Pulsed Radiofrequency on Dorsal Root Ganglion Improves Neuropathic Pain-induced Depression in SNI Rats by Regulating Hippocampal BDNF and Neuroinammation via Spinal IRF8 Inactivation

Background: Increasing evidence suggests that neuroglia, neuroimmune, and neuroinammatory processes are involved in the development of nerve injury-induced pain and depression. Interferon regulatory factor 8 (IRF8), a crucial factor for microglial activation, is essential for the development of neuropathic pain. The brain-derived neurotrophic factor (BDNF) and inammatory mediators (IL-1β, IL-6, and TNF-α) in the hippocampus contribute to the pathophysiology of neuropathic pain-depression comorbidity. Our previous study found that depressive-like behaviors induced by spared nerve injury (SNI) could be improved by applying pulsed radiofrequency (PRF) to the dorsal root ganglion (DRG) (PRF-DRG). However, the anti-depressive mechanisms of PRF-DRG therapy remain largely unknown. Methods: All rats (except for those in the sham group) were subjected to SNI. The nuclease-free water group and the IRF8 siRNA group were intrathecally injected with nuclease-free water and IRF8 siRNA on days 5 and 6 after SNI, respectively. PRF therapy on the L5 DRG was performed in the PRF group on day 7 after SNI, whereas no PRF current was delivered in the Sham-PRF group. The 50% paw withdrawal threshold, forced swimming test, and sucrose preference test were performed. The expression levels of spinal IRF8 and hippocampal BDNF were tested by molecular biochemistry, while IL-1β, IL-6, and TNF-α were tested by ELISA. Results: The depressive-like behaviors induced by SNI were remarkably developed in rats, which was indicated by a signicant reduction in the sucrose preference rate and prolonged immobility time on day 42 after SNI. Mechanical allodynia and depression-like behaviors of rats with SNI were remarkably improved after PRF-DRG or intrathecal IRF8 siRNA. Spinal IRF8 overexpression, hippocampal BDNF downregulation, and increased hippocampal IL-1β and TNF-α

Notably, it has been proved that interferon regulatory factor 8 (IRF8) can trigger a reactive microglial phenotype [11,12]. In our recent studies, we found that reversing excessive spinal IRF8 expression could inhibit microglial activation and improve neuropathic pain and pain-induced depression [13,14].
Moreover, we demonstrated that the application of pulsed radiofrequency (PRF) on the dorsal root ganglion (DRG) (PRF-DRG) in rats with PNI could ameliorate the neuropathic pain-depression dyad remarkably due to downregulation of spinal IRF8 and microglial activation [13][14][15]. However, the mechanisms of PRF-DRG via spinal IRF8 to intervene in mood-associated brain regions have not been explored systematically to date.
The hippocampus, a central component of the limbic system, is a crucial mood-regulating region of the brain involved in nociception processing [16,17]. The hippocampus plays a critical role in the development of chronic pain and depression [16]. In addition, the evidence suggests that signaling via hippocampal brain-derived neurotrophic factor (BDNF) plays an important role in the pathophysiology of depression and the therapeutic mechanisms of antidepressants [18]. Animal experiments on chronic pain demonstrated that the expression of BDNF in the hippocampus was decreased [19,20], which further indicated that abnormal hippocampal BDNF might explain the comorbid pain and depression. However, whether restoration of BDNF levels in the hippocampus could be involved in the underlying antidepression mechanism of PRF-DRG remains unclear.
In the present study, we attempted to validate whether hippocampal BDNF and pro-in ammatory factor levels mediated by the inhibited spinal IRF8 contribute to PRF-DRG therapy in improving depressive-like behaviors induced by spared nerve injury (SNI). The results would be valuable for clinicians to prefer PRF-DRG to treat chronic pain and depression comorbidity.

Animals
Male Wistar rats weighing 180-200 g were provided by the Shanghai SLAC Laboratory Animal Co., Ltd., China. All animals were kept in plastic cages, at a temperature of 22 ± 1°C, relative humidity of 50 ± 5%, and a light/dark cycle of 12/12 h, and were given food and tap water ad libitum. In addition, the animals were habituated to the environment for seven days before starting the experiment.
All animal studies (including the rat euthanasia procedure) were performed in compliance with the institutional animal care regulations and guidelines of Shengli Clinical Medical College of Fujian Medical University and conducted according to the Association for Assessment and Accreditation of Laboratory Animal Care (AAALAC) and the Institutional Animal Care and Use Committee (IACUC) guidelines.

Study Design
Two hundred and forty rats were randomly divided into six groups (n = 40/group): (1) sham, (2) SNI, (3) Sham-PRF (SPRF), (4) PRF, (5) intrathecal injection of nuclease-free water (NFW), and (6) intrathecal injection of IRF8 siRNA (IRF8 siRNA). All rats (except for those in the sham group) were subjected to SNI of the left sciatic nerve. The rats in the NFW and IRF8 siRNA groups were intrathecally treated with NFW and IRF8 siRNA on days 5 and 6 after SNI, respectively. PRF was applied to the ipsilateral L5 DRG in the PRF group on day 7 after SNI. The basal 50% paw withdrawal threshold (50% PWT) was measured before SNI and on days 3, 7, 8, 10, 14, 21, 28, 35, and 42. Sucrose preference test (SPT) and forced swimming test (FST) were performed, and the expression levels of spinal IRF8, hippocampal BDNF, IL-1β, IL-6, and TNF-α were detected on days 21 and 42 ( Fig. 1). Based on our previous experiments and other articles [21,22], all biological analyses were repeated ve times, with ve rats in each biological test method, in each group, and in each time to ensure adequate tissue materials for immunohistochemistry, RT-qPCR, Western blot, and ELISA analysis and to obtain statistical signi cance. Behavioral measurements were performed with ten rats in each group.

Model of neuropathic pain
The SNI procedure was performed as described previously [23]. Brie y, the left sciatic nerve and its three terminal branches in the upper lateral thigh were exposed. The tibial and common peroneal nerves were then transected, leaving the remaining sural nerve intact. The muscle and skin were then closed in two layers. In the sham group, only the nerves were exposed, and no cutting or ligation was performed. In all cases, utmost care was taken to neither stretch the nerve or its branches nor affect the intact nerves.

Intrathecal catheters and drug administration
An intrathecal catheter was inserted as previously described [13]. After anesthesia, a PE-10 polyethylene catheter was implanted between the L4 and L5 vertebrae to reach the spinal cord subarachnoid space before SNI. The rats that showed neurological de cits after catheter implantation were euthanized. IRF8 siRNA (140 pmol; AM16708; Abcam, Cambridge, UK) was dissolved in 20 µL of NFW. The rats were intrathecally treated with 20 µL IRF8 siRNA or 20 µL NFW via the catheter, followed by washing with 10 µL NFW on days 5 and 6 after SNI in the IRF8 siRNA group and NFW group, respectively. PRF-DRG therapy PRF-DRG was conducted as described previously [13,14]. The rats were anesthetized with sodium pentobarbital (50 mg/kg, i.p.). The left L5 DRG was exposed in the SPRF and PRF groups through laminectomy and facetectomy without injuring the dura mater. An RF electrode (type 20 G, 5 cm long, 4 mm active tip) was placed adjacent to the corresponding DRG via direct visualization using a radiofrequency device (Cosman Medical, Inc., Burlington, MA, USA). PRF waves were applied after performing a nerve stimulation test through muscle contraction of the lower extremities. The parameters for stimulating the PRF waves were as follows [24]: pulse rate of 2 Hz, voltage of 45 V, maximum temperature of 42°C, pulse width of 20 ms, and stimulated time of 6 min. No PRF current was delivered to the SPRF group.

Behavioral measurements
All behavioral measurements were performed between 8 am and 2 pm by an experimenter blinded to the grouping. 50% Paw withdrawal threshold (50% PWT) test Rats were placed in individual Plexiglas cells (10×15×20 cm) on a wire mesh oor for 15 min. Von Frey laments (Stoelting, Wood Dale, IL, USA) were applied to the plantar surface of the left hind paw using the up-and-down method as previously described [25]. The lament was subjected to a bending force for 3-4 s applied to the hind paw plantar. Avoiding further contact with the lament, quickly turning the head away, scratching the stimulated area, or attacking the lament were considered positive responses. The 50% paw mechanical withdrawal threshold (50% PWT) was measured according to the methods reported by Chaplan et al [25].

Sucrose preference test (SPT)
This test was performed on days 21 and 42, as previously described [26], with minor modi cations. A 48h two-bottle free-choice paradigm was conducted on rats, one of which contained tap water, and the other contained 1% sucrose solution. After adaptation, the animals were food-and water-deprived for 23 h, followed by the SPT, in which the rats were housed in individual cages exposed to two pre-weighed test solution bottles for 1 h. To avoid possible side effects, the positions of the water and sucrose bottles were changed after 30 min. At the end of the test, uid consumption was recorded by reweighing the bottles. The sucrose preference (%) was calculated using the following equation: sucrose preference (%) = sucrose consumption/(sucrose consumption + water consumption) × 100%.

Forced swimming test (FST)
The forced swimming test was performed on days 21 and 42 after SNI, as described in a previous study [27]. After a minimum of 1 h of habituation in the test room, the rats were placed in a Plexiglas cylinder (40 cm height × 20 cm diameter) containing water (30 cm depth) at a temperature of 25 ± 1°C. The rats were not able to touch the bottom with their hind paws. The FST consisted of two parts: pretesting for 15 min and testing for 5 min, which was conducted 24 h after the rst session. After both sessions, the rats were removed from the water and allowed to dry for 15 min at 32°C before being returned to their home cages. This session was videotaped using a digital camera for subsequent analysis.

ELISA analysis
On days 21 and 42 after SNI, the bilateral hippocampi were dissected and immediately frozen in liquid nitrogen. Samples were added to PBS (pH 7.4), homogenized by hand or grinders, and centrifuged for 20 min at 3000 rpm. The supernatant was then removed. The levels of IL-1β, IL-6, and TNF-α were determined using rat IL-1β, IL-6, and TNF-α ELISA Kit (Shanghai Mlbio Technology, China) with a commercial reagent kit following the manufacturer's instructions.
All of the above pro les (IHC, WB, RT-qPCR, and ELISA) were detected and counted by an observer blinded to the treatment.

Statistical Analysis
Statistical analysis was performed using SPSS software (version 20.0; IBM, Armonk, NY, USA). All experimental results were expressed as the mean ± standard deviation (SD). To analyze the 50% PWT, we used a repeated-measures two-way ANOVA followed by the LSD test. The sucrose preference rate, forced swimming immobility time, and the levels of IRF8, BDNF, and pro-in ammatory cytokines were analyzed using one-way ANOVA, followed by the LSD test. Statistical signi cance was set at p < 0.05.

Results
PRF-DRG improves the neuropathic pain-induced depressive behaviors in rats with SNI First, we examined the changes in the mechanical pain threshold in rats with SNI. Compared with that in the sham group, the 50% PWT in the SNI group was signi cantly reduced after SNI (P < 0.05, Additional le 1: Table 1, Table 2, and Fig. 2), and reached a peak on day 7 after SNI; thereafter, it maintained a lower level until the end of the observation. The occurrence of mechanical allodynia indicated the successful establishment of the neuropathic pain model. Second, we explored whether depressive behaviors could be induced by neuropathic pain in the present study. On day 42 after SNI, the sucrose preference rate was signi cantly reduced (P < 0.05, Fig. 3a), whereas the forced swimming immobility time was remarkably prolonged (P < 0.05, Fig. 3b) in the SNI group compared to the sham group. However, there were no signi cant differences in the sucrose preference rate and forced swimming immobility time among the six groups on day 21 after SNI. These ndings indicated that in the present study, SNI induced depressivelike behaviors in rats on day 42 but not on day 21 after SNI. Third, on day 42 after SNI, the 50% PWT (P < 0.05, Additional le 1: Table 1, Table 2, and Fig. 2) and sucrose preference rate were increased (P < 0.05, Fig. 3a), while the forced swimming immobility time was reduced (P < 0.05, Fig. 3b) in the PRF group compared with those in the SPRF and SNI groups. Thus, PRF-DRG could ease the depressive behaviors induced by SNI. As can be seen from Table 2 overall, the difference of 50% PWT in each group was statistically signi cant, and the difference of 50% PWT at each time point was statistically signi cant. At the same time, there was a cross effect between the group and time, so it could be considered that the 50% PWT was not identical at different time points, and the rate of change at different times in different groups was not identical. In contrast, we can see that sham>PRF=IRF8 siRNA>SNI=SPRF=NFW, indicating that the 50% PWT in the SNI group, PRF group, IRF8 siRNA group, SPRF group, and NFW group were signi cantly reduced after SNI, and reached a peak on day 7 after SNI, and 50% PWT was increased in the PRF group and IRF8 siRNA compared with those in the SPRF, SNI, and NFW groups. There were no signi cant differences in the 50% PWT between the PRF and IRF8 siRNA groups.
PRF-DRG therapy increases the hippocampal BDNF by inactivating the spinal IRF8 to involve in the improvement of depressive-like behaviors in rats with SNI On day 42 after SNI, we found that the 50% PWT (P < 0.05, Additional le 1: Table 1, Table 2, and Fig. 2) and the sucrose preference rate were increased (P < 0.05, Fig. 3a), while the immobility time in the FST was reduced (P < 0.05, Fig. 3b) in the IRF8 siRNA group compared with those in the NFW and SNI groups. Thus, intrathecal injection of IRF8 siRNA could protect against depressive-like behaviors induced by SNI in rats. We next evaluated the expression of BDNF in the hippocampus and IRF8 in the spinal cord of rats by IHC (Fig.4), qRT-PCR (Fig.5), and WB (Fig.6). The levels of hippocampal BDNF were remarkably decreased, and spinal IRF8 was signi cantly increased in the SNI and NFW groups compared with those in the sham group on days 21 (P < 0.05) and 42 after SNI (P < 0.05). We further examined whether hippocampal BDNF expression could be in uenced by spinal IRF8 activation. Biochemical analysis showed that intrathecal injection of IRF8 siRNA reversed the increased spinal IRF8 (P < 0.05) and decreased hippocampal BDNF expression on days 21 (P< 0.05) and 42 (P < 0.05) in the IRF8 siRNA group compared with those in the NFW and SNI groups. Therefore, these results indicate that spinal IRF8 negatively regulates hippocampal BDNF expression.
Similar to the IRF8 siRNA group, biochemical analysis ( Fig. 4-6) showed that PRF-DRG therapy inhibited the enhanced expression of spinal IRF8 (P < 0.05) while reversing the reduction in hippocampal BDNF expression in the PRF group compared to that in the SPRF group on days 21 (P < 0.05) and 42 (P < 0.05) after SNI. There were no signi cant differences in the depressive-like behaviors, the levels of spinal IRF8, and hippocampal BDNF between the PRF and IRF8 siRNA groups on days 21 and 42 after SNI. Since spinal IRF8 activation could mediate the reduction of BDNF in the hippocampus, we considered that PRF-DRG therapy could mimic the IRF8 siRNA to improve depressive-like behaviors. Thus, PRF-DRG probably increased hippocampal BDNF by inactivating spinal IRF8 to improve pain and depression comorbidity.
PRF-DRG therapy also negatively regulates hippocampal IL-1β and TNF-α by inactivating spinal IRF8 to ameliorate depressive-like behaviors in rats with SNI We also evaluated the expression levels of pro-in ammatory factors (IL-1β, IL-6, and TNF-α) in the hippocampus of rats by ELISA (Fig. 7). On day 42 after SNI, the concentrations of hippocampal IL-1β and TNF-α were increased (P < 0.05) in the SNI, SPRF, and NFW groups compared with those in the sham group (P < 0.05). There were no signi cant differences in IL-1β and TNF-α among the six groups on day 21 and in IL-6 among the six groups on days 21 and 42. These data indicate that hippocampal IL-1β and TNF-α are involved in neuropathic pain-induced depression. Compared with those in the SNI, SPRF, and NFW groups, intrathecal injection of IRF8 siRNA or PRF-DRG treatment resulted in a signi cant reduction in hippocampal IL-1β and TNF-α concentrations on day 42 after SNI (P < 0.05, Fig. 7). There were no signi cant differences in IL-1β and TNF-α levels between the PRF and IRF8 siRNA groups on day 42.
Hence, PRF-DRG therapy could mimic intrathecal injection of IRF8 siRNA to reduce the levels of hippocampal IL-1β and TNF-α. In summary, PRF-DRG decreased hippocampal IL-1β and TNF-α by inactivating spinal IRF8 to participate in pain and depression comorbidity improvement.

Discussion
In the present study, SNI induced long-lasting pain sensitization and resulted in pain-induced depressive behaviors. In addition, the overexpression of IRF8 in the spinal cord, downregulation of BDNF, and increase of IL-1β and TNF-α in the hippocampus of rats with SNI were detected. In contrast, PRF-DRG or intrathecal injection of IRF8 siRNA relieved mechanical allodynia and depression-like behaviors and reversed the spinal IRF8 overexpression, hippocampal BDNF reduction, IL-1β, and TNF-α increase after SNI. These results extend the understanding of the effective mechanisms of PRF on the DRG for treating neuropathic pain-induced depression. To the best of our knowledge, this is the rst report on the PRF-DRG to modulate hippocampal neuroin ammation to relieve neuropathic pain-associated depressive behaviors.
Effects of PRF therapy on neuropathic pain-induced depressive behaviors PRF is a minimally neurodestructive therapy used for the treatment of chronic neuropathic pain. Previous studies have suggested that the DRG is an active participant in the development and maintenance of neuropathic pain [28][29][30]. Targeted modulation of the DRG is considered a critical therapeutic option for neuropathic pain management [31]. The application of PRF on the DRG showed a better analgesic effect than the treatment of the injured peripheral nerve [15,24,32,33]. However, research on the treatment of neuropathic pain and depression comorbidity using this method is rare. In the present study, we not only observed the analgesic effects but also focused on improving depressive behaviors induced by SNI.
We observed that the 50% PWT and sucrose preference percentage were signi cantly decreased, while the immobility time of FST was signi cantly increased due to SNI. Our results reproduced the previous ndings that animals with PNI could display anhedonia and behavioral despair [34,35]. In the current study, PRF-DRG therapy resulted in a dramatic increase in 50% PWT in rats from day 8 to day 42 after SNI, which was consistent with the results reported by Liu et al [36]. These data strengthen the conclusion that the application of PRF-DRG could attenuate neuropathic pain induced by SNI. Notably, the present study recon rmed our previous ndings on the restorative effects of PRF-DRG on neuropathic paininduced depression [13]. Taken together, these results support the antidepressant-like effects of PRF-DRG.
Therapeutic mechanisms of PRF-DRG on pain-depression comorbidity at the levels of spinal cord and hippocampus

Effects of PRF-DRG on IRF8 in the spinal cord
Previous studies have provided evidence that neuroimmune responses contribute greatly to the development of nerve injury-induced neuropathic pain and depression [37]. Microglial cells are resident immune cells that monitor the central nervous system (CNS) and play an essential role in the pathogenesis of many nervous system diseases and conditions, including neuropathic pain and depression. After PNI, microglia cells transform into a reactive phenotype, initiating a progressive series of cellular and molecular changes in the spinal cord. Notably, IRF8 is a critical transcription factor that transforms microglia into a reactive phenotype to release multiple in ammatory cytokines. IRF8 de ciency or PRF-DRG treatment in PNI models can prevent microglial activation and improve pain sensitivity and depressive-like behaviors [11,13,14]. The current results reproduced our previous ndings that PRF-DRG or intrathecal IRF8 siRNA reversed the overexpression of spinal IRF8 in rats with SNI, thereby contributing to the improvement of depression-like behaviors. Since PRF-DRG can alter the spinal immune environment and exert immunomodulatory and biological effects on cell morphology, synaptic transmission, and pain signals [38], we assumed that the anti-depressive-like effects of PRF-DRG might be achieved by targeting microglial inactivation via the IRF8 signaling pathway at the spinal cord level to prevent neurotransmission between the spinal cord and depression-related brain regions.

Effects of PRF-DRG on BDNF in the hippocampus
The hippocampus is a crucial limbic area involved in anxiety, depression, learning, and memory [39]. Recent studies have reported abnormal cytokine expression in the hippocampus of animal models of persistent pain [40]. Additionally, there was a signi cant association between depressive illness and loss of hippocampal volume in the form of atrophy was reported [41]. BDNF is a major neuronal growth factor in the brain that regulates neurogenesis, neuronal maturation, survival, and synaptic plasticity. Recent studies have demonstrated that BDNF signaling in the hippocampus is signi cantly correlated with in ammation-related depression [42]. Abnormal BDNF expression changes in the hippocampus are involved in the neuropathology of comorbid neuropathic pain and depression [19,20]. Clinical studies have also shown that BDNF levels are reduced in plasma or brain tissues, especially the hippocampus, in patients with depression. In contrast, an increase in BDNF expression remarkably improved depressive disorders [43,44]. Thus, hippocampal BDNF has been suggested as a promising therapeutic target for depression. In our previous study, PRF-DRG exerted antidepressant-like effects in rats with comorbid neuropathic pain and depression. However, the impact of PRF-DRG on BDNF expression in the hippocampus has not been reported.
In the present study, we extended our previous research to focus on the role of BDNF in the hippocampus. Signi cantly decreased hippocampal BDNF levels in the SNI group were detected on days 21 and 42. The bene cial effects of PRF-DRG therapy on neuropathic pain-induced depressive behaviors were accompanied by the reversal of the reduced levels of BDNF in the hippocampus. These ndings revealed that restoration of hippocampal BDNF expression might exert anti-depressive effects. However, the aforementioned BDNF alterations preceded the changes in sucrose preference rate and immobility time in the FST on day 21. This nding is consistent with clinical observations that report a time lag spanning several weeks before the therapeutic effect. The delayed anti-depressive-like effect may result from multiple complicated factors, and hippocampal BDNF is sensitively triggered and downregulated by PNI. In contrast, in our recent study, the BDNF levels in the nucleus accumbens (NAc) and spinal cord were upregulated in rats with SNI [13,15]. The opposite ndings suggested that BDNF was differentially regulated in the hippocampus, spinal cord, and NAc after SNI, and PRF-DRG could negatively regulate BDNF in different regions.

Effects of PRF-DRG on neuroin ammation in the hippocampus
Neuroin ammation is considered a common mechanism underlying depression comorbidity of chronic pain. A meta-analysis found a signi cant elevation in TNF-α, IL-6, and IL-1β levels in patients with depression compared to subjects without depression [7]. In addition, in ammatory reactions in discrete brain areas (cingulum, hippocampus, and prefrontal cortex) have been observed in an animal model of neuropathic pain, which affects the descending regulation of pain and the development of pain-induced depression behaviors [45]. In recent years, some studies have revealed that pro-in ammatory factors are markedly increased in the hippocampus of animals with comorbid neuropathic pain and depression [8,46]. In addition, affective disturbances in neuropathic pain could be associated with neuroin ammation in the interconnected hippocampal-medial prefrontal circuitry [10].
In this study, we noted a statistically signi cant increase in IL-1β and TNF-α levels on day 42 after SNI, while IL-6 was not signi cantly changed on days 21 and 42 after SNI. These results suggest that IL-1β and TNF-α might be more sensitive than IL-6 in this model. IL-1β is a pleiotropic cytokine released from glial cells and is involved in a wide variety of disease states associated with immune and in ammatory responses, including neuropathic pain and depression [47]. Another source of IL-1β in the CNS is local ampli cation via autocrine and/or paracrine mechanisms [48]. The hippocampus is rich in IL-1β binding sites with the highest density [49]. IL-1β signaling applies to the IL-1 receptor type I on the surface of different types of cells and generates signals to the downstream pathways of mitogen-activated protein kinases and nuclear factor kappa-B, leading to global gene regulation [7]. TNF-α is a multifunctional proin ammatory cytokine produced by a variety of cell types in response to in ammatory stimulation, including neurons, glial cells, and immune cells, and plays a regulatory role in the growth, proliferation, differentiation, and apoptosis of nerve cells [50]. Almost all noxious stimulations, including pain, are associated with immune system activation and pro-in ammatory cytokine release. The pro-in ammatory cytokine cascades initially increase TNF-α expression, and TNF-α is the rst cytokine to appear in the cascade after stress activation. Animals with elevated TNF-α levels in the hippocampus are prone to chronic pain and depressive behaviors [19,51]. There is evidence that sustained nociceptive stimulation leads to increased levels of TNF-α in the hippocampus, which mediates chronic pain and pain-related affective components. Elevated TNF-α levels, in turn, activate the HPA axis, reduce glucocorticoid response, inhibit the production of norepinephrine, deregulate monoamine and glutamate levels, and reduce BDNF in the hippocampus, thus eliciting comorbid pain and depression [17,51]. These data might explain why hippocampal IL-1β and TNF-α levels were more signi cant in the current study. Thus, targeting neuroin ammation in the hippocampus is likely to be a new way to treat pain-depression comorbidity. We further observed that the bene cial effects of PRF-DRG therapy were accompanied by a signi cant reduction in hippocampal IL-1β and TNF-α levels. Therefore, in our study, PRF-DRG regulated neuroimmunity mainly through IL-1β and TNF-α signaling pathways rather than IL-6, thereby alleviating pain-induced depressive-like behaviors. This study revealed a potential anti-in ammatory mechanism, especially the negative regulation of hippocampal IL-1β and TNF-α, underlying the therapeutic effects of PRF-DRG therapy for neuropathic pain-depression dyad.
Regulation of hippocampal BDNF and pro-in ammatory cytokines via spinal IRF8 inactivation participates in anti-depressant mechanism of PRF-DRG Since IRF8 plays a crucial role in activating spinal microglia and anti-depressive-like effects could be achieved by targeting spinal IRF8 signaling, we further investigated the correlation between the expression levels of spinal IRF8 and hippocampal BDNF and pro-in ammatory cytokines. We found that the levels of hippocampal BDNF, IL-1β, and TNF-α were reversed by IRF8 siRNA intrathecally in the present study; therefore, we deduced that inactivation of spinal IRF8 expression should mediate hippocampal BDNF, IL-1β, and TNF-α levels, which contribute to the improvement of neuropathic pain-associated depressive symptoms. In our study, profound antidepressant-like behaviors were conferred by PRF-DRG and IRF8 siRNA. We observed that PRF-DRG therapy had an effect similar to that of an intrathecal injection of IRF8 siRNA. Both treatments reduced the overexpressed spinal IRF8, reversed the reduced hippocampal BDNF, and inhibited the increased IL-1β and TNF-α levels after SNI. No signi cant differences were detected between the PRF and IRF8 siRNA groups. Therefore, PRF-DRG therapy could mimic IRF8 siRNA to improve depressive-like behaviors.
BDNF protects neurons from damage caused by nerve injury and plays a key role in hippocampal plasticity and memory processes that are disrupted by the dysregulation production of IL-1β . TNF-α is most likely to modulate BDNF production in the hippocampus. According to studies on chronic pain and depression, an increase in TNF-α beyond physiological levels decreases BDNF production in the hippocampus [17]. It was found that BDNF levels were decreased in the hippocampus after SNI, and these changes were mediated by microglial cell activation and TNF receptor 1. It has been suggested that TNF-α regulates BDNF expression and modulates hippocampal synaptic plasticity in the hippocampus through a microglia-dependent mechanism [19]. Moreover, it was suggested that the interaction of BDNF and TNF-α genes might decrease the expression of BDNF in the hippocampus and cause cognitive dysfunction in patients with schizophrenia [57]. However, whether activated spinal IRF8 inhibits BDNF signaling in the hippocampus via weakening hippocampal IL-1β and TNF-α needs to be explored in future research.
In summary, reversing the decreased BDNF and enlarged IL-1β and TNF-α in the hippocampus mediated by the inactivated spinal IRF8 signaling pathway underlies PRF-DRG for improving neuropathic paininduced depression in rats with SNI. These ndings also indicate that non-pharmacological treatments could depress spinal microglial activation, thereby mitigating hippocampal neuroin ammation. However, PRF-DRG in the present study only partially alleviated the neuropathic pain-induced depression behaviors, and future investigations are imperative for the other underlying mechanisms to achieve a more satisfactory curative effect.
This study has some limitations. First, we did not use spinal IRF8 activators to further explore whether hippocampal BDNF and pro-in ammatory cytokine levels are closely related to the spinal IRF8 signal pathway. Second, we did not pretreat the SNI rats with BDNF blockers to investigate whether the antidepression effect of the PRF-DRG could be abolished. Third, no signi cant differences in hippocampal IL-6 levels were detected among the groups. These negative results may be related to the different experimental conditions. These limitations need to be addressed in future in-depth studies.
Conclusion PRF-DRG therapy plays an immunoregulatory role in weakening neuroin ammation in the spinal cord and hippocampus. Regulation of BDNF, IL-1β, and TNF-α in the hippocampus via spinal IRF8 inactivation participates in the therapy of PRF-DRG to ease comorbid pain and depression.

Consent for publication
All authors agree to the publication of this manuscript.

Availability of data and materials
The datasets used and/or analyzed in the current study are available from the corresponding author upon reasonable request.

Competing interests
Each author declares that he or she, or a member of his or her immediate family, has no commercial association that might pose a con ict of interest in connection with the submitted manuscript.

Funding
This study was supported by the Project of Medical Innovation of Fujian Province, China (2018-CX-6).
Authors' contributions XXR designed and performed the experiments, analyzed and interpreted the data, and wrote the manuscript. DZS helped with data collection and interpretation and contributed to critical manuscript revision. LRG conceived the study, obtained funding, and critically revised the manuscript. LC and LF offered lab instruments and participated in data analysis and the revised version. All the authors have read, revised, and approved the nal manuscript.
52. Patterson SL. Immune dysregulation and cognitive vulnerability in the aging brain: Interactions of microglia, IL-1β, BDNF and synaptic plasticity. Neuropharmacology. 2015;96:11-8. Change chart of 50% PWT mean ±SD for each group Figure 3 Page 21/23 Alterations of sucrose preference rate and forced swimming immobility time on day 42 (a) The reduction of sucrose preference rate induced by SNI on day 42 was partly reversed by intrathecal injection of IRF8 siRNA or PRF-DRG therapy in rats. (b) The increase of immobility time induced by SNI on day 42 was reduced after applying PRF-DRG therapy or intrathecal injection of IRF8 siRNA. Data were expressed as mean ± SD. *P < 0.05 vs. Sham group; #P < 0.05 vs. SPRF group; &P< 0.05 vs. NFW group. One-way ANOVAs followed by the LSD test, n=10 per group.  RT-qPCR analysis of hippocampal BDNF and spinal IRF8 mRNA expression on days 21 and 42 (a, b) RT-qPCR data showed that a reduction of BDNF mRNA levels induced by SNI was reversed by intrathecal injection of IRF8 siRNA or PRF-DRG therapy on days 21 and 42. (c, d) RT-qPCR data revealed that the increase of spinal IRF8 mRNA expression was downregulated by intrathecal injection of IRF8 siRNA or PRF-DRG therapy in SNI rats on days 21 and 42. Data were expressed as mean ± SD. *P < 0.05 vs. Sham group; #P< 0.05 vs. SPRF group; &P< 0.05 vs. NFW group. One-way ANOVAs followed by the LSD test, n=5 per group.

Figure 7
ELISA analysis of concentration of hippocampal IL-1β and TNF-α on day 42 after SNI The hippocampal IL-1β and TNF-α increase induced by SNI was reversed by intrathecal injection of IRF8 siRNA or PRF-DRG therapy. Data were expressed as mean ± SD. *P < 0.05 vs. Sham group; #P < 0.05 vs. SPRF group; &P < 0.05 vs. NFW group. One-way ANOVAs followed by the LSD test, n=5 per group.