Acupoint Injection of Dextrose Attenuates Chronic In ammatory Pain In Mice Through Modulation of Transient Receptor Potential V1 Channel Expression


 Background: Neuroinflammatory processes initiated by injury, infection, and other insults induce long-term alterations in the signaling characteristics of nociceptive neurons and associated non-neuronal cells, resulting in chronic inflammatory pain (CIP). Prolotherapy, also termed acupoint injection (AI) therapy, utilizes anatomically based meridians derived from Chinese Medicinal Theory to attenuate chronic pain. In fact, AI has demonstrated superior results compared to traditional acupuncture methods without significant side effects. Methods: In this study, we explored the efficacy and underlying mechanisms of AI on CIP in mice. The mouse hindpaw was injected with complete Freund’s adjuvant (CFA) to induce CIP, followed AI treatment. Results: After two weeks, indices of mechanical and thermal hyperalgesia were significantly reduced by AI treatment. CIP was associated with elevated expression levels of transient receptor potential V1 (TRPV1) channels, various downstream kinases and transcription factors, and nociception-associated voltage-gated sodium channels (Navs) in dorsal root ganglion (DRG), spinal cord, thalamus, and somatosensory cortex, while AI-induced analgesia was associated with reversal of these expression changes. Further, upregulation of TRPV1-associated signaling factors was not observed in Trpv1 knockout mice following CFA injection. In addition, the activated microglial markers Iba-1 and S100B demonstrated similar expression changes and were colocalized with TRPV1. Conclusions: These findings suggest that AI can mitigate chronic pain (and sequel such as comorbid depression) by suppressing TRPV1 overexpression in neuronal or microglial cells.


Introduction
Acupoint injection (AI or prolotherapy) is increasing in popularity as a complementary therapy for chronic pain. The most common form involves the injection of a glucose solution to induce local in ammation and initiate a wound-healing cascade. Prolotherapy is often utilized for relief pain of ligaments, tendons, and articular cartilage. After injection of concentrated glucose, the damaged tissue is repaired to form a denser, stronger, and tighter ligament or tendon [1,2]. In the clinic, prolotherapy is usually performed in 4-6 courses to ensure tissue repair and pain relief [3,4]. Several other solutions have also been used for prolotherapy, including dextrose and phenol. Although prolotherapy has demonstrated e cacy for healing local tissue damage, the underlying mechanisms of action are still unclear. One possibility is that dextrose induces osmotic damage to local neurons, which in turn attracts brain-resident in ammatory mediators to initiate the wound-healing cascade [5]. Local osmotic damage may also induce the in ltration of granulocytes followed by monocytes and macrophages [6]. However, elucidation of the actual mechanisms requires the use of appropriate animal models.
Another potential mechanism involves the transient receptor potential vanilloid 1 (TRPV1) channel responsive to thermal, mechanical, and capsaicin stimulation and expressed by peripheral sensory neurons. Activation of TRPV1 channels triggers cation in ux, including Ca 2+ , which activates downstream signaling pathways causing peripheral and central pain sensitization [7,8]. Peripheral in ammation induced by painful (potentially tissue-damaging) stimulation and local injury may also enhance signal propagation from the periphery to somatosensory cortex via the dorsal root ganglion, spinal cord, and thalamus, termed central sensitization, leading to allodynia, a condition in which even light touch and other normally non-painful stimuli are perceived as painful [7,8]. The TRPV1 channel and related  downstream signaling molecules are involved in both acute pain and central sensitization, and TRPV1 was recently shown to detect neuroin ammation in mice [9,10].
Acupuncture has been used for over 3,000 years to treat pain and associated disorders, and modern acupuncture techniques have proven to be convenient, safe, and inexpensive approaches for treating diseases [11][12][13][14][15]. These therapies are particularly bene cial for lower back pain, in ammatory pain, and muscle pain due to exercise, bromyalgia (FM), and myofascial pain [16][17][18][19][20]. Recent studies have implicated release of adenosine [21], dopamine [22], and endogenous opiates [23] in the analgesic effects of acupuncture as evidenced by elevated levels in serum or cerebrospinal uid.
In the current study, we examined the e cacy of AI against chronic in ammatory pain (CIP) and the underlying mechanisms. We hypothesized that AI of high concentration glucose would be effective for treating chronic mechanical and thermal hyperalgesia in mice by regulating TRPV1 signaling from periphery to somatosensory cortex.

Animals
We used 8-12 week old C57BL/6 mice for all experiments, which were purchased from BioLASCO Co. Ltd. (Taipei, Taiwan). The mice were randomly assigned to four groups (n = 8 per group): (1) Normal, (2) CIP, (3) CIP + AI, and (4) CIP + Trpv1 −/− . The sample size required for an alpha of 0.05 and a power of 80% was eight animals per group. After arrival, the mice were housed in a room under a 12/12 h light/dark cycle with water and food available ad libitum. All procedures were approved by the Institutional Animal

CIP induction
Mice were randomly divided into four groups and then anesthetized with 1% iso urane for CFA injections and AI treatment. Mice were next injected with 20 µl saline (pH 7.4, buffered with 20mM HEPES) or CFA 20 µl (complete Freund's adjuvant; 0.5 mg/ml heat-killed M. tuberculosis (Sigma, St. Louis, MO) suspended in oil:saline 1:1 emulsion) in the plantar surface of the hind paw. CFA was used to induce intraplantar chronic in ammation. CFA were administered twice: at baseline and day 7. Four groups and their treatments were as follows: (1) normal group: anesthesia with normal saline injection; (2) CIP group: anesthesia with CFA injections to induce CIP; (3) AI group: CIP mice treated with AI; and (4) Trpv1 −/− group: CFA injections in Trpv1 −/− mice.

Nociceptive behavioural examination
The mechanical and thermal pain behaviors were determined 3 times at day 0, day 7 and day 14 throughout the experiment after the induction of the CIP model. All mice were moved to the behavior analysis room, and were adapted to the environment for at least 30 min before behavior tests. All experiments were performed at room temperature and the stimuli were applied only when the animals were calm but not sleeping or grooming. First, the von Frey lament test was conducted. Mechanical sensitivity was measured by testing the force of responses to stimulation with 3 applications of the electronic, calibrated von Frey lament (IITC Life Science Inc., USA). Mice were placed onto a metal mesh (75x25x45 cm) and covered with a plexiglass cage (10x6x11 cm). Subjects were then mechanically stimulated by the tip of the lament at the plantar region of the right hind paw. The lament gram counts were recorded when the stimulation caused the subject to withdraw its hind paw. Second, the Hargreaves' assessment was used to measure thermal pain behavior by testing the time of response to thermal stimulation with 3 applications using Hargreaves' test IITC analgesiometer (IITC Life Sciences, SERIES8, Model 390G). The mice were placed in a plexiglass cage on top of a glass sheet. The thermal stimulator was positioned under the glass sheet and the focus of the projection bulb was aimed exactly at the middle of the plantar surface of the right hind paw. A cut-off time of 20s was set to prevent tissue damage. In the thermal paw withdrawal test, the nociception threshold was assessed using the latency of paw withdrawal upon stimulus, and was recorded when the constant applied heat stimulation caused the subject to withdraw its hindpaw.

Acupoint injection
AI was conducted in the morning (9:00-10:00 am) immediately after the induction of CIP. High concentration of glucose (20%, 20µL) was injected at the bilateral side of ST36 at baseline and day 7.
Similar to humans, the ST36 point is located longitudinally at 3 cun below the knee joint and intersects with the middle of the tibialis anterior muscle. Mice were placed into a xation machine under anesthesia with 5% iso urane for induction, which was then decreased to 1% for maintenance. Bilateral ST36 acupoint were selected, sterilized with 70% alcohol and iodine solution and the glucose solution was injected at the 3-5 mm depth.

Tissue sampling and western blot analysis
The lumbar dorsal root ganglion (DRG), spinal cord (SC), full thalamus and somatosensory cortex (SSC) neurons were excised immediately to extract proteins. Total proteins were prepared by homogenizing the tissues in lysis buffer containing 50 mM Tris-HCl (pH 7.4), 250 mM NaCl, 1% NP-40, 5 mM EDTA, 50 mM NaF, 1 mM Na 3 VO 4 , 0.02% NaNO 3 , and 1× protease inhibitor cocktail (Amresco, Solon, OH, USA). The extracted proteins (30 µg per sample according to the BCA protein assay) were subjected to 8% sodium dodecyl sulfate-Tris glycine gel electrophoresis and transferred to a PVDF membrane. The membrane was blocked with 5% non-fat milk in TBS-T buffer (10 mM Tris pH 7.5, 100 mM NaCl, 0.1% Tween 20), incubated with a primary antibodies (1:1000, Alomone, Jerusalem, Israel): anti-tubulin, anti-TRPV1, anti-Nav1.7, anti-Nav1.8, anti-RAGE, anti-pmTOR, anti-pPI3K, anti-pAkt, anti-pERK, anti-pS100B, anti-pNFκB, anti-pCREB, anti-pPKC or anti-pPKA in TBS-T and 1% bovine serum albumin, and incubated for 1 h at room temperature. A peroxidase-conjugated anti-rabbit antibody (1:5,000) was used as the secondary antibody. The bands were visualized using an enhanced chemiluminescent substrate kit (Pierce, Rockford, IL, USA) with LAS-3000 Fuji lm (Fuji Photo Film Co. Ltd., Tokyo, Japan). If appropriate, the image intensities of speci c bands were quanti ed with NIH ImageJ software (Bethesda, MD, USA). The protein ratios were determined by dividing the target protein intensities by the intensity of α-tubulin or βactin (internal control) in the same sample. The calculated ratios were adjusted by dividing the ratios from the same comparative group relative to the control.

Immuno uorescence
Mice were euthanized with a 5% iso urane via inhalation and intracardially perfused with normal saline followed by 4% paraformaldehyde. The brain was immediately dissected and post xed with 4% paraformaldehyde at 4 ºC for 3 days. The tissues were placed in 30% sucrose for cryoprotection overnight at 4 ºC. The brain was embedded in an Optimal cutting temperature (OCT) compound and rapidly frozen using liquid nitrogen before storing the tissues at -80 ºC. Frozen segments were cut at 20µm width on a cryostat then instantaneously placed on glass slides. The samples were xed with 4% paraformaldehyde, and then incubated with a blocking solution, consisting of 3% BSA, 0.1% Triton X-100, and 0.02% sodium azide, for 1 h at room temperature. After blocking, the samples were incubated with the primary antibody (1:200, Alomone), TRPV1 and Iba1, prepared in 1% bovine serum albumin solution at 4ºC overnight. The samples were then incubated with the secondary antibody (1:500), 488-conjugated A niPure donkey anti-rabbit IgG (H + L), 594-conjugated A niPure donkey anti-goat IgG (H + L) and Peroxidase-conjugated A niPure donkey anti-mouse IgG (H + L) for 2 h at room temperature before being xed with cover slips for immuno uorescence visualization. The samples were observed by an epiuorescent microscope (Olympus, BX-51, Japan) with 20 x numerical aperture (NA = 1.4) objective. The images were analyzed by NIH Image J software (Bethesda, MD, USA).

Statistical analysis
Statistical analysis was performed using the SPSS statistic program. All of the data were expressed as the mean ± standard error (SEM). Shapiro-Wilk test was performed to test the normality of data. Signi cant differences among all groups were tested using the repeated measure ANOVA, followed by a post hoc Tukey's test. p < 0.05 was considered signi cantly different.

AI treatment suppresses TRPV1-dependent mechanical and thermal hyperalgesia
We used von Frey laments to assess mechanical sensitivity and Hargreaves' test to assess thermal sensitivity at baseline, after injection of CFA to induce CIP, and following either control treatment or AI. At baseline (day 0), there were no signi cant differences in von Frey lament or Hargreaves' thermal test sensitivity among treatment groups. One week after CIP induction, however, mechanical sensitivity was signi cantly elevated in the CIP group compared to the sham treatment group, indicating hyperalgesia ( Fig. 1A, red column, n = 8). This mechanical hyperalgesia was signi cantly reversed by AI treatment (CIP + AI group) (Fig. 1A, blue column, n = 8 mice). Similar to CIP + AI group mice, Trpv1 −/− mice (Fig. 1A, green column, n = 8 mice) also exhibited little mechanical hyperalgesia following CIP. Thus, the hyperalgesic effects of CIP and the anti-hyperalgesic effects of AI appear to depend on modulation of TRPV1 channels.
3.2 Effects of CIP and AI on the expression levels of TRPV1 and associated signaling molecules in mouse dorsal root ganglia and spinal cord To examine the involvement of TRPV1 channels and related signaling factors in hyperalgesia and AImediated analgesia, changes in expression were examined by western blotting of DRG and SC tissue samples isolated after successful induction of hyperalgesia by CIP and reversal by AI (as con rmed by von Fey and Hargreaves' tests). In the DRG (Fig. 2), TRPV1 protein expression was signi cantly increased by CIP induction compared to sham-treated controls, and this augmentation was signi cantly reversed by AI. In addition, expression levels of phosphorylated (activated) PKA (pPKA) (Fig. 2B), pPI3K (Fig. 2C), and pPKC (Fig. 2D) were signi cantly elevated in the DRG after CIP, and these responses were attenuated by AI. Increased expression levels were also attenuated following CIP in Trpv1 −/− mice compared to wild type (WT) mice (CIP group). In addition, DRG expression levels of the phospho-activated mitogen activated protein kinases (MAPKs) pERK (Fig. 2E), pJNK (Fig. 2F), and pp38 (Fig. 2G) were elevated in CIP model mice, and these effects were reversed by AI and lower in Trpv1 −/− mice than WT mice following CIP induction. Expression levels of pAkt (Fig. 2H) and pmTOR Fig. 2I), two downstream effectors of pPI3K, and of the transcription factors pCREB (Fig. 2J) and NFκB (Fig. 2K) were upregulated during CIP, reversed by AI, and lower following CIP in Trpv1 KO mice than WT mice. The nociception-associated Na channel subunits Nav1.7 (Fig. 2L) and Nav1.8 (Fig. 2M) demonstrated similar patterns of expression change.
Finally, the activated microglial markers Iba1 (Fig. 2N), S100B (Fig. 2O), and RAGE (Fig. 2P) were also elevated by CIP, and these responses were reversed by AI and lower in CIP model Trpv1 −/− mice than CIP model WT mice. Collectively, these ndings suggest that CIP was induced by activated microglia. Similar changes in protein expression were also observed in tissue samples from the SC (Fig. 3).

Co-expression of TRPV1 and Iba1 in DRG and SC
To examine the cellular elements responsible for these expression changes, we conducted immuno uorescence staining of the DRG (Fig. 6) and somatosensory cortex (Fig. 7). Immuno uorescence images of the DRG revealed changes in TRPV1 immunoexpression that paralleled those measured by western blotting, with elevation following CIP induction compared to sham control mice and markedly lower expression levels in CIP + AI and CIP + KO groups (Fig. 6A). Immuno uorescence staining also con rmed elevated expression of the glial cell marker Iba1 in the DRG following CIP induction, which was attenuated by AI treatment and Trpv1 gene deletion (Fig. 6B). Further, TRPV1 and Iba1 were colocalized in CIP model mice, indicating that the elevation in tissue TRPV1 was mediated by increased expression in DRG glial cells. Consistent with western blotting results, this increase was reversed in CIP + AI and Trpv1 KO mice (Fig. 6C). Similar changes in TRPV1 and Iba1 expression were also observed in the somatosensory cortex (Fig. 7).

Co-expression of TRPV1 and Iba1 in cells of thalamus and somatosensory cortex
Finally, we examined these changes in cellular TRPV1 and Iba1 in thalamus (Fig. 8) and somatosensory cortex (Fig. 9). In accord with western blotting and other immuno uorescence results, TRPV1 immunoreactivity (Fig. 8A) and Iba1 immunoreactivity (Fig. 8B) were elevated concomitantly by CIP induction. Further, these elevated expression levels were reversed by AI ( Fig. 8A and 8B, respectively), while Iba1 expression was lower in CIP model Trpv1 KO mice than WT mice (Fig. 8C). In the somatosensory cortex as well, TRPV1, Iba1, and TRPV1/Iba1 co-staining signals were all increased in CIP model mice and reversed by AI (Fig. 9). Further, the changes in Iba1 were suppressed in Trpv1 −/− mice.

Discussion
In the present study, we present evidence that the hyperalgesia induced by CFA and the analgesic e cacy of AI are dependent on reciprocal modulation of TRPV1 channels expressed by glial cells in the DRG, SC, thalamus, and somatosensory cortex. Two weeks after CFA injection, mice demonstrated both mechanical hyperalgesia as evidence by increased von Frey lament sensitivity and thermal hyperalgesia as evidenced by Hargreaves' thermal response test, and these changes were associated with upregulation of TRPV1 channels, a myriad of downstream kinases and transcription factors, and voltage-gated sodium channels (Na v s). Conversely, both forms of hyperalgesia were suppressed by AI treatment concomitant with reduced expression of TRPV1, associated signaling molecules, and Na v s. Further, these hyperalgesia-associated increases in downstream signaling molecules and Na v s were markedly suppressed by Trpv1 KO. Expression levels of the activated microglial markers Iba1 and S100B were also upregulated by CIP induction, and these responses were reversed by AI. Collectively, these ndings indicate that TRPV1 signaling in microglia is a major driver of CFA-induced hyperalgesia and a primary therapeutic target of AI.
The TRPV1 is also implicated in ST36 acupoint treatment, suggesting crucial involvement in the neurobiology of chronic pain and clinical potential as a treatment target [16]. In ux of cations, particularly Ca 2+ , through TRPV1 channels activates the PI3K/Akt/mTOR and MAPK signaling pathways [24][25][26][27], and simultaneous overexpression of PKA, PI3K, PKC, ERK, Akt, and mTOR has been observed in various chronic pain states, while reversal of overexpression has been observed after successful acupuncture [28]. Among transcription factors implicated in clinical pain conditions, CREB and pNFκB appear to mediate the gene expression changes observed in CIP model mice. Overexpression of the voltage-gated sodium channels Nav1.7 and Nav1.8 have also been observed in the DRG and SC of acute pain model animals [29]. Conversely, suppression of Nav1.7 was associated with long-lasting analgesia [30]. Furthermore, Nav1.8 is highly expressed by nociceptive sensory afferents, and expression increases during chronic pain [31]. The present study provides evidence that AI of glucose alleviates mechanical and thermal hyperalgesia by suppressing this increase in Nav1.8 through downregulation of TRPV1 signaling in both PNS and CNS glial cells.
A recent study by He et al. (2017) provided evidence that nociceptive neuropeptides such as SP, CGRP and 5-HT are highly expressed at acupoints [32], implicating modulation of these signaling pathways in AI-induced analgesia. Increased excitability of DRG neurons has been linked to overexpression of a hyperpolarization-activated cyclic nucleotide-gated channel [33], and Zhang et al. al. also reported that AI of vitamin K at SP6 increased plasma vitamin K in dysmenorrhea patients, and that a higher vitamin K concentration was correlated with greater pain suppression [37]. A randomized, double-blinded, placebo-controlled clinical trial found that AI of onabotulinum toxin A (BoNTA) was more effective at suppressing pain than muscle injection in rats. They further suggested that BoNTA may suppress migraine through changes in CGRP and SP release within the medulla oblongata, possibly via a Our results extend the potential mechanisms underlying the bene ts of AI to include analgesia via modulation of glial TRPV1 expression and signaling.

Conclusion
In summary, the present study indicates that CFA can reliably induced CIP via upregulation of TRPV1 signaling and that AI can suppress CIP by reversing this upregulation. In addition, the activated microglial markers Iba-1 and S100B demonstrated similar expression changes and were colocalized with TRPV1. These ndings suggest that AI can mitigate chronic pain by suppressing TRPV1 overexpression in both neuronal and microglial cells.

Consent for publication
All authors have read the manuscript and consent for publication.

Availability of data and materials
The datasets supporting the conclusions of this article are included within the article.

Con ict of Interest
There are no nancial or other relationships that might lead to a con ict of interest for all authors.        (C) Overexpression of co-localization of TRPV1 and Iba1 (yellow) in the mice DRG. Scale bar is 50 μm.