Discussion
Consistent with previous studies using this model [17] we found persistent weight-bearing asymmetry and mechanical allodynia of the ipsilateral paw in animals with MIA-induced knee OA. These observations are similar to the clinical features seen in patients with knee OA, in that they typically experience pain with loading the affected joint and a subset of patients also develop hyperalgesia in tissue distant from that joint (referred pain) [25]. Both characteristics represent the hypersensitivity state, as the pain is caused by mild, normally non-noxious stimuli. Vortioxetine dose-dependently suppressed both forms of pain-related behaviour observed in MIA-treated animals, suggesting that it may counteract sensitization of pain pathways and exert efficacy against pain in OA. Previously, we noted the analgesic efficacy of vortioxetine in various models of neuropathic and inflammatory pain that share a common feature of sensitization of pain pathways [12, 14]. Duloxetine was also effective in both behavioural tests in MIA-injected animals.
We observed a pronounced increase in Ngf mRNA expression in knees injected with MIA compared with sham-injected counterparts. Increased NGF levels have been previously documented in the knees of rats receiving intra-articular MIA [26] and in synovial samples from patients with knee OA [27]. There is ample evidence for the particular importance of NGF in the pathogenesis of OA pain and treatments that inhibit its action (anti-NGF antibodies or small molecules) have been or are being developed as a new class of analgesics for the treatment of pain in OA (reviewed by Schmelz et al. [4) and Wise et al. [5]). NGF is released in the affected joint by chondrocytes, synovial fibroblasts, and macrophages in response to various stimuli (e.g. mechanical stress and pro-inflammatory mediators including IL-1β) [27-29] and participates in the development of peripheral and central sensitization, in multiple ways [4, 5].
Vortioxetine profoundly reduced Ngf mRNA expression in MIA-injected knees. Previously, vortioxetine was shown to have the ability to inhibit nuclear factor (NF)-κB in isolated human macrophages exposed to pro-inflammatory stimuli [16]. NF-κB is a transcription factor that plays a central role in the expression of a variety of genes involved in inflammatory responses [30]. It has been demonstrated (in cultured human corneal epithelial cells exposed to hyperosmolar stress) that IL-1β upregulates NGF and that the IL-1β-induced increase in NGF production is NF-κB-dependent [31]. The reduction inNgf mRNA levels induced by vortioxetine could therefore be related to the suppression of NF-κB. This effect can be attributed to the main mechanisms of action of vortioxetine: SERT inhibition, 5-HT1B receptor (partial) agonism and 5-HT3 receptor antagonism [10]. Inhibition of SERT (by fluoxetine or sertraline) has been shown to suppress NF-κB activation (in isolated brain tissue from mice with neuronal injury and in inflammatory cytokine-stimulated cultured glial cells) [32, 33]. This effect of increased extracellular 5-HT could be mediated at least in part by activation of 5-HT1B/1D receptors, as their selective agonist (sumatriptan) decreased NF-kB levels (in tongue samples from rats with oral mucositis) [34]. In agreement with this, we have shown that 5-HT1B/1D receptors are involved in the antinociceptive effect of vortioxetine in a trigeminal model of inflammatory pain in mice [14]. Vortioxetine can itself activate 5-HT1B receptors, and this direct effect could also contribute to lowering NF-kB and NGF expression levels. On the other hand, an increase in extracellular 5-HT may counteract these beneficial effects via 5-HT3 receptors, as activation of 5-HT3 receptors has been shown to upregulate the production of IL-1β (in isolated human monocytes) [35], which could subsequently increase NGF levels. By blocking 5-HT3receptors, vortioxetine could prevent a 5-HT-induced increase in IL-1β expression and thus maintain/enhance the suppressive effects of locally elevated 5-HT on NGF levels via 5-HT1B/1D receptors. The lack of direct effects on 5-HT receptors may explain why duloxetine (serotonin and norepinephrine reuptake inhibitor, SNRI) did not markedly reduce Ngf mRNA expression.
It is important to note that vortioxetine-treated MIA-injected animals had Ngf mRNA at the level measured in sham-injected, healthy animals. The major concern with anti-NGF treatment of OA pain is the risk of rapid joint destruction and osteonecrosis, which has been observed in both animals and humans [4, 5, 36, 37]. The risk appears to be dose-dependent and is thought to be due to the disruption of physiological functions of NGF, which include peripheral sensory neurons survival, cartilage repair, and load-induced bone formation [4, 5]. By maintaining basal NGF levels, vortioxetine treatment might be devoid of joint/bone deterioration risk, while providing effective analgesia.
Consistent with previous findings in animal models of OA [6, 38] we found increased markers of oxidative stress in MIA-injected joints. Oxidative stress, as a result of an imbalance between the production of reactive oxygen species and their detoxification by the antioxidant defense system [39] is highly pronounced in OA cartilage and is an important cause of chronic inflammation, and vice versa [40]. The contribution of reactive oxidative species to pain and hyperalgesia in OA thus appears to be indirect. However, an oxygen radical scavenger (rebamipide) has shown an anti-hyperalgesic effect in rats with MIA-injected knees [41]. In our study, vortioxetine reduced parameters of oxidative stress in the knees of rats treated with MIA. Accordingly, vortioxetine prevented an increase in superoxide anion (O2•-) production in isolated human monocytes and macrophages exposed to an oxidative burst-inducing substance [16]. The antioxidant effect of vortioxetine may be also related to 5-HT3 receptor antagonism, as this has been shown to protect against oxidative stress (reviewed in Gupta et al. [42]). 5-HT3 receptor antagonist (tropisetron) reduced oxidative stress and inflammation in colonic tissue in a rat model of colitis after rectal administration [43], suggesting that 5-HT3 receptor blockade at the site of inflammation may be responsible for these effects. We found a greater increase in oxidative stress markers in the MIA-injected knees of male rats than in those of female rats. Similarly, a lower defense capacity against oxidative stress was observed in synoviocytes isolated from men with knee OA than from women [44]. This may be the reason why vortioxetine was less effective in reducing oxidative stress parameters in male rats than in female rats in the present study.
Significantly increased mRNA levels of Il-1β , Tnf-α ,Ngf , Bdnf and Tac 1 (encoding substance P) were recorded in L3-L5 ipsilateral DRGs and spinal cords of the MIA-injected rats (except Ngf in the spinal cord), compared with saline controls. These mediators originate from neurons and/or non-neuronal cells (e.g. glial and T cells) that communicate with each other, and are upregulated under conditions of intense/sustained activation of primary nociceptive neurons (as in peripheral inflammation or nerve injury) and could contribute to the development of central sensitization (reviewed by Sandkühler 2013; Grace et al., 2014; Miller et al., 2014; Schmelz et al. 2019; Yang et al., 2022; Cheng et al., 2014) [4,7,8 45-47]. Vortioxetine significantly reduced the expression of all upregulated mediators in male rats, whereas in female rats the effect was mostly absent. The sex difference in this effect of vortioxetine can be at least partially explained by the finding of Sorge et al. [48], who reported that different cells are responsible for the development of central sensitization in mice with peripheral inflammation: in males it depends on glial cells, whereas in females it is independent of glia and probably depends on T cells. In a chronic pain model caused by peripheral nerve injury in male rats, SERT inhibitor (fluoxetine) suppressed the activation of microglia in the DRG/spinal cord and exerted an anti-allodynic effect [49]. In male rats with nerve injury, spinal administration of a 5-HT3 receptor agonist resulted in glial hyperactivity, neuronal hyperexcitability and pain hypersensitivity, which was attenuated by a 5-HT3receptor antagonist [50]. Inhibition of SERT and blockade of 5-HT3 receptors could therefore lead to glial cell suppression and be involved in the downregulation of pain-related mediators that we observed in male rats after treatment with vortioxetine. With peripheral tissue inflammation, T cells (mainly Th1) migrate to the spinal cord and release mediators that induce chronic pain [46]. It has been shown in human and murine T lymphocytes that among the different 5-HT receptor subtypes expressed on their surface, the 5-HT1B receptors play the most important role in their activation [51]. Thus, by directly and indirectly activating 5-HT1B receptors, vortioxetine could contribute to the activation of T cells, which could explain why it mostly lack effect on the expression of pain-related mediators in female rats. Vortioxetine was more effective in reducing oxidative stress in MIA-injected knees in female rats and conversely, it was more effective in reducing the expression of pain-related mediators in DRGs/spinal cord in male rats. This may explain why the anti-allodynic effects of vortioxetine did not differ significantly between the sexes.
In this study, it was not possible to compare the efficacy of vortioxetine with that of duloxetine as only two doses of each drug were tested. (An extension of this study with additional doses, i.e. animal groups, would not be ethically justifiable.) However, higher doses of both antidepressants did not reach the toxic dose range, as they did not impair motor functions/cause sedation, as observed in the rotarod test. Converted to human doses [21], the doses studied correlate to therapeutic doses (vortioxetine 2 and 10 mg/kg/day is equivalent to 19.5 and 106 mg/day for humans, therapeutic dose range is 5-20 mg/day [10]; duloxetine 15 and 25 mg/kg/day is equivalent to 146 and 243 mg/day for humans, therapeutic dose range is 30-120 mg/day [52]. Thus, our result suggests that vortioxetine may be as effective as duloxetine for OA pain. It is even possible to expect vortioxetine to be more effective, as it can prevent/reverse NGF upregulation in the diseased joint, whereas duloxetine cannot. Another advantage of vortioxetine over duloxetine could be its multimodal analgesic effect, i.e. the involvement of multiple pain-modulating systems in this effect (serotonergic, adrenergic, cholinergic, cannabinoid, adenosine and opioidergic) [14, 53], the improved cognitive performance and the lack of negative effects on cardiac redox status and general well-being that we observed in a previous study using the same model [54].
In conclusion, vortioxetine may be effective against chronic pain in OA. This effect appears to be mediated, at least in part, by the regulation of NGF expression in the affected joint, i.e. by reducing its expression to basal levels. It would be worthwhile to investigate the effect of vortioxetine in patients with OA, considering the beneficial effects it has shown in an animal model.