While still held up as the “gold standard” for treating chronic and post-operative pain, morphine and other opiate drugs come with serious side effects. Tolerance and drug-induced hyperalgesia top the list, and until now were thought to share a common mechanism. But new work from Yves De Koninck at Institut Universitaire en Santé Mentale de Québec, Canada, shows that they are, in fact, distinct processes.
The work “really puts another nail in the old hypothetical coffin that you can’t separate these effects of tolerance and hyperalgesia,” says Howard Gutstein, a researcher and clinician at MD Anderson Cancer Center, Houston, US, who was not involved in the current work.
The separation of hyperalgesia from tolerance, says De Koninck, means that the insidious side effect can also be separated from the drugs’ pain-killing power. “Hyperalgesia is really a separate reaction of the CNS. This opens an avenue to prevent that side effect without losing the analgesia of morphine.”
The new findings, from De Koninck and co-first authors Francesco Ferrini (also in Québec) and Tuan Trang, from Michael Salter’s lab at the Hospital for Sick Children in Toronto, Canada, appear online in the January 6 Nature Neuroscience.
In defiance of Occam’s razor—a principle dictating that among competing hypotheses, the simplest is preferable to the more complex—De Koninck’s team uncovered an intricate signaling pathway that leads to aberrant pain signaling. For reasons unknown, microglia—the nervous system’s sentinel immune cells—within the dorsal horn of the spinal cord express μ (mu)-opioid receptors. After morphine binds to those receptors, microglia increase their expression of ATP-binding P2X4 receptors, which in turn leads to microglial release of brain-derived neurotrophic factor (BDNF). BDNF then acts on lamina I projection neurons, decreasing their expression of KCC2, a transporter that maintains chloride homeostasis, which is key to inhibitory neurotransmission.
The researchers showed that interrupting any step of the pathway—depletion of spinal microglia, deletion of microglial BDNF, or inhibition of the P2X4 receptor—prevented the development of hyperalgesia in mice. However, all three manipulations left analgesia and tolerance intact. Gutstein describes the experiments as a “tour de force dissecting the signaling pathway.”
The results on tolerance stand in contrast to previous work showing a role for microglia and neuroinflammation in both opioid-induced hyperalgesia and tolerance (see PRF news story and comment below from Mark Hutchinson).
Network effects
The team demonstrated the drug-induced collapse of the chloride gradient in neurons in an elegant series of electrophysiological experiments. The ion balance disruption directly affected the directional flow of ions through open GABA-receptor channels. Whereas the inhibitory neurotransmitter usually evokes a hyperpolarizing current, the chloride disruption perturbed the reliability of that signal. The end result is that morphine appears to release inhibition of spinal cord projection neurons by handicapping chloride extrusion, affecting GABA signaling as well as the membrane's overall excitability.
In previous work, the authors described the same basic signaling mechanism as underlying the pain following nerve injury (Coull et al., 2005). Ironically, says De Koninck, the new results suggest that “morphine tends to recapitulate what happens in neuropathic pain.” Why this inflammatory cascade—initiated in microglia and culminating in neuronal chloride balance disruption—follows chronic morphine exposure presents an interesting puzzle. De Koninck asks, “Is it a typical homeostatic response of the body to drugs? Or is it just bad luck?”
In either case, the upshot is that normally painless sensory events become painful, but how? The majority of projection neurons in lamina I of the spinal cord normally respond only to noxious input, dependably signaling the brain that a potentially damaging event is happening out in the periphery. But after nerve injury—and presumably after hyperalgesia brought on by morphine—over 80 percent of the projection neurons begin to respond to painless touch. While the disruption in chloride concentration may be localized to excitatory projection neurons, De Koninck explains, those neurons are part of a larger pain-signaling network that encompasses both nociceptive and non-nociceptive sensory neurons. The connections with non-nociceptive inputs are likely indirect and normally suppressed. “This [hyperalgesia] pathway basically unmasks this existing interconnection, causing a noxious response to innocuous input.”
In terms of a therapeutic “avenue,” the new results reveal targets to combat hyperalgesia, including P2X4 receptors, BDNF signaling, and GABA receptors. Despite significant efforts, specific antagonists for the P2X4 receptor have so far eluded chemists. And while BDNF is a more tractable target, most neuroscientists would agree that shutting down that molecule might do more harm than good, because it provides neurons with critical support. Enhancing GABA transmission might also be folly, says De Koninck, because inhibitory currents in this scenario might either collapse, “or worse, switch to an excitatory signal,” he says. Instead, the group has chosen to focus on KCC2 itself with a double-pronged strategy: to stabilize the protein at the membrane, and to enhance its activity.
The corollary to the current work came from Gutstein, who last year described a mechanism for opioid tolerance involving a growth factor receptor kinase (see PRF related news story). While the side effects of morphine present significant limitations, says Gutstein, “opiates are all we’ve got” to battle some types of pain. But now, he says, it appears that we might develop ways to contend with their negative consequences, “one side effect at a time.”
Stephani Sutherland, PhD, is a freelance neuroscience writer based in Southern California.
Image Credit: © Deeboldrick | Dreamstime.com
Comments
Jürgen Sandkühler, Medical University of Vienna
This article describes a
This article describes a novel mechanism of morphine-induced hyperalgesia which distinguishes it clearly from tolerance to morphine analgesia. Mice received morphine (10 mg ∙ kg−1) subcutaneously twice daily for seven days. Morphine-induced analgesia was assessed as an increase in thermal paw withdrawal thresholds one hour after the injections. The antinociceptive effect of morphine diminished each day and was absent from day 5 on, indicating development of complete tolerance to morphine’s antinociception. In addition, thermal and mechanical paw withdrawal thresholds were measured also one hour prior to the morning injection of morphine (i.e., at a time point when morphine levels were at a minimum). Paw withdrawal thresholds decreased significantly from day 3 to day 10, indicating development of morphine-induced hyperalgesia.
Previous studies identified two mechanisms that contribute to opioid-induced hyperalgesia: 1) activation of facilitatory pathways descending from brainstem areas to the spinal cord (Ossipov et al., 2005), and 2) long-term potentiation (LTP) at synapses between C-fiber afferents and spinal dorsal horn lamina I neurons (Drdla et al., 2009). In the current study, the authors discovered an additional mechanism of morphine-induced hyperalgesia: the impairment of GABAergic inhibition due to altered Cl− homeostasis in lamina I neurons. In an extensive series of experiments in vivo and in cell cultures, the authors provide evidence that repetitive morphine applications lead to the upregulation and activation of ionotrophic purinergic receptors of the P2X4 subtype specifically on microglial cells. This triggers the release of BDNF from microglia. BDNF is known to downregulate the expression of the KCC2 co-transporter, which critically regulates Cl− homeostasis in lamina I neurons (Coull et al., 2005). The current study revealed that this signaling cascade is activated by repetitive morphine applications and results in impaired GABAergic inhibition in spinal dorsal horn lamina I neurons. Interrupting the cascade at any site (depletion of microglial cells in spinal cord; blocking P2X4 receptors or microglia-derived BDNF function) prevented morphine-induced hyperalgesia. Interestingly morphine-induced upregulation of P2X4 receptors was blocked by the opioid receptor antagonist (−) naloxone, whereas the release of BDNF was blocked by the non-opioid receptor antagonist (+) naloxone. This suggests that P2X4 upregulation requires activation of μ-opioid receptors, whereas the BDNF release from microglia by morphine involves signaling independent of opioid receptors. Given the fact that inhibition in the spinal cord dorsal horn serves five essential functions for nociception (Sandkühler, 2009), one might speculate that hyperalgesia may not be the only consequence of severely impaired spinal inhibition after repetitive morphine applications.
One would expect that morphine-induced hyperalgesia counteracts its antinociceptive efficacy. It has indeed been suggested that morphine-induced hyperalgesia and tolerance to the analgesic effects of morphine share cellular mechanisms, both in the peripheral- (Joseph et al., 2010) and in the central nervous system (Price et al., 2000). The current study revealed, however, that morphine-induced hyperalgesia, but not morphine tolerance, requires spinal microglia and microglial BDNF. Thus, in this animal model, morphine-induced hyperalgesia and morphine tolerance represent distinguishable aspects of morphine’s action involving distinct signaling cascades. There are also other reports which support the current conclusion that morphine-induced hyperalgesia and tolerance to morphine are distinct phenomena (e.g., Zissen et al., 2007). It will be interesting to learn if the current findings also hold when an opioid is given in a way that ensures continuous analgesia, as recommended for chronic pain patients.
References:
BDNF from microglia causes the shift in neuronal anion gradient underlying neuropathic pain.
Coull JAM, Beggs S, Boudreau D, Boivin D, Tsuda M, Inoue K, Gravel C, Salter MW, De Koninck Y.
Nature. 2005 Dec 15; 438(7070):1017-21. PMID: 16355225.
Induction of synaptic long-term potentiation after opioid withdrawal.
Drdla R, Gassner M, Gingl E, Sandkühler J.
Science. 2009 Jul 10; 325(5937):207-10. PMID: 19590003.
Shared mechanisms for opioid tolerance and a transition to chronic pain.
Joseph EK, Reichling DB, Levine JD
J Neurosci. 2010 Mar 31; 30(13):4660-6. PMID: 20357116.
Underlying mechanisms of pronociceptive consequences of prolonged morphine exposure.
Ossipov MH, Lai J, King T, Vanderah TW, Porreca F.
Biopolymers 80:319-324.
NMDA-receptor antagonists and opioid receptor interactions as related to analgesia and tolerance.
Price DD, Mayer DJ, Mao J, Caruso FS.
J Pain Symptom Manage. 2000 Jan; 19(1 Suppl):S7-11. PMID: 10687332.
Models and mechanisms of hyperalgesia and allodynia.
Sandkühler J.
Physiol Rev. 2009 Apr; 89(2):707-58. PMID: 19342617.
Tolerance, opioid-induced allodynia and withdrawal associated allodynia in infant and young rats.
Zissen MH, Zhang G, McKelvy A, Propst JT, Kendig JJ, Sweitzer SM.
Neuroscience. 2007 Jan 5; 144(1):247-62. Epub 2006 Oct 19. PMID: 17055659.
Mark Hutchinson, University of Adelaide
http://www.painresearchforum
This is a large body of work that very nicely characterizes a beautiful mechanism of opioid-induced hyperalgesia. However, there are conclusions that have been drawn based on suboptimal approaches. I am surprised to see that several key control experiments were not included in several of the experiments such as Saline+Mac-Saporin. It is surprising to think that depleting microglia such as this wouldn't have some nociceptive consequence.
It is also surprising that the wealth of neuroimmune data on the development of opioid tolerance has not been addressed here. For example, the work of Wilcox, Song, Cao, Shavit, and many others over the past 14 years (reviewed in Hutchinson et al., 2011) have implicated proinflammatory products from astrocytes and microglia in the development of opioid tolerance. The new mechanism presented here is very nice, but how it relates to all of this work has not been addressed.
Some key points:
1. The authors show that (+)-naloxone did block hyperalgesia and not tolerance. This contrasts with our work, which showed attenuation of both tolerance and hyperalgesia by (+)-naloxone. Critically, in our experiments we used continuous blockade with intrathecal infusions of (+)-naloxone, whilst acute administration was employed by Ferrini et al. The important distinction here is that whilst morphine has some activity at TLR4, we have demonstrated that the 3-gluconidated metabolite of morphine (M3G) has much greater activity and, owing to its pharmacokinetic profile, will "outlast" acute (+)-naloxone.
Interestingly, the (+)-naloxone regimen was capable of blocking the CD11b changes in microglia. Importantly, these CD11b changes in microglia are traditionally associated with an inflammatory phenotype, thus relating to the wealth of neuroimmune literature already available, but not examined by Ferrini.
How the data in Figure 8g and f align with the established opioid-neuroimmune literature was not investigated. This will be a key area of future research to fill in the gaps and refine the mechanism.
2. It is interesting to see that the opioid-ligand-induced BDNF responses were not inhibited by the TLR4 blocker (Figure 8j), whilst (+)-naloxone (Figure 8i) was successful in inhibiting this response. The reason for the difference in these results compared to what we have published previously in numerous manuscripts examining proinflammatory products is intriguing. If we combine our results with these data here, it suggests that (+)-naloxone and LPS-RS are biased antagonists. That is, LPS-RS blockade of morphine responses in our experiments is capable of blocking the proinflammatory response, as is (+)-naloxone, but LPS-RS is not able to block opioid-ligand-induced BDNF release. Other explanations for the differences would require the authors to examine the proinflammatory parameters we assessed. I think this highlights the complexity of the interactions of opioid ligands with opioid receptors and innate immune receptors. Biased signaling and biased antagonists are all the rage in the GPCR field now. How GPCRs interact with other GPCRs and non-GPCRs, and how these interactions bias the signaling fingerprint, will be key to understanding opioid receptor and non-opioid receptor dependent actions.
3. Most concerning here is the misrepresentation of the C3H/HeJ mice. These mice are not deficient in functional TLR4; they merely have a point mutation in TLR4 that limits their NF-κB signal. Therefore, they are still capable of non-NF-κB signals (see Goodridge et al., 2007). They are a poor method of assessment for the role of TLR4 in mechanisms and should be avoided at all cost in research. They are the result of a long-unrecognized spontaneous mutation in the TLR4 gene, and as such are a second-rate research tool. Null mutant mice are available and should be employed ahead of these mice. The complexity of the TLR4 signaling system has not been adequately addressed by the use of this strain.
4. We have examined the development of opioid-induced hyperalgesia and tolerance previously in multiple strains of mice. In
Liu et al., 2011, we saw no significant development of hyperalgesia in any of the mouse strains (see Table 1), despite seeing this phenomenon in rats before. Critically, we also examined the development of tolerance in these animals. Here we saw that TLR4-null mutant mice (truly TLR4 deficient) were protected against the development of morphine tolerance, whilst MyD88 -/- mice were not (Hutchinson et al., 2008).
Our data suggest a TLR4-dependent, but non-MyD88 (and therefore likely non-NF-κB-dependent) mechanism is involved in the tolerance in our mice.
Why are there these discrepancies? I don't know. Ferrini et al. have now given us plenty to go and look at in our mice to see if our data agree with theirs.
The neuroimmunology of nociception, pain, and opioid response is very complex. The hypothetical model of opioid ligands binding to neuronal opioid receptors to produce their response is far too simplistic. Clearly, there are multiple cell-type and multiple receptor classes involved in this response. A nociceptive threshold response likely reflects the summation of the allosteric load of the nociceptive system. That is, both pro- and antinociceptive systems are in balance. Sometimes, the magnitudes of the positive and negative forces are extreme, but still represent the same behavioral response (positive and negative are equal). Other times, the system is beneficially or pathologically out of balance, presenting as either analgesia or hyperalgesia. How neuroimmunology contributes to this does appear to be on the "amplifier" side based on the existing data. However, there is also a wealth of data examining the inhibitory actions of the neuroimmune component, such as astrocyte control of glutamate homeostasis.
The overall purpose of this system in changing pain responses has been hypothesized to be linked to the illness response. This is a well-researched area, but why opioids engage this pathway remains a mystery at this stage.
References:
Phosphorylcholine mimics the effects of ES-62 on macrophages and dendritic cells.
Goodridge HS, McGuiness S, Houston KM, Egan CA, Al-Riyami L, Alcocer MJC, Harnett MM, Harnett W.
Parasite Immunol. 2007 Mar; 29(3):127-37. PMID: 17266740.
Proinflammatory cytokines oppose opioid-induced acute and chronic analgesia.
Hutchinson MR, Coats BD, Lewis SS, Zhang Y, Sprunger DB, Rezvani N, Baker EM, Jekich BM, Wieseler JL, Somogyi AA, Martin D, Poole S, Judd CM, Maier SF, Watkins LR.
Brain Behav Immun. 2008 Nov; 22(8):1178-89. Epub 2008 Jul 02. PMID: 18599265.
Exploring the Neuroimmunopharmacology of Opioids: an Integrative Review of Mechanisms of Central Immune Signaling and Their Implications for Opioid Analgesia.
HutchinsonMR, Shavit Y, Grace PM, Rice KC, Maier SF, Watkins LR.
Pharmacol Rev. 2011 Sep; 63(3):772-810. PMID: 21752874.
Naloxone-precipitated morphine withdrawal behavior and brain IL-1β expression: comparison of different mouse strains.
Liu L, Coller JK, Watkins LR, Somogyi AA, Hutchinson MR.
Brain Behav Immun. 2011 Aug; 25(6):1223-32. Epub 2011 Apr 03. PMID: 21447380.