Inflammatory mediators potentiate pain induced by experimental tissue acidosis

Pain, 66 (1996)163-170 @ 1996International Associationfor the Study of Pain. All rights reserved0304-3959/96/$15.00 PAIN 3034

Inflammatory mediators potentate pain induced by experimental tissue acidosis Kay H. Steen*a, Astrid E. Steena, Hans-Wilhelm Kreysela and Peter W. Reehb aUniversittits-Hautkliniktad Poliklinik &r Universit5t Bonn, Dermatophysiologie, Sigmund-Freu&Str. 25, D-53105 Bonn, (Germany) and brmtitutfi Physio[ogie ~ Exwrimentelle patir@ysi&@e, Universittt Erlangen-Nrirnberg,Universifbtsstra$e IT. D-91054 Erlangen, (Germany)

(Received19June 1995;revised8 Novemlxr 1995;accepted 14November1995)

Surnnuuy Electrophysiologicalevidencefrom cutaneousnociceptorssuggesteda synergismbetweenexcitatoryactions of inflammatorymediators(IM) and low pH. In human skin it is possibleto induceconstant ongoingpain with continuous infusion of acid buffer. This method was used to study the interaction with mediators of inflammation psychophysiologically.A skin area on the palmar forearm of 6 subjects(either gender, age 22–35 years) was continuouslyinfiltratedwith a phosphate bufferedelectrolytesolution (pH 5.2)usinga motorizedsyringepump that was adjusted so as to produce constant pain of about 2W!!0on a visual analog scale (VAS; extending from ‘no’ to ‘unbearablepain’).Pain was assessedon the VAS at lo-seeintervals;the rating was calledup by means of an acoustic signal.An additionalcannulawas placedin the skin beforethe infusionof acidicbuffer started. Injectionsof an acidic combinationof IM (BK, 5-HT, HIS, PGE2)0.2 ml were then given through the cannula at intervalsof 10 min in a randomizeddoubleblind order of concentrations.The other arm was used for negativecontrol, i.e. IM in neutral solution were injectedinto normal skin continuouslyinfiltrated with a buffer solution at pH 7.4. The IM induced dosedependent,transient burning pain on both arms — markedly more intenseand prolonged, however,in the acidotic skin (P e 0.004, U-test).The differencecorresponded to a 10-foldincreasein algogenicpotency with 10-7M IM, being smallerwith 10-6and 10-5M concentration.The interaction betweenlow pH and IM was mutual: additional injectionsof plain phosphatebuffer(pH 5.2)into the acidoticskinweresignificantlymore painful(20-fold)after application of IM than under control conditions. Thus, we tend to conclude that it is the inflammatorymediators that potentate the algogeniceffectof lowpH ratherthanviceversa.Tissueacidosisappearsas a dominantfactor in inflammatory pain. Key words: Cutaneous; Hydrogen ions; Acid pH; Sensory;Bradykinin;Serotonin;Psychophysics

Introduction

In intlarned states of the skin, numerous chemicals that mediate and facilitate the inflammatory process have been identified.These mediators of inflammation include(among others) bradykinin (BK), serotonin (5HT), histamine (HIS) and prostaglandins (PGE) but also an increaseof the hydrogenion concentration.RePriv.-Doz.Dr med.Kay H. Steen, Universities-Hautklinik und Poliklinikder Universitat Bonn,KlinischeDerrnatophysiologie, Sigmund-Freud-Str. 25, D-53105Bonn,Germany.Tel.: (49)228 2876969or 5518;Fax:(49)2282874333.

“ Corresponding atdtor:

PII 0304-3959(96)03034-5

search into inflammatorypain investigatedthe role of singleand combinedhormonal mediatorsin the inflammatory environment and possible interactions of one mediatorwith another. As a resultof this, the excitatory actionsof BK and 5-HT on cutaneousnociceptorswere established (Beck and Handwerker 1974; Fock and Mense 1976;Lang et al. 1990;Mizumura et al. 1990; Khan et al. 1992),and it turned out that 5-HT can strongly enhance the response of nociceptors to BK (Fjallbrandand Iggo 1961;Hiss and Mense 1976;Lang et al. 1990),Also prostaglandins, though ineffectivein nociceptor activation on their own (Mizumura et al. 1987;Lang et al. 1990;Rueff and Dray 1993;Belmonte et al. 1994),can contributeto sensitizingnociceptorsfor

164

BK in some tissue (Mizumuraet al. 1987;Schaibleand Schmidt 1988;Neugebauer et al. 1989).For almost 70 years it has been known that low pH (down to 4.7) is a regular finding in inflamed tissues (Habler 1929)and that acidic solutions are particularly painful when injected into the skin (von ‘Gazaand-B~andi 1926).The generalresearchactivity,however,has not been focused on the role and interactionsof protons in inflammation and especiallyin nociceptor excitation and pain. Continuous administration of low pH buffered solutions into human skin causes dose-dependentsustained pain and hyperalgesiato mechanicalstimulation(Steen and Reeh 1993).This correlates with observationsthat protons selectivelyactivate nociceptors in rat skin in vitro and produce a sensitization of nociceptors to mechanicalstimuli(Steenet al. 1992).In the latter skinsaphenous nerve preparation in vitro, a potent synergism between nociceptiveactions of a combination of inflammatory mediators (IM: BK, 5-HT, HIS, PGE2) and lowpH has recentlybeen demonstrated(Steenet al. 1995b).In the investigationpresented here, the method of inducingconstant ongoingpain with continuousinfusion of acid btiers was used to study psychophysiologicallythe interaction of acid pH with mediators of inflammation injected into the acidotic skin. To this purpose,the dose-responsecurvefor the algogeniceffect of different IM concentrationswas determined in normal and in acidotic skin. Furthermore, it was asked whether the interaction of protons and IM, both being algogenic,would be reciprocal or whether the low pH would play a dominant role in causing pain, as it appeared from rat skin nociceptors(Steen et al. 1995b). A preliminary account of the tindings of the study presentedhere has been reported previously(Steenet al. 1993). Materislsandmethods The interactions of hormonal mediators (IM) with low tissue pH werestudiedpsychophysiologically byusinga methodwhichproduces localizedsustainedgradedpain fromexperimentaltissueacidosis.The latter is achievedby continuousintradermal infusionof a phosphate butTeredsolution(pH 5.2)into the pahnar forearmskinof humansubjects. The pain can be adjusted by altering the speed of a motorized syringepump to any deliberateleveland remainsconstant at a given sp@. The methodhas been previously described in detail and validated(Steen and Reeh 1993). Subjects and study conditions

In the present study, recordingsfrom 6 healthy volunteers,three males and three femaleswere made on different occasions.The subjects werebetween25and 35yearsof age. Each subjectgavehis or her written informedconsent to the proceduresused in the experiments whichwere approved by the Universityof Bonn Ethical Committee and wereperformedin accordancewith the ‘Declarationof Helsinki’. Method

The subjectswerecomfortablyseated in a dentist’schair. The right or left arm was randomly used for the studies and placed on an

upholsteredarmrest.Two27G-cannrdawereinsertedintracutaneous.

lyinto the pahnar forearm about 0.5 cm apart. The first cannrda was

connectedto a pressureinfusionsystemof phosphate buffered solution (pH 5.2),joined upstream by a sterile filter. The pressure of the irdusionwas providedby a motorized syrinm DUD which was adjustable to va~ous flow }ates (1-10 n@ xl, ‘x IOj. Recordingpain sensation, &ta processing and graphical presentation

Painwasassessedon a visualanalogscale(VAS;extendingfrom ‘no pain’to ‘unbearablepain’)at 1O-WC intervals.Thesubjectswereasked, by meansof an acousticbeep, to estimate their current pain intensity by movinga lever that controlleda horizontallydisplayedelectronic VASconsistingof 90rectangularLEDsin an unbrokenline.The lever was movedback to zero after each rating by the subjects.The rating was passedas an analog voltagesignalthrough an AD converter (12 bit AD-converter,2 Hz samplingrate), and for quantitative analysis, data werestoredin a 80386-typecomputerusinga DAP 1200interface card with built-in digital signal processor (Microstar Laboratories, Bellevue,WA).The recordingsoftwarewasa self-designeddata acquisition program (’get-analog’)which recorded the data in ASCIIforrnat. Data reduction(1:20)was performedby a filteringprogram (’get-peak’).The graphicalpresentationwas done off-lineby usingthe Sigmaplotsoftwarepackage(Jandel, Berkeley,CA). Test solutions used

The infmionsolutionwas made from modifiedsyntheticinterstitial fluid(SIF; Bretag 1969),in whichthe sodiumbicarbonate(26.2mM) normallycontainedin SIF wasreplacedby appropriateproportionsof NaH#04 and Na2HP04(SigmaChemicalCo.) to produce the bufferedpH levelsof pH 7.4(for control) or 5.2(for stimulation),respectively.This solution is referred to as phosphate buffered SIF in the followingtext. Beforethe experimentsthe pH of the solutions was finelyadjusted.MinorpH-differences(+0.05) to the valuedetermined by the butTercompositionwere titrated by adding HC1 or NaOH, respectively. The combinationof IM was made up of prostaglandin ~, histaminetriacetate,serotoninhydrochlorideand bradykinintriawtate (all chemicalswereprovidedby SIGMA)whichweredissolvedin the acidic phosphatebufferedSIF at the concentrations 10-7,10-6and 10-5 M for each constituent.Atl solutionswere made immediatelybefore applicationand were applied at room temperature. Experimentalprotocol.rfor adrnirristrationof chemicals

Duringa preliminaryrun of 5-10 rein, the flowrate of the syringe pump was adjusted so as to result in individualpain estimations of about 2(E4on the visualanalog scrde,whichwas the baselineand the backgroundcondition for the followingrecording period of 50 ruin duration. Duringthis we appliedfour intradermalbolus injectionsof 0.2ml of the acidicvehicleand of the combinationof IM (BK, 5-HT, HIS, PGE2 at pH 5.2) through the second cannrda which had previouslybeenplacedin the skin beforethe acidificationwas started. The injectionswere givento the individualsat intervals of 10tin in a randomizeddoubleblind order of concentrations(10-7 M, 10-6M, 10-5M and plain phosphate buffered SIF, pH 5.2). Experinrentalprotocols of the control sessions All experimentalseasionswerespacedby intervalsof at least 3 days betweencross-over,and their successionwas randomized. 1. In the samesubjects(n = 6), the other arm was used for negative control, i.e. IM in neutral solutionwere injectedin the same concentrations and amount into normal skin continuouslyinfiltrated with phosphate bufYeredsolution at pH 7.4 using the same flow rate as previouslyapplied. 2. The same subjects(n= 6) were asked to estimate their pain in a third sessionin whichacidifiedphosphatebutTeredsolution (pH 5.2) was infusedas baselineand three injectionsof 0.2 ml acid phosphate butTerwithout IM were made at intervals of 10ntin in order to estimate the pain evokedb this additionaldose of hydrogenions alone. 3. In one subject 10-J M IM wereinjectedfivetimesinto the acidified skin,the first four injectionsat intervalsof 10rein, the last injection spaced by an interval of 30 mitt (Fig. 5).

165 4. In another subject, lower IM concentrations(10-9 M, 1O-sM, 10-7M) wereinjectedinto the acidifiedskin (each 0.2 ml) and, tinally, an increw of pain was inducedby increasingthe flowrate in order to demonstratedose-dependency(Fig. 6). Statistical analysis

Pain ratingvaluesafter each bohrsinjectionwere summatedto a cumulativesumof consecutiveratingsduring 10ruinafter injectiononset comprising60 singleratings. This samplingperiod coveredeven the longest-lastingpain responses(e.g. to 10-5M IM into acidotic skin). To account for the pH-induced backgroundpain in tbe interaction study, the cumulativesum of the individualrating values during 10 min of acid buffer infusion before any IM injection was subtracted from the responseto IM injection.Thesecorrected cumulativesums, relatedto one particular IM concentration,wereaveragedacross subjects (n= 6) and used for statistical processing.The latter is referred to as ‘meancumulativesum of rating values’throughout the text; it representsthe ‘area under the curve’of the averagedpain responses (shownin Figs. 2 and 4A). SPSSSoftware(Munich, Germany)was employedfor statisticalanalysisof the data. Non-parametricstatistics had to be applied becauseof the small but sti]cient number of subjects that wouldnot allowto reasonablytest normalityof the data distribution. The Wilcoxonmatched pairs test was employedto analyze the dose-responsecurves of tbe IM. For all other statistical comparisons, the Marm-Whitney U-testwas usedas indicated.In general, P-valueslower than 0.05were regarded as significant.

Results Sustained acid pH stimulation as baselinefor the interaction study

Flow rates of phosphate buffered SIF (pH 5.2) between 15 mlh and 30 mUh(20 mlh, on average) were needed to achieveconstant individualpain estimations of around 20% on the visual analog scale (VAS). The subjects reported a burning pain sensation which was localizedaround the injection site and sustained (Figs. 1A, 5, 6) during the recordingperiod of about 1 h. The injectionsite swelledup with the fluid infiltration and was surroundedby a circumscriptflamingred flare reaction with a pale center. This has previously been reported in detail (Steen and Reeh 1993).

Inflammatory mediators induce cutaneouspain sensation (control 1)

When 0.2 ml of IM (BK, HIS, 5-HT, PGEZ;pH 7.4) were injected into the neutral phosphate-buffer infiltratedskin area, dose-dependentpain sensationswere induced(Fig. IB). With a concentration of 10-7M IM, a cumulativesumof rating values(seeMethods)of Z 64, on average,was achievedin the 6 subjects,and the pain lastedfor 0-213 sec(Z = 88 see);with 10-6M the mean cumulative sum was E 270 and pain ratings were deliveredfor 20-228 see@ = 128see),and with 10-5M the mean cumulative sum was )2 1558, and the pain duration was further prolonged with 156-693 sec (X= 304 see). Fig. IB shows a specimen of a typical experimentalprotocoland Fig. 2 presentsaveragedpain responses. A linear dose-responserelationship of the algogeniceffects of the different IM concentrations is evident (Fig. 3). The mean cumulative sums of pain ratings were significantly different from each other (Wilcoxontest: 10-7/10-6M: P <0.03 and 10-6/10-5 M: P < 0.03).For comparison, the acidic vehicleused in the interactionstudywasalso injectedinto the neutral infiltrated skin; the pain induced by the single 0.2-ml acid-bufferinjectionswithout IM was small and brief, the cumulativesum of ratingsbeing E 58 and lasting for 0-80 sec (X=22 see) (Fig. IB). Inflammatory mediators induceenhancedpain in acidotic tissue

Fig. 1A demonstratesthat the injectionsof 0.2 ml of IM (pH 5.2)into the acidotic skin produced much larger and prolonged burning pain.sensationsthan in neutral skirt(Fig. IB); in particular, the algogeniceffectsof low concentrationsof IM wereconsiderablyenhanced,while with the highestconcentrationof IM, a ‘ceiling’effectof

100 A%

~

APH (5.2 - 7.4)

10-7M

60 40

104M

10-6M

5

5

60 40

20 0[ o’ IIM]

1=6

V;.4

100 -A, VA:O%

O 30 40 50 0 10 20 30 40 min flo+ tIO-+04M {1O+ I1O+’M

ml.inn Hr.PB,pH5.2,15 mUh .X Injw,. PBpH 5.2

rent. !nfiMr.PB, pH7.4,43m!Ih .=,n~,,

F’B@i 5.2

Fig. 1. Raw records of two experimentalprotocolswith one subject: A, test session.B,control session.Test and control sessionsdifferonly in respect to the pH of the continuously,intraderrndy infusedsolution (PB= phosphate bdered SIF). Vertical bars represent ratings (every10see)of burningpairrsensationon the visualanalogscale.The arrows point to bolus injections(0.2 ml) of IM or of phosphate buffered SIF (*).

20 0

—— 0

5

10 0

10

min 10

Fig.2. Averageof pain responsesto IM injectionsat differentconcentrationsinto normal(lowertrace) and into acidoticskin (upper trace), respztively. For better comparison,the constant backgroundpain of 20%VASwassubtractedfromthe data at pH 5.2.The pain ratingsare considerablylarger in acidotic skin, the differences(shaded areas) beingmore pronouncedat lower concentrationsof IM.

166

z%

VAI 1000

100

L SEM Y in PH 5.2 o in pH 7.4

?/

p< IIM]

Y .004

.004

.004

10-7

10$

10-5 M

Fig. 3. Meancumulativesumsof pain ratings 10rninafter bolusinjection of differentIM concentrationsinto either neutral or acidoticskin. The data from acidotic skin are corrected for the individual backgroundpain (-20%) by subtraction.All differencesbetweensessions were significant(see P-values from Mann-Whitney U-test) as well as all differences between concentrations (Wilcoxonmatched pairs test).

the VAS becameapparent and the enhancingeffectwas mainlya prolongationof the pain. This enhancementis not due to a superimpositionof the acute pain response to IM upon the elevatedbaselineof ongoingpain from the acid buffer infusion, since it is also evident in the followingnumericalvaluesand Figs.2 and 3 whichpresent corrected data (see Methods).With a concentration of 10-7M IM, a cumulativesum of rating values of E 335,on average,was achievedin the 6 subjectsand the ratings remained above baseline for about 70-360 sec (3= 215 see); with 10-6 M the mean cumulative sum was E 668 lasting 100-510 sec (X= 267 see) and with 10-5M the cumulativesum was Z 2386,on average,at a duration of 270-810 sec (X= 460 see). Fig. 2 gives averaged data (n= 6). The differences between the responsesin neutral as comparedto acidoticskin are obvious in magnitude and duration of the pain induced. The dose-responsecurve (Fig. 3) of IM is still linear in acidoticskinbut the VASratingsare significantlyhigher in comparisonto the IM effectsin neutral skin (U-test: P <0.004 at all concentrations).IM at 10-7M induce more pain in acidoticskin than the 10timeshigherconcentration in neutral skin;the gain in algogenicpotency is less but still significantat higher IM concentrations. Also in the acidotic milieu, the algogeniceffectsof the differentIM concentrationswere significantlydifferent from each other (Wilcoxontest: 10-7/10-6M: P <0.05 and 10-6/10-5M: P < 0.03) (Fig. 3). Inj7ammatorymediators sensitize nociceptorsfor 1owpHstimulation

one injectionof plain phosphate btiered SIF (pH 5.2) wasmade into the acidoticskin after the first, secondor third IM applicationat 10-minintervals(Fig. 1). Thus, 1-3 IM injectionsof differentconcentrationspreceded this control injection. In spite of this variable background,the pH-bolusinjectionsprovoked unexpectedly high pain responses(the mean cumulative sum was Z 1204)and a prolonged response duration (50–470see) (X= 312 see) which appears even more marked when pain responses are averaged (Fig. 4). This led us to assumethat the various IM injectionshad left a sensitizationto pH stimulationin the skin.Indeed, the average effect of the plain acidic injectionsinto IM pretreated skin was signMcantlylarger (20-foldhigher cumulative rating sum) as compared to repeated bolus injections into acidoticskin without IM pretreatment, which were made in a separate session(control 2, see below). This sensitizationdid not prevent the pain ratings after IM injectionfrom returning to the background level— although it delayedthe recovery(seeabove)— but it obviouslyincreasedthe sensitivityto abrupt pH changesas induced by the bolus injections. Control experiments Acid bolus injection in acidotic skin (control 2)

Because of the unexpectedly high response to the bolusinjectionof plain phosphate buffer during the IM experiments,we undertook a further series of experimentswith our subjectsand again induced tissueacidosis as described above. Then, we made three bolus injectionsof plain phosphate butTerpH 5.2 and found much smallerpH-responses(the mean cumulativesum, n = 18, was E 60) and shorter response durations (lasting 1-287 see, Z = 47 see) than in the IM sessions (Fig. 4). The low pH-bolus injectionsinto neutral skin (see control 1) were not significantly different in algogenic effect from the bolus injections into the acidoticskin if the background pain (- 2(WOVAS) was subtracted from the latter data. However, the pain responselastedclearlylonger,on average,in acidoticthan ,“”

I

o/, A VAS SO n=6

::

01 The IM solutionshad to be dissolvedin acidicphos-

phate buffer-SIF becausewhen IM in neutral buffered solutionwere injectedinto acidotic skin, the first effect was a substantialrelieffrom the constant ongoingpain due to acute butTeringof the tissueacidosis.To control for the effect of the additional dose of hydrogen ions,

2500.

.rnn .

I bolusin-

~fd:=6

/

&M::#control 1

IOmin ‘

.,

01

●= after IM

~p <0.004i

Fig. 4. (A) Averagedpain responsesto,0.2 ml low pH injections(pH 5.2) 10min after one to three previousapplicationsof IM in different concentrationswereappliedas comparedto control responseswithout pretreatmentof the skin with IM. In both series of experimentsthe skinwas acidotic.(B)Cumrdativesumsof pain ratings in responseto lowpH injectionsfrom6 individualsubjects(Mann-Whitney U-test).

167

appeared as the lowest effectiveconcentration (Fig. 6). Thisrecordingalsoconfirmsthat no habituation is to be observedwith continuousacid buffer infusionand that increasingthe flow-rate can enhance the pain to any magnitude at any time. In this particular subject, the syringepump was stopped as the subject complained that the increasingpain would soon becomeunbearable (Fig. 6).

100

V@os 60 sl40

0 0

10

20

30

t

f

f

cont.

40

50

60

70

min

t 0,2mllM10*MI

Fig. 5. Repeated in@ions of IM demonstrate tschyphylaxisof the slgogetticsotionand its recoveryafter a prolongedintervalof 30rrtin; PB= phosphate butTeredSIF.

in neutral skin, probably due to reducedbuffercapacity of the tissue (data not shown). Repeated injections of IM in acidotic skin (control 3)

The algogenicaction of hormonalmediatorsnormally underliesprofound tachyphylaxisupon repeated application. This phenomenonwas also shown in this study when in one subject, 10-6 IM were injected four times at intervalsof 10min duringcontinuousinfusionof acid buffer.The pain rating continuouslydecreasedfrom the first to the fourth injection,but clearlyrecoveredto full magnitudeafter a 30-minintervalbeforea fifth and final IM injection(Fig. 5). Determination of threshold concentrations (control 4)

In pilot experiments,injections of 10-8 M IM into normal skin never produced sensations different from control injections.The large average pain responsesto 10-7 M IM in acidotic skin, however, suggestedthat lowerconcentrationsought to be tried again in order to see a potential decreaseof the threshold to IM stimulation. In one subject, 10-9M and 10-8M IM did not induce more pain than was to be expected-from the additional dose of hydrogen ions. Thus, 10-7 M still 100 .

Vs 4 80

UPto 70

rnllh

6040 - IIM] 1O-s 10+

10-7 M

20mllh

20 0 0

Discussion

infiltration,PB, PH 5.2, 15 mllh

10

20

30

40

50

min

Fig. 6. One subject’spain intensityratingsvia VASduringinfusionof lowpH buffer(pH 5.2)into the forearmskin and additionalIM injections at low concentrations.The infusion rate of the buffer was adjusted so as to produce 2(F%VASratings during the experiment,but at the end, the infusionrate was increasedwhichresultedin a drastic increasein pain intensity.

A mutual interaction of experimentaltissue acidosis with an ample combinationof inflammatorymediators (IM) in pain induction has been demonstrated in this study.The potentiatingeffectof the IM onto the action of lowpH wasoutlastingthe algogenicactions.The role of IM appears predominantlyto be in sensitizationto the effectsof low pH. Thus the acidosisin inflamedtissueseemsto be the decisivefactor in ongoingnociceptor excitationand inflammatorypain. Thepossible role of acidpH in inflammatory pain

The methodusedin this studywas to inducesustained graded pain in human skin by intracutaneous infusion of an acidic phosphate buffered solution (Steen and Reeh 1993).The model is based on older literature reporting low local pH levels(down to pH 4.7) that have beenmeasuredin painfulinflammatorydiseases(Habler 1929)and also on recent electrophysiologicalfindings that pathophysiologicallyrelevanthydrogenion concentrations are able to selectivelyexcite and sensitizecutaneous nociceptors, in vitro, in a sustained manner (Steen et al. 1992). Previous psychophysiologicalattempts of inducingsustainedcutaneous pain by means of hydrogen ions had been inconclusive, since with blisterbase experimentsand injectionsof acid pH solutions only transient algogeniceffects of the hydrogen ionswereto be achievedin the skin (Lindahl 1961;Keele and Armstrong1964).The latter studieswereperformed in normal skin where the bufferingcapacity of the well perfusedtissueis high, whichis not likelyto be the case in pathophysiologicalstates (e.g. ischemia, inflammation). Thus, it appears justified to overcome the physiologicalbuiTeringby continuousinfusionof acidic solutionin order to mimicthe clinicalacidosisin inflamed tissues: This procedure was shown to establish a steady-stateconditionin whicha limitedarea of the skin is subjectto pH-levelsthat are excitatoryto nociceptive nerveendings(Steenet al. 1995a).In the present study, phosphate-bufferedSIF (as in the electrophysiological work of Steen et al. 1992)was used in contrast to the previouslyappliedisotonicphosphatebuffer (Steenand Reeh 1993)whichhas a much higherbufferingcapacity. Therefore,higherinfusionrates (15-30 rnllh)wereneeded to achieve pain ratings around 20Y0 of the VAS extension.

168

Selection of hormonal mediators for this study

It is wellknown that variousinflammatorymediators are able to excite nociceptivenerve endings only transiently (Kanaka et al. 1985; Kumazawa et al. 1987; Handwerkeret al. 1990;Lang et al. 1990)becausetheir effects are restricted by tachyphylaxis or adaptation (Fig. 5). On the other hand, in clinicallyinflamedstates, multiplehormonal mediatorshave been found together in the exudate(Handwerkerand Reeh 1991)and inflammatory pain is likelyto be inducedby a combinationof IM. To imitate the effit of the multitude of agents releasedin an acute inflammation,an ample combination of mediators has been investigatedin this study: BK, 5-HT, HIS and PGE2. The same combination has previouslybeen demonstratedto excitenociceptorsin a rat skinnervepreparation (Kessleret al. 1992),however long-lastingeffectswere only achievedin high (10-5M) concentration (Reischl, Steen and Reeh, unpublished). Recently, we were able to demonstrate mutual interaction betweenlow pH and this combinationof IM (in submaximal(10-6 M) concentration)in the skin nerve preparation in vitro (Steen et al. 1995b).Although the IM do not representthe completecompositionof an inflammatoryexudate,wheree.g. leukotrienes,thromboxanes, substanceP, PAF, ATP and free radicalsmay also play a role,many importantmediatorsshouldhavebeen considered in the combination. However, the present study supports the suggestionfrom the in vitro experiment that hydrogenions surpass the algogenicpotency of this combination of IM. The algogenicpotency of inflammatory mediators and of protons

Psychophysiologicalwork on the algogeniceffectsof the combinationof inflamrnatorymediatorsused in this study, bradykinin (BK), serotonin (5-HT), histamine (HIS) and prostaglandin Ez (PGEJ, has not been previouslypublished,althoughit is known from the singlecompoundsthat someare painfulwhen injectedand that somecan inducehyperalgesia.In blisterbaseexperimentsit was demonstratedthat 5-HT and BK inducea mild to severebut transient pain (Keeleand Armstrong 1964;Richardsonet al. 1985)and HIS an itch sensation (Keeleand Armstrong 1964).Intradermal injectionsof BK also induce pain (Corrnia and Dougherty 1960; Manning et al. 1991)and a neurogenicflare reaction starting with 10-5 M concentrations (Manning et al. 1991).Intradennal injectionsof HIS induce a pure itchingsensation(Simoneet al. 1991).OnlyfollowingBK preinjectionsdid HIS induce a burning pain sensation (Koppert et al. 1993).Prostaglandinsdo not evokepain when injectedinto human skin (Crunkhom and Willis 1971)but are believedto act via sensitizationof nerve terminalsto other chemicalagents,though the evidence is controversial(Mizumuraet al. 1987;Langet al. 1990).

The present paper demonstrates a linear doseresponserelationshipfor the algogeniceffect of the IM combinationwith significantdifferencesin magnitudes of pain inducedby differentconcentrationsof IM (Fig. 3). This linearityis perfectlymaintainedif the IM injectionsare made into acidoticskin,but the pain responses are significantlylarger than in neutral skin. Under the present experimental conditions, these findings may reflecta relevantincreasein algogenicpotencyor, more psychologically,it may mean that painful stimuli applied on top of a mild preexistingpain (as present in acidoticskin,seeFig. 1)are perceivedmore painfulthan in skin free of interfering sensations. We suggest the formerinterpretationwhichis congruentwith the results of our previous electrophysiologicalstudy on primary cutaneousatTerents,in vitro (Steenet al. 1995b).In addition,the two experimentalserieswith bolus injections of plain acidic btier were both made in acidotic skin, and the subjectswereall able, under doubleblind conditions,to expressthe (significant)increasein painfulness inducedby localpretreatmentwith IM (Fig.4). This fin: ding should allow us to discard the second, more psychologicalinterpretation of the basic data. On the other hand, the intracutaneous pretreatment with IM may inducelocalvasodilatation,in addition to that due to the acidosis,and plasma extravasationand, by that, alterthe tissueconditionsunderlyingthe effectsof acidic bufferboluses.However,these‘inflammatory’reactions are more likelyto increasethan decrease the buffering capacity of the skin and the clearance of the injected protons which would not explain their enhanced algogenicaction. Possible mechanisms of the mutual potentiation between low pH and Inj7ammatoryMediators

The questionas to whichcomponentof the IM is the most likely agent to interact with acid pH cannot be answered from the present psychophysiologicalstudy. In electrophysiologicalexperimentson nociceptorsin a rat skin-nervepreparation, PGE2 and HIS have been shown not to play an important role among the IM (Lang et al. 1990).In the cutaneouspreparation, PGEZ neitherexcitesnor sensitizesnociceptors,in contrast to a testicularpreparation in vitro (Mizumuraet al. 1987) and to the cat kneejoint in vivo (Schaibleand Schmidt 1988).HIS can activate only a small population of polymodalC-fibersin the skin preparation. It did excite these fibers significantlymore after BK pretreatment, but its effectdid not add to an ongoingBK-induceddischarge(Koppert et al. 1993).BK and 5-HT, each excite a proportion of nocieeptors,and 5-HT is able to sensitizethe nociceptorsfor the action of BK (Lang et al. 1990).The excitatory effect of the IM combination seemsto depend on this sensitization,since the 5-HT3 antagonist metoclopramidecould suppress the IM responsereversiblyand dose-dependently(Handwerkeret

169

al. 1990).Thus, there are indicationsthat BK and 5-HT are the most relevantcompoundsamong the IM and the most likelycandidatesfor an interactionwith hydrogen ions, In patch clampexperimentson DRG cells,a sustained cationic inward current activated by protons has been demonstrated(Bevanand Yeats 1991).BK also evokes an unspecificcationicinwardcurrent in thesecellswhich may relate to its excitatory action on nociceptiveafferents(Burgesset al. 1989).On the other hand, BK also acts through G-protein dependentactivation of protein kinase C which may convey its sensitizingproperties (Dray et al. 1988).This may provide a future working hypothesisto explain the inflammatorypotentiation of the pH-induced nociceptor dischargeand pain. The recognitionof the excitatory and the sensitizing actionsof inflammatorymediatorsonto nociceptorsdid not really help to explain ongoing ‘spontaneous’and recurrent pain from irdlamed tissues, since tachyphylaxisturned out to limitthe effectsduring prolonged or repeatedexposure(Kanaka et al. 1985;Kumazawaet al. 1987; Handwerker et al. 1990; Lang et al. 1990; Kessleret al. 1992).The findingof sustainednociceptor dischargeand pain induced by local acidosistypical of inflamed tissue filled a logical gap (Steen et al. 1992; Steenand Reeh 1993).However,to causepain, nociceptor activitymust exceeda certain, yet undefinedquantity. The prominent synergism between inflammatory mediators and tissue acidosis, concluded from our results, may provide this critical quantity by increasing temporal as well as spatial summation of nociceptive input. Acknowledgement This project was supported by the Deutsche Forschungsgemeinschaft,grant Ste 593/1-2.The work of P.W. Reeh was supported by the DeutscheForschungsgemeinschaft,SFB 353-A3.The authors wish to thank Jorg Breuer for developingmost of the excellentprograms and hardware used in this study.

Beck,P.W. and Handwerker,H.O., Bradykininand serotonineffects onvarioustypes of cutaneousnervefibres,Ptliig.Arch., 347(1974) 209-222. Belmonte, C., Gallar, J., Lopez-Bnones, L. and Pozo, M.A., Polymodality in nociceptive neurons: experimentedmodels of

chemotransduction. In:L. Urban(Ed.),afl~sr M-~sms of Sensorypmcessin~ Vol. 79, Springer,Berlin, 1994,PP. 87-117. Bevrm,S. and Yeats, J., Protons activate a cation conductancein a sub-populationof rat dorsal mot ganglionneurones, J. Physiol. (Imnd.), 433(1991)145-161. Bretag,A.H., Syntheticinterstitial fluid for isolated rnamrmdiantissue, Life Sci., 8 (1969)319-329. Burgess,G.M., MuUaney, L, McNeill, M., Dunn, P.M. and Rang,

H.P., Secondmessengerinvolvedin the mechanismof action of bradykininin sensory neurons in culture, J. Neurosci., 9 (1989) 3314-3325. Cmnkhorn, P. and Willis,A.L., Cutaneousreactions to intraderrnal prostaglandins,Br. J. PklMUICOL, 41 (1971)49-56. Cormia, F.E. and Doughe@, J.W., Proteolyticactivity in develop ment of pain and itching:cutaneous reactions to bradykininand kallikrein,J. Invest. DerrnatoL,35 (1960)21-26. Dray, A., Bettaney,J., Forster, P. and Perkins, M.N., Bradykinininducedstimulationof tierent fibres is mediatedthrough protein kinaseC, Neurosci.Mt., 91 (1988)301-307. Fjiitlbrant,N. and Iggo,A., The effect of histamine,5-hydroxytryptarnineand acetylcholineon cutaneous atTerentfibrea, J. Physiol. (Lend.), 156(1%1) 578-590. Fock, S. and Mense, S., Excitatoryeffkctsof 5-hydroxytryptamine, histamineand potassiumionson musculargroupIV tierent units: A comparisonwith bradykinin,Brain Res., 105(1976)459-469. Hiibler,C., ~r den K+-und Ca++-Gehaltvon Eiter und Exsudaten und seineBeziehungenzum Entziindungsschrnmz,Klin. Wochenschr., 8 (1929)1569-1572. Handwerker,H.O. and Reeh, P.W., Pain and inflammation. In: M.R. Bond, J.E. Charlton and C.J. Woolf (Eds.), Proc. VIti World Congrmson Pain, Elsevier,Amster~ 1991,pp. 59-70. Handwerker,H.O., Reeh, P.W. and Steen, K.H., Effectsof 5 HT on nociceptors.In: J.M. Besson(Ed.), Serotoninand Pain, Ekevier, Amsterd~ 1990,pp. 1-15. Hiss,E. and Mense,S., Evidencefor the existenceof differentreceptor sites for algesic agents at the endings of muscular Group IV atTerentunits, PtXig.Arch., 362(1976)141-146. Kanaka, R., Schaible,H.-G. and Schmidt, R.F., Activation of fine articular atTerentunits by bradykioin, Brain Rea., 327 (1985) 81-90. Kcele,C.A.and Armstron&D., Paindue to acidsand alkalis; In: Substances producingPain and Itch, E. Arnold, London, 1964,pp. 73-ss.

Kessler,W., Kirchhoff,C., Rech, P.W. and Handwerker,H.O., Excitationof cutaneousatTerentnerveendingsin vitro by a combination of irdlammatory mediators and conditioning effect of substauceP, Exp. Brain Res., 91 (1992)467-476. Khan, A.A., Raja, S.N., Manning D.C., Campbell,J.N. and Meyer, R.A., The effectsof bradykininand sequence-relatedanalogs on the response properties of cutaneous nociceptors in monkeys, Somatosens.Mot. Rea., 9 (1992)97-106. Koppert,W., Reeh,P.W.and Handwerker,H.O., Conditioningof histaminebybradykininalters responsesof rat nociceptorand human itch sensation,Neumsci. Lett., 152(1993)117-120. K umazaw~ T., Mizurm@ K. and Sate, J., Responseproperties of polymmdalreceptorsstudiedusingin vitro testissuperiorspermatic nervepreparationsof dogs, J. Neurophysiol.,57 (19S7)702-711. Lang E., Novak,A., Rech, P.W. and Handwerker,H.O., Chemosensitivityof tine fierents from rat skin in vitro. J. Neurophysiol.,63 (1990)887-901. Lindabl,O., Experimentalskin pain, Acta Physiol.Scand., 51(Suppl. 179)(1%1) 1-90. Manning D.C., Raja, S.N., Meyer, R.A. and Campbell,J.N., Pain and hyperalgesiaafter intraderrrudinjectionof bradykininin humans, Clin. Pharmacd. Ther., 50 (1991)721-729. Mizumura,K., Minagawa,M., Tsujii,Y. and Kumazawa, T., The effectsof bradykininagonistsand antagonistson visceralpolymodal receptor activities,Pain, 40 (1990)221-227. Mizumura,K., Sate, J. and Kumszawa, T., Effects of prostaglandins

170 and other putativechemicalintermediarieson the activityof canine testicular polymodalreceptors studied in vitro, Plliig. Arch., 408(1987)565-572. Neugebauer,V., !lchaible,U.-G. and Schmidt,R.F., Sensitizationof articular atTerentsto mechanical stimuli by bradykinin, Pfliig. Arch., 415(1989)330-335. Richardson,B.P., Engel,G., Dortatsch,P. and Stadler, P.A. Identification of serotonin M-receptorsubtypesand their specificbbckade by a new class of drugs, Nature, 316(1985)126-131. Rueff, A, and Dray, A., Pharmacologicalcharacterizationof the effects of 5-hydroxytryptamineand different prostaglandins on peripheral sensory neurons in vitro. Agents Actions, 38 (1993) C13-C15. schaible,H.-G.and Schmidt,R.F., Excitationand sensitizationof tine articular afferents from cat’s knee joint by prostaglandin E2, J. Physiol.(Lend.), 403(1988)91-104. Sirnone,D.A., Alreja, M. and LaMotte, R.H., Psychophysicalstudies of the itch sensation and itchy skin (’alloknesis’)produced by intracutarreousinjection of histamine.Somatosens.Mot. Res., 8 (1991)271-279.

Steen,A.E., Reeh, P.W., Kreysel,H.W. and Steen, K.H., Experimental tissue acidosis potentates pain induced by inflammatory mediators,VIIthWorld Congresson Pain (Paris) (1993)R 39. Steen,K.H., Reeh, P.W., Anton, F. and Handwerker,H.O., Protons selectivelyinducelastingexcitationand sensitizationto mechanical stimulationof nociceptorsin rat skin, in vitro, J. Neurosci., 12(1) (1992)86-95. Steen,K.H. and Reeh, P.W., Sustainedgraded pain and hyperalgesia from harmless experimental tissue acidosis in human skin, Neurosci.L&t., 154(1993)113-116. Steen,K.H., Issbemer, U. and Reeh, P.W., Pain due to experimental acidosis in human skin: evidence for non-adapting nociceptor excitation,Neurosci,Lett., 199(1995a)29-32. Steen,K.H., Steen,A.E. and Reeh,P.W., A dominantrole of acid pH in inflammatory excitation and sensitizationof nociceptorsin rat skin, in vitro, J. Neurosci., 15(5)(1995b)3982-3989. von Gaza, W. and Brandi, B., Beziehungerrzwischen Wasserstoftionenkonzentrationund Schmerzemptindung,Klin.Wocherrschr., 5 (1926)1123-1127.