Intervertebral foramen injection of plerixafor attenuates CCD-induced low back pain in rats through down-regulation of Nav1.8 and Nav1.9


 BackgroundLow back pain, an extremely common chronic pain, is now the number one cause of disability globally. On the chronic compression of the dorsal root ganglion (CCD) model, a typical low back pain model, previous study has discovered that C-X-C chemokine receptor type 4 (CXCR4) involved in mediating low back pain. However, the underlying molecular mechanisms and whether it has the potential to serve as an applicable target for clinical treatment of low back pain remains unclear. MethodsCCD-induced low back pain model was established by inserting stainless steel L-shape rod into the intervertebral foramen at L5 level. Multiple pain behaviors including mechanical hyperalgesia and allodynia, thermal hyperalgesia and cold allodynia were evaluated. The role of CXCR4 in CCD-induced pain was assessed by intraganglionar CXCR4 siRNA injection and intervertebral foramen injection of plerixafor, a selective CXCR4 antagonist. The expression change of CXCR4, Nav1.8 and Nav1.9 was examined by immunofluorescent staining.ResultsWe showed that the CXCR4 were dramatically up-regulated in the compressed dorsal root ganglion (DRG) neurons in CCD model, and intraganglionar CXCR4 siRNA injection significantly reduced CCD-induced multiple pain behaviors, including mechanical hyperalgesia and allodynia, thermal hyperalgesia and cold allodynia. The expression of Nav1.8 and Nav1.9 in the compressed DRG were also enhanced after CCD, and which were remarkably reversed by intervertebral foramen injection of plerixafor at the compression level. Moreover, intervertebral foramen injection at the compression level of plerixafor and ropivacaine as well as A-803467, a Nav1.8 selective blocker, were all able to reverse CCD-induced multiple pain behaviors. However, the analgesic duration of plerixafor maintained at least for 24 h which was much longer than that of A-803467 and ropivacaine. Finally, we found intervertebral foramen injection of plerixafor at the adjacent non-compression level has no effect on CCD-induced pain and intervertebral foramen injection of plerixafor in normal rats did not affect their basal pain sensitivity.ConclusionsOur study provides evidence that CXCR4-Nav1.8/Nav1.9 axis in compressed DRG contributed to CCD-induced low back pain and intervertebral foramen injection of plerixafor was a potential and applicable therapeutic strategy for the treatment of low back pain.


Background
Low back pain, the most common chronic pain, is the leading cause of disability and higher health care costs over the world which severely affect the patients' quality of daily life [1]. For many patients, the pharmacotherapy of low back pain is challenging and the effective treatment methods is lacking [2].
Currently, an urgent need exists for the development of novel, more effective treatment approaches and drugs for low back pain. To narrow the gap between understanding the underlying mechanism and its clinical treatment, a chronic compression of dorsal root ganglion (CCD)-induced low back pain model in murine has been developed attempting to mimic the common clinical causes of low back pain, such as intervertebral disc laterally herniation, intervertebral foramen stenosis, or other disorders affecting the dorsal root ganglion (DRG) or its near nerve root [3][4][5]. More and more work is endeavoring to explore the new potential targets or strategy for pain relief on CCD model. The chemokine CXC motif receptor 4 (CXCR4) belonging to the well-studied G protein-coupled receptor (GPCR) family is a primary cognate receptor for the chemokine CXC motif ligand 12 (CXCL12) which plays an important role in many biological functions, including embryogenesis, immune surveillance, in ammation response, tissue homeostasis, tumor growth and metastasis [6][7][8]. Much evidence has accumulated that CXCR4 chemokine signaling contributes to the development and maintenance of chronic pain [9][10][11]. A recent study have demonstrated that intraperitoneal injection of plerixafor (AMD3100), a selective antagonist for CXCR4, which has been approved for mobilization of hematopoietic stem cells clinically by the U.S. Food and Drug Administration in 2008, signi cantly ameliorated CCD-induced mechanical and thermal hyperalgesia, suggesting CXCR4 might be an adequate analgesic target [12][13][14]. As we known, the CXCR4 is widely located in the nervous system and distributed in various kinds of cells [9,15]. Consequently, the abovementioned drug systemic administration modality will inevitably result in a number of adverse events and blur the action site of the plerixafor. Giving that the dorsal root ganglion (DRG) containing bodies of primary sensory neurons which may generate abnormal spontaneous discharges and become an important source of hyperexcitability and increased nociceptive signaling, the intervertebral foramen injection aiming at delivering drugs onto the DRG might be an appropriate alternative approach for drug delivery [16][17][18].
Emerging data veri ed that the CCD causes an increase in the excitability of the cell bodies of compressed DRG neurons and the CCD-induced spontaneous activity and/or lowered rheobase of DRG neurons was accompanied, and possibly caused by, an increased expression of voltage-gated sodium currents (VGSCs) [19][20][21]. Recent research proved that knockdown of NaV1.6 and regulatory NaVβ4 sodium channel markedly reduced spontaneous pain and mechanical and cold hyperalgesia induced by CCD, indicating that targeting the VGSCs may have therapeutic value for low back pain [22]. However, the roles of two other distinct VGSCs identi ed in DRG neurons, Nav1.8 which plays an important role in action potential generation, and Nav1.9 which contribute to setting the resting membrane potential and to neuronal excitability [23], have not yet been explored in CCD-induced low back pain. Our previous study has showed that activation of CXCR4 could directly modulate primary neuronal excitability by mediating the up-regulation of Nav1.8 under in ammatory pain state [24]. Interestingly, the CXCR4 expression also signi cantly increased in the compressed DRG from CCD mice [12]. Whether the Nav1.8 and Nav1.9 was the downstream effector target for CCD-induced CXCR4 activation remains unclear.
Considering all, we set out to investigated whether CXCR4 in the compressed DRG participated in CCDinduced multiple pain behaviors, and then tested whether delivery of plerixafor to the compressed DRG by intervertebral foramen injection was an effective approach to management CCD-induced pain. Moreover, we extended our study to demonstrate that CXCR4 activation regulated the expression of Nav1.8 and Nav1.9. The present ndings might provide preclinical evidence for targeting CXCR4 by intervertebral foramen injection as a potential therapeutic strategy for the treatment of low back pain.

Animals
Experiments were performed on male Sprague-Dawley rats (200-250 g, provided by Laboratory Animal Center of Fuzhou General Hospital). The animals were housed in plastic cage under speci c pathogenfree conditions with a 12 h light/dark cycle and with free access to water and food. This study was carried out in accordance with the recommendations of the Animal Care and Use Committee of Fuzhou General Hospital and performed in accordance with the updated Guide for the Care and Use of Laboratory Animals. The number of rats used and their suffering were minimized. During the whole experiment, the rats were randomized 2.2 Chronic compression of DRG CCD model were established according to the methods described previously [25]. Under anesthesia of sodium pentobarbital (50 mg/kg, i.p.), the rats were xed in a prone position and the left transverse process and intervertebral foramina of L5 were exposed. A stainless steel L-shape rod was implanted unilaterally into the intervertebral foramen L5. After the rod was implanted, the wound was closed by suturing muscle and skin layers. The surgical procedure of sham surgery was identical to that described above but without the rod insertion. All rats were allowed to recover in cages from anesthesia and surgery with food and water available at libitum. These rats with autophagy phenomenon, feeling de ciency, and disability were eliminated.

Intraganglionar siRNA microinjection
Intraganglionar siRNA microinjection was just prior to the insertion of metal L-shape rod. By using in vivo transfection reagent ( R0541, Thermo Scienti c Inc) as the delivery vehicle, the CXCR4 small interfering RNA (siRNA) solution (2 μl, 10 μM, sc-270577, Santa Cruz) or no-targeting control scrambled siRNA (2 μl, 10 μM, sc-37007, Santa Cruz) was injected close to L5 DRG with a glass Hamilton microsyringe. The microsyringe was remained for 5 min after injection to prevent spread of the agent followed by rod implanted into the intervertebral foramen L5.

Intervertebral foramen injection
Intervertebral foramen drug injection was implemented as we previously described 16 . Brie y, rats were anesthetized with sevo urane and the fur over the lower back were shaving. A 30 G injecting needle was inserted towards the L5 intervertebral foramen or L4 intervertebral foramen through the skin at an angle of 45° both to the horizontal and sagittal plane until the tip touched the lateral region of the vertebrae. To localize the L5 or L4 intervertebral foramen, delicate movements of the needle were made until the bone resistance was diminished and a paw inch re ex was observed. The paw inch re ex was used as a sign that the needle tip enters in the intervertebral foramen. In all intervertebral foramen injection, 10 μL of solution was injected over a period of 15 seconds and saline was served as the control for plerixafor (Abcam, 10, 50, 100 μg/10 μL), A803467 (Abcam, 30 μg/10 μL) and ropivacaine (AstraZeneca LP, 40 μg/10 μL, equivalent to 0.4% in concentration). To con rm the intervertebral foramen injection, the presence or absence of dye (DiI, 2.5%) was visually con rmed by the exposure of the lumber region after each experiment was completed

Mechanical pain sensitivity testing
The mechanical pain sensitivity testing was performed by experimenters blinded to the experimental treatments. For examination of mechanical pain sensitivity, the mechanical stimuli were applied by using a series of von Frey mono laments with different bending forces. The rats were placed on a metal mesh oor covered with plastic box, and von Frey laments were applied from underneath the metal mesh oor to the plantar area of the hind paws. Each von Frey lament was applied ve times (once every several seconds) in order to induce the withdrawal re ex. The bending force value of the von Frey lament that caused an appropriate 50% occurrence of paw withdrawal was expressed as the paw withdrawal mechanical threshold (PWMT, g).

Thermal pain sensitivity testing
The thermal sensitivity was tested by a blinded researcher and determined by measuring the withdrawal latency of the hind paws in response to radiant heat. Rats were placed in a plastic chamber on the surface of a 2-mm thick glass plate and the sensitivity of hindpaw to thermal stimuli by a radiant heat stimulator was measured. The latency was determined as the duration from the beginning of heat stimuli to the occurrence of a marked withdrawal re ex. Five stimuli were repeated, and the latter three values were averaged as mean paw withdrawal thermal latency (PWTL, s). A maximal cutoff of 30 s was used to avoid excessive tissue injury. The inter-stimulus interval for each heat test was more than 10 min.

Mechanical allodynia testing
The mechanical allodynia testing was performed by experimenters blinded to the experimental treatments For examination of mechanical allodynia, the rats were habituated on a metal mesh oor covered with transparent plastic box. A wisp of cotton pulled up from, but still attached to a cotton swab was stroked mediolaterally across the plantar surface of the hindpaws to measure the presence of a withdrawal re ex to a normally innocuous mechanical stimulus (light touch-evoked tactile allodynia) which does not evoke a response in normal animals. The percentage of withdrawal response to ten cotton stimuli were calculated.

Cold allodynia testing
The cold allodynia was tested by a blinded researcher and determined by measuring the withdrawal response of hindpaws to a drop of acetone applied to the ventral surface of the hind paw. Using a at-top syringe, we applied a droplet of acetone to the ventral surface of the hind paw of each rat through metallic mesh oor and observed their withdrawal responses. Acetone was applied alternately ten times to each hindpaw and the total percentage of withdrawal response were calculated.

Statistical analysis
Data were analyzed using GraphPad Prism version 7.0 (GraphPad,San Diego, CA, USA) and all data were expressed as means ± SD. Behavioral time course data were analyzed using a two-way repeated measures ANOVA with Holm-Sidak's post hoc test. Differences in immuno uorescence intensity of each group were tested using t-tests or one-way ANOVA followed by Tukey's posttest. A level of P < 0.05 was accepted as signi cant.

Knocking down the up-regulation of CXCR4 in the compressed DRG reduced CCD-induced multiple pain behaviors
Representative images of CXCR4 immuno uorescence staining in DRG sections from sham DRG and from compressed DRG on day 3 are shown in Fig.1A. The mean immuno uorescence intensity of CXCR4 in compressed DRGs were signi cantly increased comparing with that from sham DRGs (Fig.1B).
To determine whether the up-regulation of CXCR4 was involved in the CCD-induced pain behaviors, we knocked down the expression of CXCR4 by using intraganglionar siRNA microinjection as we reported previously [26]. As shown in the Fig.2A and Fig.2B, intraganglionar CXCR4 siRNA microinjection remarkably reduced CXCR4 expression in compressed DRG comparing with scrambled siRNA injection group. Previous studies have shown that CCD induces obvious mechanical and thermal pain sensitivity in the ipsilateral hindpaw as measured by the von Frey test and thermal stimuli, mechanical allodynia as measured by the paw withdrawal responses to a light cotton wisp stroking and cold allodynia as measured by increased withdrawal to acetone stimuli [22]. These multiple pain behaviors were also occurred as early as 1 day post CCD and maintained at least for 14 days (Fig.2). We found knockdown of CXCR4 expression in compressed DRG by intraganglionar CXCR4 siRNA microinjection signi cantly attenuated CCD-induced mechanical and thermal pain sensitivity over the whole observation period ( Fig.2C and D). Additionally, the mechanical and cold allodynia in CCD rats were also improved by intraganglionar CXCR4 siRNA microinjection ( Fig.2E and F).

Intervertebral foramenplerixafor injection reduced CCD-induced multiple pain behaviors
We used a selective and effective CXCR4 inhibitor plerixafor to assess whether focal intervertebral foramen injection was a potential new and effective route in treatment of CCD-induced low back pain.
Single plerixafor (10, 50 or 100 μg) was intervertebral foramen injection at L5 ipsilaterally on day 3 post CCD and CCD-induced multiple pain behaviors were evaluated at 0.5, 1, 3, 5, 24 and 48 hour after injection. As shown in Fig. 3A and Fig. 3B, intervertebral foramen plerixafor (50 and 100 μg) injection dose-dependently decreased CCD-induced mechanical and thermal pain sensitivity compared to saline injection, while low dose of plerixafor (10 μg) has no signi cant effect on decreased PWMT and PWTL induced by CCD comparing with saline injection. The analgesic effect of plerixafor (50 and 100 μg) initiated at 1 h, reaching peak at 5 h and lasting for 24 h after intervertebral foramen injection. The mechanical and cold allodynia induced by CCD were also attenuated by intervertebral foramen plerixafor (50 and 100 μg) injection ( Fig.3C and D). The ipsilateral paw withdrawal responding to cotton wisp stroking and acetone stimuli signi cantly decreased starting at 1 h and lasting for 24 h after plerixafor (50 and 100 μg) injection. However, the plerixafor (10 μg) did not signi cantly affect CCD-induced mechanical and cold allodynia over 48 h (Fig.3C and D). It has been demonstrated that activation of CXCR4 increased the excitability of DRG neurons through a mechanism involving voltage gated sodium channels (VGSCs) [27]. We further con rmed that Nav1.8 and Nav1.9 was a downstream effector of CXCR4 by examining the effect of plerixafor on Nav1.8 and Nav1.9 expression. As shown in Fig.4A and Fig.4B, the expression level of Nav1.8 and Nav1.9 in compressed DRGs of CCD rats remarkably increased compared to that DRGs from sham rats. However, intervertebral foramen plerixafor injection signi cantly reversed CCD-induced the up-regulation of Nav1.8 and Nav1.9 expression. These results suggesting that Nav1.8 and Nav1.9 might mediate the pronociceptive effect of CXCR4 activation under CCD condition.

Intervertebral foramen A803467 and ropivacaine injection reduced CCD-induced multiple pain behaviors
We next sought to explore the role of VGSCs in CCD-induced multiple pain behaviors by using a selective Nav1.8 blocker, A803467 and a non-selective local anesthetics, ropivacaine which was commonly used in clinic. As shown in Fig.5A and Fig.5B, CCD-induced mechanical and thermal hyperalgesia were signi cantly reversed by ipsilaterally intervertebral foramen A803467 (30 μg) injection at L5 level, and the analgesic effect of A803467 starting at 0.5 h, reaching peak at 1 h and lasting for only 3 h after injection. The mechanical and cold allodynia induced by CCD were also attenuated by A803467 (30 μg) from 0.5 h to 3 h after intervertebral foramen injection (Fig.5C and D), indicating that up-regulation of Nav1.8 in compressed DRG contributed to the maintenance of CCD-induced low back pain. Likewise, ropivacaine intervertebral foramen injection at L5 level could also remarkably reversed CCD-induced mechanical and thermal hyperalgesia, as well as the mechanical and cold allodynia (Fig.5). The analgesic effect of ropivacaine (40 μg) was onset at 0.5 h and remained at the peak level for 5 h, which was more signi cant than A803467 (Fig.5). However, at 24 h after intervertebral foramen injection, the anti-hyperalgesia and anti-allodynia effect of A803467 and ropivacaine both vanished (Fig.5).

Intervertebral foramenplerixafor injection at L4 had no effect on CCD-induced multiple pain behaviors
To address the question whether the analgesic effect of plerixafor attributed to its systemic effect, we performed plerixafor (100 μg) injection at L4 intervertebral foramen. Over the whole time after L4 intervertebral foramen plerixafor injection, the PWMT and PWTL, as well as the withdrawal responding to cotton wisp stroking and acetone stimuli remained as they had been before plerixafor injection (Fig.6), indicating that the present dose of plerixafor intervertebral foramen injection has no systemic analgesic effect.
3.6 Intervertebral foramenplerixafor injection had no effect on the basal mechanical and thermal pain sensitivity It has been reported that CXCR4 was constitutively expressed in primary afferent neurons and remain at a low expression level under physiological status while extremely increased under pathological condition. To evaluate the signi cance of this functional up-regulation of CXCR4, we applied L5 intervertebral foramen plerixafor injection in normal rats. As shown in Fig.7A, L5 intervertebral foramen plerixafor injection did not alter the ipsilateral PWMT to Von Frey test over the whole observed time. Also, there was no change in ipsilateral PWTL to thermal stimuli after plerixafor injection (Fig.7A). These results indicated that CXCR4 in DRG did not contribute to the basal pain sensitivity.

Discussion
CXCR4 is a recently reported chemokine receptor that plays a pivotal role in mediating neuropathic pain [9]. In the present study, we rst demonstrated that CXCR4 expression signi cantly increased in compressed DRG from CCD rats. Knockdown of CXCR4 remarkably attenuated CCD-induced multiple pain behaviors, including mechanical and thermal hyperalgesia, mechanical and cold allodynia, suggesting that CXCR4 contributed CCD-induced low back pain. Moreover, we showed that intervertebral foramen injection of plerixafor, a CXCR4 selective antagonist, was able to reverse CCD-induced pain behaviors by reducing CCD-induced Nav1.8 and Nav1.9 up-regulation in compressed DRG. We further found that the Nav1.8 selective blocker, A-803467 and a non-selective VGSCs blocker, ropivacaine cold both relieve CCD-induced pain behaviors after intervertebral foramen injection, but the analgesic duration was apparently shorter than that of plerixafor. Finally, we demonstrated that intervertebral foramen injection of plerixafor eased CCD-induced pain due to its local effect on compressed DRG rather than the systemic effects, and the approach of intervertebral foramen plerixafor injection has no signi cant effect on the basal pain sensitivity. Taken together, these ndings suggested that chemokine receptor CXCR4 involved in CCD-induced pain by up-regulating Nav1.8 and Nav1.9 expression, while intervertebral foramen plerixafor injection is a novel and appropriate therapeutic strategy to treat low back pain.
4.1 CXCR4-Nav1.8/Nav1.9 axis contributes to CCD-induced multiple pain behaviors Low back pain is a major health problem over the world which is becoming worse largely because of the increase in the population base and the intensi cation of aging [28]. Despite the huge medical investment in the low back pain, its global burden is still heavy. During the past few decades, the clinical management of low back pain remains unsatisfactory. Commonly, low back pain closely related to the intraforaminal stenosis, laterally herniated disk, vertebral fracture, or inflammatory disorders [2]. To develop an effective treatment, it is important to understand the underlying mechanism of low back pain, which necessarily promoted a number of preclinical models of low back pain being developed. In the present study, we used the CCD model induced by inserting a stainless steel rod between vertebrae holes (L4 and/or L5) which has been proven to be a typical model that mimics the clinical etiology and various symptoms of low back pain 3 . In line with previous studies, we identi ed four different pain behaviors in the ipsilateral hindpaw of CCD rats, including mechanical hyperalgesia, thermal hyperalgesia, mechanical allodynia and cold allodynia [22,29], which occurred as early as 1 day after surgery and maintained at least for 14 days.
Using CCD model, we found that the expression level of CXCR4 in compressed DRG dramatically increased revealed by immuno uorescent staining which was in accordance with a recent study [12]. The functional nature of the up-regulated CXCR4 was identi ed by our siRNA knockdown assay in which intraganglionar microinjection with CXCR4 siRNA obviously attenuated CCD-induced multiple pain behaviors. Additionally, blockade of CXCR4 with intervertebral foramen plerixafor injection showed the ability to reverse CCD-induced pain behaviors. All these results demonstrated that activation of CXCR4 in compressed DRG contributed to CCD-induced pain. As supporting, we previously have veri ed that intraganglionar microinjection with CXCR4 siRNA was able to relieve complete Freund's adjuvant (CFA)induced chronic in ammatory pain and spared nerve injury (SNI)-induced chronic neuropathic pain [26]. Moreover, the CXCR4 are found to be related with in ammatory pain-induced DRG neuronal hyperexcitability by our previous study in which plerixafor could inhibit the increased ring rate and reverse the lowered action potential rheobase [24]. Interestingly, it was reported that the DRG neuronal excitability extremely enhanced and the voltage-gated sodium currents in small and medium-sized DRG neurons also increased after compression [19,21,30]. It was well known that VGSCs were very important for electrogenesis and nerve impulse conduction, especially Nav1.8 which provided a signi cant contribution to action potential generation and Nav1.9 which contribute to setting the resting membrane potential and to neuronal excitability [31]. Here, we found for the rst time that the expression of Nav1.8 and Nav1.9 in the compressed DRG neurons dramatically increased and the A-803467, a selective Nav1.8 blocker, as well as the local anesthetic ropivacine was e cient to alleviate CCD-induced pain behaviors after intervertebral foramen injection, which strongly supported that Nav1.8 and Nav1.9 were involved in the pathogenesis of low back pain. Our further study showed that intervertebral foramen plerixafor injection was su cient to reducing the enhanced expression of Nav1.8 and Nav1.9 accompanying with easing pain behaviors, suggesting that the Nav1.8 and Nav1.9 were the downstream effector of CXCR4 activation for pain genesis. Actually, the contribution of CXCR4 activation-induced DRG neuronal protein synthesis to pain has been indicated in our previous study in which the protein translation inhibitor could block the acute mechanical hyperalgesia evoked by intraplantar injection of CXCL12, the ligand of CXCR4 [26]. In addition, we further have demonstrated that CXCL12-CXCR4 signaling could regulate the excitability of primary nociceptive neurons by up-regulating Nav1.8 expression and contribute to persistent pain and hyperalgesia induced by intraplantar bee venom injection [24]. Taking together, these ndings evidenced that CXCR4-Nav1.8/Nav1.9 axis in compressed DRG was involved in CCD-induced low back pain. Recently, another study demonstrated that NaV1.6 and regulatory NaVβ4 sodium channel in compressed DRG also play key roles in low back pain [22]. Besides, TRPV4, voltage-gated potassium current and hyperpolarization-activated cation current were all reported to be correlated with CCD-induced low back pain [30,32,33]. Whether these changed ion channels and currents shared the same upstream trigger mechanism (CXCR4 activation) remains unknown which are de nitely needed to be explored in the future experiments.

Blocking CXCR4 with plerixafor is a potential and e cient treatment for neuropathic pain
Neuropathic pain is considered to be the most intractable chronic pain with current therapeutic approaches being unsatisfactory and ine cient. Thus, it is urgent to explore the new potential targets for pain management. CXCR4 is a seven transmembrane G protein-coupled receptor (GPCR) which has been widely concerned in the eld of neuropathic pain research [9]. It is constitutively distributed in the somatosensory and nociceptive system [34]. In consistent with a recent study, we previously found that CXCR4 present in both non-peptidergic (IB4-positive) and peptidergic (SP-positive or CGRP-positive) primary nociceptor neurons and also co-localized with TRPV1, a thermonociceptor of primary sensory afferent [24]. In the central nervous system (CNS), spinal dorsal horn and thalamus as well as somatosensory cortex which are the primary centers for pain perception have a high level basal expression of CXCR4 [34]. All these ndings strongly suggested that CXCR4 is closely related to pain.
Here, we identi ed that the expression level of CXCR4 in normal DRG was at a low level, while it extremely enhanced under CCD state. In support of our nding, emerging evidence have demonstrated that CXCR4 expression in the nervous system was inducible under neuropathic pain state. In diabetic condition, the total number of DRG neurons expressing CXCR4 mRNA signi cantly increased comparing to normal DRG [35]. In chronic constriction injury (CCI) and SNI model, elevation of CXCR4 immunoreactivity has also been demonstrated in the DRG neuronal soma [36,37]. The expression level of CXCR4 in the CNS could also be obviously enhanced under central neuropathic pain situations such as central post stroke pain [10], spinal cord ischemia-reperfusion [38], spinal cord contusion injury [39]. This inducible expression pattern of CXCR4 enable CXCR4 as an ideal analgesic target. Actually, extensive behavioral pharmacology experiments have proven that blocking CXCR4 with its speci c antagonist, plerixafor, was su cient to prevent and reverse various neuropathic pain [9]. Importantly, plerixafor is hematopoietic stem cell mobilizer often used in clinic with its safety and e cacy has been approved by the U.S. Food and Drug Administration in 2008 [14]. It is noteworthy that intervertebral foramen plerixafor injection did not affect the basal pain response in normal rat, indicating that plerixafor could selectively relieve pathological pain without affecting the normal sensory or motor function. Additionally, we found that the analgesic duration of intervertebral foramen single plerixafor injection lasted at least for 24 h which was much longer than ropivacaine, a kind of local anesthetics commonly used for nerve blocking in clinic. We speculated that the protracted analgesic effect was largely because of the plerixafor not only reduced the expression of Nav1.8 and Nav1.9, but also has the ability to relieve in ammation in compressed DRG which was also a pivotal mechanism for neuronal hyperexcitability and chronic pain. In supporting, recent studies have demonstrated that the pro-in ammatory cytokines dramatically up-regulated accompanying with several anti-in ammatory cytokines down-regulated in compressed DRG [22], while plerixafor has been proven to be able to reducing the expression of TNF-α, IL-1β, IL-6 and in ammatory cell in ltration in the nervous system [10,35]. Collectively, the above characteristics and advantages of plerixafor make it an effective and applicable therapeutic medicine for neuropathic pain.

Intervertebral foramen injection is an appliable treatment approach for low back pain
As we known, many drugs used to treat neuropathic pain are delivered systematically and intrathecally which has a major strategic limitation. In addition to on-target effects, systemic delivery of these drugs has side effects because of off-target neural suppression in the peripheral nervous system (PNS), and even more so in the CNS. For plerixafor, the common adverse effects in clinical trials were dizziness, nausea and diarrhea which were observed in over 10% of patients, largely because CXCR4 is widely distributed in nervous system and presented in various kinds of cells [40]. An alternative approach for treating pain is delivering drugs into the DRG which has been recognized that DRG is a critical structure in sensory transduction and modulation, including pain transmission and the maintenance of persistent pain states. There are multiple changes in gene expression in DRG under pain condition, encompassing a large number of distinct family members including neuropeptides, receptors, ion channels, signal transduction molecules, synaptic vesicle proteins, and others [41,42]. The functional consequence of these changes is sensitization and hyperexcitability of DRG neurons, which then leads to persistent pain [18]. Moreover, the DRG is consistently locating at the intervertebral neural foramina within a capsule which is an easily accessible target. All these unique characteristics make the DRG an ideal target for the treatment of chronic pain [43]. Clinically, the local anesthetics and steroid were usually injecting into the intervertebral foramina to block the electrogenesis and pain conduction at DRG level to cure various kinds of pain with the assistance of visual equipment, like the ultrasound and computed tomography [44]. In the present study, we provided preclinical evidence that intervertebral foramina injection at the compressed DRG of plerixafor or local anesthetics was su cient to reverse the CCD-induced multiple pain behaviors. However, intervertebral foramina injection at the adjacent non-compressed DRG did not affect the pain behaviors, suggesting that the plerixafor dose for intervertebral foramina injection has no systemic effects. All of these data strongly indicated that intervertebral foramen plerixafor injection is an appliable treatment approach for low back pain.

Conclusion
The present study demonstrates for the rst time that CXCR4 activation-mediated Nav1.8 and Nav1.9 upregulation in compressed DRG neurons was involved in CCD-induced low back pain and intervertebral foramina injection of plerixafor was an e cient and applicable strategy for the treatment of low back pain.

Competing interests
The authors declare that they have no competing interests.

Funding
The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Ethics approval and consent to participate
This study followed the recommendations of the International Association for the Study of Pain (IASP) for the accomplishment of experimental studies in pain involving conscious animals. This study was also carried out in accordance with the recommendations of the Animal Care and Use Committee of Fuzhou General Hospital and performed in accordance with the updated Guide for the Care and Use of Laboratory Animals