The Two Faces of Adenosine
Over several decades, adenosine—a purine nucleoside involved in many physiological and pathophysiological processes—has gained a reputation as a molecule that relieves pain. But there have also been scattered reports that adenosine can produce pain. Now, new research led by Yang Xia, University of Texas Medical School at Houston, US, helps make sense of adenosine’s dual nature.
Lead author Xia Hu and colleagues find that mice with a deficiency in adenosine deaminase (ADA), an enzyme needed for the breakdown of adenosine, had sustained levels of circulating adenosine, which promoted mechanical and thermal pain sensitivity. A prolonged rise in adenosine also contributed to pain behaviors in two other models of chronic pain, including a mouse model of sickle cell disease, and the complete Freund’s adjuvant (CFA) model of inflammatory pain. In each model, the increased pain sensitivity stemmed from adenosine binding to its A2B receptor on myeloid cells, resulting in a surge in circulating levels of an inflammatory complex. This complex, composed of the cytokine interleukin-6 (IL-6) and its soluble receptor (sIL-6R), in turn activated dorsal root ganglion (DRG) neurons, increasing expression of the transient receptor potential vanilloid type 1 (TRPV1) ion channel and pain.
“There’s a nice scientific detective story here,” said Jana Sawynok, Dalhousie University, Halifax, Canada, with the authors uncovering adenosine’s effect on myeloid cells, with ensuing signaling to DRG neurons. “These findings provide a mechanistic understanding for a neuroimmune interaction in chronic pain,” she added.
The findings were published online June 16 in Cell Reports.
A dark side
In the 1970s, researchers first took notice that adenosine could relieve pain through its A1 receptor, one of four adenosine receptors. As a result, “there’s been interest for over 20 years in developing A1 receptor agonists as novel analgesics,” Sawynok told PRF (see PRF related news story), “with efforts in the last few years also heavily focused on the A3 receptor” (see PRF related news story). But how adenosine causes pain has been less clear (Sawynok, 2015). For example, A2A receptors can spur or halt pain, depending on where they are located.
Despite this uncertainty, persistently high levels of adenosine are known to promote inflammation, sickling, and tissue damage, but past studies had examined the effects only of acute increases in adenosine. Thus, Xia and colleagues wondered if sustained increases of adenosine lead to chronic pain. To test that idea, they examined homozygous knockout mice missing ADA, which exhibit increased levels of adenosine and die because of severe metabolic disturbances (as is the case in people lacking the enzyme). Polyethylene glycol-ADA (PEG-ADA), a US Food and Drug Administration-approved enzyme therapy used to treat people with ADA deficiency, metabolizes adenosine to prevent its excess accumulation and is also used to keep experimental mice alive.
The researchers found that ADA knockout animals receiving phosphate buffered saline (PBS) in the absence of PEG-ADA had circulating adenosine levels that steadily rose over a two-week period. At the same time, these mice also became more sensitive to mechanical and thermal stimulation, compared to heterozygous littermates having sufficient ADA levels, and treatment of the homozygous knockouts with continuous PEG-ADA prevented these changes. Hence, “chronically accumulated adenosine in the plasma contributed to pain,” said Xia.
The authors next tried to determine through which receptors adenosine acted to cause the observed pain sensitivity. Using wild-type animals treated with PBS as a control, they administered PBS or different antagonists selective for each of adenosine’s four receptors to ADA-deficient mice. Only the A2B receptor antagonist reversed the mechanical and thermal sensitivity seen in ADA-deficient mice. Similarly, ADA-deficient mice that genetically lacked the A2B receptor showed less sensitivity on both measures. Chronic pain seemed then to require a receptor different from those involved in analgesia.
Extending their findings beyond ADA-deficient mice, the authors also used a humanized mouse model of sickle cell disease (SCD), characterized by elevated plasma levels of adenosine (Zhang et al., 2011). The SCD mice exhibited mechanical and thermal pain sensitivity, compared to wild-type controls, which was reduced by PEG-ADA or an A2B receptor antagonist.
To determine whether adenosine acted directly on pain-signaling neurons, the researchers applied an adenosine analog called NECA to dorsal root ganglia (DRG) neurons from wild-type mice, while recording calcium influx as a readout of the cells’ responses. Surprisingly, NECA did not activate DRG neurons, nor did an agonist for the A2B receptor. Further, a low dose of capsaicin—a TRPV1 agonist—activated neurons from ADA-deficient and SCD mice but not wild-type controls, consistent with the cells having enhanced expression of the TRPV1 channel. These findings suggested that adenosine did not act directly on DRG neurons, leaving an unidentified pathway from adenosine to TRPV1 activation and pain.
A prolonged increase in plasma adenosine activates ADORA2B on myeloid cells, leading to an increase in circulating IL-6 and sIL-6R. The IL-6/sIL-6R complex transactivates gp130 on DRG cells, leading to STAT3 phosphorylation and induction of neuronal TRPV1 expression that results in increased sensitivity and chronic pain. From Hu et al., 2016.
A neuroimmune interaction
Since Xia and colleagues previously found high levels of circulating IL-6 in ADA-deficient and SCD mice (Dai et al., 2011; Zhang et al., 2011), they asked whether the cytokine tied adenosine to chronic pain. Indeed, relative to controls, both groups of animals had greater concentrations of plasma IL-6, which fell upon treatment with either PEG-ADA or the A2B receptor antagonist, but not PBS. This maintained boost in IL-6 levels seemed to play a causal role in how adenosine caused chronic pain, as an IL-6-neutralizing antibody injected into ADA-deficient mice attenuated pain sensitivity compared to mice receiving a control antibody.
However, the authors found no evidence that IL-6 directly stimulated DRG neurons, said Xia. But based on previous work, they knew that IL-6 forms a complex with its soluble receptor to activate gp130, a transmembrane glycoprotein and IL-6 receptor co-factor. Thus, they tested whether this type of signaling could explain how IL-6 caused pain. They found that ADA-deficient and SCD mice had higher circulating levels of the soluble IL-6 receptor, which were reduced with PEG-ADA treatment or with an A2B receptor antagonist. Further, treatment of ADA-deficient and SCD mice with a gp130-neutralizing antibody reduced thermal and mechanical sensitivity in the animals compared to those that received a control antibody. DRG neurons from mice treated with the gp130-neutralizing antibody also had reduced responses to a low concentration of capsaicin, as well as reduced TRPV1 expression. Additional experiments would show that the IL-6/soluble IL-6 receptor complex increased TRPV1 expression through phosphorylation of STAT3, a signaling molecule downstream of IL-6 receptor activation.
Together, the results showed that the inflammatory complex acted through gp130 and STAT3 phosphorylation in DRG neurons to increase TRPV1 expression and cause pain—but which cell type(s) were the source of IL-6 and its soluble receptor leading to that chain of events? The authors discovered that myeloid cells were the answer. To demonstrate a role for these innate immune cells, the authors turned to knockout mice missing the A2B receptor in myeloid cells, injecting them and control animals with CFA into the hindpaw to provoke inflammation. CFA led to greater levels of circulating adenosine and IL-6, for six and 24 hours, respectively, in the controls compared to the knockouts. The knockouts also showed less mechanical and thermal sensitivity. Because levels of adenosine and IL-6 were comparable in all three models examined in the study, the authors say this strengthens the case that innate immune cells release IL-6 to drive chronic pain.
“We really think that myeloid cells become sensitive to adenosine, resulting in more cytokines. These cytokines form complexes, transactivating neurons to change gene expression and cause hypersensitivity,” said Xia.
Overall, the findings could help researchers reconcile how adenosine can be both pro- and anti-nociceptive. As Sawynok points out, these opposing effects could be explained by varying amounts of adenosine likely to be present under different conditions, and adenosine’s varying affinity for different receptors. Adenosine has a lower affinity for the A2B receptor compared to other receptors, so “under somewhat perturbed, acute conditions, you might not get too much activation of that receptor … but, under conditions like chronic inflammation, adenosine would reach levels high enough to recruit the A2B receptor, producing chronic pain.”
Because PEG-ADA is an FDA-approved drug, Xia suggests that the therapy might be a safe option for driving down adenosine levels in people with chronic pain. The results also highlight the therapeutic potential for developing drugs targeting the A2B receptor.
Matthew Soleiman is a neuroscientist-turned-science writer currently residing in Nashville, Tennessee. Follow him on Twitter @MatthewSoleiman.
Image credit: Hu et al., 2016
Article Contributors:
