture mechanosensitivity can be attributed to caffeine suppressing the development of inflammation at the site of the pain, because previous reports found systemic single administration of high doses caffeine can acutely suppress the pain while the development of inflammation was not suppressed20, 26. Consistently, in our study the withdrawal of caffeine elevated mechanosensitivity to the levels in the no caffeine group by next day (Fig. 4C Before acupuncture in Caffeine withdrawal groups compared to Fig. 1C Before acupuncture in No caffeine group). Orally administered caffeine distributes throughout the whole body including the acupuncture point. However, peripheral mechanism mediated by orally administered caffeine acting at the local acupuncture point was not involved in the reduction of basal mechanosensitivity, because in situ administration of a high dose of caffeine (100 µM), which elevates caffeine concentration at the acupuncture point but not in the central and cardiovascular systems, failed to decrease basal mechanosensitivity (Fig. 3B insert, Before versus After caffeine in situ administration). It has been well described that moderate to high doses of caffeine influences components of the central and cardiovascular systems, including blood pressure, vascular tone, alertness, sensory responses, and working memory. For example, 0.3 mg/ml caffeine in the drinking water affected cognitive function in mice9; ≥10 mg/kg of caffenine altered rodent locomotor activity14; and ≥200 mg enhanced attention in humans13, 27. Accordingly, in our study a potent reduction in basal mechanosensitivity was limited to the mice that received the drinking water containing the high doses of caffeine (0.3 and 0.6 mg/ml) (Fig. 1B, Before acupuncture), implying the involvement of systemic changes in the central and cardiovascular systems. Interestingly, with the presence of a high-dose of caffeine, acupuncture increased the mechanosensitivity, which nontheless did not exceed the level of hypersensitivity observed before acupuncture in mice that did not recieve caffeine in the drinking water (Figs 1C and 2B, Before acupuncture in No caffeine group versus After acupuncture in 0.3 and 0.6 mg/ml caffeine groups). Sawynok et al. reported that the anti-nociceptive action of caffeine is mediated by norepinephrine signaling in the locus coeruleus and spinal cord20. In contrast, several studies reported that acupuncture attenuates norepinephrinergic transmission28, 29. Thus, acupuncture may counteract the anti-nociceptive action of caffeine, resulting in an apparent increase of mechanosensitivity.
Caffeine is known to induce mild physical dependency after regular consumption11,12,13, and alter gene expression30, 31. Therefore, it is possible that caffeine influences gene expression of A1 receptors, such that there may be changes in the sensitivity to A1 receptor-mediated analgesia. We found no significant alterations in Adora1 expression in both sciatic nerve DRG and skeletal muscle tissue at the acupuncture point (Fig. 4A). The A1 receptor is expressed only in vascular endothelial cells in skeletal muscle32, thus these are not likely the targets of the pain suppression by acupuncture. Moreover, tissue injury elicited increases in Adora2b and Adora3, but not Adora1, in skeletal muscle33. However, we cannot exclude the possibility that A1 receptors expressed in connective tissue fibroblasts or Adora1 gene accumulation at peroneal nerve terminals around ST36 embedded inside the muscle tissue are involved in the analgesic mechanism. Nevertheless, we failed to detect a tissue-level change in Adora1 expression by pain induction or caffeine treatment, indicating that A1 receptors that mediate acupuncture analgesia remained unchanged even after caffeine habituation. This conclusion was supported by the full restoration of acupuncture analgesia following removal of caffeine (Fig. 4C,D), in accordance with a previous observation demonstrating the full restoration of central depressant effects of adenosine agonists 24 hours after caffeine withdrawal34. These studies indicate that caffeine habituation at the doses used in the present study does not alter adenosine receptor signaling. Furthermore, the recovery of the analgesic effect showed that the underlying mechanism of acupuncture analgesia was not influenced by any other systemic or local alteration caused by chronic caffeine consumption as well as any adverse effect caused by withdrawal after habituation.
Given that anti-nociceptive effect of acupuncture was (1) suppressed by caffeine at the acupuncture point, (2) counteractive to the nociceptive action of high dose caffeine intake, and (3) fully recovered after caffeine withdrawal, we collectively identified caffeine as a key manipulator of efficacy of acupuncture analgesia. Although the analgesic effect of acupuncture is its most clinically established effect35, 36, several clinical studies have reported a lack of benefit of acupuncture compared to placebo37,38,39. Since no clinical studies on the effect of acupuncture have taken a patient’s daily caffeine consumption into consideration, caffeine intake may have been contributing to the contradictory data in the various clinical studies37, 40,41,42. Moreover, severity of caffeine habits can be an important factor when incorporating acupuncture into treatment plans for chronic pain patients. Our data suggest that even a small residual amount of caffeine in the body, hours after a consumption of a cup of coffee, could potentially reduce the efficacy of acupuncture. For some patients, a temporal withdrawal from caffeine may be associated with undesirable side effects. However, patients who stop consuming caffeine regularly may expereince substancially enhanced efficacy of acupuncture therapy. Clinical studies need to confirm if caffeine interferes with acupuncture-induced analgesia and how long patients need to stop taking caffeine to improve acupuncture effects.
Animals, caffeine administration, and experimental models
C57BL/6 male and female mice (Jackson Laboratory) aged 8–12 weeks were used in this study. For oral administration of caffeine, mice were treated with drinking water supplemented with caffeine (Sigma-Aldrich, St. Louis, MO) at concentrations of 0.1–0.6 mg/ml, whereas mice in the no caffeine group received normal drinking water. In some experiments, caffeine (5 µl, 100 µM) or vehicle (5 µl of saline) was administered at the ST36 acupuncture point at a depth of 1.5 mm using a gastight syringe with a 33 G needle instead of oral administration.
Following one-week treatment of the drinking water without or with caffeine, acute inflammatory pain was induced by an intra-articular administration of complete Freund’s adjuvant (CFA, 10 µl, Sigma-Aldrich, St. Louis, MO) into the left hind limb ankle joint using a syringe with 33 G needle43, 44. Mice were further treated with either normal or caffeine-containing drinking water for 3–4 days (Figs 1A and 4A,B). Chronic inflammatory pain was induced by an intra-articular administration of monosodium iodoacetate (MIA, 20 mg/ml, 5 µl, Sigma-Aldrich, St. Louis, MO) into the left hind limb knee joint using a syringe with a 33 G needle45,46,47,48. All mice were returned to standard care for 3 weeks. During the last 3–4 days, the drinking water was changed to either normal or caffeine-containing water (Fig. 2A). No significant weight loss or gain was observed in animals that received caffeine, pain induction, or acupuncture treatment. The animals were housed in a room with a 12 h day/night cycle. All experiments with live animals were performed in accordance with institutional guidelines at the University of Rochester. All procedures were approved by the University Committee on Animal Resources (UCAR-2014-038).
Pain evaluation and manual acupuncture
Mechanical sensitivity was measured using the Semmes-Weinstein von Frey Aesthesiometer (Stoelting Co, Wood Dale, IL) touch test, as previously described with slight modification1. A flowchart of the experimental procedure is shown in Supplemental Materials. Before beginning the measurement session, animals were acclimated to the acrylic chamber (IITC Life Science, Woodland Hills, CA) for 60 min per day for 2 days. Baseline mechanosensitivity was measured twice on different days to validate the stable, low mechanosensitivity in all mice prior to pain induction administered on the day of the second measurement. The average of the two measurements was recorded as the baseline sensitivity value. Pain-induced hyper-mechanosensitivity was evaluated 3 days (CFA) or 3 weeks (MIA) after administration. For some experiments, measurement sessions were repeated on two consecutive days (3 and 4 days after CFA injection). In the experiments testing local caffeine administration (Fig. 3A), mechanical sensitivity was first measured twice, before and after the acupuncture treatment, without caffeine on the first day. On the next day, mechanical sensitivity in the same mice was re-evaluated twice but with administration of caffeine at the acupuncture point 5 min prior to the acupuncture. As a control group, vehicle (5 µl of saline) was administered prior to the acupuncture instead of caffeine. Similarly, caffeine withdrawal experiments (Fig. 4B) were conducted on two consecutive days during which mechanical sensitivity was evaluated twice, before and after acupuncture treatment, on each day with caffeine treatment (first day) and 24 hours after exchange to normal drinking water (second day).
The evaluation was performed by gently applying a thin filament of 0.04 g force to the hind paw of the animals, and the rapid retraction or tapping of the foot was counted as a positive response1, 49 (see Supplemental Materials). A total of six or ten trials were given to a paw in one measurement session with an interval of at least 5 min between each trial. Positive responses in the all trials were recorded, and the data are represented as the percent of positive responses. Percent change in mechanosensitivity by acupuncture treatment was obtained by subtracting values of percent responses before acupuncture from values after acupuncture for each animal. The evaluator of mechanosensitivity did not participate in the behavioral experimental design, oral caffeine assignments, local caffeine administrations, and acupuncture procedure, and thus had no prior knowledge of the animal groups. All behavioral measurements were done at the same time of the day during the 12 hour light cycle.
For acupuncture treatment, animals were placed in a restrainer with the hind leg ipsilateral to the inflammation, secured and loosely stretched (see Supplemental Materials). An acupuncture needle (0.20 mm; DBC Spring, Lhasa OMS, Weymouth, MA) was gently inserted to the ST36 Zusanli point at a depth of 1.5 mm, then slowly rotated for 1 min every 3–4 min for a 20 min session while the mice were awake1. For sham treatment, the acupuncture needle was inserted the same way, but then left untouched for 20 min without the rotational manipulation. The animals were immediately transferred back to the chamber and acclimated for at least 15 min before post-treatment mechanosensitivity measurements.
Measurement of caffeine in plasma
Blood samples collected by submandibular bleeding were immediately mixed with 10 mM EDTA, and plasma was isolated by centrifugation at 1,000 g for 10 min. Plasma caffeine concentrations were quantified using an enzyme linked immunosorbent assay (ELISA) kit (Neogen, Lansing, MI) according to the manufacturer’s instructions. Standard curves were generated from serial dilutions of caffeine reference solution (1.0 mg/ml caffeine, Cerilliant, Round Rock, TX) in assay buffer containing mouse plasma. ELISA plates were measured at 650 nm, or 450 nm after supplemental addition of 1 N HCl, with a microplate reader (Benchmark Plus, Bio-Rad, Hercules, CA). Dose-response curve of plasma caffeine versus acupuncture-induced changes in mechanosensitivity was calculated with Prism (GraphPad Software, La Jolla, CA) using Hill slope.
Real-time PCR for adenosine receptor A1 gene expression
Tissues were harvested from five animal groups: (1) Naïve, (2) four days after CFA administration, (3) same as group 2 but with acupuncture treated on the third day, (4) same as group 3 but with 0.3 mg/ml caffeine oral administration, and (5) same as group 3 but with 0.6 mg/ml caffeine oral administration. Groups 4 and 5 were chronically treated with systemic caffeine as described in the “caffeine administration” section above. Ipsilateral L3-L5 dorsal root ganglia (DRG) and tissue surrounding ST36 acupuncture point (roughly 3 × 3 × 3 mm3 of tibialis anterior and extensor muscles attached to tibia, without dermis) were dissected from the mice following perfusion fixation with RNAlater solution (Ambion, Foster City, CA) to preserve RNA integrity. The isolated tissues were submerged in RNAlater solution overnight at 4 °C and then, stored at –80 °C. Tissues were homogenized using FastPrep-24 5 G with CoolPrep Adapter (MP Biomedicals, Santa Ana, CA), which separates tibia from acupuncture point tissue samples. Total RNA was extracted with RNeasy Plus Universal Kit (Qiagen, Valencia, CA). Expression of the adenosine A1 receptor gene, Adora1, was assessed using TaqMan Gene Expression Assays (Mm01308023_m1), and the expression levels were normalized to multiple reference genes50: glyceraldehyde-3-phosphate dehydrogenase (Gapdh), acidic ribosomal protein (Arbp) and hypoxanthine-guanine phosphoribosyl transferase (Hprt) for DRG, and 18 S ribosomal RNA (Rn18S), beta-actin (ActB), Gapdh and Hprt for Acupoint, before calculating relative expression.
One-way ANOVA with Tukey’s multiple comparison procedure was used to compare differences between groups. Paired t-test was used to compare before and after acupuncture within each group. The significance level was set at 0.05 for all comparisons. The analyses were implemented with SAS 9.4 (SAS Institute Inc., Cary, NC). The complete statistical data are presented in Supplemental Materials.
Goldman, N. et al. Adenosine A1 receptors mediate local anti-nociceptive effects of acupuncture. Nat Neurosci 13, 883–888 (2010).
Takano, T. et al. Traditional acupuncture triggers a local increase in adenosine in human subjects. J Pain 13, 1215–1223 (2012).
Frary, C. D., Johnson, R. K. & Wang, M. Q. Food sources and intakes of caffeine in the diets of persons in the United States. J Am Diet Assoc 105, 110–113 (2005).
Lelo, A., Birkett, D. J., Robson, R. A. & Miners, J. O. Comparative pharmacokinetics of caffeine and its primary demethylated metabolites paraxanthine, theobromine and theophylline in man. Br J Clin Pharmacol 22, 177–182 (1986).
Newton, R. et al. Plasma and salivary pharmacokinetics of caffeine in man. Eur J Clin Pharmacol 21, 45–52 (1981).
Perera, V., Gross, A. S., Xu, H. & McLachlan, A. J. Pharmacokinetics of caffeine in plasma and saliva, and the influence of caffeine abstinence on CYP1A2 metrics. J Pharm Pharmacol 63, 1161–1168 (2011).
Muller, C. E. & Jacobson, K. A. Recent developments in adenosine receptor ligands and their potential as novel drugs. Biochim Biophys Acta 1808, 1290–1308 (2011).
Sawynok, J. Pharmacological rationale for the clinical use of caffeine. Drugs 49, 37–50 (1995).
Arendash, G. W. et al. Caffeine protects Alzheimer’s mice against cognitive impairment and reduces brain beta-amyloid production. Neuroscience 142, 941–952 (2006).
Kang, S. H., Lee, Y. A., Won, S. J., Rhee, K. H. & Gwag, B. J. Caffeine-induced neuronal death in neonatal rat brain and cortical cell cultures. Neuroreport 13, 1945–1950 (2002).
Griffiths, R. R. et al. Low-dose caffeine physical dependence in humans. J Pharmacol Exp Ther 255, 1123–1132 (1990).
Griffiths, R. R. & Woodson, P. P. Caffeine physical dependence: a review of human and laboratory animal studies. Psychopharmacology (Berl) 94, 437–451 (1988).
Smith, A. Effects of caffeine on human behavior. Food Chem Toxicol 40, 1243–1255 (2002).
Fredholm, B. B., Battig, K., Holmen, J., Nehlig, A. & Zvartau, E. E. Actions of caffeine in the brain with special reference to factors that contribute to its widespread use. Pharmacol Rev 51, 83–133 (1999).
Sawynok, J. & Yaksh, T. L. Caffeine as an analgesic adjuvant: a review of pharmacology and mechanisms of action. Pharmacol Rev 45, 43–85 (1993).
McPherson, P. S. et al. The brain ryanodine receptor: a caffeine-sensitive calcium release channel. Neuron 7, 17–25 (1991).
Wakui, M., Osipchuk, Y. V. & Petersen, O. H. Receptor-activated cytoplasmic Ca2+ spiking mediated by inositol trisphosphate is due to Ca2(+)-induced Ca2+ release. Cell 63, 1025–1032 (1990).
Choi, O. H., Shamim, M. T., Padgett, W. L. & Daly, J. W. Caffeine and theophylline analogues: correlation of behavioral effects with activity as adenosine receptor antagonists and as phosphodiesterase inhibitors. Life Sci 43, 387–398 (1988).
Pohanka, M. & Dobes, P. Caffeine inhibits acetylcholinesterase, but not butyrylcholinesterase. Int J Mol Sci 14, 9873–9882 (2013).
Sawynok, J., Reid, A. R. & Doak, G. J. Caffeine antinociception in the rat hot-plate and formalin tests and locomotor stimulation: involvement of noradrenergic mechanisms. Pain 61, 203–213 (1995).
More, A. O. et al. Caffeine at Moderate Doses Can Inhibit Acupuncture-Induced Analgesia in a Mouse Model of Postoperative Pain. J Caffeine Res 3, 143–148 (2013).
Daly, J. W. & Fredholm, B. B. Caffeine–an atypical drug of dependence. Drug Alcohol Depend 51, 199–206 (1998).
Nikodijevic, O., Jacobson, K. A. & Daly, J. W. Locomotor activity in mice during chronic treatment with caffeine and withdrawal. Pharmacol Biochem Behav 44, 199–216 (1993).
Nehlig, A., Lucignani, G., Kadekaro, M., Porrino, L. J. & Sokoloff, L. Effects of acute administration of caffeine on local cerebral glucose utilization in the rat. Eur J Pharmacol 101, 91–100 (1984).
Shen, W. L. et al. Function of rhodopsin in temperature discrimination in Drosophila. Science 331, 1333–1336 (2011).
Wu, W. P., Hao, J. X., Fredholm, B. B., Wiesenfeld-Hallin, Z. & Xu, X. J. Effect of acute and chronic administration of caffeine on pain-like behaviors in rats with partial sciatic nerve injury. Neurosci Lett 402, 164–166 (2006).
Einother, S. J. & Giesbrecht, T. Caffeine as an attention enhancer: reviewing existing assumptions. Psychopharmacology (Berl) 225, 251–274 (2013).
Wen, G., He, X., Lu, Y. & Xia, Y. Effect of Acupuncture on Neurotransmitters/Modulators. in Acupuncture therapy for neurological diseases: a neurobiological view (eds Xia, Y., Cao, X., Wu, G. & Cheng, J.) 120–142 (Tsinghua University Press; Springer, Beijing; Heidelberg, 2010).
Zhao, Z. L. et al. Hypothalamic Norepinephrine Mediates Acupunctural Effects on Hypothalamic-Pituitary-Adrenal Axis During Ethanol Withdrawal. J Acupunct Meridian Stud 9, 4–10 (2016).
Stonehouse, A. H., Adachi, M., Walcott, E. C. & Jones, F. S. Caffeine regulates neuronal expression of the dopamine 2 receptor gene. Mol Pharmacol 64, 1463–1473 (2003).
Fang, X., Mei, W., Barbazuk, W. B., Rivkees, S. A. & Wendler, C. C. Caffeine exposure alters cardiac gene expression in embryonic cardiomyocytes. Am J Physiol Regul Integr Comp Physiol 307, R1471–1487 (2014).
Lynge, J. & Hellsten, Y. Distribution of adenosine A1, A2A and A2B receptors in human skeletal muscle. Acta Physiol Scand 169, 283–290 (2000).
Warren, G. L., Hulderman, T., Liston, A. & Simeonova, P. P. Toll-like and adenosine receptor expression in injured skeletal muscle. Muscle Nerve 44, 85–92 (2011).
Finn, I. B. & Holtzman, S. G. Pharmacologic specificity of tolerance to caffeine-induced stimulation of locomotor activity. Psychopharmacology (Berl) 93, 428–434 (1987).
Vickers, A. J. & Linde, K. Acupuncture for chronic pain. JAMA 311, 955–956 (2014).
Kelly, R. B. Acupuncture for pain. Am Fam Physician 80, 481–484 (2009).
Manheimer, E., Linde, K., Lao, L., Bouter, L. M. & Berman, B. M. Meta-analysis: acupuncture for osteoarthritis of the knee. Ann Intern Med 146, 868–877 (2007).
Moore, M. E. & Berk, S. N. Acupuncture for chronic shoulder pain. An experimental study with attention to the role of placebo and hypnotic susceptibility. Ann Intern Med 84, 381–384 (1976).
Petrie, J. P. & Hazleman, B. L. A controlled study of acupuncture in neck pain. Br J Rheumatol 25, 271–275 (1986).
Lin, X. et al. The Effects of Acupuncture on Chronic Knee Pain Due to Osteoarthritis: A Meta-Analysis. J Bone Joint Surg Am 98, 1578–1585 (2016).
Selfe, T. K. & Taylor, A. G. Acupuncture and osteoarthritis of the knee: a review of randomized, controlled trials. Fam Community Health 31, 247–254 (2008).
Tukmachi, E., Jubb, R., Dempsey, E. & Jones, P. The effect of acupuncture on the symptoms of knee osteoarthritis–an open randomised controlled study. Acupunct Med 22, 14–22 (2004).
Butler, S. H., Godefroy, F., Besson, J. M. & Weil-Fugazza, J. A limited arthritic model for chronic pain studies in the rat. Pain 48, 73–81 (1992).
Chillingworth, N. L. & Donaldson, L. F. Characterisation of a Freund’s complete adjuvant-induced model of chronic arthritis in mice. J Neurosci Methods 128, 45–52 (2003).
Moilanen, L. J. et al. Monosodium iodoacetate-induced inflammation and joint pain are reduced in TRPA1 deficient mice–potential role of TRPA1 in osteoarthritis. Osteoarthritis Cartilage 23, 2017–2026 (2015).
Ogbonna, A. C., Clark, A. K., Gentry, C., Hobbs, C. & Malcangio, M. Pain-like behaviour and spinal changes in the monosodium iodoacetate model of osteoarthritis in C57Bl/6 mice. Eur J Pain 17, 514–526 (2013).
Rahman, W. & Dickenson, A. H. Osteoarthritis-dependent changes in antinociceptive action of Nav1.7 and Nav1.8 sodium channel blockers: An in vivo electrophysiological study in the rat. Neuroscience 295, 103–116 (2015).
van Osch, G. J., van der Kraan, P. M. & van den Berg, W. B. Site-specific cartilage changes in murine degenerative knee joint disease induced by iodoacetate and collagenase. J Orthop Res 12, 168–175 (1994).
Khan, J. et al. Interleukin-10 levels in rat models of nerve damage and neuropathic pain. Neurosci Lett 592, 99–106 (2015).
Qureshi, M. A. A. et al. Validation of Reference Genes for mRNA Quantification in Adjuvant Arthritis. Open J Rheumatol Autoimmune Diseases 2, 64–72 (2012).
The study was funded by NIH R01-AT007945. We thank the UR Genomics Resource Center for their research support for gene expression analysis. We thank Dr. Eli Eliav and Dr. Elaine Smolock for their critical comments on the manuscript.
The authors declare that they have no competing interests.
Publisher's note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Electronic supplementary material
About this article
Cite this article
Fujita, T., Feng, C. & Takano, T. Presence of caffeine reversibly interferes with efficacy of acupuncture-induced analgesia. Sci Rep 7, 3397 (2017). https://doi.org/10.1038/s41598-017-03542-x
This article is cited by
Peripheral receptors and neuromediators involved in the antihyperalgesic effects of acupuncture: a state-of-the-art review
Pflügers Archiv - European Journal of Physiology (2021)