Astrocytes: Dual Roles in Chronic Pain and Memory Formation

Astrocytes are glial cells, abundant in the central nervous system, that form a highly interconnected and organized network of cells.  They make simultaneous close contact with thousands of synapses, allowing them to rapidly respond to, and even orchestrate, neuronal activity in homeostatic and pathologic conditions. It is now clear, however, that the specific contribution of astrocytes is location- and context-specific. For example, in the spinal cord, astrocytes can become “activated” after pain-producing injuries and negatively contribute to increased neuronal excitability through the release of gliotransmitters such as adenosine triphosphate, D-serine, and L-lactate.  These complex cells are also, however, key positive contributors to processes such as memory formation. Astrocytes in the CA1 region of the hippocampus release L-lactate, which is taken up by neurons and triggers a series of downstream synaptic structural changes requisite for long-term memory. 

In this issue, Han et al. perform a series of elegant experiments providing a unifying view of how astrocytes in the hippocampus contribute to memory deficits after pain-producing neuropathic injury in mice and rats. In the first part of the paper, the authors demonstrate poor short-term spatial memory and reduced excitability of pyramidal neurons in CA1 of the hippocampus after spared nerve injury, confirming comorbid pain and memory deficits, similar to human studies. Coincident with this memory impairment, L-lactate levels and calcium responses from astrocytes were also suppressed. Importantly, infusion of L-lactate into the CA1 region attenuated pain behaviors and improved memory deficits, while inhibiting lactate synthesis in astrocytes, or transport into neurons, induced paw sensitivity in naïve mice. While these experiments provided evidence connecting L-lactate to pain and memory deficits, the contributory role of astrocytes remained unclear. The authors thus took advantage of a highly novel and specific approach to modulate astrocytes only in the CA1 region: viral-based Designer Receptors Exclusively Activated by Designer Drugs (DREADDs). DREADDs are recently developed tool G-protein–coupled receptors linked to excitatory or inhibitory G proteins, often referred to as “chemogenetics.” DREADDs are solely agonized by the drug clozapine-N-oxide, allowing for the specific activation or suppression of G-protein–coupled receptor–mediated signaling in the target cell. When paired with viruses with tropism for specific cell types (for example, AAV2/8 for astrocytes) that include cell-type specific promoters (such as GfaABC1D for astrocytes), the result is the ability to express DREADDs with high temporal and spatial control. Using DREADD-containing viruses with astrocyte promoters injected into the CA1 region, the authors demonstrate that activation of astrocytes increased hippocampal L-lactate and CA1 neuronal firing, improved pain-like behaviors, and reversed the spared nerve injury–induced memory deficit. In contrast, chemogenetic inhibition of astrocytes in naïve rodents caused pain-like behaviors and memory impairment. Together these experiments nicely demonstrate that CA1 astrocytes, and their L-lactate release, are both necessary and sufficient for pain modulation and memory formation.

The contribution of astrocytic L-lactate to memory formation is not a new concept, but rather has been demonstrated by several groups during the past decade.  One discrepancy between previous work and the current study is that inhibition of astrocytic L-lactate production or release or uptake in CA1 has previously been linked only to long-term, and not short-term, memory formation. Han et al. utilize just one assay for memory, novel object recognition, which is a test for short-term spatial memory, and demonstrate that astrocytic L-lactate directly affects this outcome. One intriguing possibility for this contradictory finding is that in the context of pain-producing injury, astrocyte L-lactate may have expanded functions. The addition of tests to assay long-term memory in the context of pain would further solidify this prospect.

In the second part of the article, Han et al. expand their studies to the locus coeruleus, the origin of noradrenaline-releasing fibers projecting to the hippocampus, among other areas of the brain. The authors demonstrate that activation of locus coeruleus neurons using a viral DREADD approach, or the administration of a β2-adrenergic agonist, lead to decreased pain-like behaviors, improved memory deficits, and enhancement of astrocytic L-lactate release in CA1. What remains unclear is how β2-adrenergic receptor activation leads to L-lactate release from astrocytes, especially given that recent transcriptome profiling of astrocytes fails to identify the transcript for this receptor in hippocampal astrocytes and there is variability in the immunostaining presented in the current article. It is therefore possible that locus coeruleus–derived noradrenaline release in the hippocampus acts on local neurons, which definitively express the β2-adrenergic receptor, and that the effect on astrocytes is indirect, through a change in neuronal excitability.

The most novel aspect of the current work by Han et al. is the connection to chronic pain, a condition that anesthesiologists are uniquely positioned to diagnose and treat. Astrocytes may in fact have a dual function in the case of chronic pain: in the spinal cord, their activation increases pain behaviors through the release of algesic mediators, while in the hippocampus, their activation enhances memory formation through the release of L-lactate. This suggests that astrocytes may both influence and respond to the excitability state of neurons in their direct environment. For example, it is possible that primary afferent neuron hyperexcitability, triggered by pain-producing peripheral injury, results in activation of astrocytes and feed-forward enabling of further neuronal firing. In brain circuits tied to memory formation, such as in the hippocampus, it is possible that suppressed CA1 neuron excitability leaves astrocytes in a more quiescent or homeostatic state, unable to initiate glycogenolysis and thus L-lactate release. The crucial next step highlighted by the current work, and the key to future translation, would be to identify specific pharmacologic agents that activate astrocytes to enhance memory formation and trigger downstream neuronal circuits to suppress pain, without affecting local spinal cord signals. Such a treatment would have important clinical implications as a lack of abuse liability would be inherent to targeting astrocytes, and not neurons.

Once thought of as mere support cells, astrocytes are now taking center stage as orchestrators of neuronal excitability and circuit dynamics. The current work by Han et al. makes a definitive step toward understanding how these cells contribute to memory deficits and supraspinal control of chronic pain. The integration of modern neuroscience tools into the study of astrocytes, as demonstrated by the current work, will no doubt continue to rapidly expand our knowledge on the breadth of unique location and context-specific functions of these complex cells.