Our findings demonstrate that a neuron-glia-cytokine-neuronal signaling cascade is involved in the mechanisms underlying spinal 5-HT3 receptor-mediated hyperalgesia. The development of persistent pain after inflammation and nerve injury appears to be dependent, in part, upon 5-HT pathways originating from the rostral ventromedial medulla leading to activation of 5-HT3 receptors at the spinal level [5, 16, 19, 22, 24, 25, 37]. Consistent with these views and a recent study , our data showed that blockade of spinal 5-HT3 receptor function by intrathecal Y25130, a selective 5-HT3 receptor antagonist, attenuated mechanical and thermal hypersensitivity following L5 SNL in rats. Interestingly, some studies reported that intrathecal injection of 5-HT3 receptor antagonists such as CGP35348  or ondansetron  had no preventive effects on mechanical allodynia and/or thermal hyperalgesia in a rat with L5/6 SNL, which conflicts with the study with the same drug ondansetron in the same SNL model  and our results. However, we noticed that there were no expected plastic changes of both 5-HT immunoreactive intensity and 5-HT3 receptor innervation in the lumbar spinal dorsal horn at 14 d after L5/L6 SNL in the study reported by Peters and colleagues . In contrast, we found a robust increase of tissue Tph-2 level in the RVM  at 14 d and a progressive enhancement of tissue 5-HT3 receptor expression in the spinal dorsal horn from 1 d to 28 d after nerve injury when compared with that in the sham group (unpublished observations). We suspect that the discrepancies between our positive findings and those reported by Peters et al.  could be attributed to the utilization of different neuropathic pain models and different 5-HT3 receptor antagonists as well as the absence of more quantitative measures used for the 5-HT3 receptor expression and fewer time points measured after injury in their study. In the present study, our data indicate that the effective dose of intrathecal Y25130 for attenuation of behavioral hypersensitivity following SNL did not alter thermal and mechanical thresholds in the sham animals at 14 d after surgery. Thus, we propose that increased descending 5-HT drive and spinal 5-HT3 receptor expression after tissue and nerve injury contribute to the maintenance of central sensitization, including glial hyperactivity and neuronal hyperexcitability at the spinal level underlying the development of persistent pain.
We have determined that a number of chemical mediators contribute to the spinal 5-HT3 receptor-induced novel spinal signaling cascade that includes the chemokine, fractalkine released from 5-HT3 receptor-containing neurons, cytokine IL-18 released from microglia, IL-1β released mainly from astrocytes, enhanced phosphorylation of spinal NMDA receptors, and ultimately behavioral hyperalgesia. Moreover, the mechanisms by which these events are sequentially activated through multiple signaling cascades to link neuron-microglia-astrocyte-neuronal interactions (Figure 8D) is unexpected and novel, and highlights how cellular circuitry and molecular signaling interact in the spinal dorsal horn response to 5-HT3 receptor activation. The findings indicate that spinal hyperexcitability or central sensitization underlying the development of hyperalgesia not only depends on the initiation of nociceptive input from primary afferent neurons after tissue and nerve injury [35–37], but also requires the maintenance of descending facilitation from the RVM 5-HT-spinal 5-HT3 receptor systems [25, 37]. Our study supports the growing evidence that spinal 5-HT3 receptors play a crucial role in the cellular and molecular mechanisms of the development and maintenance of persistent pain states.
Our results demonstrate that there are at least three active signaling cascades, including fractalkine and its receptor, CX3CR1, for mediating spinal neuron-to-microglia signaling; IL-18 and its receptor for microglia-to astrocyte signaling; and IL-1β and its receptor for astrocyte-to-neuron signaling, as important components involved in the functional intercellular transduction in the dorsal horn after 5-HT3 receptor activation. These findings do not rule out the role of other chemical mediators released from the same neurons, or different subpopulations of neurons (excitatory or inhibitory neurons) or glial cells in the regulation of spinal nociceptive processes. It has been reported that some central terminals of primary afferent neurons express 5-HT3 receptors [20, 21]. In a recent study, we also showed that 5-HT3 receptors in the central terminals of primary afferent neurons are involved in enhanced primary nociceptive afferent activity and excitatory signaling input by increasing TRPV1 function during the maintenance of neuropathic pain . Intrathecal injection of 5-HT3 receptor agonists may excite these central terminals to release fractalkine, glutamate and ATP, and directly activate glial cells and even directly enhance NMDA receptor function in dorsal horn neurons. Although these findings suggest other signaling cascades, the converging data in the present study suggest that spinal neuron-glia-neuronal interaction may be particularly important in the 5-HT3 receptor-mediated central sensitization associated with intra-RVM 5-HT-dependent descending pain facilitation. Thus, up-regulation of 5-HT3 receptor expression in the spinal dorsal horn, following enhanced descending 5-HT drive after nerve injury, may play an important role in glial hyperactivity involved in the maintenance of persistent pain.
Although spinal glial hyperactivity has been reported in acute and persistent pain models [26, 29, 30, 53, 54], few studies have investigated the involvement of spinal 5-HT3 receptors in spinal glial hyperactivity. In the present study, intrathecal injection of the selective activation of spinal 5-HT3 receptors by intrathecal injection of the receptor agonist induced significant up-regulation of GFAP and Iba1. Although Western blot analysis did not show up-regulation of CD11b in the dorsal horn after single i.t. injection of SR57227, immunostaining for CD11b exhibited hypertrophic status of spinal microglia after SR57227, similar to Iba1 labeling. Interestingly, up-regulation of CD11b expression in spinal dorsal horn tissue was observed at 1d after hindpaw inflammation but not 2 h after intrathecal injection of SR57227, suggesting that increase of CD11b expression may require longer-lasting excitatory or nociceptive input on microglia. Molecular depletion of the descending 5-HT system significantly attenuated peripheral inflammation-produced glial hyperactivity in the spinal dorsal horn. These data provide the first evidence that either exogenous or endogenous activation of the 5-HT3 receptor results in spinal glial hyperactivity. Moreover, we were interested in the mechanisms by which quiescent spinal glia alter their function in response to 5-HT3 receptor activation. Recent studies have demonstrated special expression patterns for chemokines, cytokines and their receptors in spinal cord cells. For example, fractalkine exists in spinal neurons [39, 40] and its receptor CX3CR1 is selectively expressed in microglia [40, 55]. IL-18 and its receptor are present in spinal microglia and astrocytes, respectively . Consistent with our previous study on the RVM [34, 56], we found that IL-1β is mainly expressed in astrocytes but not microglia in the spinal dorsal horn. Its receptor IL-1RI is present in dorsal horn neurons expressing GluNRs. These proteins have been demonstrated to play a role in spinal nociceptive modulation and the development of persistent pain after injury [26, 27, 29, 30]. However, previous studies have not shown a relationship between these proteins and 5-HT3 receptor activation in the spinal cord. In addition, extending our recent findings [16, 35, 56], we showed the colocalization of IL-1RI with the GluNR subunit GluN1R in dorsal horn neurons and with IL-1RI-mediated facilitation of GluN1R phosphorylation after 5-HT3 receptor activation. Thus, the IL-1β-mediated amplified signaling from spinal astrocytes further enhances neuronal excitability through signaling coupling with GluNRs in the spinal cord, which plays an important role in neuronal hypersensitivity. Our findings also suggest that activation of spinal 5-HT3 receptor is sufficient to induce glial hyperactivity and cytokine release which are necessary for neuronal and behavioral hypersensitivity after 5-HT3 receptor activation. The activated glia-mediated positive signaling amplification then sensitizes spinal nociceptive neurons, leading to further neuronal activation and behavioral hyperalgesia. These findings offer new insights into the cellular and molecular mechanisms in the spinal level responsible for descending pain facilitation during the development of persistent pain after tissue and nerve injury.
In the present study, we directly activated the spinal 5-HT3 receptor to mimic 5-HT release through descending pain facilitation pathways . We found that intrathecal injection of 10 pmol of the 5-HT3 receptor agonist SR 57227 produced thermal and mechanical hypersensitivity that lasted for 4 hours. This observation provides direct evidence that the spinal 5-HT3 receptor plays a role in pain facilitation. Activation of 5-HT3 receptors in the spinal cord by 5-HT is mediated by the descending excitatory drive from the RVM to the spinal cord [5, 24, 37]. Consistent with studies with another 5-HT3 receptor agonist 2-Me-5H [57, 58], we found that intrathecal injection of higher doses of SR 57227 (10 nmol) induced transient analgesia. The different doses used in our experiments may reflect different mechanisms that depend on specific cellular circuits or the particular proteins involved. It has been shown that 5-HT3 receptors are predominantly localized in terminals of excitatory axons in the rat superficial dorsal horn and some of these originate from dorsal horn neurons [20, 21, 59, 60]. Although cell bodies expressing these receptors in the dorsal horn are further identified as excitatory neurons [59, 61], some 5-HT3 receptor-labeled neurons in rat dorsal horn express glutamate decarboxylase (GAD), a marker for GABAnergic neurons . Recent studies have demonstrated in the mouse that some dorsal horn neurons sensitive to 5-HT3 receptor agonists were GAD positive  and that some 5-HT3 receptor mRNA-containing dorsal horn neurons were GAD positive . Thus, 5-HT3 receptors appear to be expressed in both excitatory and inhibitory intrinsic neurons and terminals in the spinal dorsal horn. Synaptic plasticity of 5-HT3 receptor expression and function in the spinal dorsal horn neurons and the terminals of primary afferent fibers during the development of persistent pain will require further study.