|Home | About | Journals | Submit | Contact Us | Français|
Although skeletal pain can have a marked impact on a patient’s functional status and quality of life, relatively little is known about the specific populations of peripheral nerve fibers that drive non-malignant bone pain. In the present report, neonatal male Sprague Dawley rats were treated with capsaicin or vehicle and femoral fracture was produced when the animals were young adults (15–16 weeks old). Capsaicin treatment, but not vehicle, resulted in a significant (>70%) depletion in the density of calcitonin-gene related peptide positive (CGRP+) sensory nerve fibers, but not 200 kD neurofilament H positive (NF200+) sensory nerve fibers in the periosteum. The periosteum is a thin, cellular and fibrous tissue that tightly adheres to the outer surface of all but the articulated surface of bone and appears to play a pivotal role in driving fracture pain. In animals treated with capsaicin, but not vehicle, there was a 50% reduction in the severity, but no change in the time course, of fracture-induced skeletal pain related behaviors as measured by spontaneous flinching, guarding and weight bearing. These results suggest that both capsaicin-sensitive (primarily CGRP+ C-fibers) and capsaicin-insensitive (primarily NF200+ A-delta fibers) sensory nerve fibers participate in driving skeletal fracture pain. Skeletal pain can be a significant impediment to functional recovery following trauma-induced fracture, osteoporosis-induced fracture and orthopedic surgery procedures such as knee and hip replacement. Understanding the specific populations of sensory nerve fibers that need to be targeted to inhibit the generation and maintenance of skeletal pain may allow the development of more specific mechanism-based therapies that can effectively attenuate acute and chronic skeletal pain.
Painful musculoskeletal conditions are the most common cause of chronic pain and physical disability in both developing and developed countries (Lubeck, 2003, Woolf and Pfleger, 2003, Brooks, 2006, Kidd, 2006, Rosemann et al., 2007). Although non-malignant skeletal pain can be caused by a very diverse group of disorders and injuries including trauma-induced fracture, osteoarthritis, osteoporosis, low back pain and orthopedic procedures, what they share in common is frequent impairment of physical function and long-term disability (Morrison et al., 2003, Woolf and Pfleger, 2003, 2005, Brooks, 2006, Kidd, 2006).
In general, the prevalence of skeletal pain is higher in women than men and is increased following trauma and with aging (Rollman and Lautenbacher, 2001, Woolf and Pfleger, 2003, Leveille et al., 2005, Woolf and Pfleger, 2005). Recently it has been shown that lifestyle factors including obesity and lack of physical activity also impact the prevalence of skeletal disorders and chronic skeletal pain (Lamb et al., 2000, Woolf and Pfleger, 2003, Sawatzky et al., 2007, Tukker et al., 2008). Importantly, the pain and physical disability brought about by skeletal disorders frequently has significant secondary effects including decreased mobility, loss of bone and muscle mass and a reduction in cardiovascular and cognitive health, all of which can diminish the individual’s functional status and quality of life (Woolf and Pfleger, 2003, Dominick et al., 2004, Sawatzky et al., 2007). As the world’s population ages, becomes heavier and adopts a more sedentary lifestyle, skeletal pain is expected to exert an increasing economic and medical burden on individuals and society (Lubeck, 2003, Woolf and Pfleger, 2003, Stewart, 2005, Gruber et al., 2006, Reynolds and Himes, 2007, Lutz et al., 2008, McCarthy et al., 2008, Tukker et al., 2008).
Given the impact of skeletal pain on affected individuals and society, what is surprising is how little is known about the specific populations of sensory nerve fibers and mechanisms that drive non-malignant skeletal pain. In large part, this lack of knowledge is due to the lack of available animal models of non-malignant skeletal pain which closely mirror the human condition and the difficulty of working with nerves in calcified tissues (Mullink et al., 1985, Schwei et al., 1999, Honore et al., 2000, Sevcik et al., 2005, Jimenez-Andrade et al., 2007, Freeman et al., 2008). In the present report, we explore the specific populations of primary afferent nerve fibers that are involved in driving bone fracture pain. To accomplish this we use a recently developed model of orthopedic surgery induced-fracture pain (Freeman et al., 2008) in rats that received neonatal capsaicin treatment. Capsaicin is the active ingredient of the pungent capsicum pepper and previous results have shown that systemic administration of a single dose of capsaicin in neonatal rats causes permanent degeneration of a significant fraction of unmyelinated primary sensory neurons, with no significant change in myelinated afferent fibers (Nagy et al., 1981, Nagy et al., 1983).
All procedures were approved by the Institutional Animal Care and Use Committee at the University of Minnesota and the University of Arizona (Minneapolis, MN and Tucson, AZ) and were in accordance with the National Institutes of Health guidelines for care and use of laboratory animals. All efforts were made to minimize the suffering and number of animals used.
Male Sprague Dawley rats (Harlan, Indianapolis, IN, USA) at postnatal day 1–3 (7–8 g) were anesthetized with 2% isofluorane and subcutaneously injected in the neck region with 3ml/kg of 50 mg/kg capsaicin (Sigma Chemical Co., St. Louis, MO, USA) dissolved in vehicle solution (10% ethanol, 10% Tween-20 and 80% saline) or vehicle only, as originally described by Jancso et al. (Jancso et al., 1977). This administration protocol has been reported to result in 90–95% depletion of the unmyelinated fibers of the third lumbar dorsal root of the adult rat (Nagy et al., 1981, Nagy et al., 1983).
The rat pups were returned to their home cages with their dams and housed in groups of 5–8 in plastic cages with soft bedding under a reversed 12 hour light/dark cycle until they were 21 days old. At that time, the rats were separated and housed in groups of 2–3 in plastic cages with soft bedding and were given free access to food and drinking water until they were 12 weeks old.
Determination of the effectiveness of capsaicin-induced sensory denervation was performed by evaluating the chemosensitivity of the cornea by application of capsaicin solution (0.01%) to the cornea when neonatal capsaicin and vehicle treated rats were 12 weeks old. In the neonatal capsaicin group, only animals that blinked three or fewer times were used in the study (Tsuda et al., 2000, Nakagawa et al., 2007).
To provide mechanical stabilization of the fracture site during and following fracture, an orthopedic procedure was performed by inserting a stainless steel pin into the medullary cavity of the femur. Pin placement surgery was performed on capsaicin or vehicle-treated rats at 12 weeks of age.
Rats received an intraperitoneal injection of 100 mg/kg ketamine and 10 mg/kg xylazine to provide anesthesia for the following orthopedic procedure: an incision of approximately 6mm was made in the skin, then the proximal patellar ligament of the left femur was severed revealing the synovial space of the knee joint (Manigrasso and O’Connor, 2004, Freeman et al., 2008). A 20-gauge needle was used to core between the condyles and into the medullary canal of the left femur. Rats were immediately radiographed to ensure proper medullary coring. A pre-cut 0.8mm diameter (27 mm long) stainless steel wire pin (Small Parts Inc., Miami Lakes, FL, USA) was inserted into the medullary space and dental amalgam was used to secure the pin and seal the hole. Wound clips (MikRon Precision Inc., Gardena, CA, USA) were used to close the incision and were removed 7 days post-pin placement.
A closed mid-diaphyseal fracture of the left femur was produced 21 days post-pin placement in anesthetized rats (100 mg/kg ketamine and 10 mg/kg xylazine given intraperitoneally) as originally described by Bonnarens and Einhorn (Bonnarens and Einhorn, 1984). The 3-point impactor device (BBC Specialty Automotive Center, Linden, NJ, USA) used to produce fractures was based on the original design of Bonnarens and Einhorn (Bonnarens and Einhorn, 1984) and subsequently adapted by Simon et al (Simon and O’Connor, 2007). The left femur of the anesthetized rat was secured between two lower supports and an upper impactor head. A guillotine-like effect was created by dropping a rod guided 411 g weight from a height of 20 cm onto the spring loaded upper impactor head resulting a femoral fracture (Freeman et al., 2008). Immediately following fracture, rats were radiographed to ensure localization of a mid-diaphyseal fracture. Exclusion criteria were fractures located in the methaphysis (the wider part at the end of the shaft of a long bone, adjacent to the epiphyseal disk), dislodged pins and non-visible fracture following impact (Gerstenfeld et al., 2007). Following recovery from anesthesia, rats were allowed unrestricted movement and weight bearing of the fractured limb.
Pain related behaviors were evaluated before fracture (day 0) and at days 1, 3, 7, 14, and 21 following fracture to assess ongoing (spontaneous) fracture pain-related behaviors (guarding and flinching) as previously described (Jimenez-Andrade et al., 2007, Koewler et al., 2007, Freeman et al., 2008). Briefly, the number of hind limb flinches and time spent guarding over a two minute observation period were recorded as measures of ongoing pain, as these endpoints are similar to observations in patients who protect their fractured limb (Santy and Mackintosh, 2001). The investigator was blinded as to the experimental condition of the animals.
Fracture-induced pain was also assessed by differences in weight distribution of the left hind limb (fractured or pin hind limb) as compared to the right hind limb (intact hind limb) using an incapacitance meter as previously described (Medhurst et al., 2002, Freeman et al., 2008). Weight-bearing was used as an endpoint in this study as this has been widely used in humans to evaluate bone healing following fracture (Oni et al., 1988, Skak and Jensen, 1988, Corrales et al., 2008). Briefly, the mean force applied during 3 sec by each hind limb was measured in five trials. Weight bearing on the affected left hind limb was calculated as percentage of total weight bearing on both hind limbs by the following equation:
[weight on affected hind limb / (weight on affected hind limb + weight on intact hind limb)] × 100.
Our experimental protocol consisted of four different groups: naïve (n=4), pin (n=4), pin+fracture in rats treated with vehicle (n=9) and pin+fracture in rats treated with capsaicin (n=9).
In the present study, evaluation of the effects of capsaicin treatment on the periosteal sensory innervation was performed in whole-mount preparations. This is because it is relatively easy to visualize and quantify the density of the mesh-like neuronal networks within periosteum relative to cross sectional analysis of the periosteum attached to the bone.
Immunohistochemical analysis was performed on the whole mount preparations of periosteum from adult rats (12 weeks old) treated with capsaicin (n=6) or vehicle (n=6) without fracture to determine the changes in the innervation of the periosteum following capsaicin treatment. Rats were perfused intracardially with 200 ml of 0.1M phosphate buffered saline (PBS) followed by 200 ml of 4% formaldehyde/12.5% picric acid solution in 0.1M PBS. The femurs were removed, post-fixed for 4 hours in the perfusion fixative and placed in PBS solution. Periosteum from the left diaphyseal shaft was removed as a whole mount and processed for immunohistochemistry according to the following procedure adapted from previous studies (Mach et al., 2002, Martin et al., 2007). Excess muscle was removed from the femur using surgical scissors without disturbing the bone and attached periosteum. Periosteum was obtained from the distal end growth plate region to immediately below the third trochanter (Fig 1A) as all of the experimental fractures were localized within this anatomical region. The periosteum was removed from the bone by tracing the lower and upper limits of the desired area with a micro scalpel blade and a vertical cut was then performed along the posterior surface of the bone. Under a dissecting microscope, the periosteum was removed by gently scraping against the bone using the edge of a forceps (Brownlow et al., 2000). In our hands, the technique above described results in maximal preservation and minimal damage of both the cambium and fibrous layers of the periosteum. During periosteum removal, femurs were continually irrigated with PBS to prevent tissue dehydration. Once the periosteum was removed, it was cut on the vertical axis to produce two samples of the whole mount preparation which have both the mid-diaphyseal and metaphyseal regions. The size of the periosteal whole mount preparation and its attached thin muscle layer used for immunohistochemistry was approximately: width=5 mm; length=12 mm; thickness=0.5 mm.
The harvested whole mount preparations of periosteum were then washed in PBS three times for 10 minutes each wash (3×10 minutes), incubated for 60 minutes at room temperature (RT) in a blocking solution of 3% normal donkey serum in PBS with 0.3% Triton-X 100 and then incubated overnight at RT with primary antibodies. Unmyelinated primary afferent sensory nerve fibers were labeled with polyclonal rabbit anti-rat calcitonin gene related peptide (CGRP) (1:15,000 dilution; Sigma Chemical Co, St. Louis, MO, USA). Myelinated primary afferent sensory nerve fibers were immunostained for 200 kD neurofilament H (NF200) (polyclonal chicken anti-mouse NF200, 1:1,000, Chemicon, Temecula, CA, USA). Preparations were then washed in PBS 3×10 minutes and incubated for three hours at RT with secondary antibodies conjugated to fluorescent markers (Cy3: 1:600, Cy2: 1:200; Jackson ImmunoResearch, West Grove, PA, USA). Preparations were counterstained with DAPI (4'-6-diamidino-2-phenylindole, 1:20000, Invitrogen, Carlsbad, CA, USA) for 5 minutes and washed with PBS. Finally, tissue was washed in PBS and dehydrated through an alcohol gradient (70, 80, 90 and 100%), cleared in xylene, mounted (attached muscle layer in contact with the slide) on gelatin-coated slides and coverslipped with di-n-butylphthalate-polystyrene-xylene (Sigma Chemical Co., St. Louis, MO, USA).
Two random images of periosteal whole-mount preparations (400x magnification, 323µmX285µmX40µm observation field) were obtained from mid-diaphysis periosteal whole mounts from each left rat femur using an Olympus Fluoview FV1000 laser scanning confocal imaging system (Olympus America Inc, Melville, NY, USA software v. 5.0). Digital confocal grayscale images were compiled from 40 optical sections spaced 1.0µm apart in the z-plane. The z-series of each field of view were then used to generate a single topographic image (two dimensional projection). Images of periosteal whole mount preparations were obtained from the interphase between cambium and fibrous layers using DAPI (nuclear marker) as a reference.
Quantification of CGRP positive (CGRP+) and NF200 positive (NF200+) sensory nerve fibers in rat periosteal whole mount preparations was performed using two different approaches. The number of intersections between nerve fibers and vertical grids has been used previously to quantify density of nerve fibers in periosteum (Hukkanen et al., 1993, Martin et al., 2007). Briefly, digital confocal photomicrographs were acquired as described above. Images were viewed on a high-resolution monitor and the number of intersections between nerve fibers and the vertical grids (7.35 µm spacing, Adobe Photoshop software v. 7.0, San Jose, CA, USA) was quantified. Results were expressed as the mean number of intersections per mm2. Additionally, a second technique for quantification of CGRP+and NF200+ sensory nerve fibers in rat periosteal whole mount preparations was adapted from previous studies (Knyihar-Csillik et al., 2000, Mechawar et al., 2000, Tsukagoshi et al., 2006, Xie et al., 2007). The confocal images were viewed on a high-resolution monitor and the length of nerve fibers was determined using Image Pro Plus v. 6.0 image analysis software. The intensity threshold was set to exclude nonspecific background fluorescence and applied to all sections analyzed.
In order to determine the thickness of the femoral periosteum in capsaicin vs. vehicle treated rats, cross sectional analysis of the right femurs from rats treated with capsaicin and vehicle with the periosteum attached to the bone were processed as follows. Right femurs were post-fixed, washed in PBS and minimal bone decalcification was performed as previously described (Mach et al., 2002). After complete bone demineralization, determined radiographically, bones were cryoprotected in 30% sucrose at 4° C for at least 48 hours and serially sectioned on the longitudinal axis at a thickness of 40µm. Five sections at least 150µm apart were counter stained with DAPI and used to determine the thickness of the mid-diaphysis periosteum. Two images of sections at the mid-diaphysis were digitally captured at 20x using a SPOT II digital camera with SPOT image capture software (Diagnostic Instruments, Sterling Heights, MI, USA) attached to an Olympus BX51 microscope (Olympus America Inc, Melville, NY, USA). These images were analyzed using Image Pro Plus software v. 6.0 (Media Cybernetics Inc, Silver Spring, MD, USA) to determine the thickness of the periosteum.
Radiograph images (Specimen Radiography System Model MX-20, Faxitron X-ray Corporation, Wheeling IL, USA; Kodak film Min-R 2000, Rochester NY, USA) of fractured femurs were obtained immediately post-fracture and at days 1, 3, 7, 14, 21, 28 post-fracture. Fracture-induced callus formation was radiographically assessed at 3X magnification (Bergstrom et al., 2000, Koivukangas et al., 2003). The radiographs were scanned (Scanjet XPA, Hewlett Packard, Palo Alto, CA, USA) at 600 dpi resolution and the area of the callus was determined by an investigator blind to the experimental condition using Image Pro-Plus v. 6.0 software. Additionally, the images were used to evaluate bone bridging of the fractured femurs at all time points. The mean bone bridging score was determined using a 0 to 4 scale (Bergenstock et al., 2005). Briefly, points were assigned according to the bridging across the two cortices of the fractured femur (1 point each) and across the two peripheral sides of the callus (1 point each). This procedure was performed by two investigators who were blind to the experimental condition. Results were averaged.
SPSS (version 15) statistics package (SPSS, Chicago, IL, USA) was used to perform statistical analyses. Frequency distributions of the dependent variables appeared markedly non-normal, therefore the groups were compared at each timepoint using a non-parametric ANOVA (Kruskal-Wallis), followed by Mann-Whitney two-group comparisons; p<0.05.
Radiographic evaluation indicated no significant bone remodeling following intramedullary pin placement. Age-matched naïve and pin rats were radiographically similar in appearance at all time points examined (Fig 1A, B). The three-point fracture protocol resulted primarily in transverse or slightly oblique fractures (Fig 1C). All experimental fractures were located at mid-diaphysis of the femur. Thus, analysis of the effects of capsaicin neonatal treatment on periosteal innervation using whole mount preparations was restricted to the mid-diaphysis. The brackets in Fig 1A indicate the region where periosteal whole mounts were obtained.
The periosteum is a fibrous and cellular sheath which covers the outer surface of all the bones of the body (Allen et al., 2004) (Fig 2A–C). Three dimensional reconstruction of the femoral periosteum shows a high density of sensory CGRP+ nerve fibers in the periosteum which form a mesh-like network (Fig 2B–C) that may be involved in detecting algogenic substances as well as mechanical distortion of the underlying mineralized bone following fracture.
Confocal photomicrographs of whole mount diaphyseal periosteum preparations show that CGRP+ and NF200+ nerve fibers have a linear and bifurcating pattern of fibers. Sensory fibers in the periosteum can be found as single nerve fibers or nerve bundles (Fig 3A–D). These sensory fibers form a mesh-like network which envelops the naïve, unfractured bone (Fig 3A, 3C).
Neonatal capsaicin treatment did not alter the thickness of the periosteum compared to vehicle-treated animals as determined by cross sectional analysis of the periosteum attached to bone (data not shown). Quantitative analysis of the density of CGRP+ and NF200+ nerve fibers using two different methodologies revealed that capsaicin treatment resulted in a significant reduction in the density of CGRP+ fibers in the mid-diaphysis femoral periosteum as compared to rats treated with vehicle (937±105 vs. 4142±255 CGRP+ fiber intersections per mm2; p<0.05). Likewise, the total length of CGRP+ nerve fibers in the mid-diaphysis of the periosteum (volume examined: 323µmX285 µmX40µm) was significantly reduced in femurs from rats treated with capsaicin as compared with vehicle (538±92 vs. 3200±397µm CGRP+ nerve fibers; p<0.05, Fig 3A–B).
In contrast to the effect on CGRP+ nerve fibers by capsaicin treatment, there was no significant difference in the density of NF200+ fibers in the periosteum of rats treated with capsaicin and rats treated with vehicle (3630±180 vs. 3878± 113 NF200+ fiber intersections per mm2). This result was also confirmed when the total length of NF200+ nerve fibers was determined in femoral periosteal sections (volume examined: 323µmX285µmX40µm) from capsaicin and vehicle-treated rats (3095±235 vs. 2818±471 µm NF200+ nerve fibers, Fig 3C–D).
No capsaicin-treated or vehicle-treated rats were excluded due to protruding pins following the surgical protocol or fractures localized in the metaphysis. One capsaicin-treated rat was excluded due to a broken pin.
Spontaneous guarding, spontaneous flinching and weight bearing in the left hind limb were analyzed in naïve, pin alone, pin+fracture rats treated with vehicle and pin+fracture rats treated with capsaicin from day 1 through day 21 post-fracture. Rats with an intramedullary pin alone (pin, Fig 4A–C) exhibited minimal flinching and guarding as well as normal hind limb weight bearing from day 1 through day 3 post-fracture. All three behaviors were not significantly different from those observed in naïve rats (baseline values). Rats with a femoral fracture and treated with vehicle exhibited a greater time spent guarding, an increased number of flinches, and marked reduction in weight bearing as compared to naïve and pin alone rats (Fig 4A–C) from day 1 through day 14 post-fracture. These spontaneous pain-related behaviors peaked at day 1 post-fracture, decreased gradually and returned to baseline values at day 21 post-fracture (Fig 4A–C). Following fracture, a significant reduction in weight bearing was observed from day 1 to day 14 post-fracture, with the greatest reduction being detected at day 1 post-fracture (Fig 4C). At day 21 post-fracture pain-related behaviors in rats with femoral fracture were not significantly different than those in pin alone rats.
The depletion of sensory nerve fibers with capsaicin resulted in reduction of the magnitude (50%) of quantified pain-related behaviors following fracture in nearly all time points evaluated as compared to the vehicle-treated rats. However, this treatment did not modify the onset and duration of these pain-related behaviors. Rats treated with capsaicin displayed pain related behaviors which peaked at day 1 post-fracture and returned to baseline values at day 21 post-fracture (Fig 4A–C).
In the present model, femoral calcified callus formation and bridging of the cortical bone in fracture+capsaicin and fracture+vehicle groups were examined at days 7, 14, 21 and 28 post-fracture. Total calcified callus area, defined by the total radiopaque area within the outermost boundary of the fracture callus, was measured in these groups. A radiopaque callus was first evident at day 7 post fracture in both groups and it increased progressively until day 21. Calcified callus area was significantly increased in the fracture+capsaicin group when compared to the fracture+vehicle group at days 14 and 28 days post-fracture (Fig 4D). In contrast to these changes in the calcified callus formation, depletion of sensory fibers by capsaicin did not significantly modify the bone bridging scores in rats with femoral fracture (Fig 4E).
Injury and age related degeneration of the musculoskeletal system are one of the most common reasons individuals seek medical care and pain associated with these conditions is among the leading causes of chronic pain and long term disability (Woolf and Pfleger, 2003, Weevers et al., 2005, Brooks, 2006, Kidd, 2006, Rosemann et al., 2007). A major reason why skeletal pain remains a significant health problem is that the few analgesics available to treat this skeletal pain have adverse effects that negatively affect bone healing or the central nervous system. For example, studies in rodents and humans have suggested that many non-steroidal anti-inflammatory drugs and selective cyclo-oxygenase-2 inhibitors hinder callus formation and effective bridging of the fracture site resulting in delayed bone healing and increased incidence of nonunion of the fractured cortical bone (Giannoudis et al., 2000, Simon et al., 2002, Gerstenfeld et al., 2003, Bhattacharyya et al., 2005, Koester and Spindler, 2006, Murnaghan et al., 2006, Simon and O’Connor, 2007).
Opioids are also frequently used to control moderate to severe skeletal pain (McCann and Stanitski, 2004, Ekman and Koman, 2005, Mahowald et al., 2005, Mamaril et al., 2007). While there are relatively few studies that have examined the effect of sustained administration of opioid agonists or antagonists on bone injury/repair/healing, the few publications that have addressed this issue are conflicting. Some studies suggest that opiate agonists are associated with accelerated bone and tissue healing while others suggest the opposite (Poonawala et al., 2005, King et al., 2007, Petrizzi et al., 2007, Zagon et al., 2007). However, one area where there is significant agreement is that opioids do have a variety of non-skeletal side effects that can indirectly inhibit bone healing (Ensrud et al., 2003). Thus, opioids, as a class, frequently cause somnolence, agitation, constipation, dizziness and cognitive impairment which can reduce mobility resulting in loss of bone and muscle mass and, especially in the elderly, affect the ability of these patients to engage effectively in the rehabilitation required for effective bone healing (Ensrud et al., 2003).
The periosteum is a thin, cellular and fibrous tissue that is densely innervated by sensory nerve fibers and is tightly adherent to the outer surface of cortical bone and there are several lines of evidence suggesting that sensory nerve fibers located in the periosteum play a significant role in driving acute bone fracture pain. First, in human volunteers, acute mechanical stimulation of the periosteum resulted in intense sharp pain that was referred to the stimulated bone (Inman, 1944). In contrast, when orthopedic surgeons drill or “core” the marrow and the adjacent mineralized bone to insert a stabilizing rod, many patients report minimal pain (Moed et al., 1998). Second, in the small group of elderly individuals who appear to lack an observable periosteum in the proximal neck of the femur, fracture of this site (which is commonly referred to as a hip fracture) can be relatively painless and only diagnosed by radiological examination (Brunner et al., 2003). Third, repositioning the fractured bone and tightly adherent periosteum to its original non-fractured orientation, results in a marked attenuation of the fracture pain (Yates and Smith, 1994, Piermattei and Flo, 1997, Hedequist et al., 1999, Greenfield, 2006, Pape and Giannoudis, 2006, Mudge and Bramlage, 2007). Lastly, in both humans and animals, further movement of the fractured bone and distortion of the tightly adherent periosteum (which can occur when the patient is moving out of bed or participating in rehabilitation) frequently results in the reappearance of sharp, highly intense pain that subsides when the movement ceases and the periosteum is returned to its original non-fractured orientation (Yates and Smith, 1994, Santy and Mackintosh, 2001).
In the present report we demonstrate that capsaicin-induced depletion of CGRP+ sensory nerve fibers resulted in 50% reduction in the severity of bone fracture pain behaviors but not a significant change in the time course of fracture pain behaviors in adult rats. It is interesting to note that neonatal capsaicin-induced depletion of CGRP+ sensory nerve fibers also reduced pain behaviors by 50% in a rat model of inflammatory muscle pain (Kehl et al., 2000). Recent clinical studies have reported positive findings on the ability of capsaicin-based analogs applied topically or instilled into the surgical site to attenuate but not abolish osteoarthritis-related knee pain and postoperative pain in knee replacement surgeries (McCarthy and McCarty, 1992, McCleane, 2000, Wong et al., 2008). Together these preclinical and clinical studies suggest that there are two populations of nerve fibers, capsaicin sensitive C-fibers and capsaicin-insensitive A-delta fibers, that are normally involved in signaling fracture. Presumably, the presence of either population of nerve fibers alone will signal a component of skeletal pain but for the full intensity of fracture pain to develop requires the presence and activation of both the C- and A-delta populations that normally innervate the bone.
Previous results have suggested that sensory nerve fibers may play a role in bone remodeling in the normal, non-injured state (Hill et al., 1991, Offley et al., 2005). While the primary endpoint of the present study was fracture pain, not bone healing, we did monitor the effect that capsaicin administration had on bone healing as assessed by the formation of the calcified callus and the bridging of the fracture site. In the present report, we observed that at two time points (day 14 and day 28 post-fracture) formation of the calcified callus was significantly higher in capsaicin compared to vehicle-treated animals. It should be noted that although capsaicin-induced depletion of sensory nerve fibers produced an increase in calcified callus formation at days 14 and 28 post-fracture, it has previously been shown that callus does not begin to contribute significantly to stabilizing the bone until 7–10 days post-fracture (Einhorn, 1998, 2005). Thus, the reduction in fracture-induced pain in capsaicin-treated animals was evident at day 1, 3 and 7 days post-fracture when callus formation had yet to play a significant role in stabilizing the fractured bone. These present results suggest that while sensory nerve fibers may modulate fracture-induced bone remodeling, the reduction in fracture-induced pain behaviors in the capsaicin-treated animals at early time points is not simply due to an increase in callus formation in these animals.
The permanent ablation of unmyelinated sensory nerve fibers that is induced by neonatal capsaicin treatment has been widely used to evaluate the role of capsaicin-sensitive nerve fibers in the development and/or maintenance of inflammatory, visceral, neuropathic and muscle pain (Levine et al., 1986, Meller et al., 1992, Friese et al., 1997, Kehl et al., 2000). However, there are potential limitations of this technique including the potential neurochemical and anatomical reorganization of the remaining A-delta nerve fibers in the dorsal horn of the spinal cord and increased tissue concentrations of histamine and 5-hydroxytryptamine in the dorsal skin of the hindpaw (Holzer et al., 1981, Nagy and Hunt, 1983, Rethelyi et al., 1986, Shortland et al., 1990). While the functional significance of these alterations in pain transmission remains unclear, neonatal capsaicin treatment remains a valuable tool to evaluate the role of capsaicin-sensitive sensory nerve fibers in the development and maintenance of pain in a variety of preclinical models of acute and chronic pain (Khan et al., 2004, Bellinger et al., 2007, Nakagawa et al., 2007).
Capsaicin therapies may be particularly effective in reducing skeletal pain in that, compared to skin, bone may be innervated by a unique repertoire of sensory nerve fibers. The skin is innervated by multiple classes of sensory nerve fibers including thickly myelinated nerve fibers (A-beta), thinly myelinated sensory nerve fibers (A-delta), and two classes of unmyelinated C-fibers: the peptidergic rich CGRP+ expressing nerve fibers and peptidergic poor nerve fibers (Hunt and Mantyh, 2001, Julius and Basbaum, 2001, Zylka et al., 2005). Interestingly, the bone appears to be innervated primarily by NF200+ (A-delta) and CGRP+ (A-delta or peptidergic C-fibers) sensory nerve fibers, while largely lacking a significant innervation by either the thickly myelinated A-beta nerve fibers or the non-peptidergic C-fibers that express Isolectin B4 (IB4), the purinergic receptor P2×3, and mas-related G-protein-coupled receptor member D (Mrgprd) (Ozawa et al., 2003, Aoki et al., 2005, Zylka et al., 2005, Ivanusic et al., 2006, Ozawa et al., 2006, Ohtori et al., 2007).
One reason why many chronic pain states are difficult to fully control is that there is an inherent redundancy in nociceptors that innervate peripheral tissues such as the skin. Thus, tissues like the skin receive an innervation by CGRP+ and IB4 P2×3/Mrgprd+ populations of C-fiber nociceptors and these two populations of nociceptors express a different repertoire of receptors and respond to different noxious stimuli. For example, most CGRP+ sensory nerve fibers express the TrkA receptor and are directly excited and sensitized by NGF (McMahon et al., 1994, Averill et al., 1995, Molliver et al., 1995, Bennett et al., 1996, Shu and Mendell, 2001). In contrast, the IB4+ population does not express TrkA, does not respond to NGF but rather it responds to GDNF and ATP (Averill et al., 1995, Molliver et al., 1995, Molliver et al., 1997, Bogen et al., 2008, Dussor et al., 2008). If it is definitely shown that bone lacks significant innervation by the IB4/P2×3/Mrgprd population of nerve fibers, this may offer a unique opportunity to block skeletal vs. skin pain. Due to the lack of peptidergic rich and peptidergic poor C-fiber “redundancy” in bone, fewer classes of nociceptors would need to be blocked to attenuate skeletal pain vs. skin pain. Indeed, studies using an anti-NGF strategy (that targets the primary afferent nerve fibers that express TrkA+, CGRP+ fibers) to reduce pain suggest that this therapy is more effective in reducing mechanical hyperalgesia developed in bone pain models as compared to skin pain models (Zahn et al., 2004, Jimenez-Andrade et al., 2007, Koewler et al., 2007).
The present results suggest that both capsaicin-sensitive C-fibers and capsaicin-insensitive A-delta sensory nerve fibers that innervate the bone are involved in driving fracture pain. Whether the C-fibers and A-delta nerve fibers that innervate the bone transmit different types of bone pain (i.e. the dull, aching, throbbing pain vs. the sharp, intense, movement-evoked pain) remains unclear. Understanding the unique populations, organization, phenotype and plasticity of sensory and sympathetic nerve fibers that innervate the skeleton and how these nerve fibers change with trauma, disease and aging should provide insight into the development of mechanism based therapies to treat acute and chronic musculoskeletal pain.
This work is supported by the National Institutes of Health (NS23970) and by the Department of Veterans Affairs, Veterans Health Administration, Rehabilitation Research and Development Service. The authors thank Magdalena Kaczmarska and Kate Lentz for their helpful comments on the manuscript. None of the authors has any conflict of interest to declare.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.