Brain networking pain and anxiety: From basic mechanism to future treatment

Abstract

It is well known that pain and anxiety can enhance each other in both animals and humans. In case of chronic pain, patients often suffer anxiety and depression. Animal experiments provide important basic mechanisms for the interaction between chronic pain and anxiety. At cortical level, recent studies have consistently indicated that anterior cingulate cortex (ACC) and insular cortex (IC), two critical cortical regions for pain-related unpleasantness and suffering, are also involved in the process of emotional anxiety. At synaptic level, long-term potentiation (LTP), a key cellular mechanism for memory and chronic pain, has also been found to contribute to emotional anxiety in animal models of chronic pain. In a recent study published in Neuron by the group of Prof. Xu, it has been found that at subcortical level, anterior and posterior paraventricular nucleus of the thalamus (PVT) contribute to pain and anxiety through distinct projections to the basolateral amygdala (BLA) and central amygdala (CeA). In this review, I will first introduce the recent work by Prof Xu, and then discuss possible mechanisms at different levels for pain and anxiety in the condition of chronic pain, including chronic visceral pain. Some of medicines used in the current treatment will be analyzed, and potential future treatment for pain and anxiety in chronic pain conditions will be discussed.

Introduction

It has been long known that pain and emotion are highly interacted.^1,2 As human experience, physiological pain or acute pain can easily trigger anxiety and fear in both human and animals. Such anxiety and fear often serve as protective functions, and prevent potential injuries. ³ Furthermore, acute anxiety often contributes the motivation for future actions. However, in case of chronic pain caused by injuries, human often experience long-lasting fear and anxiety, especially when the medical treatment failed to control chronic pain. While animal models of anxiety and fear are often limited to completely mimic changes of emotional responses in humans, animal models of chronic pain have recently reproduced some enhanced anxiety and fear. The integrative neurological approaches such as genetic knockout, optogenetic methods, neuropharmacological and behavioral methods, have helped us to understand molecular insights for the long-term changes in the brain. In this minireview and commentary for a recent paper ⁴ from Prof. Xu, I will first briefly review recent progress of cortical and amygdala roles in chronic pain and anxiety, and then discuss special roles of posterior paraventricular nucleus of the thalamus (PVT) in relaying pain and anxiety. Synaptic and intracellular mechanism will be also discussed.

Cortical circuits involved in different types of pain

Brain imaging studies in humans have provided critical information about the cortical regions involved in different types of pain.^1,2,5 The anterior cingulate cortex (ACC) and other cortical areas, including insular cortex (IC), primary somatosensory cortex (S1), secondary somatosensory cortex (S2), and prefrontal cortex (PFC), are activated by noxious stimuli (Figure 1(a)). In healthy human subjects, the ACC is active and the amount of activation is correlated with the intensity of pain. ¹ Furthermore, activation of ACC is also related to emotional pain. This strongly suggests that the ACC is involved in signaling the unpleasantness associated with physical and emotional pain. In chronic pain conditions, ACC is also activated and excitatory synapses were potentiated. Animal studies expand our knowledge of such activation, and we also discovered that ACC is highly plastic in adult brains.^1,5 In animal models of phantom pain or neuropathic pain, ACC neurons and network are sensitized after peripheral injuries, and biochemical studies revealed that ACC glutamate signaling pathways are upregulated after injury. ¹ Furthermore, inhibiting ACC plasticity produces analgesic effects in different animal models of chronic pain.^6,7 Recently, increasing evidence have also suggested that these cortical areas contribute to chronic visceral pain as well as emotional pain. ⁸ These findings consistently indicate that ACC and its excitatory circuits are important for pain perception, and enhancement of excitatory synaptic transmission contribute to different types of unpleasant sensation after injuries.

Figure 1.

Brain pathways for pain and anxiety. (a) A simplified diagram shows the key pathways and cortical areas for pain (in red) and amygdala for anxiety (Blue) and (b) PVT network proposed for chronic pain and its related anxiety based on Prof. Xu’s work. Schematic diagram of the differential neural mechanisms by which the PVT mediates chronic visceral pain and anxiety. Different subregions of the PVT mediate visceral pain and anxiety-like behaviors through distinct neural circuits and molecular targets, with visceral pain indirectly promoting anxiety through the regulation of the CeA^GABA.

ACC: anterior cingulate cortex; BLA: basolateral amygdala; CeA: central amygdala; DRG: dorsal root ganglion; IC: insular cortex; PAG: periaqueductal gray; PVT: posterior paraventricular nucleus of the thalamus; RVM: rostroventral medulla; SDH: spinal dorsal horn.

In addition to ACC, the IC has also been identified as a key cortical region for pain. ⁹ In human brain imaging studies, the IC is also activated along with the ACC in various pain conditions.^1,2,10 Direct electrical stimulation of the IC in humans can elicit painful and somatic sensations, supporting the critical roles of the IC in pain and sensory perception. ⁹ Similar to the ACC, damage in the IC produce less pain suffering or smaller level of empathic pain in patients. ⁹ In animal experiments, it has been reported that lesions in the IC, increasing inhibitory transmission within the IC or unilateral microinjection of morphine into the IC produced analgesic in both acute and chronic pain conditions. Recent studies have also suggested that anterior IC may be necessary for empathic pain perception.^11,12 Human studies reveal that IC can be involved in enhanced painful responses to a noxious stimulus following pain and fear observation. ¹³ These results indicate that both ACC and IC are likely involved in somatic and emotional pain.

Amygdala: Central role in emotional anxiety

Investigation of brain mechanisms for anxiety have been the focus of many neuroscientists (see Pessoa, ¹⁴ LeDuke et al., ¹⁵ Phelps and LeDoux, ¹⁶ Johansen et al., ¹⁷ and Zhuo ¹⁸ for review; Figure 1(a)). Human and animal studies consistently indicate that amygdala is a key structure for emotional responses, such as anxiety and fear. Lesions in the amygdala result in an inability to recognize fearful stimuli, and electrical stimulation of the amygdala in humans generates feelings of fear and anxiety. Consistently, hyperexcitability of the amygdala in response to negatively stimuli has been observed in patients with several types of anxiety disorders, and this is reversed following successful treatment with cognitive behavioral therapy. Imaging studies have also shown engagement of the amygdala in anxiety disorders, and behavioral studies confirm the involvement of the basolateral amygdala (BLA) in anxiety-like traits in rodents. Thus, inactivation of the amygdala can block anxiogenic responses, and optogenetic stimulation of BLA projections to the ventral hippocampus can evoke anxiety-like states.

The amygdala consists of multiple subdivisions, and the BLA and central amygdala (CeA) are particularly critical for anxiety. Inactivating the BLA disrupts aversive and appetitive responses as well as anxiety behaviors, suggesting that BLA activity has a role in driving these distinct behavioral responses. The BLA receives sensory information from the thalamus, cortical association areas and PFC through the lateral nucleus (LA), processes this information in the basal nucleus (BA), and sends it to the lateral subdivision of the CeA. In parallel, inputs from the BLA directly excite the medial subdivision of the CeA. In response to excitation by the BLA, projection neurons of the lateral and medial CeA target and regulate multiple regions implicated in anxiety, including the periaqueductal gray (PAG), hypothalamus and dorsal vagal complex, to give rise to autonomic and motor responses. Thus, the excitatory output of the BLA to the CeA is translated into a behavioral reaction to aversive stimuli, including avoidance and freezing.

Roles of ACC and IC in anxiety

In addition to the amygdala, the involvement of ACC in anxiety has been also noted in some of previous studies ² (Figure 1). Recently human and animal studies have suggested the potential important contribution of ACC and IC to emotional anxiety and mood disorders. ⁹ Inhibiting ACC activity by microinjection of the inhibitory receptor agonist mucimol produced significant reduction of anxiety-like behaviors in mice. ¹⁹ Creson et al. ²⁰ reported that the ACC ERK pathway contributes to regulation of behavioral excitement and hedonic activity. Studies of oxytocin suggest that ACC synapses may serve as one central target for oxytocin in reducing anxiety. By applying oxytocin directly to the ACC, behavioral anxiety caused by nerve injury is significantly reduced. ⁷ By using electrophysiological and optogenetic approaches, Sellmeijer et al. ²¹ reported that hyperactivity of ACC neurons is critical for chronic pain-related anxiodepressive behaviors. Recently, there are reports of ACC and IC involvement in migraine-related anxiety. Liu et al. ²² reported that inhibition of AC1 activity within the ACC produced significant inhibition of behavioral anxiety in an animal model of migraine. Inhibition of AC1 in the IC also produced inhibitory effects on behavioral anxiety triggered by migraine. ²³ Interestingly, there are reports of contribution of ACC and IC to emotional anxiety. Benassi-Cezar et al. ²⁴ found that both ACC and IC may contribute to anxiety induced by emotional contagion in animals cohabiting with a conspecific suffering pain.

Cortical projections from ACC and IC: Roles in pain and anxiety

Recent studies using optogenetic approaches have revealed functional contribution of various cortex-related networks to pain and anxiety. For example, Li et al. ²⁵ reported that there is a direct synaptic projection from one side ACC to the contralateral ACC. Glutamate is the major excitatory transmitter for bilateral ACC connection, including projections to pyramidal cells in superficial (II/III) and deep (V/VI) layers of the ACC. Behaviorally, light activation of the ACC-ACC connection facilitated behavioral withdrawal responses to mechanical stimuli and noxious heat. In an animal model of neuropathic pain, light inhibitory of ACC-ACC connection reduces both primary and secondary hyperalgesia. However, emotional anxiety in different animal tests were not affected. These results suggesting the ACC-ACC projections preferentially affect pain transmission.

Furthermore, Hao et al. ²⁶ reported that direct excitatory glutamatergic projections from the RSC to the ACC in mice. This excitatory transmission is predominantly mediated by the postsynaptic AMPA receptors. Furthermore, the activation of the retrosplenial cortex (RSC)-ACC projections through opto-/chemogenetics significantly facilitated the behavioral responses to both mechanical and thermal nociceptive stimuli in adult mice. Notably, this activation did not anxiety-like or aversive behaviors. These findings indicate that the RSC-ACC glutamatergic pathway modulates nociceptive perception primarily at the supraspinal cortical level.

There are several recent studies related to the IC projections in pain and depression. Zhang et al. ¹¹ reported that glutamatergic projection from the IC to the BLA is critical for the formation of observational pain. The selective activation or inhibition of the IC-BLA projection pathway strengthens or weakens the intensity of observational pain, respectively. Recently in an animal model of neuropathic pain, Chen et al. ²⁷ found that synaptic potentiation of the IC→BLA and IC→thalamus projections may be important pathophysiological bases for hyperalgesia and depression-like behavior in neuropathic pain.

Thalamus: A relay for somatic and visceral pain

Thalamus is a key subcortical nucleus to convey somatosensory sensory information to the amygdala and cortical areas, including inputs from body surface (somatosensory areas) and internal organs (visceral).^28–30 The thalamus, located in the middle of the brain, and can be anatomically subdivided into several major nuclear regions, including the anterior, ventral, medial, lateral, midline, intralaminar, medial and lateral geniculate body nuclei. One of important cells that are critical for ascending pain transmission is spinothalamic tract (STT) cells. ²⁸ Many of these cells show specific responses to peripheral noxious stimuli, including visceral stimuli.^31,32

Previous studies have consistently demonstrated that the projection from the thalamus to the ACC plays important in pain perception. Electrophysiological experiments using in vivo thalamic stimulation in rats have confirmed the existence of thalamic-anterior cingulate projections. Electrical stimulation of mediodorsal (MD), midline, and intralaminar thalamic nuclei induced short-term plastic changes in layers II/III of the ACC that transmit nociceptive information to the ACC in early stage of chronic pain. Using anatomic dye labeling methods, it has been reported that the ACC receive a large proportion of projection fibers from the thalamus, including anteromedial, ventroanterior, mediodorsal, and submedial, ventrolateral, and medial dorsal thalamic nuclei. Recently, Xue et al.^33,34 reported that cortical neurons in the ACC receive direct afferents from these thalamic nuclei in adult mice; suggesting that the thalamus plays important role in relaying sensory nociceptive information to the ACC (Figure 1(a)).

PVT: A possible sorting center for chronic visceral pain and anxiety

Among several subnuclei of thalamus, recent studies suggest that the paraventricular nucleus of the thalamus (PVT) is a critical component of the neural pathways related to sensory modulation (including pain) and emotional responses such as anxiety and depression. ³⁵ The PVT can be divided into two distinct subregions, anterior PVT (aPVT) and posterior PVT (pPVT). However, it is unclear how they may act differently when pain and anxiety coexist. Recently, Li et al. ⁴ published a pivotal study in Neuron, making significant progress in understanding the comorbidity of chronic visceral pain and anxiety. They discovered that offspring mice subjected to prenatal maternal stress (PMS) exhibited both chronic visceral pain and anxiety-like behaviors, more closely mirroring the symptoms observed in clinical patients. Their findings identified the PVT as a central “sorting center” for processing chronic visceral pain and anxiety-like behaviors. They found that the PVT exhibits subregion-specific responses to visceral pain and anxiety, along with distinct molecular expression profiles.

Similarly in the case of somatic neuropathic pain, Deng et al. ³⁶ reported that aPVT-NAC projection contribute to injury induced depression like behaviors, while pPVT-PAG projection contribute to the possible descending modulation of pain-like behaviors responses. There two studies consistently indicate that aPVT and pPVT could contribute to pain and emotion (anxiety or depression) via distinct projections. However, a recent study from Tang et al. ³⁷ found that PVT glutamatergic projections to BLA contribute to both pain and anxiety in a mouse model of neuropathic pain. These studies indicate that aPVT and pPVT neurons may act differently in the pain and anxiety in different animal models of chronic pain (such as chronic visceral pain vs neuropathic pain). Interestingly, a report by Mindaye et al. ³⁸ also found that aPVT projections to BNST and NAC could contribute to affective and sensory aspects of pain in an acute inflammatory model.

Different projections from PVT: A new finding

Li et al. ⁴ further showed the downstream projection patterns of PVT neurons activated by these two conditions differ markedly (Figure 1(b)). Visceral pain activates glutamatergic neurons in the anterior PVT (aPVT^Glu), which predominantly project to the basolateral amygdala glutamatergic neurons (BLA^Glu), forming the aPVT^Glu-BLA^Glu circuit that modulates visceral pain behavior. In contrast, anxiety activates the posterior PVT glutamatergic neurons (pPVT^Glu), which primarily project to the central amygdala GABAergic neurons (CeA^GABA), establishing the pPVT^Glu-CeA^GABA circuit responsible for mediating anxiety-like behavior. Importantly, they showed that the visceral pain-associated aPVT-BLA neural circuit can indirectly modulate anxiety-like behavior through CeA^GABA neurons. These findings provide critical insight into the neurobiological mechanisms by which visceral pain may exacerbate anxiety.

Synaptic LTPs as potential synaptic model for pain and anxiety after injury

LTP of sensory synapses in the spinal cord dorsal horn and cortical synapses in the ACC are thought to contribute to behavioral sensitization during chronic pain in different animal models. Especially, postsynaptic amplification of excitatory responses by different subtype of glutamate receptors are thought to contribute to allodynia and hyperalgesia.^1,5 In addition to postsynaptic potentiation (post-LTP), the enhancement of transmitter release from presynaptic terminals have also been reported after the injury, and presynaptic form of LTP (pre-LTP) has been found to contribute to injury-related emotional changes such as anxiety^1,35 (Figure 2). Thus, the investigation of signaling mechanisms for pre-LTP in the ACC (maybe IC as well) provides basic mechanisms for long-term anxiety in different pathological conditions including chronic pain.

Figure 2.

Synaptic mechanisms and signaling mechanisms for chronic pain and anxiety.

There are at least two forms of LTP in the ACC: presynaptic LTP and postsynaptic LTP. The two forms may occur at the same synapses. Induction of postsynaptic LTP requires the activation of NMDARs, and the maintenance of postsynaptic LTP expression requires the postsynaptic activity of an atypical PKC isoform (most probably PKMζ). AC1 is also critical for the induction of postsynaptic LTP in the ACC. Activation of cAMP-dependent protein kinase (PKA) drives the insertion of CP-AMPARs (GluA1 homomers). PKA also translocates to the nucleus where it phosphorylates the transcription factor CREB, leading to the synthesis of several downstream plasticity proteins, including FMRP, and BDNF. PKMζ may maintain LTP by upregulating GluA1/2 heteromers. In addition, increased FMRP may enhance the postsynaptic function of AMPARs as well as NMDARs. BDNF may bind to its receptor TrkB located postsynaptically (maybe presynaptically as well) to contribute to the amplification of AMPARs. AC1 activity is found to be required for BDNF produced potentiation. Presynaptic LTP is not dependent on NMDARs; instead, presynaptic kainate (KA) receptors and AC1 activation are necessary for its induction, and the expression of presynaptic LTP may require the activity of HCN channels. cAMP binds to the HCN channel to increase its sensitivity, and PKA enhances vesicle fusion. Oxytocin could contribute to the regulation of pre-LTP through inhibitory neurons in the ACC.

Potential brain circuit mechanisms for pain and anxiety

The current study by Prof. Xu and recent work using similar optogenetic approaches have increasingly reveal our understand of brain network that are involved in different types of pain and anxiety. For example, for chronic visceral pain, Li et al. ⁴ found that aPVT-ACC selectively contribute to pain sensitization. Considering a recent work that revealing ACC-spinal descending facilitation, ³⁵ it is quite likely that this positive feedback loop will serve a brain network for prolonged neuronal excitation during chronic visceral pain (see Figure 3). Furthermore, the bilateral monosynaptic excitatory connections between ACCs in two hemispheres will add additional complexity of this network. For emotional anxiety, Li et al. ⁴ found that pPVT-amygdala contribute to anxiety in chronic visceral pain. It has been reported that amygdala also form excitatory connection with the ACC. Thus, it is potentially possible that ACC-spinal descending facilitation (a pathway that is mainly for pain modulation) can be also recruited to enhance neuronal excitation in pPVT; and contribute to the maintenance of long-term anxiety (see Figure 3).

Figure 3.

ACC/IC/amygdala neuronal circuits for pain and anxiety.

Simplified network diagram showing bilateral pathways for pain and anxiety, from the spinal cord to the cortex. ACC neurons receive nociceptive inputs from the thalamus. Studies found that ACC neurons also receive inputs from the amygdala and somatosensory cortices (not shown). By using optogenetic method, recent studies showed that neurons in the thalamus and amygdala also receive excitatory inputs from the ACC, suggesting feedback control loops exist among these areas. Furthermore, left ACC neurons also receive excitatory innervation from right ACC; and vice versa. These bidirectional positive feedback coops contribute to pain facilitation, and possibly emotional anxiety and fear as well. Work from Boshakov’s group found similar feedback loops between two side of amygdala. ACC neurons located in the deep layers send their projections directly or indirectly through the RVM to the dorsal horn of the spinal cord. This top-down circuit allows cortical neurons to directly modulate the sensory input into the central nervous system. While top-down modulation from the ACC is mostly facilitatory, descending modulation from the RVM is biphasic, and the PAG (not shown) may contribute to descending inhibition of spinal nociceptive transmission.

Intracellular signaling mechanisms

The discovery of important role of pre-LTP in the ACC in anxiety has provided several drug targets for future interventions^18,39 (Figure 2). The first candidate is kainate (KA) receptors. Inhibiting KA GluK1 receptors may reduce chronic pain-related anxiety by reducing or inhibiting pre-LTP. A possible concern is the side effect of inhibiting KA receptors in other systems. KA receptors have also been implicated in amygdala regulation and plasticity. ³ Thus, depending on the location of excitatory vs inhibitory synapses, the effects of general inhibition of GluK1 by systemic application may be mixed. Second, the adenylyl cyclase subtype 1 (AC1) inhibitor NB001 is another attractive candidate. ⁴⁰ AC1 and other signaling molecules serve as potentially novel, neuronal selective targets for pain and anxiety. At the molecular level, AC1 has been found to be a key second messenger enzyme. NB001 has also already been shown to reduce chronic pain in different animal models. Furthermore, NB001 has also been shown to reduce IBS induced anxiety and spontaneous pain. ⁴¹ Thus, inhibiting AC1 may serve to reduce chronic pain as well as anxiety. Third, HCN channel inhibitors can erase pre-LTP, ¹³ and may be useful even in the late stage of chronic anxiety. Considering the existence of different subtypes of HCN channels, it would be useful to have selective inhibitors targeted at subtypes that are mainly expressed in neurons. Finally, Li et al. ⁷ reported that microinjections of oxytocin into the ACC attenuate nociceptive responses and anxiety-like behavioral responses in animals with neuropathic pain. Application of oxytocin selectively blocks the maintenance of pre-LTP but not post-LTP. In addition, oxytocin enhances inhibitory transmission and excites ACC interneurons. It is thus likely that oxytocin acts on central synapses and reduces chronic-pain-induced anxiety by reducing pre-LTP.

Current drugs for the treatment of chronic pain and anxiety

Current drugs for treating anxiety and pain are mostly based on the mechanisms to inhibit excitatory transmission or enhancing inhibitory transmission within the central nervous systems. Some of these drugs act through inhibiting calcium channels or enhancing GABA receptor mediated functions. In addition, several G protein couple receptor targeted drugs act through regulating excitatory synaptic transmission or regulating the release of transmitters. For example, Selective Serotonin Reuptake Inhibitors (SSRIs) or Serotonin-Norepinephrine Reuptake Inhibitors (SNRIs) are used to treat anxiety, such as Fluoxetine (Prozac), Venlafaxine (Effexor XR), and Duloxetine (Cymbalta). For those drugs acting through GABAergic inhibitory transmission, they are benzodiazpins such as Clonazepam (Klonopin) and Diazepam (Valium). Related to this group of medicine, Gabapentin is also commonly used to treat anxiety as well as neuropathic pain.

On the preclinical and early clinical studies, a selective inhibitor for AC1 has been continuously found to be useful for the treatment of chronic pain and anxiety in preclinical studies. ⁴² For example, AC1 inhibitor NB001 has been reported to reduce spontaneous pain in a animal model of irritable bowel syndrome (IBS). ⁴¹ In animal model of migraines, AC1 inhibitor NB001 applied orally or locally into the ACC produce significant inhibition of behavioral anxiety and migraine-related pain.^22,23 In support of these behavioral results, NB001 is found to inhibit pre-LTP in both ACC and IC.^39,43 In an animal model of Parkinson’s disease, NB001 is also found to produce significant inhibition of pain and anxiety. ⁴⁴ These results consistently suggest that by targeting at anxiety- and injury-related synaptic plasticity (pre- and post-LTPs), we may be able to effectively control and pain and anxiety with less side effects on other brain functions.

Future direction

In summary recent progress made in neuroscience of pain and anxiety by using integrative experimental approaches, such as genetic manipulation, electrophysiology, pharmacological, anatomic, optogenetic approaches have greatly increased out knowledge of basic mechanisms of pain and anxiety in different chronic pain conditions. The current work by Prof. Xu added a new information for separate pathways from PVT to ACC and amygdala respectively. By combining recent findings from other laboratories including some of my own laboratories, we now have a better picture of different, and often complicated brain networks that contribute to pain and anxiety. This information is essential for our understanding of chronic pain pathology in humans. Animal experiments will be always important for our understanding of pain, emotion, depression, and consciousness. AI and model nerve cells alone cannot replace these classic animal experimental approaches.

Future studies are greatly needed to understand how these brain network work in physiological and pathological conditions. Molecular and cellular mechanisms at each synapse of these brain networks are also needed. We hope that the combination of these experimental approaches and recent findings will facilitate our understanding of chronic pain, and help us to design better medicines or instruments for patients who suffer from pain and anxiety.