Substance P expression inhuman tooth pulp in relationto caries and pain experienceRodd HD, Boissonade FM. Substance P expression in human toothpulp in relation to caries and pain experience. Eur J Oral Sci 2000; 108: 467±474.# Eur J Oral Sci, 2000

The neuropeptide substance P (SP) is found within a subpopulation of nociceptivea�erent nerve ®bres and has been shown to be upregulated in a variety of sitesfollowing peripheral in¯ammation. The aim of this study was to investigate theexpression of SP within human teeth, both in health and disease, and to seek acorrelation between reported pain history and SP expression within pulpal nerves.Coronal pulps were removed from 62 permanent mandibular molars with a knownpain history. Teeth were categorised as intact, moderately or grossly carious. Usingindirect immuno¯uorescence, sections were double-labelled for the generalneuronal marker protein gene product 9.5 (PGP 9.5) and for SP. Image analysiswas then used to quantify the percentage area of PGP 9.5-labelled tissue which wasalso labelled for SP. Throughout the pulp, the expression of SP was found to besigni®cantly increased with the progression of caries. Furthermore, SP expressionwas signi®cantly greater in grossly carious painful specimens than in grossly cariousasymptomatic specimens. These datawould suggest that the expression of SPwithinpulpal nerves undergoes dynamic changes following caries, which may have animportant clinical signi®cance in terms of in¯ammation and pain experience.

Helen D. Rodd, Fiona M. Boissonade

Department of Oral and Maxillofacial Surgery,School of Clinical Dentistry, University ofSheffield, Sheffield, UK

Dr. Helen Rodd, Department of Oral andMaxillofacial Surgery, School of ClinicalDentistry, Claremont Crescent, SheffieldS10 2TA, UK

Telefax: z44±114±2717843E-mail: h.d.rodd@sheffield.ac.uk

Key words: substance P; tooth pulpinnervation; caries; pain

Accepted for publication September 2000

Substance P (SP) belongs to a family of neuropep-tides, known as the tachykinins, and is widely dis-tributed throughout the central and peripheralnervous systems (1). It has been implicated in avariety of biological mechanisms, including neuro-genic in¯ammation (2), modulation of immune cellfunction (3), and regulation of endothelial cellgrowth (4). In addition, SP is thought to play amajor role in the neurotransmission of nociceptiveinput (5).SP was the ®rst neuropeptide to be identi®ed

in dental tissues (6), and a number of immunocyto-chemical studies have subsequently demonstratedthe relative abundance of SP-immunoreactive nerve®bres within the tooth pulp and dentine of severalspecies, including man (7±13). The trigeminalorigin of SP-expressing intradental nerves hasbeen well-established by denervation studies (9).Furthermore, administration of capsaicin, a select-ive irritant of sensory unmyelinated ®bres, has beenshown to attenuate pulpal SP immunoreactivity (14).There is unequivocal evidence that, together

with other neuropeptides, SP expression undergoes

dynamic changes within the in¯amed dental pulp(15±18). The abundance of SP-immunoreactiveintradental nerve ®bres, together with the apparentupregulation of SP during pulpal injury, havestimulated considerable interest as to the functionalsigni®cance of this peptide within dental tissues.There is good experimental evidence to supportthe role of SP in vasodilation and neurogenicplasma extravasation within tooth pulp (19, 20),but much speculation remains as to its relevanceto dental pain.

Nonetheless, there is some indirect evidenceto support an interrelationship between SP anddental pain. An immunocytochemical investigationof teeth from a child with hereditary sensory andautonomic neuropathy, a condition associatedwith an inability to perceive pain, revealed a totalabsence of SP-immunoreactive pulpal nerve ®bres(21). Furthermore, teeth subject to experimentalforces have demonstrated signi®cant increases inpulpal levels of SP, which may be consistent withthe pain symptoms commonly reported by patientsundergoing orthodontic treatment (22).

Eur J Oral Sci 2000; 108: 467±474Printed in UK. All rights reserved

To date, there have been no quantitative immuno-cytochemical studies comparing SP expressionin healthy and carious human teeth. Furthermore,the role of SP in dental pain has not been wellestablished. Therefore, the aim of the present studywas to investigate the expression of SP withinhuman teeth, both in health and disease, and toseek a correlation between reported pain historyand SP expression.

Material and methods

Experimental material

The experimental material comprised 62 mandib-ular ®rst permanent molars which were obtainedfrom children requiring dental extractions undergeneral anaesthesia. These subjects all had oneor more ®rst permanent molars of extremely poorprognosis. In some cases, opposing or contralateral®rst permanent molars (which may or may nothave been carious), was also extracted for ortho-dontic reasons. Ethical approval for the study wasgranted by the South She�eld Research EthicsCommittee.

Immediately prior to surgery, the child andaccompanying guardian were interviewed to ascer-tain a simple dental pain history. A positive painhistory was recorded either when the child per-sonally reported that they had experienced somespontaneous toothache during the previous week,or when the parent stated that the child hadsu�ered sleep loss, attributed to dental pain, overthe past few nights. Thus the pain history wasconsidered consistent with that of irreversiblepulpitis (23). Teeth were excluded from the studyif there was any associated pathology other thancaries.

Immediately following simple forceps extrac-tion, a groove was cut with a diamond disc onthe buccal aspect of each crown. Teeth were thensplit longitudinally using an osteotome and asurgical mallet. The mesial half of the tooth wasretained and placed in ®xative (4% paraform-aldehyde and 0.2% picric acid in 0.1 M phosphatebu�er, pH 7.4) for 24 h at 4³C. The coronal pulpwas carefully removed from the pulp chamber andplaced in phosphate-bu�ered saline (PBS). Thedegree of caries in each tooth was then assessedvisually under a dissection microscope at 620magni®cation. Each tooth half was categorised asintact (no colour change within dentine but withpossible staining con®ned to enamel), moderatelycarious (colour changes did not extend beyondhalf the dentine thickness), or grossly carious(colour changes extended beyond half the dentinalthickness). This approach to caries assessment has

been widely used in similar studies and was found tobe highly reproducible (24, 25).

Tissue preparation for immunocytochemistry

The coronal pulps were left in PBS for 24 hat 4³C before placing in 0.1 M PBS containing30% sucrose solution for cryoprotection (5 hat 4³C). The pulp tissue was then embedded inTissue-Tek OCT compound (Bayer Diagnostics,Basingstoke, UK) and three 10 mm longitudinalsections (200 mm apart) were cut from each toothpulp and collected on poly-D-lysine-coated glassslides.

Tissue sections were processed for indirectimmuno¯uorescence as follows. Slides were washedin PBS containing 0.2% Triton X-100 (PBST)(2610 min) and then incubated in 10% normalgoat serum (Vector Laboratories, Peterborough,UK) diluted in PBST for 30 min at room temper-ature. Sections were subsequently incubated witha mixture of the following: a polyclonal antibodyraised in rabbit against human SP (dilution 1:800;Genosys, Cambridge, UK) and a monoclonalantibody raised in mouse against human proteingene product 9.5 (PGP 9.5) (dilution 1:1000;Ultraclone, Isle of White, UK) diluted in PBSTcontaining 5% normal goat serum for 24 h at 4³C.Antiserum to PGP 9.5 has been widely appliedfor immunocytochmeical demonstration of nerve®bres and is considered an excellent generalneuronal marker (26, 27).

Slides were then washed again in PBS(2610 min) before incubating, for a further 90 minat room temperature, with a mixture of goat anti-rabbit IgG conjugated to ¯uorescein isothiocyanate(dilution 1:20, Vector) and horse anti-mouse IgGconjugated to Texas red (dilution, 1:100; Vector).The secondary ¯uorescent antibodies were dilutedin PBST containing 2% normal goat serum. Slideswere ®nally washed again in PBS (2610 min), andsections were carefully dried and mounted inVectashield (Vector).

Immunohistochemical controls for SP wereperformed by incubating sections with antiserumto SP which had been preabsorbed with an excess(10 nmol/ml) of the peptide. Controls for PGP 9.5labelling were carried out by incubating sectionswith the antibody dilutent alone. No positive label-ling was seen in any of the controls. The cross-reactivity of SP antiserum with other structurallyrelated peptides and proteins was not investigatedthus immunoreactive material should strictly beconsidered to demonstrate SP-like immunoreact-ivity. Whilst this point is recognised, the presentstudy will, for brevity, simply refer to the labellingas showing SP-immunoreactivity.

468 Rodd & Boissonade

Analysis of immunolabelling

Sections were viewed using a Zeiss axioplan¯uorescent microscope and all analyses wereperformed blind. Descriptive ®ndings were basedon the examination of all three sections fromeach tooth pulp. However, quantitative data wereobtained from the analysis of only one tissuesection from each specimen, which was the ®rstsection to be collected on each slide. Preliminaryinvestigations had established that there were nosigni®cant di�erences in PGP 9.5- or SP-labellingbetween three sections each cut 200 mm aparttherefore quantitative analysis of the ®rst sectionwas considered to be representative of the entirecoronal pulp.Three di�erent ®elds were subject to quantitative

analysis: the mesio-buccal pulp horn, the buccalsubodontoblastic nerve plexus and the mid-coronalpulp region (containing large nerve trunks). Each®eld was viewed with the 620 objective andrepresented 0.22 mm2 of pulp tissue. In total,approximately 7% of the overall coronal pulptissue was sampled.Computer-assisted image analysis software

(Image-Pro Plus v3.0; Media Cybernetics, SilverSpring, MD, USA) was used to create a digitalimage from the microscopic image. The digitalimage was recorded in 8-bit monochrome with7686576 pixel resolution and 256 grey levels(0~black, 255~white). Regions of the digitalimage which were positively labelled appearedwhite and the background black. Interactivethresholding of this monochrome image was thenperformed. This involved setting the grey valuesin order to de®ne the range of intensities of theobjects which were to be included in subsequentautomatic measurements. The upper grey level wasalways set at 255 and the lower value was setsubjectively according to the intensity of theimmunolabelling. All highlighted areas, represent-ing positive staining, were then automaticallycalculated. The total areas of both PGP 9.5- andSP-labelled tissue were determined for each ®eldand the percentage of PGP-labelled tissue that wasalso positively labelled for SP was calculated. Forbrevity, throughout the subsequent text the term``SP expression'' is used to describe the percentagearea of PGP 9.5-labelled tissue which was alsolabelled for SP.Image analysis has found increasing application

in quantitative histology and has enabled greaterobjectivity, accuracy and e�ciency than was pos-sible with previous manual counting techniquesor semi-quantitative assessment (28, 29). However,it should be appreciated that the data do notrepresent absolute measures of peptide expression,

as is the case with radioimmunoassays for example,but rather provide valid data for comparativeinvestigations.

Statistical analysis

A number of post-eruptive changes in intra-dental nerve density and distribution have beendescribed in human teeth (30, 31), thus it wasnecessary to ensure that age did not have any e�ecton neural density (PAS for PGP 9.5) or SPexpression in the present experimental material.Data for the PAS for PGP 9.5 and the PASof PGP 9.5-labelled tissue that was also labelledfor SP were plotted against age in each of thethree sampling regions, and Pearson correlationcoe�cients were determined for each association.

One-way analysis of variance was employedto test for statistically signi®cant di�erences inmean SP expression, according to the three cat-egories of caries. Further pair-wise comparisonswere undertaken using Tukey's honestly statist-ically signi®cant test. In addition, an independentsample t-test was used to test for any statisticallysigni®cant di�erence in mean SP expressionbetween grossly carious asymptomatic and report-edly painful specimens. All statistical analyseswere performed on logarithmically transformed(Log10) data, but the data are presented graph-ically in their raw form. The signi®cance levels wereset at P50.05.

Results

In total, 62 permanent tooth pulps were analysedand comprised 19 intact specimens, 22 moderatelycarious specimens and 21 grossly carious speci-mens. Of the grossly carious specimens, 11 hadbeen reportedly asymptomatic and 10 had beenpainful.

The mean age of patients from whom theexperimental material was obtained was 9.7 yr(¡2.02, range 6.1±14.1). There was no signi®cantdi�erence in the mean age of subjects or the male tofemale ratio within each of the caries subgroups.Furthermore, no signi®cant correlation was foundbetween age and neural density or SP expression inany of the three ®elds analysed. Thus, caries wasconsidered to have a real e�ect on the neuralvariables.

Qualitative analysis of SP expression

In intact specimens, a moderate number of SP-immunoreactive ®bres were seen throughout thecoronal pulp (Fig. 1A, C, E). In some cases,®ne SP-immunoreactive ®bres could be seen

Substance P expression in human teeth 469

extending into the odontoblast layer. Within thepulp horn and the subodontoblastic nerve plexus,SP-immunoreactive ®bres were predominantly vari-cose, butwithin the nerve trunks SP-immunoreactive®bres demonstrated two di�erent morphologies:varicose and smooth-surfaced. There was a closespatial relationship between some pulpal bloodvessels and SP-immunoreactive ®bres, which

appeared to form a varicose network around thevessel walls.

Overall, there was an increased SP expressionthroughout the coronal pulp of carious specimens,and this appeared to be due to a relative increasein the proportion of SP-expressing ®bres (Fig. 1B,D, F). This increase in the number of ®bresimmunoreactive for SP was particularly evident

A B

DC

E FFig. 1. Double-exposure photomicrographs showing labelling for SP within neural tissue labelled for PGP 9.5 in intact (A, C, E) andcarious (B, D, F) teeth. PGP 9.5-immunoreactivity is represented by the red staining and any SP-immunoreactivity occurring withinPGP 9.5 labelled tissue is represented by the yellow areas. A small amount of green staining (B, D, F) can also be seen, representingareas where SP-immunoreactivity occurs in tissue with very faint or absent PGP-labelling. (A) The pulp horn region of an intacttooth showing few SP-immunoreactive ®bres and (B) a greater proportion of SP-immunoreactive ®bres in a carious specimen.(C) The subodontoblastic region of an intact tooth showing minimal SP expression and (D) a marked increase in SP expression ina carious specimen. (E) A nerve trunk in the mid-coronal region of an intact tooth with few SP-immunoreactive ®bres and (F) anincrease in SP-immunoreactive ®bres within the nerve trunk of a carious specimen. Scale bar~30 mm.

470 Rodd & Boissonade

within the nerve trunks, where there was no con-comitant increase in overall neural density. In a fewof the grossly carious specimens with a positive painhistory, very dense areas of SP-immunoreactivetissue were seen in the pulp horn region, andthere was a dramatic increase in the expression ofSP within all nerve trunks.

Quantitative analysis of SP expression

Fig. 2 shows the mean (¡ SEM) SP expressionfor each sampling region according to the degreeof caries. It can be seen that SP expression wasgreatly increased in grossly carious teeth in allthree regions. Statistical analysis con®rmed thatcaries had a highly signi®cant e�ect on SPexpression within the pulp horn region (P50.001,ANOVA) and also had a signi®cant e�ect withinthe subodontoblastic nerve plexus and mid-coronalregion (P50.05, ANOVA). Further analysisrevealed that, in the pulp horn and the mid-coronalpulp region, SP expression was signi®cantly higherin grossly carious specimens than in both intact andmoderately carious specimens (P50.05, Tukey'stest). However, within the subodontoblastic nerveplexus, SP expression was only signi®cantly di�er-ent between grossly carious and intact specimens(P50.05, Tukey's test).There was a signi®cantly greater SP expression

in reportedly painful teeth (n~10) as compared toasymptomatic teeth (n~11), and this was true forall three sample regions (P50.05, independentsample t-test). These data are presented graphicallyin Fig. 3.

Discussion

Our descriptive ®ndings for the anatomical dis-tribution of SP-immunoreactive nerve ®bres withinintact human teeth are consistent with those fromprevious investigations (8, 10, 11). It is di�cult tocorroborate our observations for the distributionof SP-immunoreactive ®bres within carious teethas there are little pre-existing data. However, ouranatomical ®ndings for SP-immunoreactive ®bresin carious teeth are comparable with thoseseen in experimental models of rat pulpal in¯am-mation (16, 17). In essence, it is apparent thatSP-immunoreactive ®bres are not static during thein¯ammatory process but undergo dynamic struc-tural changes, arborising extensively and formingdense ®bre networks.In addition to the tooth pulp, several other

models of in¯ammation have been developed toassess associated morphological and cytochemicalneural changes. These studies, mostly involving

Fig. 2. Bar graphs showing mean (¡SEM) percentage area ofprotein gene product 9.5-immunoreactive tissue which was alsolabelled for substance P (SP) within di�erent sample regionsand according to the degree of caries. 1* indicates signi®cantdi�erence between grossly carious and intact specimens2* indicates signi®cant di�erence between grossly carious andmoderately carious specimens (P50.05, Tukey's test for pair-wise comparisons).

Substance P expression in human teeth 471

experimentally-induced arthritis, have conclusivelyshown that during chronic in¯ammation there isa local increase of neuropeptides, including SP,within the sensory nerves innervating the a�ectedtissues (32±34). Of particular clinical relevance isthe ®nding that SP expression is increased withinthe crevicular ¯uid of patients with gingival andperiodontal disease (35). It would therefore appearthat the overall increase in SP expression found incarious teeth by the present study is consistent withpeptide increases described in other models ofin¯ammation.

The regulatory mechanisms underlying thisapparent upregulation of SP expression are likelyto involve nerve growth factor, as this is greatlyincreased in in¯amed pulp tissue (36, 37) and isknown to promote the neural synthesis of SP (38).However, it is not known whether the neuralincreases in SP, seen in the present caries model,are due to an enhanced biosynthesis by neuronsthat normally express the peptide (32), or resultfrom a de novo synthesis of SP by nervesubpopulations that do not normally express thepeptide (34). Increases in SP expression arisingsimply as a result of increased axonal transport (39)or reduced peripheral release (40) seem less likely.

Numerous investigators have attempted, andfailed, to ®nd any correlation between reportedpatient symptoms and the histopathological statusof pulpal innervation (41±43). Recent studies in thislaboratory have revealed a signi®cant increase

in pulpal neural density with caries progression,however no correlation was found between reportedpain experience and overall neural density (unpub-lished data). Thus, it was very interesting that, ata biochemical level, we were able detect a signi-®cant di�erence in the expression of a speci®cneuropeptide between painful and asymptomaticgrossly carious teeth. There is now a growingconsensus, based on several lines of experimentalevidence, that the upregulation of SP in injuredtissues may be associated with the development ofhyperalgesia (44). Thus, our ®nding that increasedSP expression correlated with a positive dental painhistory would lend further support to the view thatSP has an important role in in¯ammatory painmechanisms. However, some con¯icting evidencefor the role of SP in the development of hyper-algesia stems from the recent ®nding that tachykininknockout mice appear to maintain lowered thermaland mechanical withdrawal re¯exes following tissueinjury (45).

It would seem reasonable to hypothesise that,during pulpal in¯ammation, there is a concom-itant increased synthesis and peripheral release ofSP from primary a�erent intradental neurons. Itis considered unlikely that SP directly sensitisesnociceptive a�erents, but it may exert an indirecte�ect via its numerous vasodilatory and pro-in¯ammatory interactions (45). Thus, peripheralsensitisation is likely to contribute to the hyper-algesia which is often clinically associated with

Fig. 3. Bar graph showing mean (¡SEM) percentage area of protein gene product 9.5-immunoreactive tissue which was also labelledfor substance P (SP) for grossly carious specimens within di�erent sample regions and according to pain history. * indicatessigni®cant di�erence between asymptomatic and painful specimens (P50.05, independent sample t-test).

472 Rodd & Boissonade

pulpal in¯ammation. In addition, central changesaccompanying peripheral in¯ammation are alsolikely to have important signi®cance in termsof pain experience. An obvious limitation ofthe present study is that information regardingin¯ammation-related changes in SP immuno-reactivity is limited to the peripheral nervoussystem. Nonetheless, numerous studies in experi-mental animal models have shown that, followingperipheral in¯ammation, there is also an increasein SP expression in the central terminals, wherenociceptive transmission takes place (32, 46, 47, 48).Thus, it is speculated that the peripheral changesin SP expression seen within intradental nervesmay also re¯ect equivalent changes in the centralterminals.Having established a positive correlation between

dental pain experience and SP expression in pulpalnerves, it is tempting to speculate that the regulationof intradental SP secretion may provide a newtherapeutic approach for the management of dentalpain and in¯ammation. Whilst this may be a validpremise, it is important to recognise that changes inSP expression represent just one of a host ofprofound in¯ammation-induced changes in neuralcytochemistry (49). Therefore, it is highly likely thatother neural changes may also have importantrelevance in terms of altered sensory nerve excit-ability following tissue injury and in¯ammationwithin di�erent systems. Nonetheless, continuedresearch in this exciting ®eld may lead to a greaterinsight into the complex nature of dental pain andmay direct the development of novel analgesicstrategies.

Acknowledgement ± This investigation was supported by aclinical training fellowship grant awarded by The MedicalResearch Council (UK).

References

1. STRAND F. Neuropeptides: regulators of physiologicalprocesses. Cambridge, MA: MIT Press, 1999; 365±381.

2. LEMBECK F, HOLZER P. Substance P as neurogenic medi-ator of antidromic vasodilation and neurogenic plasmaextravasation. Arch Pharmacol 1979; 310: 175±183.

3. MCGILLIS JP, ORGANIST ML, PAYAN DG. Substance Pand immunoregulation. Fed Proc 1987; 46: 196±199.

4. NILSSON J, VON EULER AM, DALSGAARD C. Stimulation ofconnective tissue cell growth by substance P and substanceK. Nature 1985; 315: 61±63.

5. FLEETWOOD-WALKER SM. Nonopioid mediators and modu-lators of nociceptive processing in the spinal dorsal horn astargets for novel analgesics. Pain Rev 1995; 2: 153±173.

6. OLGART L, GAZELIUS B, BRODIN E, NILSSON G. Release ofsubstance P-like immunoreactivity from the dental pulp.Acta Physiol Scand 1977; 101: 510±512.

7. GROÈ NBLAD M, LIESI P, MUNCK AM. Peptidergic nerves inhuman tooth pulp. Scand J Dent Res 1984; 92: 319±324.

8. WAKISAKA S, ICHIKAWA H, NISHIMOTO T, MATSUO S,YAMAMOTO K, NAKATA T, AKAI M. Substance P-likeimmunoreactivity in the pulp-dentine zone of humanmolar teeth demonstrated by indirect immuno¯uorescence.Arch Oral Biol 1984; 29: 73±75.

9. WAKISAKA S, NISHIKAWA S, ICHIKAWA H, MATSUO S,TAKANO Y, AKAI M. The distribution and origin ofsubstance P-like immunoreactivity in the rat molar pulpand periodontal tissues. Arch Oral Biol 1985; 30: 813±818.

10. CASASCO A, CALLIGARO A, CASASCO M, SPRINGALL DR,POLAK JM, POGGI P, MARCHETTI C. Peptidergic nerves inhuman dental pulp. Histochemistry 1990; 95: 115±121.

11. LUTHMAN J, LUTHMAN D, HOÈ KFELT T. Occurrence anddistribution of di�erent neurochemical markers in thehuman dental pulp. Arch Oral Biol 1992; 37: 193±208.

12. HEYERAAS KJ, KVINNSLAND I, BYERS MR, JACOBSEN EB.Nerve ®bers immunoreactive to protein gene product 9.5,calcitonin gene-related peptide, substance P, and neuropep-tide Y in the dental pulp, periodontal ligament, and gingivain cats. Acta Odontol Scand 1993; 51: 207±221.

13. TUISKU F, HILDEBRAND C. Combined retrograde tracingand immunohistochemistry of trigeminal ganglion neuronsprojecting to gingiva or tooth pulps in the lower jaw of thecichlid Tilapia mariae. J Neurocytol 1997; 26: 33±40.

14. JACOBSEN EB, HEYERAAS KJ. E�ect of capsaicin treatmentor inferior alveolar nerve resection on dentine formationand calcitonin gene-related peptide- and substance P-immunoreactive nerve ®bres in rat molar pulp. Arch OralBiol 1996; 41: 1121±1131.

15. GRUTZNER EH, GARRY MG, HARGREAVES KM. E�ect ofinjury on pulpal levels of immunoreactive substance P andimmunoreactive calcitonin gene-related peptide. J Endodon1992; 18: 553±557.

16. SWIFT ML, BYERS MR. E�ect of ageing on responses ofnerve ®bres to pulpal in¯ammation in rat molars analysedby quantitative immunocytochemistry. Arch Oral Biol 1992;11: 901±912.

17. HONG D, BYERS MR, OSWALD RJ. Dexamethasone treat-ment reduces sensory neuropeptides and nerve sproutingreactions in injured teeth. Pain 1993; 55: 171±181.

18. HARGREAVES KM, SWIFT JQ, ROSZKOWSKI MT, BOWLES W,GARRY MG, JACKSON DL. Pharmacology of peripheralneuropeptide and in¯ammatory mediator release. Oral SurgOral Med Oral Pathol 1994; 78: 503±510.

19. ROSELL S, OLGART L, GAZELIUS B, PANOPOULOS P, FOLKERS

K, HOÈ RIG J. Inhibition of antidromic and substanceP-induced vasodilation by a substance P antagonist. ActaPhysiol Scand 1981; 111: 381±382.

20. KEREZOUDIS NP, OLGART L, EDWALL L. Involvement ofsubstance P but not nitric oxide or CGRP on neurogenicplasma extravasation in rat incisor pulp and lip. Arch OralBiol 1994; 39: 769±774.

21. RODD HD, LOESCHER AR, BOISSONADE FM. Immunocyto-chemical and electron-microscopic features of tooth pulpinnervation in hereditary sensory and autonomic neuro-pathy. Arch Oral Biol 1998; 43: 445±454.

22. VANDEVSKA-RADUNOVIC V. Neural modulation of in¯am-matory reactions in dental tissues incident to orthodontictooth movement. A review of the literature. Eur J Orthod1999; 21: 231±247.

23. TROWBRIDGE HO. Review of dental pain ± histology andphysiology. J Endodon 1986; 12: 445±452.

24. IZUMI T, KOBAYASHI I, OKAMURA K, SAKAI H. Immuno-whistochemical study on the immunocompetent cells of thepulp in human non-carious and carious teeth. Arch OralBiol 1995; 40: 609±614.

25. SAKURAI K, OKIJI T, SUDA H. Co-increase of nerve ®bresand HLA-DR- and/or factor-XIIIa-expressing dentritic

Substance P expression in human teeth 473

cells in dentinal caries-a�ected regions of the human dentalpulp: an immunohistochemical study. J Dent Res 1999; 78:1596±1608.

26. RODE J, DHILLON AP, DORAN JF, JACKSON P, THOMPSON RJ.PGP 9.5, a new marker for human neuroendocrinetumours. Histopathology 1985; 9: 147±158.

27. FRISTAD I, HEYERAAS KJ, KVINNSLAND I. Nerve ®bres andcells immunoreactive to neurochemical markers in devel-oping rat molars and supporting tissues. Arch Oral Biol1994; 39: 633±646.

28. JAGOE R, STEEL JH, VUCICEVIC V, ALEXANDER N, VAN

NOORDEN S, WOOTTON R, POLAK JM. Observer variationin quanti®cation of immunocytochemistry by image ana-lysis. Histochem J 1991; 23: 541±547.

29. SPRINGALL DR, GENTLEMAN SM, TERENGHI G, POLAK JM.Quantitative pathology. In: WOOTON R, SPRINGALL DR,POLAK JM, eds. Image analysis in histology. Cambridge:Cambridge University Press, 1995: 403±412.

30. BERNICK S. E�ect of aging on the nerve supply to humanteeth. J Dent Res 1964; 43: 406±411.

31. JOHNSEN DC, HARSHBARGER J, RYMER HD. Quantitativeassessment of neural development in human premolars.Anat Rec 1983; 205: 421±429.

32. DONALDSON LF, HARMAR AJ, MCQUEEN DS, SECKL JR.Increased expression of preprotachykinin, calcitonin gene-related peptide, but not vasoactive intestinal peptidemessenger RNA in dorsal root ganglia during the devel-opment of adjuvant monoarthritis in the rat. Mol Brain Res1992; 16: 143±149.

33. CARLESON J, ALSTERGREN P, APPELGREN A, APPELGREN B,KOPP S, SRINIVASAN GR, THEODORSSON E, LUNDEBERG T.E�ects of adjuvant on neuropeptide-like immunoreactivityin experimentally induced temporomandibular arthritis inrats. Arch Oral Biol 1996; 41: 705±712.

34. NEUMANN S, DOUBELL TP, LESLIE T, WOOLF CJ. In¯am-matory pain hypersensitivity mediated by phenotypicswitch in myelinated primary sensory neurons. Nature1996; 384: 360±364.

35. LINDEN GJ, MCKINNELL J, SHAW C, LUNDY FT. SubstanceP and neurokinin A in gingival crevicular ¯uid in perio-dontal health and disease. J Clin Periodontol 1997; 24:799±803.

36. BYERS MR, WHEELER EF, BOTHWELL M. Altered expressionof NGF and P75 NGF-receptor by ®broblasts of injuredteeth precedes sensory nerve sprouting. Growth Factors1992; 6: 41±52.

37. WHEELER EF, NAFTEL JP, PAN M, VON BARTHELD CS, BYERS

MR. Neurotrophin receptor expression is induced in a sub-population of trigeminal neurons that label by retrograde

transport of NGF or Fluoro-gold following tooth injury.Mol Brain Res 1998; 6: 23±38.

38. LINDSAY RM, LOCKETT C, STERNBERG J, WINTER J.Neuropeptide expression in cultures of adult sensoryneurons: modulation of substance P and calcitonin gene-related peptide levels by nerve growth factor. Neuroscience1989; 33: 53±65.

39. DONNERER J, SCHULIGOI R, STEIN C, AMANN R. Upregula-tion, release and axonal transport of substance P andcalcitonin gene-related peptide in adjuvant in¯ammationand regulatory function of nerve growth factor. Regul Pept1993; 46: 150±154.

40. HARGREAVES KM, SWIFT JQ, ROSZKOWSKI MT, BOWLES W,GARRY MG, JACKSON DL. Pharmacology of peripheralneuropeptide and in¯ammatory mediator release. Oral SurgOral Med Oral Pathol 1994; 78: 503±510.

41. ENGLAND MC, PELLIS EG, MICHANOWICZ AE. Histopatho-logical study of the e�ect of pulpal disease upon nerve ®bresof the human dental pulp. Oral Surg Oral Med Oral Pathol1974; 38: 783±790.

42. TORNECK CD. Changes in the ®ne structure of the humandental pulp subsequent to carious exposure. J Oral Pathol1977; 6: 82±95.

43. MENDOZA MM, READER A, MEYERS WJ, FOREMAN DW.An ultrastructural investigation of the human apical pulpin irreversible pulpitis. 1. Nerves. J Endodon 1987; 13:267±276.

44. LEVINE JD, REICHLING DB. Peripheral mechanisms ofin¯ammatory pain. In: MELZACK R, ed. Textbook of pain,4th edition. London: Churchill Livingstone, 1999; 59±84.

45. BASBAUM AI. Distinct neurochemical features of acute andpersistant pain. Proc Natl Acad Sci USA 1999; 96:7739±7743.

46. SMITH GD, HAMAR AJ, MCQUEEN DS, SECKL JR. Increasein substance P and CGRP, but not somatostatin content ofinnervating dorsal root ganglia in adjuvant monoarthritis inthe rat. Neurosci Lett 1992; 137: 257±260.

47. POLGAÂ R E, SZUÃ CS P, URBAÂ N L, NAGY I. Alterations ofsubstance P immunoreactivity in lumbar and thoracicsegments of rat spinal cord in ultraviolet irradiationinduced hyperalgesia of the hindpaw. Brain Res 1998;786: 248±251.

48. HUTCHINS B, SPEARS R, HINTON RJ, HARPER RP. Calcitoningene-related peptide and substance P immunoreactivity inrat trigeminal ganglia and brainstem following adjuvant-induced in¯ammtion of the temperomandibular joint. ArchOral Biol 2000; 45: 335±345.

49. BESSON JM. The neurobiology of pain. Lancet 1999; 353:1610±1615.

474 Rodd & Boissonade