**3.3.2 A2A receptor signaling**

630 Pharmacology

very low and it has been found in great density only in bowel and bladder. A3 receptors (A3R)are widely distributed in several mammals, however, few studies have indicated specific roles for this receptor (Dixon et al., 1996; Ralevic & Burnstock, 1998; Salvatore et al.,

All adenosine receptors are coupled to G-protein. Nevertheless there are many kinds of Gprotein and each one may activate a distinct pathway. Thus, the four adenosine receptors can stimulate or inhibit several pathways and consequently exert many physiological actions (Jacobson & Gao, 2006; Ralevic & Burnstock, 1998). We show below the main

A1R is coupled to Gi/0 protein family which is pertussis toxin-sensitive. Most of the biological effects induced by A1R activation are due to inhibition of cAMP second messenger (Burnstock, 2007; Jacobson & Gao, 2006; Ralevic & Burnstock, 1998; Sawynok, 1998) (Figure

Fig. 3. Adenosine A1 receptors and its main pathways. A1R, adenosine A1 receptor; AC, adenylate cyclase; cAMP, cyclic adenosine monophosphate; DAG, diacylglicerol; ERK1/2, extracellular signal-regulated kinases 1 and 2; IP3, inositol triphosphate; K+, potassium channels; MAPK, mitogen-activated protein kinase; MEK, MAPK and ERK kinase; PKA,

Beta and gamma subunities of A1R, when activated stimulate phospholipase C (PLC). Activation of PLC cleaves phosphatidylinositol 4,5-bisphosphate (PIP2) into diacylglycerol (DAG) and inositol 1,4,5-trisphosphate (IP3) leading to increased levels Ca+2. Moreover, the enhancement of intracellular calcium can induce some enzymes such as protein kinase C

1993).

**3.3 General adenosine receptor signaling** 

**3.3.1 A1 receptor signaling** 

protein kinase; PLC, phospholipase C.

3 and Table 1).

signaling characteristics of each adenosine receptor.

A2AR is coupled to Gs (the most part) and Golf protein (mainly in striatum). The main intracellular event after activation of these proteins is adenylate cyclase activation followed by cAMP production enhancement (Table 1).


Table 1. Adenosine receptors signaling. Adapted from Ralevic and Burnstock, 1998; Sawynok and Liu, 2003; Jacobson and Guao, 2006.

The increasing of cAMP stimulates cAMP-dependent kinase (PKA). Thus, PKA becomes able to activate several pathways through PKC, calcium channels, potassium channels, cAMP responsive element-binding (CREB), MAPK, PLC activation (Burnstock, 2007; Cunha et al., 2008; Fredholm et al., 2003, 2007; Jacobson & Gao, 2006; Ralevic & Burnstock, 1998).

#### **3.3.3 A2B receptor signaling**

Through Gs and Gq activation A2B receptor induces adenylate cyclase and PLC. In humans, A2BR can increase intracellular calcium by IP3 activation. Moreover, the arachidonic acid pathway is also involved in A2BR signaling (Feoktistov & Biaggioni, 2011; Jacobson & Gao, 2006; Peakman and Hill, 1994).

#### **3.3.4 A3 receptor signaling**

A3R like A1R is coupled to Gi/0 and also to Gq/11 protein. Its main signaling transduction is the inhibition of adenylate cyclase and stimulation of PLC, IP3, DAG, PKC and PLD. Also, like other adenosine receptors, A3R activates MAPK pathway, mainly ERK1/2 (Abbracchio et al., 1995; Armstrong & Ganote, 1994; Palmer et al., 1995; Shneyvays et al., 2004).

The Involvement of Purinergic System in Pain:

cAMP production (Sawynok, 1998).

**3.4.5 A1 agonists/antagonists** 

**3.5 A2A receptors and pain** 

listed in Table 2.

**3.4.3 Interactions between A1 receptor and opioid system** 

serotoninergic systems in pain modulation (Sawynok, 1998).

Table 2. Principal adenosine A1 receptor agonists and antagonists.

Adenosine Receptors and Inosine as Pharmacological Tools in Future Treatments 633

There is a strong relationship between opioidergic and adenosinergic systems in pain modulation. Morphine and other opioids are able to release adenosine (Sawynok et al., 1989; Sweeney et al., 1987a,b; Sweeney et al., 1989). Adenosine or analog administration combined with opioids enhances the analgesic effect of the latter (DeLander & Hopkins, 1986). However, administration of methylxanthines, adenosine receptor antagonists, can augment, decrease or have no effect on analgesic activity of opioids (Sawynok, 2011). In addition, A1R can undergo dimerization with µ receptor in afferents neurons and induce the decrease of

**3.4.4 Interactions among A1 receptor and other receptors in pain mechanisms** 

A1R is able to dimerize with alpha-2-adrenergic receptor. Also, it has been shown that serotonin releases adenosine from primary afferents and that A1R receptor antagonist blocks serotonin analgesic actions, suggesting a close involvement between adenosinergic and

Methylxanthines (caffeine included) are natural antagonists of adenosine A1R and A2AR. However, the main selective agonists/antagonists to A1R are synthetics. Some of them are

Taiwo & Levine (1990) demonstrated clear distinct effects of A1R and A2AR activation in peripheral pain. They showed that, in peripheral sites, A1R mediates analgesia while A2AR facilitates painful perception. In addition, other studies also showed that peripheral A2AR activation induced pain (Doak & Sawynok, 1995; Li et al., 2010; Taiwo & Levine, 1990). However, systemically and spinally, the role of A2AR is not entirely clear. Some publications have demonstrated that A2AR activation induces pain (Bastia et al., 2002; Hussey et al., 2007). On the other hand, other authors have showed a reduction of pain when A2AR is activated (By et al., 2011; Borghi et al., 2002; Yoon et al., 2005). These controversial results might be associated with A2AR intracellular signaling. A2AR activation induces cAMP increased

#### **3.4 A1 receptors and pain**

A1R is the main responsible for inducing analgesia among the adenosine receptors (Burnstock et al., 2011; Sawynok, 1998; Sawynok & Liu, 2003). It has been shown that A1R is widely distributed in the dorsal spinal cord, mainly in lamina II (substantia gelatinosa) (Choca et al., 1988; Horiuchi et al., 2010; Sawynok, 1998). In this site, many afferent sensory nerve have connections with post-synaptic neurons. Also, A1R are localized in the descending projection within dorsal horn (Choca et al., 1988).

#### **3.4.1 Effects of A1 receptor activation in acute pain**

There are several published data showing the important effect of A1R in controlling acute pain (review see Jacobson & Gao, 2006; Sawynok, 1998). Moreover, it has been shown that systemic administration of various A1R agonists can produce analgesic effect in several models of acute pain in animals (Gong et al., 2010; Sawynok, 1998). Probably, these effects are caused by peripheral, supraspinal and mostly by spinal A1R. In mice lacking A1R (knockout animals) a lower pain threshold was observed in hyperalgesia tests (Wu et al., 2005). Further, analgesic effect induced by intrathecal adenosine was abolished as well as the increase of thermal hyperalgesia in A1R knockout mice (Johansson et al., 2001).

Several studies published show that intrathecal injection of A1R agonists cause analgesia in various animal models of acute pain, including tail flick, tail immersion, hot-plate, formalin, acetic acid, capsaicin models and others (Nascimento et al., 2010; Song et al., 2011; Zahn et al., 2007).

A1R is also found in primary afferent neurons and in the cell body and the dorsal root ganglia (Lima et al., 2010; Sawynok, 2009), then it has an important role in modulation of peripheral pain. Many studies have shown that administration of A1R agonists into the paw of animals causes an analgesic effect in several animal models of pain. It has been shown that A1R activation in the periphery inhibits formalin-induced pain and reduces hyperalgesia induced by PGE2 (Karlsten et al., 1992; Taiwo & Levine, 1990). Further, when the peripheral adenosine A1R is activated, it triggers the NO/cGMP/PKG/KATP intracellular signaling pathway and then inhibits pain (Lima et al., 2010). Moreover, A1R agonists reduce the thermal hyperalgesia, but not mechanical allodynia, caused by sciatic nerve injury. The thermal hyperalgesia is mediated by C fibers and mechanical allodynia, in turn, is mediated by A fibers, which demonstrates the presence of A1R in C but not in A fibers (Sawynok, 2009).

#### **3.4.2 A1 receptor and chronic pain**

Several authors have shown that different agonists of A1R are able to reduce distinct kinds of chronic pain. The agonist R-PIA inhibits mechanical allodynia induced by spinal nerve ligation in rats (Hwang et al., 2005; Song et al., 2011). In addition, R-PIA also reduces thermal pain threshold in rats that underwent an injury in the spinal cord (Horiuchi et al., 2010). Moreover, other mechanisms are generally involved in analgesic effect in chronic pain induced by A1R activation, such as inhibition of glutamate release. Another A1R agonist, CPA, when given to rats, inhibits pain induced by arthritis and pain induced by neuropathy (Curros-Criado & Herrero, 2005). Also, in experiments with A1R knockout mice it has been observed that these animals present a lower pain threshold than wild-type animals in inflammatory and neuropathic pain models (Wu et al., 2005).

#### **3.4.3 Interactions between A1 receptor and opioid system**

There is a strong relationship between opioidergic and adenosinergic systems in pain modulation. Morphine and other opioids are able to release adenosine (Sawynok et al., 1989; Sweeney et al., 1987a,b; Sweeney et al., 1989). Adenosine or analog administration combined with opioids enhances the analgesic effect of the latter (DeLander & Hopkins, 1986). However, administration of methylxanthines, adenosine receptor antagonists, can augment, decrease or have no effect on analgesic activity of opioids (Sawynok, 2011). In addition, A1R can undergo dimerization with µ receptor in afferents neurons and induce the decrease of cAMP production (Sawynok, 1998).

## **3.4.4 Interactions among A1 receptor and other receptors in pain mechanisms**

A1R is able to dimerize with alpha-2-adrenergic receptor. Also, it has been shown that serotonin releases adenosine from primary afferents and that A1R receptor antagonist blocks serotonin analgesic actions, suggesting a close involvement between adenosinergic and serotoninergic systems in pain modulation (Sawynok, 1998).

#### **3.4.5 A1 agonists/antagonists**

632 Pharmacology

A1R is the main responsible for inducing analgesia among the adenosine receptors (Burnstock et al., 2011; Sawynok, 1998; Sawynok & Liu, 2003). It has been shown that A1R is widely distributed in the dorsal spinal cord, mainly in lamina II (substantia gelatinosa) (Choca et al., 1988; Horiuchi et al., 2010; Sawynok, 1998). In this site, many afferent sensory nerve have connections with post-synaptic neurons. Also, A1R are localized in the

There are several published data showing the important effect of A1R in controlling acute pain (review see Jacobson & Gao, 2006; Sawynok, 1998). Moreover, it has been shown that systemic administration of various A1R agonists can produce analgesic effect in several models of acute pain in animals (Gong et al., 2010; Sawynok, 1998). Probably, these effects are caused by peripheral, supraspinal and mostly by spinal A1R. In mice lacking A1R (knockout animals) a lower pain threshold was observed in hyperalgesia tests (Wu et al., 2005). Further, analgesic effect induced by intrathecal adenosine was abolished as well as

Several studies published show that intrathecal injection of A1R agonists cause analgesia in various animal models of acute pain, including tail flick, tail immersion, hot-plate, formalin, acetic acid, capsaicin models and others (Nascimento et al., 2010; Song et al., 2011; Zahn et

A1R is also found in primary afferent neurons and in the cell body and the dorsal root ganglia (Lima et al., 2010; Sawynok, 2009), then it has an important role in modulation of peripheral pain. Many studies have shown that administration of A1R agonists into the paw of animals causes an analgesic effect in several animal models of pain. It has been shown that A1R activation in the periphery inhibits formalin-induced pain and reduces hyperalgesia induced by PGE2 (Karlsten et al., 1992; Taiwo & Levine, 1990). Further, when the peripheral adenosine A1R is activated, it triggers the NO/cGMP/PKG/KATP intracellular signaling pathway and then inhibits pain (Lima et al., 2010). Moreover, A1R agonists reduce the thermal hyperalgesia, but not mechanical allodynia, caused by sciatic nerve injury. The thermal hyperalgesia is mediated by C fibers and mechanical allodynia, in turn, is mediated by A fibers, which demonstrates the presence of A1R in C but not in A fibers (Sawynok,

Several authors have shown that different agonists of A1R are able to reduce distinct kinds of chronic pain. The agonist R-PIA inhibits mechanical allodynia induced by spinal nerve ligation in rats (Hwang et al., 2005; Song et al., 2011). In addition, R-PIA also reduces thermal pain threshold in rats that underwent an injury in the spinal cord (Horiuchi et al., 2010). Moreover, other mechanisms are generally involved in analgesic effect in chronic pain induced by A1R activation, such as inhibition of glutamate release. Another A1R agonist, CPA, when given to rats, inhibits pain induced by arthritis and pain induced by neuropathy (Curros-Criado & Herrero, 2005). Also, in experiments with A1R knockout mice it has been observed that these animals present a lower pain threshold than wild-type animals in

the increase of thermal hyperalgesia in A1R knockout mice (Johansson et al., 2001).

descending projection within dorsal horn (Choca et al., 1988).

**3.4.1 Effects of A1 receptor activation in acute pain** 

**3.4 A1 receptors and pain** 

al., 2007).

2009).

**3.4.2 A1 receptor and chronic pain** 

inflammatory and neuropathic pain models (Wu et al., 2005).

Methylxanthines (caffeine included) are natural antagonists of adenosine A1R and A2AR. However, the main selective agonists/antagonists to A1R are synthetics. Some of them are listed in Table 2.


Table 2. Principal adenosine A1 receptor agonists and antagonists.

#### **3.5 A2A receptors and pain**

Taiwo & Levine (1990) demonstrated clear distinct effects of A1R and A2AR activation in peripheral pain. They showed that, in peripheral sites, A1R mediates analgesia while A2AR facilitates painful perception. In addition, other studies also showed that peripheral A2AR activation induced pain (Doak & Sawynok, 1995; Li et al., 2010; Taiwo & Levine, 1990). However, systemically and spinally, the role of A2AR is not entirely clear. Some publications have demonstrated that A2AR activation induces pain (Bastia et al., 2002; Hussey et al., 2007). On the other hand, other authors have showed a reduction of pain when A2AR is activated (By et al., 2011; Borghi et al., 2002; Yoon et al., 2005). These controversial results might be associated with A2AR intracellular signaling. A2AR activation induces cAMP increased

The Involvement of Purinergic System in Pain:

enzyme to regulate adenosine endogenous levels.

Sawynok, 1999), but not in others (Keil & DeLander, 1992, 1994).

**3.8.2 Analgesic effect by supply of purinergic substrates** 

**3.8.3 Involvement of adenosine receptors in acupuncture pain relief** 

acupuncture analgesic effect (for review see Zylka, 2010).

**4.1 Inosine within of purinergic system** 

**4. Inosine and pain** 

review Sawynok & Liu, 2003).

Adenosine Receptors and Inosine as Pharmacological Tools in Future Treatments 635

inhibitors of ADA itself are able to cause analgesia in determined animal models (Poon &

Coadministration of AK and ADA inhibitors potentiates the analgesic effect of the former (Poon & Sawynok, 1999). Also, adenosine effect is synergically augmented by coadministration of ADA inhibitor (Keil & DeLander, 1994). Distinct effects between AK inhibitors and ADA inhibitors might be because adenosine has a higher affinity by AK than ADA (Arch & Newsholme, 1978). In this point of view, AK seems to be the most important

Ecto-5'-nucleotidase (NT5E) is an enzyme located in the cell membrane that catalyzes the extracellular conversion of adenosine monophosphate (AMP) into adenosine in several tissues, included dorsal root ganglia (DRG) and substantia gelatinosa (Zylka, 2011). Recent studies have demonstrated that the analgesic effect of AMP combined with an AK inhibitor halved in NT5E knockout animals and is totally reversed in A1R knockout animals (Sowa et al., 2010a). These results inspired another study that evaluated whether exogenous supply of NT5E (increases supply of adenosine) could induce a long lasting analgesic effect. NT5E presented effects that lasted for 2 days in models of inflammatory and chronic pain. Both effects were dependent on A1R (Sowa, 2010b). Hence, the supply of enzymes that generate adenosine is a new interesting approach that may be used in studies to treat chronic pain.

In an elegant study published in 2010, Goldman and colleagues showed that analgesia induced by acupuncture depends on purine release, such as ATP, ADP, AMP and adenosine. In addition, it has been showed that A1R agonist replicates the acupuncture effect. Also, in A1R knockout animals, acupuncture did not present analgesia. Moreover, inhibition of adenosine deaminase prolonged the analgesic effect of acupuncture in mice (Goldman et al., 2010). It is interesting to mention that caffeine, the most widely used drug across the world in beverages such as teas, coffee, *mate*, soft drinks, energy drinks and others is an antagonist of adenosine receptors. Therefore, patients in treatment with acupuncture should not drink these caffeine beverages, because caffeine might reduce the

ATP is the main molecule of purinergic system. Inside the cell, ATP may be bi-directionally converted into AMP. AMP is broken down into adenosine. Adenosine may be converted back into AMP through phosphorylation by AK. Moreover, adenosine might leave the cell by nucleoside transporter (NT). Inside the cell, adenosine deaminase is responsible for the conversion from adenosine to inosine. Outside the cell, this conversion is performed by ectoadenosine kinase or even adenosine deaminase. Inosine is a substrate to purine nucleoside phosphorylase (PNP), leading to hypoxanthines as its products. Hypoxanthines are converted into xanthines and afterwards to uric acid by xanthine oxidase (Figure 4) (See

production (it can cause pain) and also K+ channels opening (it can inhibit pain) (Jacobson & Gao, 2006; Regaya et al., 2004; Sawynok, 1998).

A2AR knockout animals are less sensitive to pain, suggesting that A2AR is a pain facilitator in acute (Hussey et al., 2007) and chronic pain (Bura et al., 2008). Bura and coworkers (2008) also demonstrated that microglia and astrocytes expression was higher in wild-type A2AR animals than in A2AR knockout animals. Also, A2AR located in glial cells is responsible for the release of inflammatory mediators that induce and maintain chronic pain (Boison et al., 2010). Thus, A2AR blockade might be an interesting approach for future treatments of neuropathic and chronic pain. Meantime, also in chronic pain exists distinct results about A2AR. A report showed that only one spinal injection of A2AR agonist was able to induce analgesia during several days in rats undergoing neuropathic pain (Loram et al., 2009). After all, it is clear that A2AR is involved in pain modulation. However, more studies are necessary to precisely explain how this receptor works in distinct situations, only then it will be possible to make clinical approaches.

#### **3.6 A2B receptors and pain**

Few studies have been evaluating the A2BR role in pain. Most part of these studies showed that A2BR facilitates pain transmission, because A2BR antagonists have reduced pain (Abo-Salem et al., 2004; Bilkei-Gorzo et al., 2008; GodFrey et al., 2006). A2BR antagonist reduced thermal hyperalgesia and was able to potentiate the analgesic effect caused by morphine and acetaminophen (Abo-Salem et al., 2004; Godfrey et al., 2006). Also, the blockade of A2BR presented an analgesic effect in inflammatory pain (Bilkei-Gorzo et al., 2008).

#### **3.7 A3 receptors and pain**

Similar to A2BR, adenosine A3R is not an interesting target to pain relief. However, A3R is implicated in pathological conditions such as ischemic diseases and in inflammation (for review see Borea et al., 2009). Regarding pain, there are few studies evaluating A3R role. Sawynok and colleagues (1997) showed that A3R activation causes pain and paw oedema through release of histamine and serotonin. A3R knockout animals presented an increased pain threshold in some models of pain but not difference in others (Fedorova et al., 2003; Wu et al., 2002). A3R might be an interesting target to inflammatory and autoimmune diseases, but not to pain states.

#### **3.8 Novel approaches in pain management involving adenosine receptors**

#### **3.8.1 Management of adenosine receptors by metabolism modulation**

The first report showing that adenosine kinase (AK) inhibition reduces behaviour associated to pain was published by Keil and DeLander, 1992. AK inhibitors are able to decrease pain levels when given peripherally or sistemically (Kowaluk et al., 1999; Lynch et al., 1999; Sawynok, 1998). Moreover, these inhibitors are efficacious against acute and chronic pain (Kowaluk et al., 2000; Lynch et al., 1999; McGaraughty et al., 2005; Poon & Sawynok, 1998, 1999; Suzuki et al., 2001). Another enzyme that regulates adenosine level is adenosine deaminase (ADA), that converts adenosine to inosine (Sawynok, 1998). However, the analgesic effect caused by ADA inhibition is not so clear yet. It has been showed that inhibitors of ADA itself are able to cause analgesia in determined animal models (Poon & Sawynok, 1999), but not in others (Keil & DeLander, 1992, 1994).

Coadministration of AK and ADA inhibitors potentiates the analgesic effect of the former (Poon & Sawynok, 1999). Also, adenosine effect is synergically augmented by coadministration of ADA inhibitor (Keil & DeLander, 1994). Distinct effects between AK inhibitors and ADA inhibitors might be because adenosine has a higher affinity by AK than ADA (Arch & Newsholme, 1978). In this point of view, AK seems to be the most important enzyme to regulate adenosine endogenous levels.

## **3.8.2 Analgesic effect by supply of purinergic substrates**

Ecto-5'-nucleotidase (NT5E) is an enzyme located in the cell membrane that catalyzes the extracellular conversion of adenosine monophosphate (AMP) into adenosine in several tissues, included dorsal root ganglia (DRG) and substantia gelatinosa (Zylka, 2011). Recent studies have demonstrated that the analgesic effect of AMP combined with an AK inhibitor halved in NT5E knockout animals and is totally reversed in A1R knockout animals (Sowa et al., 2010a). These results inspired another study that evaluated whether exogenous supply of NT5E (increases supply of adenosine) could induce a long lasting analgesic effect. NT5E presented effects that lasted for 2 days in models of inflammatory and chronic pain. Both effects were dependent on A1R (Sowa, 2010b). Hence, the supply of enzymes that generate adenosine is a new interesting approach that may be used in studies to treat chronic pain.

#### **3.8.3 Involvement of adenosine receptors in acupuncture pain relief**

In an elegant study published in 2010, Goldman and colleagues showed that analgesia induced by acupuncture depends on purine release, such as ATP, ADP, AMP and adenosine. In addition, it has been showed that A1R agonist replicates the acupuncture effect. Also, in A1R knockout animals, acupuncture did not present analgesia. Moreover, inhibition of adenosine deaminase prolonged the analgesic effect of acupuncture in mice (Goldman et al., 2010). It is interesting to mention that caffeine, the most widely used drug across the world in beverages such as teas, coffee, *mate*, soft drinks, energy drinks and others is an antagonist of adenosine receptors. Therefore, patients in treatment with acupuncture should not drink these caffeine beverages, because caffeine might reduce the acupuncture analgesic effect (for review see Zylka, 2010).
