The peripheral and central expression of CGRP and IB4 in chronic pain from MIA-induced TMJOA rats

Chronic pain is the prevalent symptom that drives temporomandibular joint osteoarthritis (TMJOA) patients to ask for medical care, yet present alleviator remain less effective. This study aimed to investigate the actual TMJOA related chronic pain and the peripheral and central response in a TMJOA animal model. This study rstly determined appropriate MIA dose based on pain behavior assessment with automated electronic von frey in rats. TMJOA pain correlated condylar structure alteration was evaluated by histological staining and Micro-CT. Then, the period of TMJOA chronic pain was further explored by assessing the alteration of glial brillary acidic protein (GFAP) positive astrocytes and ionized calcium binding adaptor molecule 1 (Iba1) positive microglia numbers in trigeminal spinal nucleus (TSN) and carrying out non-steroidal anti-inammatory drugs (NSAIDS) pharmacological ecacy experiment. Finally,


Background
Chronic pain is the prevalent symptom that drives temporomandibular joint osteoarthritis (TMJOA) patients to ask for medical care, yet present alleviator remain less effective. This study aimed to investigate the actual TMJOA related chronic pain and the peripheral and central response in a TMJOA animal model.

Methods
This study rstly determined appropriate MIA dose based on pain behavior assessment with automated electronic von frey in rats. TMJOA pain correlated condylar structure alteration was evaluated by histological staining and Micro-CT. Then, the period of TMJOA chronic pain was further explored by assessing the alteration of glial brillary acidic protein (GFAP) positive astrocytes and ionized calcium binding adaptor molecule 1 (Iba1) positive microglia numbers in trigeminal spinal nucleus (TSN) and carrying out non-steroidal anti-in ammatory drugs (NSAIDS) pharmacological e cacy experiment.

Background
Chronic pain is one of the common symptoms drives temporomandibular joint osteoarthritis (TMJOA) patients to ask for medical care from oral-facial pain clinics (1). Treatment of TMJOA is directed at relieving pain, decelerating the progress of the disease, and restoring TMJ function (2). Nevertheless, current analgesics remain less effective. Neither pharmacological therapies like non-steroidal anti-in ammatory drugs (NSAIDs) nor non-pharmacological therapies like occlusal splinting have su cient evidence to support the long-term pain treatment (3)(4)(5). Therefore, it is urgent to nd out the mechanism of TMJOA pain.
In order to investigate the mechanism in TMJOA progression, pathology, and pain symptoms, animal models have been established by means of chemical agents' injection, surgical manipulation, or customized functional appliance (4, 6, and 7). Mono-iodoacetate (MIA) injection has been regarded as an appropriate strategy to explore osteoarthritis (OA) pain recently, as it is easy to handle and can keep the joint cavity relatively intact. In a MIA-induced knee OA model, for example, nerve growth factor antibody have been thought to suppressed the pain behavior of knee OA model and provide new strategy for OA pain treatment (4,6,7). In TMJOA patients, chronic pain is the major reason for medical care (9).
However, there is no literature in the past to distinguish whether the pain in TMJOA animal model is acute or chronic. Previous researches related to OA pain investigated the pain mechanism and assessed the analgesic effect immediately after the accomplishment of OA animal models, without considering whether the assessment was carried out over the chronic pain period (10,11). Notably, animals frequently presented severe pain behavior without or with only slight joint pathological changes in the early period after TMJOA induction (12,13). Therefore, it may be not appropriate to investigate mechanism or treatment of OA related chronic pain in this early period, as typical OA structural alteration have not developed and pain behavior may be partly due to the local swelling or trauma.
In a preclinical knee OA model, NSAIDs failed to block the persistent pain (14), which consistent with ineffectiveness of NSAIDs on long-term pain in advanced knee OA patients (15). As NSAIDs also have insu cient evidence to support the long-term pain treatment on TMJOA patients, it is reasonable to hypothesis that NSAIDs may be ineffective in TMJOA model when the chronic pain emerges. Besides, gliosis plays a vital role in the genesis and maintenance of chronic pain (16), However, it has never been investigated whether the gliosis occurs in MIA induced TMJOA model. Peripheral sensory neurons in small diameter, with un-myelinated axons(C ber) are classi ed into peptidergic neuron, which is calcitonin gene-related peptide (CGRP) positive, and non-peptidergic neuron, which is isolectin B4 (IB4) positive. In contrast, sensory neurons in large or medium diameter, with myelinated or thinly myelinated axons (A ber), are marked with neuro lament 200 (NF200). Repetitive activation of primary neurons leads to a synaptic strengthening of nociceptive transmission and alters central projection in spinal cord, thereby increasing excitability of central nervous system, that is, central sensitization (16). Both CGRP and IB4 positive afferents have been proposed to be involved in peripheral and central sensitization in migraine (18,19). In neuropathic pain model, mechanical allodynia can be inhibited by TLR5-mediated NF200 positive A-ber blockade both peripherally and centrally (20).However, the distribution and alteration of CGRP, IB4 or NF200 positive neurons (afferents) in peripheral and central system in the chronic pain from TMJOA are still unclear.
Therefore, this study was to investigate the emerging period of chronic pain from TMJOA induced by MIA injection, as well as the distribution and alteration of CGRP, IB4 or NF200 positive neurons (afferents) in peripheral and central system.

Animal model
In this study, a total 72 male Sprague Dawley (SD) rats (seven-week-old) were purchased from the Experimental Animal Centre of Hubei Province and housed in a SPF laboratory condition. All experiments in this study were conducted in accordance with the Ethics Committee for Animal Research, School and Hospital of Stomatology, Wuhan University, China.
After a week of quarantine, rats were randomly divided into saline-injected group (control) and MIAinjected group, and anesthetized with iso urane (Abbott Laboratories, North Chicago, IL, USA) in oxygen. MIA dose course and pain behavior assessment To con rm the appropriate concentration of MIA for TMJOA related chronic pain research, different doses of MIA (0.3, 0.5, 1.0, and 2.0 mg/50 µl; n = 4 rats) dissolved in the saline were bilaterally injected into the TMJ cavity, while the identical 50 µl of saline in the control group (n = 4 rats). Pain behavior was assessed by measuring head withdrawal threshold (HWT) with an automated electronic von frey (IITC, CA, U.S.A.) as previously described (22). Brie y, rats were acclimatized to the test environment a week before the measurement. Over the HWT measurement, rats' heads were not restrained, but their movements were still restricted. Von frey laments were used to assess the TMJ mechanical sensitivity.
The plastic probe was applied to the TMJ region, with the midpoint between the tragus and the outer rim. Head inching, characterized by sudden quick head withdrawal, or shaking were considered as positive response of pain. The minimum applied force causing head withdraw was recorded as HWT. Each TMJ was measured ve times with a few seconds interval, and the mean values of ten recordings were the nal data of HWT. The measurement was performed on 0, 1, 3, 5, 7, 10, 14, 17, 21, 24, 28, 31, 35 days post-injection.

Tissue Preparation
Based on pain behavior assessment, the appropriate dose of MIA was con rmed and further used in the subsequent experiments. The con rmed MIA dose was injected (7, 14, 21, 28 days, n = 5 rats per group), while 50 µl of saline into control groups (7, 14, 21, 28 days, n = 5 rats per group). Rats were sacri ced by overdose of iso urane. The TMJ tissues with the mandibular condyle disc, retrodiscal area, fossa, trigeminal ganglion (TG) and trigeminal spinal cord in each group were carefully dissected from rats and xed in 4% paraformaldehyde for 24 hours. One TMJ of each rat was demineralized in 10% EDTA, which was changed every 3 days for histopathology, while the other non-demineralized TMJ for a micro-CT examination. TG and TSN tissues were dehydrated in 30% sucrose overnight.

Histology
To assess the pain behavior related joint structure degeneration of MIA-induced rat TMJOA, the demineralized TMJs (MIA, n = 5 joints; saline, n = 5 joints) were embedded with para n and sagittally cut at a 5 µm thickness per slide. Hematoxylin and eosin (H&E) stain was conducted to evaluate TMJ morphology, cartilage thickness, chondrocyte number. Safranin-O (S.O) stain was performed to assess the proteoglycan in cartilage matrix as previous (23). Brie y, condyle was divided into anterior, middle and posterior part by three dotted squares based on the average of fan center angle ( Fig. 2A). The three parts of proteoglycan area proportion in S.O staining condyle and thickness of cartilage, numbers of chondrocytes in HE staining condyle, were averaged and analyzed.

Micro-CT examination and 3D reconstruction
For further assessing the pain behavior related joint structure degeneration, the condylar morphological characteristics were examined by Micro-CT (Bruker SkyScan1176). The specimens were scanned at 58 k V and 431 µA with a thickness of 9 µm per slice. Then 3D reconstruction of condyle was carried out and represented by a superior and lateral view.
Immuno uorescence TG and TSN tissues were embedded with optimal cutting temperature compound (OCT) (Sakura, USA) and cut with serial section at a 10 µm thickness. Immuno uorescence staining of TG for CGRP, NF200, IB4, and TSN for CGRP, NF200, IB4, GFAP, and IBA-1 was performed.

Statistical analysis
Statistical analysis was performed using Graph Pad Prism 6.0. All data were presented as mean ± SD and were analyzed using two-way analysis of variance (ANOVA) Tukey's test. Statistical signi cance was considered at p < 0.05.

MIA induced dose-and time-dependent pain behavior
A rat model for TMJOA pain was induced by injection of MIA followed by behavioral tests. Consecutive HWT measurements reveal that MIA induced-pain behavior occurs in a dose-and time-dependent manner (Fig. 1B).  Fig. 2A-D). Speci cally, cartilage thickness decreased from 1 to 4 weeks ,while chondrocyte numbers and percentage of proteoglycan area reduced from 2 to 4 weeks, which indicates the histopathologic change mainly began in 2 weeks post MIA injection (*p < 0.05, **p < 0.01, *** p < 0.001, ****p < 0.0001). Consistently, Micro-CT results exhibited that condylar structure gradually degenerated overtime (Fig. 3A). Compared with control group, slight bone lesion occurred at 1 week, then exacerbated at 2 and 3 weeks in MIA -injected rats. Subchondral bone remodeling, larger bone marrow cavity and even osteophyte could be observed in MIA-injected rats on 4 weeks. These data suggested that pain related condylar degeneration was induced by 1 mg/50 µl MIA and started in 2 weeks post injection.

Establishment of chronic pain from MIA-induced TMJOA
Gliosis plays a vital role in the genesis and maintenance of chronic pain (16). Hence, gliosis has been used to be a sign of chronic pain. Gliosis were observed in the MIA induced TMJOA on 2, 3, 4 weeks post injection (Fig. 4. A-C). In MIA injection group, the quantity of GFAP positive astrocytes increased 98%, 133%, 112% respectively on 2,3 and 4 weeks compared with control group. Meanwhile, the IBA-1 positive microglia numbers also increased in 2, 3 and 4 weeks and with 57%, 67%, and 46% respectively.
Interestingly, at 1 week, the quantity of these two cells had no statistic difference between the MIA injection group and control group (*p < 0.05, **p < 0.01, *** p < 0.001, NS: no signi cance).
To further con rm the chronic pain period of MIA-induced TMJOA, we utilized one of NSAIDs, celecoxib, as former reports demonstrated NSAIDs exert an analgesic effect in early period of OA (24). Indeed, in the early period (1, 3, 5, 7 days), celecoxib effectively improved the percentage of HWT baseline and achieved analgesia effect when compared with saline injection or vehicle group (Fig. 4D). From 10 to 35 days, celecoxib lost its analgesic e cacy (Fig. 4D) (*p < 0.05, ****p < 0.0001). These ndings indicate that the chronic pain from MIA-induced TMJOA started in 2 weeks after MIA injection.
Sensitization of CGRP and IB4 positive neurons/projection in peripheral and central neural system The alteration of neurochemical markers of NF200, CGRP and IB4 was analyzed by immuno uorescence to assess the peripheral and central nociceptive sensitization in MIA induced TMJOA rats. Typical uorescence image of NF200 positive (green) (Fig. 5A a-c), CGRP positive (red) (Fig. 5A d-f), IB4 positive (chartreuse) (Fig. 5A g-i) neurons showed the relative expression outline. Both percentages of CGRP positive and IB4 positive TG neurons in MIA injection group were increased in 2 and 4 weeks in contrast to those of control group (Fig. 5C, D), while the ratio of NF200 positive neurons had no statistical signi cance between these two groups ( Fig. 5B) (**p < 0.01, *** p < 0.001, ****p < 0.0001 VS control, NS: no signi cance). Similarly, typical uorescence image of NF200 positive (green) (Fig. 6A a-c), CGRP positive (red) (Fig. 6A d-f), IB4 positive (chartreuse) (Fig. 6A g-i) central projection showed the relative expression outline. The mean uorescence intensity of CGRP and IB4 signi cantly increased in both 2 and 4 weeks post MIA injection compared with control group (Fig. 6C, D). However, mean intensity of NF200 (Fig. 6B) had no difference between them (**p < 0.01, ****p < 0.0001 VS control, NS: no signi cance). These ndings suggest that CGRP, IB4 positive afferents in peripheral and central nervous system is involved in chronic pain from TMJOA.

Discussion
This study showed that intra-articular injection of MIA induced a dose-and time-dependent mechanical allodynia and OA-like pathological alteration in TMJOA rats. Even though previous research reports that 0.5 mg/50 µL also induced both TMJ degeneration and pain behavior (24), we de ned 1 mg/50 µL as the appropriate dose of MIA. Because this dose induced mechanical allodynia in a longer period (over 5 weeks), which is consistent to the typical OA pain model validation researches that OA pain behavior lasted more than 4 weeks. Besides, preclinical research of analgesic e cacy and safety ask for long period follow-up, which supported by the preclinical research of OA treatment with bone marrow-derived mesenchymal stromal cells that analgesic e cacy and safety was assessed beyond 4 weeks (27). In addition, pathological and Micro-CT examination further demonstrates that time-dependent condylar structural alteration is closely related to TMJOA pain behavior in relatively late period(2,3 4 weeks).This agrees with previous MIA induced OA models of TMJ or knee joint (27). However, within one week of MIA injection, there was obvious discordance between the severity of pain behavior and TMJ structural alteration, as HWT went down to the lowest while only slight condylar bone destruction occurred within one week. This discordance suggests MIA induced pain behavior within one week may result from damage of other TMJ structures like synovium, as reported that MIA causes an initial in ammatory in ltrate of monocytes, neutrophils and basophils in joint, peaking around day 3 (29,30). Therefore, this pain behavior occured in relatively early pain period may not mimic the OA chronic pain symptom in human. So, it is necessary to identify the exact chronic pain period in animal TMJOA model. Gliosis is a key mechanism underlying the genesis and maintenance of chronic pain (16). Mounting evidence shows that activation of glia contributes to the plasticity of synaptic transmission that is associated with chronic pain states (31). In this study, gliosis results demonstrate that chronic pain emerged in MIA induced TMJOA rats in a relatively late period (from 14 to 28 days) .This result is similar with MIA induced knee OA in rats that the proliferation of IBA1 + microglia occurred in 14, 21, 28 days post MIA injection and GFAP + astrocyte in 28 days (32), while in this study, GFAP + astrocyte proliferated earlier in 14, 21, 28 days. This difference may be due to the smaller joint cavity that leads to anabatic damage of TMJ condyle and sooner activation of astrocytes. The lost analgesic effect of celecoxib and gliosis suggest that MIA induced TMJOA chronic pain emerged in relatively late period post injection (10-14 days). Besides, the lost analgesic effect of celecoxib is in an agreement with fact that NSAIDS are usually effective only for TMJ in ammation or muscle pain in patients (33).This study is consistent with previous study that NSAIDs are effective at alleviating allodynia in early period MIA animals, they are ineffective in late period MIA animals (4,34). This phenomenon can be explained by the initial in ammatory in ltration caused by MIA neurotoxicity in the early period mentioned above (29,30). Based on our knowledge, this is the rst time to verify the exact chronic pain period of TMJOA in animal model.
It is well accepted that chronic pain results from peripheral sensitization and central sensitization, which means the altered neuronal activity of peripheral and central nervous system (35). Peripheral sensitization is the result of peripheral primary neurons activation (36). By means of immuno uorescence, we found up-regulation of CGRP, IB4 TG and TSN over OA chronic pain period (2 and 4 weeks).This can be explained by that CGRP and IB4 positive nociceptors are small diameter unmyelinated C bers, which usually involved in pain perception(36). CGRP was reported to be expressed in sensory neurons innervating osteoarthritic joints in rats, and involved in peripheral and central sensitization of primary neurons in in ammatory arthritis (39,40). An antagonist of CGRP reduced the sensitization of spinal neurons to both innocuous and noxious mechanical stimulation applied to the in amed joint (41). Similarly, IB4 positive nociceptors mediate persistent muscle pain induced by glial cell line-derived neurotrophic factor (GDNF) (42). Not surprisingly, the expression of NF200 in both TG and TSN remains unaltered, as NF200 positive neurons are involved in proprioception such as position sense (43,44), but not involved in the typical spared nerve injury induced chronic pain model (43,44). Thus, the immuno uorescence results revealed that CGRP, IB4 positive afferents in peripheral and central nervous system are involved in TMJOA related chronic pain in rats.
There are some limitations in this study. Firstly, unlike knee OA pain that have other pain assessment like weight distribution of hind paw(46), we determined the ultimate does of MIA by assessing pain behavior with electronic von frey alone, which may be a little bit monotonous. However, it is acceptable that von frey is most the typical device for mechanical sensitivity assessment. Meanwhile, our pathological assessment demonstrated that TMJOA induced by 1 mg/50 µl of MIA mimics the development of human TMJOA. Secondly, we could not excluded the possibility that the expression of NF200 may increase after 4 weeks, as we had no examationation after 4 weeks.
In conclusion, chronic pain from MIA induced -TMJOA starts at 2 weeks in rats. CGRP, IB4 positive afferents in peripheral and central nervous system are involved in TMJOA related chronic pain in rats.

Availability of data and materials
The data used and analyzed in this article are available from on reasonable request.

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

Funding
This study was supported by the National Science Foundation of China (NO. 81771100 and 81870789).