PAIN�

148 (2010) 473–480

w w w . e l se v i e r . c o m / l o c a t e / p a i n

Human nerve growth factor sensitizes masseter muscle nociceptors in female rats

Peter Svensson a,b, Mian Wei Wang c, Xu-Dong Dong c, Ujendra Kumar c, Brian E. Cairns c,*

a Department of Clinical Oral Physiology, School of Dentistry, Faculty of Health Sciences, University of Aarhus, DK-8000 Aarhus C, Denmarkb Department of Oral and Maxillofacial Surgery, Aarhus University Hospital, DK-8000 Aarhus C, Denmarkc Faculty of Pharmaceutical Sciences, The University of British Columbia, Vancouver, Canada V6T 1Z3

a r t i c l e i n f o

Article history:Received 26 June 2009Received in revised form 9 December 2009Accepted 15 December 2009

Keywords:Craniofacial painNeurotrophic factorsTemporomandibular disordersMuscle pain models

0304-3959/$36.00 � 2009 International Associationdoi:10.1016/j.pain.2009.12.009

* Corresponding author. Address: Faculty of PhUniversity of British Columbia, 2146 East Mall, VCanada V6T 1Z3. Tel.: +1 604 822 7715; fax: +1 604 8

E-mail address: brcairns@interchange.ubc.ca (B.E.

a b s t r a c t

Injection of nerve growth factor (NGF) into the masseter muscle is not painful but does induce a localized,quick onset (�1 h) and long-lasting mechanical sensitization in healthy human subjects. We tested thehypothesis that human NGF mechanically sensitizes masseter muscle nociceptors by increasing the sen-sitivity of peripheral N-methyl-D-aspartate (NMDA) receptors. Co-expression of the NR2B subunit of theNMDA receptor with P75 and TrkA NGF receptors by trigeminal ganglion neurons that innervate the mas-seter muscle was investigated immunohistochemically. Nociceptor activity was recorded extracellularlyfrom the trigeminal ganglion of anaesthetized female rats. Nociceptor mechanical threshold was assessedbefore and every 30 min for 3 h after injection of human NGF (25 lg/ml, 10 ll, n = 12), and in subsequentexperiments NGF with TrkA (n = 12) or P75 (n = 11) receptor antibodies. Glutamate (1 M, 10 ll) wasinjected at the end of each experiment. Approximately 85% of NR2B positive masseter ganglion neuronsco-expressed P75 or TrkA receptors, suggesting the potential for interaction. When compared with thevehicle control, it was found that injection of NGF into the masseter muscle did not evoke significantnociceptor discharge but did significantly reduce nociceptor mechanical threshold (�30%). There wasno effect of NGF on glutamate-evoked nociceptor discharge or glutamate-induced mechanical sensitiza-tion. Additional experiments indicated that NGF-induced mechanical sensitization could be partiallyattenuated with TrkA receptor antibodies, but not P75 receptor antibodies. These findings indicate thathuman NGF-induced sensitization of masseter nociceptors results, in part, from the activation of TrkAreceptors, but does not appear to be mediated through enhanced peripheral NMDA receptor activity.

� 2009 International Association for the Study of Pain. Published by Elsevier B.V. All rights reserved.

1. Introduction

Nerve growth factor (NGF) is a neurotrophic protein moleculeessential for the growth and survival of sympathetic and small diam-eter sensory afferent neurons [1,29]. It has been speculated that NGFmight contribute to muscle pain in conditions such as myofascialtemporomandibular disorders (TMDs) based on the previous obser-vations that systemic administration of recombinant human NGF asa treatment for diabetic neuropathy induced myalgia, particularly inthe jaw and eye muscles [2–4,16,30]. Recently, it has been shownthat single injections of NGF (�0.1 lg/kg) made into the human mas-seter (jaw-closer) muscle although not acutely painful, could pro-duce, within hours of injection, local mechanical sensitization inthe masseter muscle with little or no ongoing muscle pain [37–39]that was associated with pain on chewing and yawning. The magni-tude of mechanical sensitization produced after injection of NGF intothe masseter muscle of women appears to be significantly greater

for the Study of Pain. Published by

armaceutical Sciences, Theancouver, British Columbia,22 3035.

Cairns).

than that found in men [37–39] and lasted for about 2–3 weeks postinjection. The symptoms produced by this model are quite similar tothose reported by patients suffering from myofascial TMD and therelatively long duration of these symptoms make this an attractivemodel for the study of experimental masticatory muscle pain thatis more chronic in nature [37–39]. However, the mechanismsresponsible for the onset and maintenance of NGF-induced massetermuscle sensitivity have yet to be determined.

We speculated that the sensitizing effects of NGF might be due toa local effect of this neurotrophin on masseter nociceptors and resultfrom an interaction between NGF and peripheral N-methyl-D-aspar-tate (NMDA) receptors. We have previously investigated the effect ofglutamate on masseter muscle nociceptor activity and have shownthat activation of peripheral NMDA receptors mediates excitationand mechanical sensitization of masseter nociceptors producedwhen glutamate concentrations become elevated in the massetermuscle [6,7,9]. Masseter nociceptor discharge evoked by activationof peripheral NMDA receptors is greater in females than in malesand this sex-related difference appears to be due in part to an estro-gen-mediated increased expression of NR2B subunit-containingNMDA receptors by masseter nociceptors in females [15]. Althoughthe interaction between NGF and NMDA receptors is incompletely

Elsevier B.V. All rights reserved.

474 P. Svensson et al. / PAIN�

148 (2010) 473–480

understood, in vitro studies of central neurons suggest that NGF canrapidly increase NMDA receptor currents in cultured hippocampalcells [5]. NGF exerts its biological effects by acting on tropomyo-sin-related kinase (TrkA) and P75 receptors and there is also evi-dence that increased NGF levels result in the phosphorylation ofNR2B subunits of the NMDA receptor, possibly through activationof these receptors [11–13,21,34]. Taken together these results sug-gest that elevated levels of NGF could influence the excitability ofmasseter nociceptors acutely by altering the functional propertiesof peripheral NMDA receptors. However, nothing is presently knownabout the potential for an interaction between NGF and peripheralNMDA receptor mechanisms. Therefore, we tested the hypothesisthat NGF would act in female rats to decrease the mechanical thresh-old of masseter muscle nociceptors in part through an interactionwith peripheral NMDA receptors.

2. Materials and methods

2.1. Surgical preparation

Adult female Sprague–Dawley rats (n = 71, 225–400 g) were pre-pared for acute in vivo recording of masseter nociceptors under iso-flurane anesthesia as has been previously described in detail[6,7,9,15]. Upon completion of all surgical procedures, the isofluranelevel was reduced to 2.0–2.5% to maintain a continued absence of re-flex response to noxious toe pinch. Heart rate, mean blood pressure,expired CO2 and core body temperature were continuously moni-tored throughout the whole experiment. A vaginal lavage was per-formed and a sample of epidermal cells was collected on a glassslide to determine the estrus stage using a microscope. All proce-dures were performed in adherence with the principles of the Cana-dian Council on Animal Care and were approved by the University ofBritish Columbia Animal Care Committee.

2.2. Immunohistochemistry

To identify ganglion neurons which innervate the masseter mus-cle, the fluorescent dye fast blue (2%, 10 ll, Polyscience, USA) was in-jected bilaterally into the masseter muscle of four isoflurane-anesthetized female rats. Seven days after the fast blue injection,the rats were euthanized (pentobarbital 100 mg/kg) and then per-fused with 120 ml of cold saline followed by 120 ml of paraformalde-hyde (4% in distilled water). The right and left trigeminal gangliawere removed and cut into sections (40 lm) with a vibratome.Free-floating sections were treated with 5% normal goat serum(NGS) in phosphate-buffered saline (PBS) for 1 h. Every second sec-tion was incubated with 5% normal goat serum (NGS) in phos-phate-buffered saline (PBS) for 1 h and then incubated overnight at4 �C with either anti-rat TrkA receptor antibody (1:40; goat poly-clonal, R&D Systems, Minneapolis, MD, Lot # VFA01), anti-rat P75receptor antibody (1:225; rabbit polyclonal, Sigma–Aldrich, St.Louis, MO, Lot # N3908), anti-rat NR2B subunit antibodies (1:300;mouse monoclonal, Chemicon, Billerika, MA, Lot # MAB5780), or acombination of NR2B and P75 or TrkA antibodies in PBS containing1% normal goat serum. The specificity of these antibodies has beeninvestigated by Western blot by the manufacturers and in the caseof the NR2B and P75 antibodies by independent investigators[23,26]. The next morning the sections were washed several timeswith PBS and then incubated for 1 h at room temperature in the darkin the presence of FITC and CY3-conjugated IgG selective for the ani-mal species from which the primary antibodies were derived (forTrkA: donkey-anti-goat (FITC), for P75: goat-anti-rabbit (FITC), forNR2B: goat-anti-mouse (CY3)). After several washes in buffer, thesections were mounted on slides and visualized with a Leica DMfluorescent microscope. Four to 11 sections per animal (from twoganglia) were sampled. In control experiments, the tissue sections

were incubated without primary antibody to confirm specificity ofreceptor-like immunoreactivity. Fast blue positive ganglion neurons(masseter ganglion neurons) were counted and photographed forthe estimation of cell area using NIH Image J freeware. The numberof images depended on the number of fast blue positive neurons ineach slice. This ranged from 4 to 50 images viewed with the 63�oil lens. On average,�22 images were taken from each slice. Neuronswere considered positive when the intensity of the fast blue, FITCand CY3 signals, respectively, exceeded the 95% confidence intervalof the mean background intensity.

2.3. Stimulation and recording techniques

Single trigeminal nociceptors were recorded with a parylene-coated tungsten microelectrode (A-M Systems, Carlsborg, WA,USA) that was lowered into the trigeminal ganglion. Afferent fibersthat responded to mechanical stimuli applied with a blunt probeto the masseter muscle and failed to respond to application of brush,pinch, or pressure stimuli directly to the overlying skin surface wereconsidered to innervate muscle. Confirmation that the mechano-receptive field of the afferent fiber indeed lay within the jaw-closermuscle was further obtained by observing discharge in response toinjection of glutamate at the end of each experiment. All afferent fi-bers examined in the present study were required to project to thecaudal brainstem, since previous work has indicated a strong projec-tion of putative jaw-closer muscle nociceptors to this region[6,7,9,15]. Orthodromic action potentials evoked by mechanicalstimulation of the jaw-closer muscle receptive field were collidedwith antidromic action potentials evoked by electrical stimulationof the caudal brainstem to confirm this projection [6,7,9,15]. Thestraight-line distance between the stimulating and recording elec-trodes was divided by the antidromic latency to permit the estima-tion of the conduction velocity (CV) of each nociceptor.

2.4. Experimental protocol

Experiments were carried out to determine the effect of intra-muscular injection of human NGF on masseter muscle nociceptorexcitability and mechanical sensitivity. Before the injection ofany substance, baseline nociceptor mechanical threshold was mea-sured with an electronic von Frey Hair (model 160IC, IITC, Califor-nia, USA) at one-min intervals for 10 min [6,7,9]. Each nociceptoridentified was randomly assigned to injection of one of the follow-ing groups: (i) vehicle control (phosphate-buffered isotonic saline,n = 12) or NGF (Sigma, St. Louis, MO; 25 lg/ml in phosphate-buf-fered saline, n = 12), (ii) TrkA antibody alone (R&D Systems, Minne-apolis, MD; 2.0 lg/ml in phosphate-buffered saline, n = 12) orNGF + TrkA antibody (n = 12), (iii) P75 antibody alone (Sigma, St.Louis, MO; 1.2 lg/ml in phosphate-buffered saline, n = 12)) orNGF + P75 antibody (n = 11). The concentration of TrkA and P75antibodies was chosen by performing immunoblots to determineantibody-binding saturation concentration in the masseter muscletissue [18]. The concentration of antibodies used were also in thesame range as those used in the previous in vitro studies to blockTrkA and P75 receptors [35,42]. As a unique feature of this study,the investigator conducting the experiment was blinded to thecontent of the injections. In all experiments, the volume of eachsingle injection was 10 ll and was administered intramuscularlyinto the mechano-receptive field of the masseter muscle nocicep-tor with a 26-gauge needle connected by polyethylene tubing toa 25-ll Hamilton syringe. Prior to injection, baseline nociceptoractivity was recorded for 10 min. After injection, evoked activitywas monitored for 10 min, and then beginning 30 min post injec-tion, mechanical threshold was assessed at 1-min intervals for10 min every 30 min for a total of 3 h. At the end of the 3-h period,glutamate (1 M, Sigma, USA) was injected into the nociceptor mec-

P. Svensson et al. / PAIN�

148 (2010) 473–480 475

hano-receptive field. Nociceptor discharge in response to injectionof glutamate was used to confirm the location of the receptive fieldwithin the masseter muscle. Ten min after injecting glutamate, themechanical threshold was assessed at 1-min intervals for 10 min.

2.5. 17b-Estradiol concentration

Plasma estrogen (17b-Estradiol) levels were measured with acommercially available ELISA kit (GenWay, San Diego, CA). Plasmaconcentration of the unconjugated form of estradiol was measuredaccording to the protocol provided by the manufacturer. The mean(±SEM) estradiol concentration measured was 77 ± 9 pg/ml.

2.6. Data analysis

The activity of the identified nociceptors was amplified and fedinto a computer equipped with a 1401 Plus board and Spike 2 anal-ysis software (Cambridge Electronic Design). Recorded nociceptoractivity was stored electronically and analyzed off-line. The evokedresponse of each nociceptor was calculated by subtracting the totalnumber of spikes during the 10-min period before the injection(baseline activity) from the total number of spikes during the 10-min period after the injection.

A (10x)

B (63x)

TrNe

Fast Blue NR2B

Fig. 1. (A) The paired low magnification images show examples of fast blue positive trigand/or P75 receptor antibodies. (B) The inset box indicates the area examined at higheimmunoreactive with both NR2B and P75 antibodies. White calibration bars indicate 50

Mechanical threshold was determined by averaging 10 con-secutive mechanical stimuli at each time epoch. The percentchange in mechanical threshold (DMT(%)) was calculated bythe equation:

DMTð%Þ ¼ 100�ðMTt �MTBÞ=MTB

where MTt is the mechanical threshold at time t (30–180 min) andMTB is the baseline mechanical threshold.

2.7. Statistics

Evoked discharge in each group of experiments was assessedby Students t-test. Two-way repeated measures ANOVA andpost-hoc Holm-Sidak method were used to assess the effect ofvarious substances on MT. Pearson product moment correlationwas used to investigate the relationship between plasma estro-gen levels and nociceptor excitability changes. The Fisher exacttest was used to examine the frequency of sensitization betweenthe NGF and NGF + TrkA antibody groups at 1 h post injection.Kruskal–Wallis one-way ANOVA on ranks was used to test fordifferences in the MTB between female rats in different stagesof the estrous cycle. In all tests, the level of significance wasset at p < 0.05.

Masseter Muscle Afferent Fiber

igeminal Ganglionuron

P75

eminal ganglion (masseter ganglion) neurons that were immunoreactive with NR2Br magnification. The white arrow indicates a fast blue-labeled neuron which waslm.

476 P. Svensson et al. / PAIN�

148 (2010) 473–480

3. Results

3.1. Expression of TrkA and P75 receptors and co-expression withNMDA receptors

A total of 882 fast blue positive neurons were examined forTrkA expression, and 783 neurons for P75 expression from the fourrats. More than 70% of masseter ganglion neurons examined ex-pressed TrkA or P75 receptors, respectively (Figs. 1 and 2). Theexpression of these receptors was uniformly distributed amongsmall, medium and large ganglion neurons (Fig. 3). Consistent withprevious findings, �50% of masseter ganglion neurons expressedthe NR2B subunit of the NMDA receptor [15]. About 85% of theNR2B expressing ganglion neurons co-expressed either P75 or TrkAreceptors, which confirmed the potential for a receptor interaction(Figs. 1 and 2).

3.2. Effect of NGF on masseter nociceptors

Injection of NGF did not evoke significantly more discharge thaninjection of vehicle (Fig. 4) in accordance with our previous study[25]. Injection of glutamate resulted in substantial nociceptor dis-charge, however, the magnitude of cumulative discharge evokedby glutamate was similar in the NGF and vehicle groups. Therewere no significant correlations between estrogen concentration,

ReceptorsP75+NR2B TrkA+NR2B

Freq

uenc

y of

Co-

expr

essi

on (%

)

0

20

40

60

80

100

ReceptorP75 TrkA NR2B

Freq

uenc

y of

Exp

ress

ion

(%)

0

20

40

60

80

100A

B

Fig. 2. (A) The bar graph illustrates the mean frequency of expression of P75, TrkAand NR2B receptors in four female rats. (B) A large number of NR2B positivemasseter ganglion neurons co-expressed P75 (85%) or TrkA (85%) receptors, whichconfirms the potential for a receptor interaction. Error bars indicate standard errorof the mean (SEM).

baseline mechanical threshold or conduction velocity and gluta-mate-evoked nociceptor discharge.

Injection of NGF significantly decreased the mechanical thresh-old of masseter nociceptors compared to vehicle (Fig. 5). Two-wayrepeated measures ANOVA revealed a significant effect of treat-ment (p = 0.022) and time (p < 0.001) on the relative mechanicalthreshold, but no significant interaction between treatment andtime. Post-hoc tests indicated that overall NGF significantly de-creased nociceptor mechanical threshold compared with vehicle(p = 0.022). There were no significant correlations between estro-gen concentration, baseline mechanical threshold or conductionvelocity and peak NGF-induced mechanical sensitization.

The change in mechanical threshold relative to the pre-gluta-mate injection mechanical threshold (at 3 h post NGF) revealedno significant difference in the magnitude of glutamate-inducedmechanical sensitization between the two groups (NGF:�29 ± 10%, vehicle: �30 ± 12%). Taken together, these resultsstrongly indicate that NGF did not increase glutamate-evoked noci-ceptor discharge or enhance glutamate-induced mechanicalsensitization.

3.3. Effect of TrkA and P75 receptor antibodies

Overall, co-administration of TrkA antibody with NGF appearedto attenuate the sensitizing effect of NGF (Fig. 6A). In contrast, co-administration of P75 antibody with NGF appeared to enhance thesensitizing effect of NGF (Fig. 6A). The scatter plot of Fig. 6B illus-trates the effect of NGF, NGF + TrkA antibody and NGF + P75 anti-body on the mechanical threshold of individual fibers 1 h postinjection. Injection of NGF decreased the mechanical threshold ofall 12 nociceptors tested at this time point, whereas only sevenof 12 nociceptors in the NGF + TrkA antibody group had a de-creased mechanical threshold after 1 h (p = 0.037, Fisher Exacttest). The addition of P75 antibody to NGF did not attenuate thesensitizing effect of NGF on individual fibers. There was no signif-icant correlation between estrogen concentration and the changein mechanical threshold in either the NGF + TrkA antibody orNGF + P75 antibody groups.

3.4. Influence of estrus stage

Of the 71 female rats used for the electrophysiology experi-ments, 48 were in diestrus, five in proestrus, 10 in estrus andeight in metestrus. There was no significant difference in the med-ian ([interquartile range]) MTB when diestrus (18 [10–39] g), pro-estrus (18 [14–30] g), estrus (20 [16–30] g) and metestrus (23[13–43] g) were compared. There was no significant correlationbetween estrogen concentration and MTB in this group of femalerats.

4. Discussion

Injection of human NGF into the masseter muscle at the sameconcentration that caused mechanical sensitization in humansdid not evoke significant nociceptor discharge but did cause a sig-nificant reduction in mechanical threshold that began 30 min postinjection. Despite finding a reasonably high co-expression ofNMDA receptor NR2B subunits with TrkA and P75 receptors, therewas no effect of NGF treatment on glutamate-evoked nociceptordischarge or glutamate-induced mechanical sensitization, whichsuggests that NGF does not alter the sensitivity of peripheralNMDA receptors as part of its effects on masseter nociceptors.Co-administration of TrkA but not P75 antibodies with NGF atten-uated NGF-induced mechanical sensitization in a subgroup of mas-seter nociceptors which indicates that at least some of thesensitizing effect of NGF is mediated through activation of this

Num

ber o

f Neu

rons

0

50

100

150

200

250

300

Area (µm2)

0-300300-600

600-900

900-1200

1200-1500

1500-1800

1800-2100

Num

ber o

f Neu

rons

0

50

100

150

200

250

300

Freq

uenc

y (%

)

0

20

40

60

80

100

Area (µm2)

0-300300-600

600-900

900-1200

1200-1500

1500-1800

1800-2100

Freq

uenc

y (%

)

0

20

40

60

80

100

120

TrkA

P75

BA

DC

Fig. 3. The histograms in (A) and (C) show the total number of masseter ganglion neurons assessed in each size category for TrkA and P75 receptor expression, respectively.The histograms in (B) and (D) illustrate the frequency of TrkA and P75 receptor expression, respectively, in these masseter ganglion neurons. Both receptors were uniformlydistributed among small, medium and large ganglion neurons.

P. Svensson et al. / PAIN�

148 (2010) 473–480 477

receptor. Taken together, these results suggest that a peripheralmechanism which involves sensitization of masseter nociceptorsthrough activation of TrkA receptors may underlie the relativelyrapid mechanical sensitization observed when NGF is injected intothe masseter muscle of human subjects.

Depending on the laboratory, the expression of TrkA receptorson trigeminal ganglion neurons has been reported to vary between

NGF Veh Glu(NGF)

Glu(Veh)

Cum

ulat

ive

Dis

char

ge

0

40

80

120

160

Fig. 4. The bar graph shows nociceptor discharge evoked by the initial injection ofNGF or vehicle (Veh) and also by the subsequent injection of glutamate. Nosignificant differences between NGF and vehicle evoked discharge or effects oftreatment on glutamate-evoked nociceptor discharge were found.

4% and 70% [20,32,33]. The expression of TrkA receptors also seemsto vary depending on the tissue innervated, such that 6%, 10% and15% of ganglion neurons that innervate the whisker pad [24], man-dibular incisor and cornea [28], respectively, are positive for thisreceptor. In marked contrast to these other craniofacial tissues,greater than 85% of trigeminal mesencephalic nucleus neurons,large diameter ganglion neurons that innervate spindle afferent fi-bers and other low threshold mechanoreceptors in the masseter

Time (hrs)Baseline 0.5 1 1.5 2 2.5 3

Cha

nge

in M

echa

nica

l Thr

esho

ld (%

)

-75

-50

-25

0

25

50

VehicleNGFNGF or

Vehicle

Fig. 5. The line and scatter plot illustrates changes in the mechanical threshold ofnociceptor after injection of NGF or vehicle into the masseter muscle. Error barsindicate SEM.

Cha

nge

in M

echa

nica

l Thr

esho

ld (

%)

NGF+TrkA antibody

Time (hrs)

NGF + AntibodyAntibody

NGF+P75 antibody

NGF NGF+TrkA NGF+P75

Rel

ativ

e C

hang

e in

Mec

hani

cal T

hres

hold

(%

)

-100

-50

0

50

100

B 0.5 1 1.5 2 2.5 3-75

-50

-25

0

25

50

B 0.5 1 1.5 2 2.5 3

A

B

Fig. 6. (A) The line and scatter plots illustrate changes in the mechanical threshold of a nociceptor after injection of NGF with TrkA antibody and NGF with P75 antibody. Co-injection of NGF and TrkA antibody attenuated NGF-induced mechanical sensitization whereas co-injection of NGF and P75 antibody appeared to enhance NGF-inducedmechanical sensitization. (B) The scatter plot illustrates the effect of NGF, NGF with TrkA antibody and NGF with P75 antibody on the mechanical threshold of individualnociceptors 1 h post injection. Injection of NGF decreased the mechanical threshold of all 12 nociceptors tested at this time point, whereas only seven of 12 nociceptors in theNGF and TrkA antibody group had a decreased mechanical threshold after 1 h.

478 P. Svensson et al. / PAIN�

148 (2010) 473–480

muscle express the TrkA receptor [41]. Consistent with the highexpression of these receptors by masticatory muscle spindle affer-ent fibers, the present results indicate that more than 70% of mas-seter afferent fibers also express TrkA and P75 receptors. The highlevel of expression of TrkA receptors by masticatory muscle affer-ent fibers may supply one rationale for specific complaints of jawmuscle sensitivity after systemic administration of NGF to humansubjects as well as the significant mechanical sensitization inducedin human subjects when NGF is injected intramuscularly [2–4,16,30].

Elevated levels of NGF have previously been reported to rapidlyincrease the excitability of deep tissue nociceptors and their re-sponse to mechanical stimulation. For example, injection of NGFor over expression of NGF (induced by viral gene transfer) in therat bladder wall caused bladder over-reactivity as measured byan increased micturation response to instillation of warm salineinto the bladder of awake rats [17,22]. Bladder Ad and C fibers wereexcited and sensitized to mechanical stimulation by instillation ofNGF into the bladder, however, at the effective concentrations of 1and 10 lg/ml NGF, there was also significant tissue inflammation[14]. The mechanical sensitization of bladder nociceptors appearedto be mediated primarily through activation of P75 receptors, as it

could be blocked in vitro by P75 receptor antibodies [42]. In a sin-gle study conducted on rat skeletal muscle nociceptors, it was re-ported that while injection of human NGF into thegastrocnemius-soleus muscle did not significantly affect mechani-cal sensitivity, it did evoke discharge in a subpopulation of C fiberswith high mechanical thresholds [19]. The receptor mechanismresponsible for NGF-evoked gastrocnemius-soleus muscle afferentdischarge was not determined. Application of human NGF to ratbrainstem slices in vitro has also been shown to rapidly (within10 s) depolarize trigeminal mesencephalic nucleus neurons (jaw-muscle spindle afferent fibers) to increase the number of actionpotentials that can be evoked by depolarizing current pulses fora period of more than 5 min [40]. This effect of NGF may be med-iated by the P75 receptor, as it is not blocked by a TrkA receptorantagonist [40]. Taken together, these results suggest that in dee-per tissues of the rat, NGF can both excite and sometimes mechan-ically sensitize nociceptors, however, these effects of NGF appearto be mediated through P75 rather than through TrkA receptoractivation. In contrast, the results of the current study indicate thatblockade of TrkA but not P75 receptors was effective in attenuatingmechanical sensitization induced by NGF injection into the masse-ter muscle in a subpopulation of masseter nociceptors. Neverthe-

P. Svensson et al. / PAIN�

148 (2010) 473–480 479

less, the finding that TrkA receptor blockade was ineffective inattenuating NGF-induced mechanical sensitization in many of thenociceptors examined suggests that additional mechanisms likelycontribute to this effect. It is unlikely that inflammation is contrib-utory to NGF-induced mechanical sensitization of masseter noci-ceptors, as previous work indicates that injection of NGF atconcentrations up to 25 lg/ml does not result in gross inflamma-tion of skeletal muscle [25]. Thermal sensitization of rat cutaneousnociceptors by human NGF has been associated with mast celldegranulation [31]. In particular, there is some evidence to suggestthat serotonin, which is released during mast cell degranulation,contributes to NGF-induced cutaneous nociceptor sensitization[27]. It is possible that this mechanism also contributes to NGF-in-duced mechanical sensitization of masseter nociceptors, as injec-tion of serotonin into the masseter muscle evokes nociceptordischarges and induces mechanical sensitization [36].

We speculated that the sensitizing effects of NGF might resultfrom an interaction between NGF and peripheral NMDA receptors.NGF can rapidly increase NMDA currents through phosphorylationof the NR2B subunit in cultured mouse and isolated rat hippocam-pal neurons through a mechanism that does not appear to involveTrkA receptor activation [13,21]. Glutamate evokes discharge andinduces a prolonged mechanical sensitization of masseter musclenociceptors through activation of NR2B subunit-containing NMDAreceptors [7,8,15]. These factors suggested that it was possible thatboth glutamate and NGF might act on the NMDA receptor to in-duce mechanical sensitization. However, pre-administration ofNGF failed to modify glutamate-evoked nociceptor discharge orglutamate-induced mechanical sensitization, which suggests thatNGF and glutamate do not appear to share a common mechanismwith regard to masseter muscle nociceptor sensitization.

The extent to which the menstrual cycle influences nociceptiveprocessing from the craniofacial region has yet to be fully eluci-dated [8,10]. Masseter muscle afferent fiber mechanical thresholdin female rats has previously been reported to be positively corre-lated with plasma estrogen concentration [25]. In the currentstudy, an estrous cycle stage-related difference in baseline noci-ceptor mechanical threshold or a significant correlation betweenplasma estrogen concentration and baseline nociceptor mechanicalthreshold was not identified. Further, no significant correlation be-tween NGF-induced mechanical sensitization and plasma estrogenconcentration was found, which suggests that fluctuations in estro-genic tone did not exert a major influence on NGF-inducedmechanical sensitization. Masseter muscle nociceptor dischargein response to NMDA receptor activation appears to be increasedduring proestrus compared to other stages of the rat estrous cycleand is positively correlated with blood estrogen levels [8,15]. Nev-ertheless, the magnitude of these changes in nociceptor excitabilitycan be fairly small, and thus detection of this effect requires thatrelatively large samples (e.g. 50–100 fibers) be assessed. The rela-tively small group sizes in the present study may have preventeddetection of small, estrous cycle-related influences on nociceptorexcitability.

We have previously shown that direct administration of smalldoses of NGF (5 lg) into the masseter muscle of healthy men andwomen is associated with a consistent and long-lasting increasein mechanical sensitivity in the NGF-injected masseter muscleand pain associated with jaw functions but without spontaneouspain or any systemic side effects [37–39]. In men, the mechanicalsensitivity was significant 3 h post injection whereas in women,significant mechanical sensitivity occurred 1 h after NGF injection.Comparison of the maximum decrease in pressure pain thresholdsone day after NGF injection in women (66.3 ± 4.2%) and men(39.4 ± 6.7%) indicated a significantly greater amount of sensitiza-tion in women [37,39]. The rapid onset of this NGF-inducedmechanical sensitization of the masseter muscle in humans

suggests that this effect is probably mediated by a peripheralmechanism consistent with our current finding of a significantreduction in nociceptor mechanical threshold in rats that began30 min post injection. We have also investigated whether pre-injection of NGF into the masseter muscle alters pain responsesor mechanical sensitization induced by injection of glutamate intothe same site 24 h later in healthy men [38]. NGF did not alter theperceived pain intensity of the glutamate injection, although theglutamate-evoked pain drawing areas were slightly, but signifi-cantly larger for the NGF-pretreated masseter muscle. The overallmagnitude of mechanical sensitization induced by glutamateinjections was greater in the NGF group than the control group[38]. However, the additional challenge of the NGF-pretreatedmuscle with glutamate did not indicate a conspicuous sensitizationto noxious chemical stimuli in healthy humans [38]. These humanfindings suggest that intramuscular administration of NGF could bean interesting model or proxy of TMD-like pain [37,38]. As thefindings in the current study conducted in rats appear consistentwith the aforementioned findings in healthy humans, we suggestthat injection of human NGF into the rat masseter muscle mightbe used as an animal model of myofascial TMD-like mechanicalsensitization. We propose that injection of NGF should be furtherevaluated in both animals and humans for its potential to modelvarious craniofacial pain conditions.

Acknowledgements

This work was supported by a Canadian Institutes of Health Re-search Catalyst Grant (CGR-86827; B.E.C.), Canadian Pain Society/Astra-Zeneca Biology of Pain Award (B.E.C.), Institute for Musculo-skeletal Health and Arthritis Visiting Professor Award (B.E.C.), aDanish Research Council Visiting Professor Award (P.S.) and anIASP Collaborative Research Grant (P.S. & B.E.C.). B.E.C. is the reci-pient of a Canada Research Chair. The authors do not have any con-flicts of interest related to this work.

References

[1] Aloe L, Bracci-Laudiero L, Bonini S, Manni L. The expanding role of nervegrowth factor: from neurotrophic activity to immunologic diseases. Allergy1997;52:883–94.

[2] Apfel SC, Arezzo JC, Brownlee M, Federoff H, Kessler JA. Nerve growth factoradministration protects against experimental diabetic sensory neuropathy.Brain Res 1994;634:7–12.

[3] Apfel SC, Kessler JA, Adornato BT, Litchy WJ, Sanders C, Rask CA. Recombinanthuman nerve growth factor in the treatment of diabetic polyneuropathy. NGFstudy group. Neurology 1998;51:695–702.

[4] Apfel SC, Schwartz S, Adornato BT, Freeman R, Biton V, Rendell M, Vinik A,Giuliani M, Stevens JC, Barbano R, Dyck PJ. Efficacy and safety of recombinanthuman nerve growth factor in patients with diabetic polyneuropathy: arandomized controlled trial. rhNGF Clinical Investigator Group rhNGF ClinicalInvestigator Group. JAMA 2000;284:2215–21.

[5] Bi R, Broutman G, Foy MR, Thompson RF, Baudry M. The tyrosine kinase andmitogen-activated protein kinase pathways mediate multiple effects ofestrogen in hippocampus. Proc Natl Acad Sci 2000;97:3602–7.

[6] Cairns BE, Gambarota G, Svensson P, Arendt-Nielsen L, Berde CB. Glutamate-induced sensitization of rat masseter muscle fibers. Neuroscience2002;109:389–99.

[7] Cairns BE, Svensson P, Wang K, Hupfeld S, Graven-Nielsen T, Sessle BJ, BerdeCB, Arendt-Nielsen L. Activation of peripheral NMDA receptors contributes tohuman pain and rat afferent discharges evoked by injection of glutamate intothe masseter muscle. J Neurophysiol 2003;90:2098–105.

[8] Cairns BE. The influence of gender and sex steroids on craniofacial nociception.Headache 2007;47:319–24.

[9] Cairns BE, Dong XD, Mann MK, Svensson P, Sessle BJ, Arendt-Nielsen L,McErlane KM. Systemic administration of monosodium glutamate elevatesintramuscular glutamate levels and sensitizes rat masseter muscle afferentfibers. Pain 2007;132:33–41.

[10] Cairns BE, Gazerani P. Sex-related differences in pain. Maturitis2009;63:292–6.

[11] Chao MV, Hempstead BL. P75 and Trk: a two-receptor system. Trends Neurosci1995;18:321–6.

[12] Chao MV. Neurotrophins and their receptors: a convergence point for manysignaling pathways. Neuroscience 2003;4:299–309.

480 P. Svensson et al. / PAIN�

148 (2010) 473–480

[13] Di Luca M, Gardoni F, Finardi A, Pagliardini S, Cattabeni F, Battaglia G, MissaleC. NMDA receptor subunits are phosphorylated by activation of neurotrophinreceptors in PSD of rat spinal cord. Neuroreport 2001;12:1301–5.

[14] Dmitrieva N, McMahon SB. Sensitisation of visceral afferents by nerve growthfactor in the adult rat. Pain 1996;66:87–97.

[15] Dong XD, Mann M, Kumar U, Svensson P, Arendt-Nielsen L, Hu JW, Sessle BJ,Cairns BE. Sex-related differences in NMDA-evoked rat masseter muscleafferent discharge result from estrogen-mediated modulation of peripheralNMDA receptor activity. Neuroscience 2007;146:822–32.

[16] Dyck PJ, Peroutka S, Rask C, Burton E, Baker MK, Lehman KA, Gillen DA,Hokanson JL, O’Brien PC. Intradermal recombinant human nerve growth factorinduces pressure allodynia and lowered heat-pain threshold in humans.Neurology 1997;48:501–5.

[17] Farquhar-Smith WP, Jaggar SI, Rice ASC. Attenuation of nerve growth factor-induced visceral hyperalgesia via cannabinoid CB1 and CB2-like receptors.Pain 2002;97:11–21.

[18] Hakim AW, Dong X-D, Svensson P, Kumar U, Cairns BE. TNF mechanicallysensitizes masseter muscle afferent fibers of male rats. J Neurophysiol2009;102:1551–9.

[19] Hoheisel U, Unger T, Mense S. Excitatory and modulatory effects ofinflammatory cytokines and neurotrophins on mechanosensitive group IVmuscle afferents in the rat. Pain 2005;114:168–76.

[20] Jacobs JS, Miller MW. Expression of nerve growth factor, p75, and the highaffinity neurotrophin receptors in the adult rat trigeminal system: evidence formultiple trophic support systems. J Neurocytol 1999;28:571–95.

[21] Jarvis CR, Xiong ZG, Plant JR, Churchill D, Lu WY, MacVicar BA, MacDonald JF.Neurotrophin modulation of NMDA receptors in cultured murine and isolatedrat neurons. J Neurophysiol 1997;78:2363–71.

[22] Lamb K, Gebhart GF, Bielefeldt K. Increased nerve growth factor expressiontriggers bladder overactivity. J Pain Symptom Manage 2004;5:150–6.

[23] Laurie DJ, Bartke I, Schoepfer R, Naujoks K, Seeburg PH. Regional,developmental and interspecies expression of the four NMDAR2 subunits,examined using monoclonal antibodies. Brain Res Mol Brain Res1997;51:23–32.

[24] Lu J, Zhou XF, Rush RA. Small primary sensory neurons innervating epidermisand viscera display differential phenotype in the adult rat. Neurosci Res2001;41:355–63.

[25] Mann MK, Dong XD, Svensson P, Cairns BE. Influence of intramuscular nervegrowth factor injection on the response properties of rat masseter muscleafferent fibers. J Orofacial Pain 2006;20:325–36.

[26] Matrone C, Marolda R, Ciafrè S, Ciotti MT, Mercanti D, Calissano P. Tyrosinekinase nerve growth factor receptor switches from prosurvival to proapoptoticactivity via Abeta-mediated phosphorylation. PNAS 2009;106:11358–63.

[27] Mendell LM, Albers KM, Davis BM. Neurotrophins, nociceptors, and pain.Microsc Res Tech 1999;45:252–61.

[28] Mosconi T, Snider WD, Jacquin MF. Neurotrophin receptor expression inretrogradely labeled trigeminal nociceptors – comparisons with spinalnociceptors. Somatosens Mot Res 2001;18:312–21.

[29] Pasero C. Pathophysiology of neuropathic pain. Pain Manag Nurs 2004;5:3–8.[30] Petty BG, Cornblath DR, Adornato BT, Chaudhry V, Flexner C, Wachsman M,

Sinicropi D, Burton LE, Peroutka SJ. The effect of systemically administeredrecombinant human nerve growth factor in healthy human subjects. AnnNeurol 1994;36:244–6.

[31] Rueff A, Mendell LM. Nerve Growth Factor and NT-5 induce increased thermalsensitivity of cutaneous nociceptors in vitro. J Neurophysiol 1996;76:3596.

[32] Shinoda M, Honda T, Ozaki N, Hattori H, Mizutani H, Ueda M, Sugiura Y. Nerveterminals extend into the temporomandibular joint of adjuvant arthritic rats.Eur J Pain 2003;7:493–505.

[33] Shoemaker SE, Kudwa AE, Isaacson LG. Sympathetic ingrowth to thetrigeminal ganglion following intracerebroventricular infusion of nervegrowth factor. Brain Res 2002;956:136–48.

[34] Shu XQ, Mendell LM. Neurotrophins and hyperalgesia. Proc Natl Acad Sci USA1999;96:7693–6.

[35] Skoff AM, Adler JE. Nerve growth factor regulates substance P in adult sensoryneurons through both TrkA and p75 receptors. Exp Neurol 2006;197:430–6.

[36] Sung D, Dong X, Ernberg M, Kumar U, Cairns BE. Serotonin (5-HT) excites ratmasticatory muscle afferent fibers through activation of peripheral 5-HT3receptors. Pain 2008;134:41–50.

[37] Svensson P, Cairns BE, Wang K, Arendt-Nielsen L. Injection of nerve growthfactor into human masseter muscle evokes long lasting mechanical allodyniaand hyperalgesia. Pain 2003;104:241–7.

[38] Svensson P, Castrillon EE, Cairns BE. Nerve growth factor-evoked massetermuscle sensitization and perturbation of jaw motor function in healthywomen. J Orofac Pain 2008;22:340–8.

[39] Svensson P, Wang K, Arendt-Nielsen L, Cairns BE. Effects of NGF-inducedmuscle sensitization on proprioception and nociception. Exp Br Res2008;189:1–10.

[40] Yamuy J, Pose I, Pedroarena C, Morales FR, Chase MH. Neurotrophin-inducedrapid enhancement of membrane potential oscillations in mesencephalictrigeminal neurons. Neuroscience 2000;95:1089–100.

[41] Zhang FX, Lai CH, Li JL, Shum DK, Chan YS. Tyrosine kinase receptorimmunoreactivity in trigeminal mesencephalic and motor neurons followingtransection of masseteric nerve of the rat. Neuroscience 2006;139:921–30.

[42] Zhang YH, Nicol GD. NGF-mediated sensitization of the excitability of ratsensory neurons is prevented by a blocking antibody to the p75 neurotrophinreceptor. Neurosci Lett 2004;366:187–92.