Calcitonin gene‐related peptide: An intra‐articular therapeutic target for TMJ disorders

Abstract Objectives The goal of this project was to evaluate the role of calcitonin gene‐related peptide (CGRP) in the development of arthritis. Methods Herein, we employed somatic mosaic analysis in two different joints by FIV(CGRP) intra‐articular inoculation in the knees or temporomandibular joints (TMJ) of young adult male C57/BL6 mice. FIV(CGRP) is a feline immunodeficiency virus over‐expressing full‐length CGRP. Joint pathology and function were evaluated at the histopathological and behavioral levels. In addition, CGRP signaling was inhibited by intra‐articular inoculation using FIV(CGRP8‐37), such that the inhibitory peptide CGRP(8‐37) was overexpressed 4 weeks after induction of joint inflammation in the TMJ of IL‐1βXAT transgenic mouse model. The mice were evaluated for behavior and killed for evaluation of knee and TMJ pathology. Results Overexpression of CGRP in the joints of wild‐type mice induced the development of joint anomalies, including meniscal hypertrophy and articular pathology, associated with nocifensive behavior. Intriguingly, overexpression of the CGRP(8‐37) inhibitory peptide in the knee and TMJ of IL‐1βXAT transgenic mice with joint inflammation resulted in partial amelioration of the attendant joint pathology. Conclusions The results of this study suggest that CGRP is sufficient and necessary for the development of joint pathology and may serve as an intra‐articular therapeutic target using gene therapy or monoclonal antibody‐based therapies.

of migraines (Russell et al., 2014). CGRP can further activate and sensitize trigeminal primary afferent neurons (Romero-Reyes et al., 2015). To this end, work by our group and others has shown the involvement of increased CGRP levels in animal models of joint disorders (Kido et al., 1993;Lai et al., 2006). In addition, Romero-Reyes and colleagues (2015) administered the small molecule MK8825, a selective CGRP receptor antagonist, in a mouse model of acute orofacial masseteric muscle pain that was employed as a surrogate of acute temporomandibular joint disorders. Here, mice pretreated with MK8825 showed alleviated orofacial pain behaviors and reduced neuronal activation in the trigeminal nucleus in response to complete Freund's Adjuvant into the masseter muscle.
Furthermore, CGRP is elevated in human TMJ specimens harvested from patients with TMJ disorders and increased in the TMJ of individuals with an open bite (Alstergren et al., 1995;Cady et al., 2011;Haeuchi et al., 1999;Kopp, 2001;Romero-Reyes et al., 2015;Sato et al., 2004). Taken together, the aforementioned studies demonstrate that CGRP is involved in pain physiology, suggesting that therapeutic modalities targeting CGRP may translate into effective therapeutic strategies for painful joint-related disorders.
We previously demonstrated the role of neuroinflammation in TMJ disorders. Specifically, Fiorentino and colleagues (Fiorentino et al., 2008) demonstrated that centrally induced neuroinflammation involving CGRP in the TMJ contributed to marked histopathological changes of the articular cartilage consistent with initial stages of osteoarthritis. In these studies, inhibition of central neuroinflammation restrained the development of articular pathology in a mouse model of TMJ osteoarthritis (Fiorentino et al., 2008;Lai et al., 2006).
Lastly, Kyrkanides et al. (2007) demonstrated that inhibition of afferent sensory signals from the TMJ ameliorated the attendant articular pathology in the mouse model of TMJ inflammation (Lai et al., 2006). Taken together, these studies demonstrate that neuroinflammation in the TMJ is sufficient and necessary for the development of articular pathology in the mouse.
Although the literature suggests that CGRP may contribute to the development of joint pathology, there is a lack of direct proof for the role of CGRP in arthritis. Therefore, the goal of this preclinical study was to elucidate whether CGRP is necessary or sufficient for the development of joint pathology. This is significant, due to the recent FDA approval of monoclonal antibody-based therapies against CGRP for the management of migraines, which could seamlessly segue into a drug-repositioning agent for the management of pain-related osteoarthritic joint disorders.

| Vector construction and packaging
The FIV(CGRP) and FIV(CGRP8-37) transfer vectors expressing fulllength CGRP and the inhibitory peptide CGRP(8-37), respectively, were constructed as follows. Two plasmids were constructed that contained the rat proCGRP complementary DNA (cDNA), modified to express either wild-type CGRP or the CGRP(8-37) antagonist, which had the first 7 aa deleted by polymerase chain reaction (PCR) using the primer sets described in Table 1. CGRP is expressed as a propeptide that must undergo two proteolytic cleavages followed by carboxyl amidation to achieve bioactivity (Rosenblatt & Dickerson, 1997). To simplify expression, the prohormone convertase-1 (PC-1) site in both constructs was replaced at the NH2-terminus of CGRP with a furin cleavage site to facilitate correct processing in a wide range of tissues (Seidah & Prat, 2012). At the carboxyl end of CGRP, 8 aa are usually removed by a second cleavage event to expose a carboxyl glycine, which is a substrate for subsequent amidation. Amplification of both constructs T A B L E 1 The PCR primer sequences used to generate proCGRP8-37 and full-length CGRP DNA sequences that were cloned into the cloning site of the viral vector used to inoculate the animals annealed, and used as a substrate for a third PCR reaction using just the two outside primers, resulting in two amplimers encoding the cDNA for either full-length proCGRP or proCGRP(8-37), which will produce the antagonist CGRP(8-37) when introduced into cells ( Figure 1). Additionally, the outside primers encoded XbaI (upstream) and BamHI (downstream) restriction sites to facilitate cloning into the expression plasmid pBluescript ( Table 1). The two constructs were then subcloned into the routinely ranged between 10 7 and 10 8 infectious particles/ml.

| Animal studies
All animal procedures described were reviewed and approved by the University of Rochester Institutional Animal Care and Use Committee (University Committee on Animal Resources) for compliance with federal regulations before the initiation of the study (OLAW/PHS Assurance A3292-01). All mice were maintained in an AAALACaccredited specific pathogen-free barrier facility. All procedures followed the AVMA guide per institutional policy. All mice were males and they were housed with five animals per cage. The mice were routinely anesthetized via intraperitoneal injection of ketamine (40 mg/kg). The mice did not experience any adverse events.
The ARRIVE guidelines for animal research were followed.
Specifically, this was a comparative study between CGRP-treated, CGRP8-37-treated, and control (gfp-treated) mice. The sample size is mentioned below. All male mice in the litter were included and there were no exclusion criteria. The mice were randomized into various groups. The operators that handled the mice were not aware of the mice groupings. The outcome measures were again collected blinded to the mice groupings. The statistical method used is described in detail below. Details on the experimental animals and F I G U R E 1 cDNA constructs used for expression of full-length CGRP and CGRP(8-37) antagonist. Wild-type CGRP is expressed as a propeptide, which after removal of its signal peptide undergoes two posttranslational endo-proteolytic events cleaved by prohormone convertase 1 (PC1) at paired basic amino acids (KR), and furin at tetrabasic sites (RRRR or RKRR). The carboxyl amino acid is glycine, which is converted to a carboxyl amide. ProCGRP furin replaces the neuroendocrine-specific PC1 cleavage site with a more frequent furin cleavage site and eliminates the propeptide downstream of the carboxyl glycine. ProCGRP(8-37) furin is processed similarly and has had the first 7 aa of mature CGRP deleted. Numbering refers to amino acid positions in the mature CGRP molecule. cDNA, complementary DNA; CGRP, calcitonin gene-related peptide.
procedures are detailed below. The results are described in detail below.
Pertaining to CGRP overexpression in the knees, 8-week-old wild-type C57BL/6 mice (N = 10) were randomly assigned to receive either FIV(CGRP). An additional group (N = 6) was randomly assigned to revive FIV(gfp) and served as controls. All mice were anesthetized with ketamine (40 mg/kg), and a 10 µl aqueous solution containing a total of 10 5 FIV(CGRP) infectious particles was injected into the right and left knee joints of the hind limbs. The knee area was located by palpation. A 27.5G needle was inserted into the joint space from a lateral approach and viral vector solution was injected intraarticularly. After injection, the mice were returned to their cages and monitored for full recovery from anesthesia. No inclusion/ exclusion criteria were applied for these animals. The operator was only aware of the animals' group assignments.
For the experiments pertaining to CGRP overexpression in the TMJs, 3-month-old wild-type C57BL/6 mice (N = 5) were randomly assigned to receive 10 µl containing 10 5 FIV(CGRP) infectious particles in the right and left TMJ under the surgical plane of anesthesia and returned to their cages. Mouse behavior was subsequently evaluated when the mice were killed 8 weeks later, and their TMJ histology was compared to that of wild-type controls (N = 5). No inclusion/exclusion criteria were applied for these animals.
JHM was only aware of animals' group assignments.
For the experiments pertaining to CGRP(8-37) overexpression in the TMJs, 3-month-old Col1-IL-1β XAT mice (Lai et al., 2006)  TMJ function was evaluated by resistance to jaw opening based on the principles of the Pain Adaptation Model, which suggests that pain reduces muscle force (Kyrkanides et al., 2007). These data were compared to data collected during a previous experiment derived from 7 Col1-IL-1β XAT transgenic mice were injected with a control FIV vector and 8 Col1-IL-1β XAT transgenic mice that received FIV (Cre) injections to activate the joint inflammation process in the TMJ (Kyrkanides et al., 2007). No animals were excluded from this evaluation. We did not control for any cofounding conditions as there were none.
At the end of each experiment, the mice were killed via intraperitoneal injection of pentobarbital (100 mg/kg) followed by decapitation.
The animal data are available upon request.

| Histological-Immunohistochemical (IHC) studies
Following fixation in 10% formalin, the mouse heads were dissected, defleshed, and decalcified by immersion in an ethylenediaminetetraacetic acid solution for 7 days at 4°C under constant agitation.
The TMJs were then processed with an RHS-1 microwave tissue processor, after which the samples were embedded in paraffin, cut on a microtome as 3 µm thick sections, and collected on glass slides.
Overall TMJ histopathology was evaluated in sections stained by Alcian blue-orange G and Safarin-O histochemistry (Lai et al., 2006).
Microphotographs were captured using a Spot CCD digital camera (Diagnostic Imaging, Sterling Heights, MI) attached to the microscope.

| Statistical analysis
Analysis of rotarod data was performed with a repeated measure two-way analysis of variance with Tukey's Multiple Comparisons (GraphPad Prism, v. 9.1, San Diego, CA). An adjusted p < .05 was set as significant.

| CGRP overexpression in the TMJ results in nocifensive behavior in mice
Eight weeks following inoculation of the TMJ with FIV(CGRP), we observed the development of intra-articular overexpression of CGRP in the TMJ resulted in a reduction of resistance to mouth opening (Table 2) when compared to previously published wildtype mice values (Figure 2b in Kyrkanides et al., 2007). Specifically, the average value of resistance to mouth opening in the CGRP-overexpressing mice was 1.712 µNT (SD = 0.38), compared to 3.22 µNT (SD = 0.22) for wild-type mice. Interestingly, the CGRP-overexpressing mice's average value was similar to that of mice suffering from TMJ inflammation with 1.74 µNT (SD = 0.23) (Kyrkanides et al., 2007).
3.6 | Overexpression of CGRP in the mouse knee resulted in join pathology and nocifensive behavior in the knee joints also resulted in significant nocifensive behavior, as detected by rotarod analysis (Figure 4).

| DISCUSSION
The goal of this preclinical study was to elucidate whether CGRP is necessary or sufficient for the development of articular pathology in mice. To this end, we used viral delivery to acutely increase CGRP expression in the mouse knee joints or TMJ. Our results demonstrate that intra-articular CGRP overexpression is sufficient in the development of meniscal hypertrophy, with loss of proteoglycans in the articular cartilage and spurring in joints. In addition, intra-articular CGRP overexpression resulted in the development of nocifensive behavior, as detected by decreased time spent on a rotarod. These results are in agreement with previous studies where we observed decreased articular pathology following competitive inhibition of CGRP signaling in the Col1-IL1β XAT mouse model of TMJ inflammation (Lai et al., 2006).
The potential role of CGRP in joint disorders has been previously suggested by a number of descriptive studies analyzing specimens T A B L E 2 Resistance to mouth opening (Kyrkanides et al., 2007) raw data was recorded from mice that received intra-articular administration of FIV(CGRP) into the TMJ harvested from human patients, as well as small laboratory animal models (Alstergren et al., 1995;Bullock et al., 2014;Cady et al., 2011;Haeuchi et al., 1999;Sato et al., 2004). However, this is the first study to demonstrate that CGRP is sufficient to induce articular pathology in the mouse. Moreover, our results build on the evidence produced by Romero-Romero-Reyes et al. (2015) whereby the small molecule MK8825, a selective CGRP receptor antagonist, alleviated orofacial pain behaviors, and reduced neuronal activation in the trigeminal nucleus in response to complete Freund's Adjuvant into the masseter muscle. Benschop et al. (2014) generated a neutralizing antibody to CGRP, namely LY2951742, which was tested in preclinical in vivo models of osteoarthritis pain in the rat. Neutralization of CGRP significantly reduced pain behavior as measured by a weight-bearing differential in the rat mono-iodoacetate model in a dose-dependent manner. In addition, pain reduction after CGRP neutralization was independent of prostaglandins. Importantly, neutralization of CGRP also provided dose-dependent and prolonged (>60 days) pain reduction in the rat meniscal tear model of osteoarthritis after only a single injection of LY2951742. However, when this antibody was administered subcutaneously to patients suffering from moderate to the severe knee joint pain due to osteoarthritis, as part of a doubleblind, double-dummy, placebo, and active-controlled phase-II clinical trial, it failed to produce analgesia after 4 months of treatment (Jin et al., 2016). The trial was consequently discontinued due to inadequate efficacy. We believe that this is a result of the following limitation: The administration of LY2951742 to patients was subcutaneous, versus intra-articular, and therefore exposed the antibody to host defense mechanisms that likely neutralized its function. Moreover, as our results show, the effect of CGRP is intra-articular in nature; conversely, subcutaneously administered antibodies will have limited access to intra-articular tissues due to the presence of the joint capsule.
Taken together, our results described herein and together with the available literature, lend to the development of a model whereby continuous peripheral injury and/or inflammation in the TMJ will result, over time, in the antidromic stimulation and release of CGRP by small diameter C and Aδ fibers through a sensory dorsal root reflex. In turn, CGRP will have a direct effect on the articular chondrocytes and meniscus, thereby contributing to the development of knee joint or TMJ pathology. Based on our results, we conclude that inhibition of CGRP signaling in the TMJ via intraarticular administration of antibodies, or other small molecule inhibitors, has the potential to provide lasting alleviation against tissue pathology and attendant symptomatology, especially since inflamed joints are susceptible to joint loading and imbalanced (rightleft) function (Piancino et al., 2015). Kyrkanides designed and carried out the experiments, collected and analyzed the data, and composed the manuscript.

ACKNOWLEDGMENTS
We would like to thank Dr. Ross H. Tallents for performing the intraarticular injections and evaluation of TMJ behavior. We would also like to thank Ms. Jen-nie H. Miller for cloning and preparation of the FIV viral vectors described herein, and for the evaluation of knee joint behavior. The project was funded in part with NIH grant DE017765 and support from the University of Rochester.
F I G U R E 4 Intra-articular CGRP overexpression in the knee joints induces nocifensive behavior. Changes in locomotion, as a measure of nocifensive behavior, were evaluated by rotarod in mice inoculated with FIV(CGRP) versus FIV(gfp). Eight weeks following intra-articular inoculation, mice over-expressing CGRP in their knee joints showed a significant decline in locomotion, which persisted at 8 weeks in the CGRP group. ***p < .001. CGRP, calcitonin gene-related peptide.