Kate L. White School of Veterinary Medicine and Science, University of Nottingham, Nottingham, UK Pain is a complex, multidimensional experience involving sensory and affective components. In layperson terms, “pain is not just about how it feels, but how it makes you feel,” and it is the unpleasant feelings that cause the suffering we associate with pain. Pain is a uniquely individual experience in humans, the pain that one individual feels associated with an injury may differ greatly from that experienced by another, both in its intensity and in how it is perceived and felt. This is evidenced from almost every clinical trial report of a new analgesic regimen, even when confounding factors are well controlled. Pain is officially defined by the International Association for the Study of Pain (IASP) as an unpleasant sensory and emotional experience associated with, or resembling that associated with, actual or potential tissue damage. The definition is expanded upon with the addition of six key notes and etymology of the word pain for further context [1]. Furthermore, the nature of pain is variable across many situations. The stories of pain experienced in traumatic situations clearly illustrate that the time course of pain and its impact on our feelings and how we behave are not directly linked in time. This is well explored by Patrick Wall in Pain: The Science of Suffering [2]. Among many illustrations of the complexities of pain, he describes the experience of Harry Beecher, a young medical officer treating wounded troops admitted to a hospital on the beachhead at Anzio in 1944 during the Second World War. Seriously wounded soldiers admitted to the hospital over a period of 4 months were asked “are you in pain; do you want something for it.” The answer was “no” to both questions in 70% of cases. Later, after the War, he asked the same two questions of a group of age‐matched men who had undergone surgery at the Massachusetts General Hospital in Boston and 70% answered “yes” to both questions. He concluded that something about the context in which tissue damage occurred influenced the degree of pain suffered. The lack of pain in the early time course of some traumatic injuries is often followed by pain reports within 24 hours; pain is rarely absent over time. The well‐known phenomenon of phantom limb pain reveals the contrary position – in a large percentage of people who have amputations, pain is present chronically in the area of the amputated limb, although healing is complete. Clearly, the onset of disease and pain as a consequence of this is less dramatic and immediate than in the traumatic situations described above and reflects the more common triggers and causes of pain in humans. The nature of pain is equally complex in animals, although all aspects of its experience and expression are not likely to be identical. The physiology and pathophysiology of pain are remarkably similar and well conserved across mammalian species, and the capacity of animals to suffer as sentient creatures is well established and enshrined in law in many countries. The majority of pain experiences stem from activity in nociceptive primary afferent (sensory) neurons caused by thermal, mechanical, or/and chemical stimuli. The membrane of the peripheral terminal is populated with transduction proteins capable of responding to the stimuli. Transduction is defined as the process where a harmful stimulus is converted into an electrical signal [3–6]. The transducers or ionotropic receptors can crudely be described by the type and origin of stimulus (mechanical, thermal, and chemical), their tissue type association, and whether they are extrinsic or intrinsic to the primary afferent. However, thereafter the classification breaks down because these transducers lack a single criterion to reliably identify them [7,8]. Additionally, as a population they share anatomical, biochemical, physical, and functional heterogeneity, which may have hindered efficacy of therapy and contributed to a failure of translation of preclinical data to clinical use. So, for example, a selection of the mechanosensory and thermosensory transduction channels in mammalian skin is outlined in Table 46.1. The activators and modality of the channels is characterized in the table but in vivo the activator may also initiate other types of cells to release chemicals and the functional implications of these concurrent events are still to be properly understood. This area of research is rapidly expanding, and the polymodality of receptors makes classification challenging. Chemical or chemo tranducers are well characterized, numerous, and can respond to exogenous and endogenous compounds. The direct form of chemotransducer is an ion channel that has a binding site for a ligand; also known as ionotropic receptors. An indirect form known as a metabotropic receptor also exists, which is slower, with the chemotransduction driven by an intracellular signaling cascade [9]; see below. The cloning of the transient receptor potential cation vanilloid subfamily member 1 (TRPV1) was pivotal not only in furthering the understanding of chemotransduction in nociceptive afferents, but also in thermotransduction [10,11]. Other mammalian transient receptor potential (TRP) family receptors, TRP subfamily ankyrin (TRPA), and TRP subfamily melastatin (TRPM), share a thermotransduction role with a number of the two‐pore potassium channels (K2P) [12,13], as well as the Nav1.8 channel [14]. Despite the overwhelming evidence that mechanosensation is the most common component of somatosensation, it is the least well understood. The transducers known to play a role in mammalian mechanotransduction include the Piezo channels, TRPA1 and TRPV4, transient Ca2+ channels, and at least three K2P channels [15–20]. There will be overlap of the contribution of the mechanosensitive properties with the chemosensitive properties as chemicals are released from cells in response to mechanical stimuli. This overlap and lack of transducer specificity makes interpretation from intact preparations difficult. The definition of sensitization is a decreased threshold for response, and/or can also be characterized by an increased response to a suprathreshold stimulus. Additionally, an increase in spontaneous activity can occur in the nociceptors, and an increase in receptor field size in the dorsal horn can be detected [21,22]. Peripheral sensitization refers to an increased responsiveness in the nociceptive neurons in the periphery, whereas central sensitization refers to an increased responsiveness of nociceptive neurons in the central nervous system (CNS) to their normal or subthreshold afferent input. It is within these spinal neurons that “wind‐up” (the progressive repeated stimulation of C fibers) occurs [23,24]. All primary afferents release glutamate, and the three inotropic receptors for glutamate (N‐methyl‐D‐aspartate [NMDA], α‐amino‐3‐hydroxy‐5‐methyl‐4‐isoxazoleproprionic acid [AMPA], and kainate) are present in the dorsal horn. Wind‐up can be considered a product of the temporal summation of NMDA and neurokinin‐1 (NK1) receptor‐mediated cumulative depolarizations caused by primary afferent release of glutamate and the excitatory neuropeptide substance P, respectively [25–27]. Temporal summation describes an increased perception of pain in response to repetitive painful stimuli. Sensitization is a neurophysiological construct that can only definitively be determined if both the neural input and output are known. Clinically, the term sensitization is inferred in cases of reported allodynia (pain due to a stimulus that would not normally provoke pain) and hyperalgesia (increased pain from a stimulus that normally provokes pain). Both these terms do not infer a mechanism per se [28,29]. In inflammatory states, a large number of endogenous factors released in the vicinity of the free endings of the nociceptors can cause sensitization [25]. The nociceptors express characteristic patterns of the different inflammatory mediator receptors; guanine nucleotide binding protein‐ (G‐protein) coupled receptors (GPCRs), ligand‐gated ion channels, tyrosine kinase, or cytokine receptors in response to the inflammation. These indirect forms of chemotransducers known as metabotropic receptors operate more slowly (compared to the inotropic receptors), with the chemotransduction driven by an intracellular signaling cascade [9]. The commonest example of this metabotropic receptor is the GPCR, which includes the μ‐, κ‐, and δ‐opioid receptors, which can be targeted for pain relief [30]. There is plentiful evidence of chemicals being released from thermally and mechanically stimulated tissue too, making this an overlapping and complicated area of study that is being fervently researched. There is an ever‐increasing list of metabotropic receptors in the sensory neurons. These include receptors for cytokines, chemokines, and neurotrophins, albeit the number normally active is small, including those for bradykinin (B1 and B2) [31], histamine (H1) [32], ATP [33,34], endothelin‐1 [35], proteases (PAR‐2) [36], and prostacyclin (PGI2) [37]. This area of study is further complicated by the fact that a neuron can express ionotropic and metabotropic receptors for the same ligand [38]. Table 46.1 A selection of mammalian mechanosensory and thermosensory transduction channels associated with the skin. TRP, transient receptor potential; TRPA, transient receptor potential subfamily ankyrin; TRPC, transient receptor potential subfamily canonical; TRPM, transient receptor potential subfamily melastatin; TRPV, transient receptor potential subfamily vanilloid; DEG/ENaC, degenerin/epithelial sodium channel; ASIC, acid‐sensing ion channel; VGNC, voltage‐gated sodium channel; PUFA, polyunsaturated fatty acid; 4αPDD, 4α‐phorbol 12,13‐didecanoate; K2P channel, two‐pore domain potassium channel; TREK, TWIK subfamily K (where TWIK, tandem of P‐domains in a weak inward rectifying channel); TRAAK, TWIK subfamily K member 4; H+, protons. Source: Adapted from Lumpkin and Caterina [7]. The resting potential of the nociceptive afferent is −40 mV, decreasing to between −50 and −75 mV at the cell body. The action potential (AP) threshold range is between −35mV and −55mV [72]. For the sensory information to be propagated to the CNS, an AP must be generated by membrane depolarization caused by the transduction process. This event is termed a “generator potential” (GP) and can be initiated in three main different ways. First, via the opening of an ion channel which pushes the membrane potential above threshold and causes an AP that can be propagated toward the CNS. One such example of this is the TRPV1 channel, which can be activated by both thermal and chemical stimuli [7,73]. Second, the GP can be indirectly elicited by the closing of K+ channels (potassium family subfamily K member 2 (TREK‐1), potassium family subfamily K member 3 (TASK‐1), and others) normally responsible for hyperpolarizing currents. TREK‐1 channels can be activated by mechanical stimulation, hydrogen ions, and warm temperature demonstrating their polymodal behavior [63]. Third, also indirectly, via activation of low‐threshold voltage‐gated ion channels that are in close association with ion channels capable of driving membrane depolarization [74]. Transduction at the nociceptor terminal results in a GP that travels passively to the spike initiation zone. At this point, an AP can be generated that is transmitted via nociceptive afferents. Propagation of the AP along the axon of the nociceptor is caused by the initial depolarization that spreads in one direction as adjacent voltage‐gated sodium channels (VGSCs) open. Primary sensory neurons exhibit a rich heterogenous morphology and the most common way to classify these neurons is based upon their speed of conduction [75]. The Aα/β group are myelinated axons with the fastest speed of conduction (33–75 m/s). The Aδ group are thinner myelinated fibers with an intermediate speed (5–30 m/s), and the C fibers are the smallest unmyelinated fibers with the slowest conducting speed (0.5–2 m/s). Most of the neurons conducting within the Aα/β group do not encode noxious stimuli but respond to innocuous mechanical stimuli and are classified as low‐threshold mechanoreceptors. Most of the Aδ and C fibers encode noxious stimuli and are classified as nociceptors. The dorsal horn is divided into parallel laminae based on the packing density of the neurons [76]. The termination of the primary afferents in the dorsal horn laminae is determined by fiber diameter and receptive field modality [77,78]. Smaller diameter nociceptive afferents synapse with nociceptive‐specific (NS) cells in laminae I–II of the superficial dorsal horn, with a small number terminating deeper in the spinal cord. Lamina I neurons contribute to the spinothalamic and spinoreticular tracts both responsible for processing pain and temperature information [79,80]. Lamina II (substantia gelatinosa) is the major site for nociceptive inputs (C and Aδ fibers) arriving at the spinal cord [81] and consists mainly of densely packed unmyelinated interneurons [82]. Lamina II has been shown to have explicit organization of modules of neurons that differentially modify and transmit the afferent input from the Aδ and C fibers [83]. Lamina II central neurons, with C fiber input, were shown to excite lamina II vertical neurons with Aδ input. Lamina II outer vertical neurons with Aδ input excited lamina I neurons [84]. The function of this arrangement is postulated to be an ability of the afferent input to be amplified or modulated. C fibers can be divided into two groups, based on phenotype and sensitivity to neurotrophins, although functionally the differentiation is less clear. One group is sensitive to nerve growth factor (NGF) and usually contains substance P, calcitonin gene‐related peptide (CGRP), and galanin (peptidergic fibers); the second group is distinguished by binding sites for lectin isolectin B4 and usually possesses the purinergic receptors P2X3 and P2Y1 (non‐peptidergic fibers) [85–87]. The central projections of these two populations of C fibers differ in their arborization, the peptidergic C fibers branching in lamina I and the outer portion of lamina II with a small number extending to deeper terminals. The non‐peptidergic fibers project to the central portion of lamina II [88]. The non‐peptidergic C fibers are chiefly associated with the superficial layers of the skin [89], whereas peptidergic fibers tend to innervate other tissues in addition to the deeper layers of the skin [90]. In summary, the simplistic classification of C and A fiber nociceptors on morphology and conduction speed fails to capture the extensive diversity of primary afferents. Laminae III–VI are known as the deep dorsal horn and receive less nociceptive input. Transmission of innocuous stimuli is predominantly through large diameter, myelinated Aβ fibers, which terminate in laminae III–VI, hence within these laminae are proprioceptive neurons responding exclusively to touch [78]. In addition to the NS and the proprioceptive neurons, a third class of spinal neurons is described, known as wide dynamic range (WDR). The WDR neurons can receive noxious or non‐noxious input from Aδ, Aβ, or C fibers and respond in a graded manner (i.e., frequency of APs) from low through to high threshold noxious input. The WDR neurons (also known as trigger, lamina V type, Class 2, multireceptive, or convergent neurons) are found in superficial layers but more so in lamina V and include interneurons involved in the polysynaptic reflexes and projection neurons [91]. For Aδ fibers terminating in lamina V, the interneurons they synapse with decussate and ascend in the spinothalamic tract, while those terminating more superficially synapse with interneurons forming the spinoparabrachial tract (Fig. 46.1). The ventral and intermediate laminae are less important in terms of inputs from primary afferents. Laminae VII and X receive a proportion of nociceptive input that is mainly visceral in origin [92,93]. Muscle afferents feed into laminae V, VI, and VII. The primary afferents form excitatory synapses with either WDR projection cells (convey information to various parts of the brain) or interneurons (responsible for local circuitry within the spinal cord; an example is represented in Fig. 46.1 connecting the WDR neurons of lamina V with the superficial dorsal horn). All primary afferents use glutamate as their primary excitatory synaptic transmitter, although arrangements do differ. Interneurons can be classified as either excitatory (predominantly glutaminergic) or inhibitory (mainly γ‐aminobutyric acid‐ (GABA‐) or glycinergic). The morphology of the interneurons has been most closely studied in lamina II, with a classification scheme that includes islet, central, radial, and vertical interneurons [83,94–96]. An alternative approach to classification is to use immunocytochemistry and stratify the interneurons based on neuropeptide and protein localization [97,98]. Many neuropeptides have been identified, and consequently, several neuropeptide receptors expressed by the interneurons are being studied, including the tachykinin receptors, neurokinin‐1 (NK1) and neurokinin‐3 (NK3); the μ‐opioid receptor‐1 (MOR‐1); the somatostatin‐2A receptor (sst2A); and the neuropeptide Y‐1 receptor (Y1). It is not possible to use the neuropeptides to definitively define populations of interneurons, although some markers show a restricted distribution. The dorsal horn also receives input from axons descending from various parts of the brain (Fig. 46.1). All three glutamate receptors (AMPA, NMDA, and kainate) are found in the dorsal horn. The GABAA and glycine receptors are distributed throughout the spinal cord and are most likely expressed on all dorsal horn neurons. Many studies do not distinguish between the interneurons and the projection neurons when describing receptors for glutamate, GABA, and glycine. GABA is present in ~25%, 30%, and 40% of neurons in laminae I, II, and III, respectively [99]. Spinal cord wiring is complex and little understood, the main spinal signaling pathways in the rodent are comprehensively reviewed elsewhere [100]. It is within the spinal cord that substantial transformation and modulation of the nociceptive signal can occur before it ascends to higher centers [101], due to the discrete populations of intrinsic interneurons that can alter responses of NS and WDR neurons, with astrocytes and microglia also modulating, particularly in disease states [102–104]. Superficially in lamina I of the dorsal horn, there is a large population of projection neurons, which preferentially express the NK1 receptor for substance P [105,106] and are thereby nociceptive neurons. These neurons project to the periaqueductal gray (PAG), thalamus, and particularly the parabrachial area (PA). These cells also project to the rostral ventral medulla (RVM) and can influence descending controls [107,108]. The ascending tracts are typically named according to where they terminate in the brain [109]. The spinothalamic tract (terminating in the thalamus) integrates the thalamic traffic (and other signals) and is responsible for the discriminative/localization component of pain via projections to the sensorimotor cortex, insular cortex, and the anterior cingulate originating mainly from laminae I, IV, V, and X [110,111]. The spinobulbar tract (terminating in both the hindbrain and midbrain regions associated with pain processing) conveys the affective/intensity component, and projects to the amygdala and hypothalamus via the parabrachial nucleus [112]. This spinobulbar pathway can influence and recruit descending pathways via the PAG, pontine locus coeruleus (LC), and RVM, thereby dictating the output passing through the spinal cord [113,114] (Fig. 46.1). The PAG is one of the most important structures associated with the descending pain control system [115,116] and receives information from the somatosensory and cingulate cortices, the thalamus, amygdala, and hypothalamus, and from the ascending pathways. The PAG–RVM exerts a degree of selective inhibition of C fiber‐mediated nociceptive impulses but preserves A fiber messages coding sensory and discriminatory information [117–119] and the RVM can be considered the final relay point through which facilitation or inhibition of the nociceptive message passes [120,121]. The degree of inhibition or facilitation is controlled in the RVM by at least two different types of neurons known as “ON cells” and “OFF cells.” Briefly, the ON, OFF, and neutral cells are defined by their response to nociceptive input [123]. The ON cells show a burst of activity prior to nociceptive input and are considered a component of excitatory drive. The OFF cells show a short‐lived decrease in firing rate before nociceptive input and are considered inhibitory. Neutral cells do not respond to nociceptive input. The role of these cells may be more complicated than previously thought [124–126]. The RVM can be considered a relay station but requires a forebrain loop [127]. Furthermore, the anterior cingulate cortex (ACC) projections regulating spinal neurons [128,129] can also modulate the pain experience. The ACC potentially offers a further target for manipulating the pain signature but understanding its role requires more sophisticated paradigms. The role of the ACC and pain processing has been comprehensively described [130]. Figure 46.1 Schematic diagram of pain pathway from periphery to brain including sagittal midline T2 weighted MRI images from a dog brain with selected transverse images at the level of the (1) thalamus, (2) midbrain, and (3) rostral medulla annotated to indicate the anatomical localization of regions involved in descending modulation. Primary afferent fibers (Aβ, Aδ, and C fibers) transmit impulses from the periphery (osteoarthritic hip joint) to the dorsal horn of the spinal cord. Secondary NS cells are predominantly located in the superficial dorsal horn (laminae I–II), with most WDR neurons located in the deeper dorsal horn (e.g., lamina V). Lamina V neurons in particular project to the thalamus (via the spinothalamic tract) where there are subsequent neuronal projections to various cortical regions that form part of a “pain matrix” (primary and secondary somatosensory, insular, anterior cingulate, and prefrontal cortices). The descending pathways originate from brainstem nuclei such as the rostral ventromedial medulla (RVM) and the locus coeruleus (LC), which are directly influenced by the ascending spinobulbar projections and higher centers to modulate spinal processing. PAG, periaqueductal gray; DRG, dorsal route ganglion; RVM, rostroventromedial medulla; LC, locus coeruleus; LS, limbic system; NS, nociceptive‐specific; PA, parabrachial nucleus; WDR, wide dynamic range. Source: White et al. [122], with permission from Elsevier. In the mature adult rat, descending inhibition is targeted to spinal neurons with a strong afferent C fiber input. However, in the first few weeks postnatally, the system is controlled differently with greater descending facilitation particularly targeted to the A fiber input [131]. The evolutionary reason proposed for this A fiber input is to provide the dorsal horn with low‐level, non‐noxious, tactile input, thereby promoting the development of the animal’s sensory networks. The switch from facilitation to inhibition as the animal matures is primarily dependent on endogenous opioid levels in the RVM [132] with GABA and endocannabinoid levels also playing a role [133,134]. It is possible that immature nervous systems could be at risk of excessive sensory overload and peripheral injury in the first few weeks of life where facilitation is favored [135]. In summary, this balance of inhibition and facilitation of descending pathways is a dynamic product of the afferent evoked activity, the age of the animal, but also the excitability of the dorsal horn cell [131]. The descending pathways appear to lack specificity, influencing all portions of the spinal cord and it is this widespread and intrinsic influence that is responsible for the ability to facilitate or inhibit transmission but other senses too, potentially explaining why some chronic pain states have clinical signs distinct from pain alone but can manifest with debilitating effects on sleep and other emotions [136,137]. Moreover, the perception of pain is sensitive to many mental processes therefore not exclusively driven by or maintained by the noxious input [138]; heightened anxiety and fear, for example, can then often exacerbate the suffering of pain [139]. The monoamines norepinephrine (noradrenaline) and 5‐hydroxytryptamine (serotonin, 5HT) are the major transmitters in the descending controls [127,140]. These bidirectional monoaminergic systems modulate the outputs of the dorsal horn neurons. The inhibition or facilitation has been shown to be mediated by distinctly different receptors [141–143], but the output is determined by the type of noxious input and the response to it. The pharmacology involved in the descending controls is comprehensively covered elsewhere [127]. Norepinephrine is the major neurotransmitter released by sympathetic postganglionic nerve fibers and is involved in autonomic regulation of numerous organs. These noradrenergic cell groups were mapped in 1964 and are prefixed by the letter A (for aminergic) [144]. Areas A1 to A7 inclusive contain norepinephrine; areas A8 to A14 contain dopamine. Group A1 is located in the caudal ventrolateral part of the medulla and is responsible for aspects of body fluid metabolism [145], group A2 plays a role in stress and food intake and is located in the solitary nucleus [146]. Group A5 (close to the superior olivary complex) and A7 (in the pontine reticular formation) project to the spinal cord [147]. Norepinephrine plays a major role in modulating nociception [148–150]; the majority of descending projections originating from the pontine nucleus LC contain norepinephrine (areas A6 and adjoining A4) and microstimulation of this nucleus was shown to produce antinociceptive effects via spinal α2‐adrenergic receptors [151]. In healthy subjects, the noradrenergic system serves to regulate pain thresholds, with its role becoming more prominent in cases of injury or inflammation [150]. Norepinephrine released from the LC is also involved in the regulation of vigilance, attention, and cognitive functions but it is still not possible to surmise the net effect of the noradrenergic system on supraspinal structures; evidence exists for both antinociceptive and pronociceptive actions [152]. However, in the LC‐spinal descending pathway, the central noradrenergic system inhibits pain, and drugs acting on the α2‐adrenergic receptor (alone or as an adjuvant) have proven effective to varying degrees as analgesics. In humans, topically applied α2‐adrenergic receptor agonists may have a role in treating neuropathies [153], and intra‐articular administration has reduced postoperative knee pain [154]. Patients with intractable cancer pain have benefited from intrathecal administration [155]. The α2‐adrenergic receptor agonists are widely utilized in wild, domestic, and laboratory animals for sedation, analgesia, and muscle relaxation. In general, spinally administered α2‐adrenergic receptor agonists have an enhanced antinociceptive potency in animal models with a persistent injury [156–158], and the intensity, duration, and type of noxious injury will ultimately determine the response [159–162]. Likewise, altered noradrenergic activity (for example, a reduction in the inhibitory influences) can lead to the hypersensitivity seen in models of both acute inflammation and nerve injury [158,163–165], hence the interest in targeting this system. This central role is further evidenced by studies that have demonstrated restoration of diminished noradrenergic control after use of intrathecal reboxetine, a selective norepinephrine reuptake inhibitor, following the development of a neuropathic pain phenotype [166]. Serotonin is commonly considered one of the most important neurotransmitters. Studies have shown that ~20% of the neurons in the RVM are serotonergic, and there was early evidence for 5‐HT involvement in descending modulation [167]. These descending pain modulatory pathways influence nociception through activation of different serotonergic receptors in the spinal cord [168]. Both acute and chronic noxious stimuli can activate these RVM 5‐HT neurons and increase the expression of 5‐HT receptors in the spinal cord [169,170]. The 5‐HT neurons can also influence non‐serotonergic neurons. The collaterals of the 5‐HT neurons have been shown to regulate the bidirectional control from the ON and OFF cells in the RVM [171]. The serotonergic/5‐HT system is more complex than the adrenergic system with pain inhibition or facilitation attributable to different subtypes of 5‐HT receptors [172–174] with facilitation assuming more influence than inhibition. Importantly, tonic activation of 5‐HT (and non‐5‐HT)‐mediated brainstem facilitatory influences is one of the contributors to the development and maintenance of central sensitization [175,176]. The monoamine dopamine is also involved in descending controls [140] with mesolimbic, mesocortical, and nigrostriatal dopaminergic pathways able to inhibit primarily the affective component of nociception [177,178]. The dopamine system is embedded in the pain matrix circuitry [179] but is also needed for movement [180] and is associated with reward [181], learning, and cognition [182,183]. Studies have shown that dopamine therapies may offer promise as analgesics but efforts to develop therapies are complicated by its intrinsic role in so many other modalities [184,185]. The descending pathways can also be considered components of an opioid‐sensitive matrix. Experiments have shown that numerous drugs interacting with components of the descending pathways can influence or mimic the production of endogenously produced opioids, indeed there is evidence of a bulbospinal opioidergic pathway [186]. The PAG is one of the major sources of opioid‐mediated inhibition of ascending nociceptive impulses [187]. Cortical inputs synapse at the PAG mediating a “top‐down” endogenous pain inhibition system. The projections from the PAG extend to the RVM and noradrenergic pontine nuclei which in turn modulate nociceptive input at the spinal cord through the release of norepinephrine and serotonin. The dorsal‐dorsolateral portions of the PAG play a role in stress‐induced analgesia, which is independent of opioids but depends on endocannabinoids, whereas the lateral‐ventrolateral portions of the PAG are implicated in analgesia induced by opioids and non‐steroidal anti‐inflammatory drugs (NSAIDs) [188]. The placebo effect is one example of this “top‐down” modulating pathway. This refers to an analgesic construct that can be elicited in certain people that has its mechanisms associated with activation of the μ‐opioid receptor and changes in blood flow to the rostral anterior and pregenual cingulate cortices, the dorsolateral prefrontal cortex, and anterior insular cortex [189,190]. In the case of placebo, it is now apparent that a complex web of interwoven processes is responsible rather than the once simplistic view that the μ‐opioid receptor was solely responsible [191]. The opioidergic descending tracts are also implicated in distraction and hypnosis and offer potential for developing further analgesic interventions. Functional magnetic resonance imaging (fMRI) studies have shown (particularly in placebo and nocebo studies) that the descending pathways are the conduits through which the cognitive influences affect a pain experience. This experience is also subsequently modulated by the individual’s control and ongoing response to the pain experience exerted by prefrontal limbic brain regions [138,192]. The exact mechanisms involved in the descending controls are not yet fully understood, but a consensus on their importance is without doubt, and research into recruitment of their endogenous modulating attributes is ongoing. The endocannabinoids are involved in the processing of nociceptive inputs [193]. The expression of endocannabinoid receptors and the presence of ligands and metabolites are dynamic and dependent on the type of pain being expressed. The system offers attractive targets for analgesic opportunities, although the isolation of the analgesic properties of molecules without psychoactive side effects is a challenging conundrum. The analgesia produced by NSAIDs in the descending pain control system also requires activation of the cannabinoid‐1 (CB1) receptor [188]. Furthermore, numerous experiments suggest that opioids, NSAIDs, and the cannabinoids in the PAG and RVM interact to potentially decrease GABAergic inhibition and thus enhance the descending flow of impulses that inhibit pain as a mechanism of analgesia [194–196]. The endocannabinoid system can adapt in the face of persistent pain states such as osteoarthritis offering a potential target of spinal hyperexcitability [197]. However, the drugs trialed so far modulate the affective component but not the sensory component of the pain [198]. GABA and glycine are inhibitory transmitters and influence nociception via their presence within interneurons [199–201]. These GABAergic or glycinergic projections can inhibit noxious inputs into the dorsal horn of the spinal cord and the neurons can express either or both neurotransmitters [202]. Studies have also demonstrated that the ON, OFF, and neutral cells have varying expression of GABA and glycine, but the expression is determined by the context of the pain experience; for example, in one study using a model of peripheral inflammation, the gene expression and the phenotype of the ON and OFF cells was altered [203]. In the absence of nociceptive input, spinal nociceptive neurons are under both tonic and stimulus‐evoked (phasic) descending controls [204,205]. Characterization of the receptors involved in the tonic control mechanisms has been undertaken [206–208]. One important caveat to these experiments involving discrete lesions and anesthesia is they may inadvertently overlook the contribution of the plasticity of the spinal cord in response to inflammation, or indeed the differences in the free‐ranging conscious animal compared to the anesthetized animal, and these should be considered when drawing conclusions. Pain consists of a somatosensory component and a psychological, affective component. The term “nociception” refers to the neural activity in the peripheral and central nervous systems caused by a painful stimulus and the term “pain” itself is used to describe both this and the emotional and autonomic responses to the insult. These different components of pain are processed in separate, discrete areas of the brain. The nociceptive insult is the cause of pain, but in some cases this insult is absent, and its magnitude is not linearly related to the pain that is reported or behaviors that are displayed [209]. This is, in part, a consequence of a feedback loop between the brain and spinal cord. This spino‐bulbo‐spinal loop can alter the extent to which pain signals are amplified or inhibited within the spinal cord. Although significant advances have been made in the basic understanding of pain processing and modulation in recent years, large gaps in our knowledge remain, particularly in the fields of anatomic, biochemical, and physiologic mechanisms of pain. The current situation in animals with respect to adequacy of pain management is less well documented. Drugs with proven efficacy for the treatment of acute surgical pain are available for companion animals (e.g., licensed opioids and NSAIDs). However, with the exception of NSAIDs, the repertoire of licensed drugs for the management of chronic pain (e.g., pain caused by degenerative joint disease) in companion animals is very limited. Acute and chronic pain remain poorly treated in farm animals and exotic species. It is widely accepted that research in animals is pivotal to an increased understanding of nociceptive and pain mechanisms and the development of new analgesic drugs, for both humans and animals. Pain studies in humans generally focus on characterizing pain states, with few studies investigating underlying pain mechanisms, and studies in humans are inevitably hampered by ethical constraints. Although in humans the advent of advanced neuroimaging techniques allows the in vivo study of patterns of CNS activity concurrent with self‐report of pain perception, ethical considerations surrounding induced pain models in humans, combined with current technological limitations, means that neuroimaging cannot replace the need for animal pain models. In addition, neuroimaging lacks cellular resolution and poor temporal resolution; fMRI may not distinguish between very high levels of neuronal activity in the brain (and therefore may not be sensitive to different pain intensities) because the blood oxygen level‐dependent (BOLD) signal can reach a ceiling. Neuroimaging does not allow the interrogation of small areas of the CNS such as the dorsal horn of the spinal cord or the peripheral nervous system with any precision, as the technique relies on integrating electrical signals across large sampling regions such as the brain. Two main approaches have been adopted when studying pain in animals. The first approach is to study responses to brief noxious stimuli (defined as stimuli that are damaging to, or threaten to damage, normal tissues) in naive animals. In other words, brief phasic noxious thermal, electrical, or mechanical stimuli are delivered to healthy animals and the magnitude of the stimulus threshold required to elicit a response is measured (e.g., nociceptive threshold testing), or changes in neuronal activity, animal behavior, or body systems are studied during delivery of the stimulus. Response to brief phasic stimuli is commonly utilized as the outcome measure in analgesic drug testing (i.e., as an integral component of pain models), although it should be viewed as having limited applicability owing to the previously naive (normal) state of the nervous system and singular nociceptive modality (e.g., thermal nociceptors), which differs from most clinical pain syndromes. The second approach is to induce pain in the animal by delivery of a tonic, sustained noxious stimulus that induces peripheral or central sensitization and subsequently to study pain mechanisms (e.g., by recording changes in behavioral or neuronal activity) or test analgesic drugs. In this chapter, the broad term “pain model” is used to refer to the second approach. There is interest in using companion animals with naturally occurring disease conditions that cause pain (e.g., spontaneous osteoarthritis [OA] in dogs) [210] to study pain mechanisms and evaluate analgesic drug efficacy for both humans and animals. However, there are disadvantages when using nonhuman subjects in analgesic studies, such as the lack of verbal feedback about the “feelings” of the patient’s pain and the vulnerability to observer bias. Within any pain model, there are three distinct components that must be differentiated [211]. First, there is the study subject (variables include species, strain, sex, and age); second, there is the stimulus or type of tissue damage that is used to initiate pain (e.g., an injection of an irritant substance into a joint versus measuring cutaneous thermal threshold); and third, there is the outcome measure (e.g., behavior or physiological parameter) used as the surrogate biomarker for pain. Significant interest has arisen recently in trying to improve pain biomarkers in animal models with the recognition that endpoints such as a tail flick in response to a noxious stimulus may be largely reflexive and not an indicator of higher brain perception. This is of particular relevance in the development of pain models for analgesic drug development. Despite significant efforts in analgesic drug discovery, there has been limited success in developing and marketing new analgesic drugs with efficacy and acceptable adverse effect profiles for clinical patients with pain. Putative new analgesic drugs undergo screening and preclinical testing in animal models; therefore, it is important that the performance of a new analgesic in animal models is successfully translated to the human pain model and clinical patients. Although the reasons for failure of new drugs at the preclinical–clinical interface are multifactorial, there is a general agreement that limitations of available animal models play a major role in the current bottleneck in analgesic drug development. Recognized general limitations of animal pain models occur at the level of the experimental animal (subject) and the pain assay or biomarker used to assess pain. The majority of preclinical pain research is carried out in young adult, healthy, intact male laboratory mice and rats of a specific strain. Although useful as a cost‐effective and rapid screening method for in vivo testing of candidate molecules, this does not always translate well to the clinical population of pain patients. Therefore, it has been recommended that pain studies include more diverse and heterogeneous groups of animals comprising both genders and a variety of strains [212]. Many nociceptive (pain) assays are currently available that aim to model nociceptive, inflammatory, and neuropathic pain. However, it is immediately apparent that there is a mismatch between the underlying pathophysiologic changes induced in experimental assays and the etiology of clinical pain conditions. For example, nerve ligation is commonly used as a model of neuropathic pain, yet diseases associated with neuropathic pain in humans rarely result from complete nerve ligation or nerve compression. There are often co‐existing abnormalities that also contribute to the perceived pain, such as neuropathy and inflammation. Most pain assays do not mimic the complexity of clinical pain states. Induced models of OA in rats and mice usually focus on the stifle joint and are induced by sterile inflammation or surgical disruption of the joint, leading to a rapid progression of OA that does not consider the effect(s) of aging itself on perceived pain [213]. Until recently, most behavioral biomarkers used to measure analgesic drug effect in animal pain models relied on either evoked spinal reflexes (e.g., limb withdrawal from a von Frey filament) or innate behaviors such as vocalization or guarding that can also be performed by decerebrate animals [214]. Evoked withdrawal responses detect hyperalgesia and allodynia (although the two are difficult to distinguish from each other in animals) and therefore only provide information on the sensory and discriminative components of pain. They do not provide information about the emotional (affective) component, which is critical to the experience of pain in humans and animals, although more difficult to define in the latter. Although many patients with chronic neuropathic pain experience hyperalgesia and allodynia, they commonly report spontaneous pain (i.e., non‐evoked pain) as the most debilitating and distressing aspect of their condition [215]. Therefore, unless biomarkers that evaluate spontaneous pain are employed in an assay, the effect on spontaneous pain may go undetected. New biomarkers sensitive to spontaneous pain that can detect changes in behavior associated with emotion are being developed. In recent years, the importance of translational approaches to animal research, which aim to bridge gaps between basic animal research and medical practice, has been recognized. As such, efforts are under way to develop assays and biomarkers that directly translate to experimental human pain models. This increases the likelihood that analgesics found to be effective in preclinical studies will remain efficacious in human clinical trials. Currently, there are very few assays that are translatable, although intradermal or cutaneous application of capsaicin and the ultraviolet B (UVB) pain model are notable exceptions. The use of spontaneous disease models in companion animals is another approach to improving translatability of pain assays between humans and animals. Experimental pain studies in humans rely largely on psychophysics to quantify pain and therefore analgesic drug efficacy, but self‐reporting of pain is inherently subjective. Commonly used behavioral biomarkers such as reflex withdrawal are also vulnerable to observer subjectivity. Considerable variability of reported results among researchers and laboratories is therefore common. Clearly, objective biomarkers directly translatable to humans that reflect the underlying neurobiology of perceived pain are needed to improve pain models. In the following sections, the term “pain” is used to describe all assays carried out in awake animals; if the animal is anesthetized during the assay, the term “nociception” is used. Brief noxious (phasic) stimuli are widely used to elicit responses in order to study nociceptive pathways and increase the understanding of the neurobiology of pain in experimental animals and also for the purposes of analgesiometry (i.e., measuring changes in biomarkers to study the action of analgesic drugs). Broadly, “nociceptive threshold testing” in animals and “quantitative sensory testing” (QST) in humans are the terms used to describe the application of tests utilizing brief phasic stimuli. The stimuli might be applied to naive animals or animals with induced pain, thus allowing the study of nociceptive pathways and drug efficacy in normal animals and animals with altered pain sensitivity caused by central and/or peripheral sensitization. For a complete review of phasic pain tests, see Le Bars et al. [216]. Beecher [217] was one of the first authors to set out ideal criteria for producing acute pain experimentally, and some of the optimal characteristics are as follows: Four modalities of noxious stimulation are commonly used in phasic acute nociceptive tests: electrical, thermal, mechanical, and chemical. The advantages and disadvantages of these different modalities when used to generate acute pain are shown in Table 46.2. Chemical stimulation is considered separately because it causes a slower, progressive, and non‐escapable noxious stimulus. Commonly used phasic tests in awake animals are described in Table 46.3 (thermal stimuli), Table 46.4 (mechanical stimuli), and Table 46.5 (electrical stimuli). For practical reasons, phasic nociceptive stimuli are usually applied to the skin to activate cutaneous nociceptors, although stimuli are also applied to viscera in some models. The predominance of tests involving cutaneous nociceptors reflects the ease of access to the skin and the ability to stimulate the skin with minimal restraint of the animal. In awake laboratory animals, the plantar surface of the hind paw is often used as the site of stimulation because it is readily accessible for application of heat and mechanical stimuli. Awareness of the differences in cutaneous sensitivity between haired (found on most of the body) and glabrous skin (the plantar paw) in laboratory animals is important as it alters the translatability of pain tests between rats and humans [218]. The tail is commonly used as the site of stimulus application in acute pain tests in rats (e.g., thermal nociceptive stimuli). However, the tail is also essential for thermoregulation and balance, which can influence threshold responses measured following stimulation [216]. Concurrent noxious stimulation of more than one body part at the same time recruits endogenous inhibitory mechanisms and confounds measured threshold responses, so is best avoided [219]. Another consideration is the potential for repeated stimulation of the same body site to cause peripheral and central sensitization, thereby causing a reduction in threshold over time. Laboratory animals generate a range of behavioral responses to delivery of phasic noxious stimuli classified by the underlying nociceptive pathways that are activated. The endpoints classically used for each assay are detailed below; however, a few general comments are outlined here. Many phasic tests rely on the detection of motor withdrawal responses; therefore, impaired motor function (e.g., during administration of analgesic drugs that have concurrent effects on locomotion) will confound threshold responses. It is important to differentiate between withdrawal responses that are reflex (governed predominantly by spinal mechanisms) and more complex behaviors such as escape, avoidance, or licking of the body part where the stimulus was applied. The complexity of the endpoint behavior will to some extent reflect recruitment of underlying nociceptive or pain mechanisms and therefore are commonly differentially sensitive to analgesic drugs [220]. Endpoint behaviors used in nociceptive tests should ideally be nociceptive specific, reliable, reproducible, and sensitive to administration of analgesic drugs. Neurophysiologic techniques are increasingly important in studies to elucidate ascending pain pathways and cortical representation of pain [221–223]. Techniques that afford a direct window on the function of the CNS, such as electroencephalography, provide a unique insight into pain processing and how activation of nociceptive pathways results in pain perception in a conscious animal. The electroencephalogram (EEG) is the electrical activity recorded from electrodes placed at various locations on the scalp (human) or head (other species) [224–226]. It consists of the summated electrical activity of populations of neurons together with a contribution from the glial cells. Neurons are excitable cells with intrinsic electrical properties that result in the production of electrical and magnetic fields. These fields may be recorded at a distance from their sources and are termed “far‐field potentials.” Fields recorded a short distance from their source are termed “near‐field” or “local field potentials.” Activity recorded from the surface of the cortex is described as the electrocorticogram (ECoG), whereas electrical activity recorded from the scalp is the EEG. The EEG and ECoG are both far‐field potentials. An evoked potential (EP) is generated by recording the EEG time‐locked to a sensory stimulus, such as presentation of an auditory tone. Averaging the EEG over a sequence of responses allows the electrical activity specific to the stimulus to be extracted and presented by a plot of voltage (amplitude) against time. Somatosensory‐evoked potentials (SEPs) recorded from the dura at various loci in animals in response to repetitive noxious stimuli are used extensively to study pain and analgesia. Table 46.6 outlines the advantages and limitations of EEG and EPs in phasic pain tests. Nociceptive withdrawal reflexes (NWRs) are an established neurophysiologic measure used in rodent models as a direct measure of spinal cord hyperexcitability and thus a biomarker of central sensitization [227,228]. They have also been translated to humans to quantify spinal cord excitability in patients with chronic pain, including pain caused by OA [229]. The NWR threshold is defined as the magnitude of the stimulus (e.g., electrical current) that is required to elicit an electromyographic (EMG) response remembering that central sensitization causes a decrease in this threshold. Temporal summation (amplification of the magnitude of the EMG signal in response to repetitive noxious stimulation) of the NWR is also measured to probe changes in the spinal nociceptive processing. Use of these techniques in rodents requires anesthesia. NWR thresholds and temporal summation have been measured in awake dogs and horses using electrical stimuli and used to characterize the antinociceptive effects of analgesic drugs [230–232]. The third major neurophysiologic endpoint used in phasic pain tests in laboratory animals is direct recording of neuronal activity during stimulus application, for example, recording from peripheral afferent sensory fibers or dorsal horn neurons using in vivo electrophysiology. These studies are usually carried out under terminal anesthesia. Table 46.2 Advantages and disadvantages of different stimulus modalities used to study acute pain.
46
Physiology and Pathophysiology of Pain
Introduction
Receipt, encoding, and transmission of nociceptive information
Biology of sensory transduction
Sensitization
Family
Identity
Modality
Activators
Temperature range
Location
References
TRPA
TRPA1
Thermal,
Mechanical
Isothiocyanates, Ca2+, icilin
< 18 °C
C fibers
[39,40]
TRPC
TRPC1
TRPC5
Mechanical
Thermal
Receptor/store‐operated
Thioredoxin
N/A
25–37 °C
C and Aδ fibers
[41,42]
[43]
TRPM
TRPM3
TRPM8
Thermal
Pregnenolone sulfate
Menthol, icilin
> 30 °C
< 28 °C
C and Aδ fibers
C fibers
[44]
[45–48]
TRPV
TRPV1
TRPV2
TRPV3
TRPV4
Thermal, Osmotic
Thermal, Osmotic, Mechanical
Thermal
Thermal,
Osmotic,
Mechanical (in injury)
Capsaicin, H+
Endocannabinoids, diphenyl compounds 2‐aminoethoxydiphenyl borate
Camphor, carvacrol, diphenyl compounds
PUFAs, 4αPDD, epoxyeicosatrienoic acids
> 42 °C
> 52 °C
> 34–39 °C
> 27–34 °C
C and Aδ fibers, keratinocytes
Aδ and Aβ fibers, immune cells
C fibers, keratinocytes
C and Aδ fibers, keratinocytes, Merkel cells
[49–51]
[52,53]
[54]
[15,55]
DEG/ENaC
ASIC1
ASIC2
ASIC3
Mechanical (touch)
H+
N/A
Aδ, Aβ, and C fibers
[56,57]
[58,59]
[60]
K2P channel
TREK‐1
TREK‐2
TRAAK
Thermal,
Mechanical
PUFAs, H+
Noxious cold and noxious heat
40–46 °C
20–25 °C
Noxious cold and noxious heat
Aδ and C fibers, Aβ?
C fibers
Dorsal root ganglia
C fibers
Dorsal root ganglia
[61–63]
[47,64]
[13,65]
VGNC
Nav 1.8
Thermal
No subfamily activator
Noxious cold
C fibers
Dorsal root ganglia
[66–68]
Piezo
Piezo 1
Piezo 2
Mechanical
Pyrazine?
N/A
Dorsal root ganglia (less than Piezo 2)
Merkel cells
Dorsal root ganglia
Aδ fibers
[69–71]
Initiation of a generator potential
Transmission of the action potential
Anatomy of the dorsal horn
Pharmacology of descending controls
Norepinephrine
Serotonin (5‐HT)
Dopamine
The role of the opioids
Endocannabinoids
GABA and glycine
Tonic control
The spino‐bulbo‐spinal loop
Models of nociception and analgesic testing
Limitations of current animal models
Experimental animal (subject)
Nociceptive assay
Biomarkers
Translational pain models
Nociceptive and pain assays
Phasic pain tests
Stimulus modalities
Site of stimulus application
Outcome measures
Behavior
Neurophysiologic endpoints
Stimulus modality
Characteristics of the stimulus
Advantages
Disadvantages
Electrical
Direct activation of all nerve fibers, thereby bypassing transduction mechanisms at the peripheral nociceptor.
Thermal: general comments
Selective activation of cutaneous thermosensitive and nociceptive fibers.
Rate of heating can be altered to selectively activate Aδ or C fibers.
Thermal: radiant heat source
Conventional radiant heat sources emit light in the visible or adjacent infrared spectra.
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
Get Clinical Tree app for offline access
