Marijuana soothes a variety of ills, including pain, but it also wreaks havoc with cognition. According to a study published November 21 in Cell, this unwanted side effect and its attendant changes in the brain may be avoided with inhibitors of cyclooxygenase-2 (COX-2), an enzyme that spurs inflammation. Led by Chu Chen of Louisiana State University Health Sciences Center, New Orleans, US, the study reports that, in mice, disruptions to memory, synaptic plasticity, and dendritic spines induced by marijuana’s psychoactive ingredient, Δ9-tetrahydrocannabinol (Δ9-THC), could be prevented by inhibiting COX-2. The researchers also found that COX-2 inhibition did not undermine the beneficial effects of Δ9-THC in a mouse model of Alzheimer's disease.
Though the relevance of these findings to pain is unclear, they raise the tantalizing possibility that COX-2 inhibitors could mitigate problematic side effects of cannabinoids given for pain relief. “It’s probably too simple to think that COX-2 inhibitors can block everything that is bad,” Alfonso Romero-Sandoval of Presbyterian College, Clinton, South Carolina, US, told PRF. “But there are many things in this paper to think about for pain.”
For example, Romero-Sandoval has found that tolerance develops to pain-relieving doses of cannabinoids in rats (Alkaitis et al., 2012). Noting how Δ9-THC raised COX-2 levels in the brain in the new study, he hypothesizes that blocking COX-2 might prolong pain relief by cannabinoids. “I think tolerance is one of the main unwanted effects of pain drugs,” he said.
Cannabinoids and COX-2 have independent, yet interacting, roles in pain. Activation of cannabinoid receptors (CBRs), either by exogenous cannabinoids such as Δ9-THC or by endogenous versions such as anandamide and 2-arachidonoyl-glycerol (2-AG), dampens activity in neurons transmitting pain information. COX-2 produces inflammation-promoting prostaglandins, and common painkillers such as ibuprofen inhibit COX-2 activity.
But COX-2 metabolizes endocannabinoids, too, so that inhibiting COX-2 produces pain relief both by lowering prostaglandins and by boosting endocannabinoids (see PRF related news story). Cannabinoids and COX-2 inhibitors can also act synergistically to produce pain relief greater than that when given alone (Ahn et al., 2007).
Though Chen’s study did not address pain, it did track several changes in the brain that may follow any therapeutic use of medical marijuana. “An over-the-counter painkiller such as ibuprofen, which blocks COX-2, might be used to eliminate these side effects,” Chen said.
Others are concerned that the high doses of Δ9-THC used in the study limit its translational potential. “For these kinds of studies where a link is being made with a human condition, it becomes imperative to use doses of Δ9-THC that are as close as possible to those used for recreational or therapeutic purposes,” said Vincenzo Di Marzo, who studies endocannabinoids at the National Research Council in Pozzuoli, Italy. Di Marzo noted that the mice received doses five to 10 times higher than a heavy recreational user would be exposed to, and the in-vitro experiments used Δ9-THC concentrations 100 times higher than the affinity for CB1Rs.
Chen said he wanted to see whether COX-2 inhibitors could mitigate extreme effects of Δ9-THC before exploring the realm of lower, therapeutic doses. “It is an important issue that we will address in future experiments,” he said.
Pot pitfalls
In the study, first authors Rongqing Chen, Jian Zhang, Ni Fan, Zhao-qian Teng, and Yan Wu initially found that injecting rats with Δ9-THC increased COX-2 expression in neurons and glia in the hippocampus, a memory center. A twofold increase was maintained after seven days of daily Δ9-THC injections, and this depended on activation of CB1Rs.
This weeklong regimen of Δ9-THC injection had several downsides in the brain. Synaptic plasticity declined, with recordings in hippocampal slices revealing decreased synaptic responses to a long-term potentiation (LTP) protocol in Δ9-THC-treated compared to vehicle-treated mice. Dendritic spines, the bulbous processes that receive messages from other neurons, were sparser in Δ9-THC-treated mice and contained fewer glutamate channels than those in controls. Behaviorally, Δ9-THC-treated mice showed disrupted fear and spatial memory: Prior to Δ9-THC injections, the mice learned to associate a stimulus with a foot shock, or the location of a hidden platform in the Morris water maze, but after a week of Δ9-THC treatment, they did not remember what they learned as well as controls did.
Inhibiting COX-2 in combination with Δ9-THC treatment, however, averted these changes: LTP, spine density, and glutamate channel subunit clusters were higher than in mice treated with Δ9-THC alone. The combination-treated mice also showed relatively normal memory. The researchers also found that COX-2 knockout mice did not show these deficits when treated with Δ9-THC.
COX-2 inhibition did not degrade the benefits of Δ9-THC in a mouse model of Alzheimer's disease that develops amyloid plaques in the brain. Daily treatment with Δ9-THC alone for a month decreased the density of plaques and dying neurons in the cortex and hippocampus compared to vehicle treatment. Combining Δ9-THC treatment with Celebrex®, a commercially available COX-2 inhibitor, resulted in similar decreases in plaque and dying neuron density, which suggests that having COX-2 inhibitors on board would not interfere with Δ9-THC’s effect.
The study also highlights a contribution by glia to Δ9-THC’s reworking of neuronal connections: Astroglia showed COX-2 levels four times that found in neurons when treated with Δ9-THC. COX-2 activity in both cell types would increase prostaglandins, which are known to elevate extracellular glutamate (e.g., Bezzi et al., 1998). This deluge of glutamate could trigger an adaptive reduction in glutamate channels and dendritic spines, thus laying the groundwork for cognitive difficulties.
Michele Solis is a science writer and former neuroscientist who lives in Seattle, Washington, US.
Image: ©iStockphoto.com/skydie